diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/activations.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/activations.rst.txt
deleted file mode 100644
index 269e6491e7ebe3899c3fb24fca756a393043473b..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/activations.rst.txt
+++ /dev/null
@@ -1,108 +0,0 @@
-===========
-Activations
-===========
-
-BaseActivation
-==============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: BaseActivation
-    :noindex:
-    
-AbsActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: AbsActivation
-    :noindex:
-    
-ExpActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: ExpActivation
-    :noindex:
-    
-IdentityActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: IdentityActivation
-    :noindex:
-    
-LinearActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: LinearActivation
-    :noindex:
-
-LogActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: LogActivation
-    :noindex:
-    
-SquareActivation
-================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SquareActivation
-    :noindex:
-    
-SigmoidActivation
-=================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SigmoidActivation
-    :noindex:
-    
-SoftmaxActivation
-=================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SoftmaxActivation
-    :noindex:
-    
-SequenceSoftmaxActivation
-=========================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SequenceSoftmaxActivation
-    :noindex:
-    
-ReluActivation
-==============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: ReluActivation
-    :noindex:
-    
-BReluActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: BReluActivation
-    :noindex:
-    
-SoftReluActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SoftReluActivation
-    :noindex:
-    
-TanhActivation
-==============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: TanhActivation
-    :noindex:
-    
-STanhActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: STanhActivation
-    :noindex:
diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/attrs.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/attrs.rst.txt
deleted file mode 100644
index ac63127bf7d9db6351365ab7b58f43db12347a8e..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/attrs.rst.txt
+++ /dev/null
@@ -1,5 +0,0 @@
-Parameter Attributes
-=======================
-
-..  automodule:: paddle.trainer_config_helpers.attrs
-    :members:
diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/data_sources.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/data_sources.rst.txt
deleted file mode 100644
index b9dd4dda01ae59d1260356aff50ddf298d02c87f..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/data_sources.rst.txt
+++ /dev/null
@@ -1,7 +0,0 @@
-..  _api_trainer_config_helpers_data_sources:
-
-DataSources
-===========
-
-..  automodule:: paddle.trainer_config_helpers.data_sources
-    :members:
diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/evaluators.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/evaluators.rst.txt
deleted file mode 100644
index 11dc735164284d6ed1d661fab1e7690d263b3a7c..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/evaluators.rst.txt
+++ /dev/null
@@ -1,108 +0,0 @@
-..  _api_trainer_config_helpers_evaluators:
-
-==========
-Evaluators
-==========
-
-Base
-====
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: evaluator_base
-    :noindex:
-
-Classification 
-==============
-
-classification_error_evaluator
-------------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: classification_error_evaluator
-    :noindex:
-
-auc_evaluator
--------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: auc_evaluator
-    :noindex:
-
-ctc_error_evaluator
--------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: ctc_error_evaluator
-    :noindex:
-
-chunk_evaluator
----------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: chunk_evaluator
-    :noindex:
-
-precision_recall_evaluator
---------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  precision_recall_evaluator
-    :noindex:
-
-Rank
-====
-
-pnpair_evaluator
-----------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  pnpair_evaluator
-    :noindex:
-
-Utils
-=====
-
-sum_evaluator
--------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: sum_evaluator
-    :noindex:
-
-column_sum_evaluator
---------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: column_sum_evaluator
-    :noindex:
-
-Print
-=====
-
-classification_error_printer_evaluator
---------------------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  classification_error_printer_evaluator
-    :noindex:
-
-gradient_printer_evaluator
---------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  gradient_printer_evaluator
-    :noindex:
-
-maxid_printer_evaluator
------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  maxid_printer_evaluator
-    :noindex:
-
-maxframe_printer_evaluator
----------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  maxframe_printer_evaluator
-    :noindex:
-
-seqtext_printer_evaluator
--------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  seqtext_printer_evaluator
-    :noindex:
-
-value_printer_evaluator
------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  value_printer_evaluator
-    :noindex:
-
diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/layers.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/layers.rst.txt
deleted file mode 100644
index 75c1b35246486528524fd5baa04704249d5a9275..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/layers.rst.txt
+++ /dev/null
@@ -1,514 +0,0 @@
-..  _api_trainer_config_helpers_layers:
-
-======
-Layers
-======
-
-Base
-======
-
-LayerType
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: LayerType
-    :noindex:
-
-LayerOutput
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: LayerOutput
-    :noindex:
-
-Data layer
-===========
-
-..  _api_trainer_config_helpers_layers_data_layer:
-
-data_layer
-----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: data_layer
-    :noindex:
-
-Fully Connected Layers
-======================
-
-..  _api_trainer_config_helpers_layers_fc_layer:
-
-fc_layer
---------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: fc_layer
-    :noindex:
-
-selective_fc_layer
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: selective_fc_layer
-    :noindex:
-
-Conv Layers
-===========
-
-conv_operator
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: conv_operator
-    :noindex:
-
-conv_projection
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: conv_projection
-    :noindex:
-
-conv_shift_layer
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: conv_shift_layer
-    :noindex:
-
-img_conv_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: img_conv_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_context_projection:
-
-context_projection 
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: context_projection
-    :noindex:
-
-Image Pooling Layer
-===================
-
-img_pool_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: img_pool_layer
-    :noindex:   
-
-spp_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: spp_layer
-    :noindex:
-
-maxout_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: maxout_layer
-    :noindex:
-
-Norm Layer
-==========
-
-img_cmrnorm_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: img_cmrnorm_layer
-    :noindex:
-
-batch_norm_layer
----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: batch_norm_layer
-    :noindex:
-
-sum_to_one_norm_layer
----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: sum_to_one_norm_layer
-    :noindex:
-    
-Recurrent Layers
-================
-
-recurrent_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: recurrent_layer
-    :noindex:
-
-lstmemory
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: lstmemory
-    :noindex:
-
-grumemory
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: grumemory
-    :noindex:
-
-Recurrent Layer Group
-=====================
-
-memory
-------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: memory
-    :noindex:
-
-recurrent_group
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: recurrent_group
-    :noindex:
-    
-lstm_step_layer
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: lstm_step_layer
-    :noindex:
-
-gru_step_layer
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: gru_step_layer
-    :noindex:
-
-beam_search
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: beam_search
-    :noindex:
-    
-get_output_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: get_output_layer
-    :noindex:
-    
-Mixed Layer
-===========
-
-..  _api_trainer_config_helpers_layers_mixed_layer:
-
-mixed_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: mixed_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_embedding_layer:
-
-embedding_layer
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: embedding_layer
-    :noindex:
-
-scaling_projection
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: scaling_projection
-    :noindex:
-
-dotmul_projection
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: dotmul_projection
-    :noindex:
-
-dotmul_operator
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: dotmul_operator
-    :noindex:
-
-full_matrix_projection
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: full_matrix_projection
-    :noindex:
-
-identity_projection
--------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: identity_projection
-    :noindex:
-
-
-table_projection
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: table_projection
-    :noindex:
-
-trans_full_matrix_projection
-----------------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: trans_full_matrix_projection
-    :noindex:
-    
-Aggregate Layers
-================
-
-..  _api_trainer_config_helpers_layers_pooling_layer:
-
-pooling_layer
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: pooling_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_last_seq:
-
-last_seq
---------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: last_seq
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_first_seq:
-
-first_seq
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: first_seq
-    :noindex:
-
-concat_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: concat_layer
-    :noindex:
-
-seq_concat_layer
-----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: seq_concat_layer
-    :noindex:
-
-Reshaping Layers
-================
-
-block_expand_layer
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: block_expand_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_expand_layer:
-
-expand_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: expand_layer
-    :noindex:
-
-repeat_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: repeat_layer
-    :noindex:
-
-rotate_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: rotate_layer
-    :noindex:
-
-seq_reshape_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: seq_reshape_layer
-    :noindex:
-
-Math Layers
-===========
-
-addto_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: addto_layer
-    :noindex:
-
-linear_comb_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: linear_comb_layer
-    :noindex:
-
-interpolation_layer
--------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: interpolation_layer
-    :noindex:
-
-bilinear_interp_layer
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: bilinear_interp_layer
-    :noindex:
-
-power_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: power_layer
-    :noindex:
-
-scaling_layer
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: scaling_layer
-    :noindex:
-
-slope_intercept_layer
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: slope_intercept_layer
-    :noindex:
-
-tensor_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: tensor_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_cos_sim:
-
-cos_sim
--------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: cos_sim
-    :noindex:
-
-trans_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: trans_layer
-    :noindex:
-
-Sampling Layers
-===============
-
-maxid_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: maxid_layer
-    :noindex:
-
-sampling_id_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: sampling_id_layer
-    :noindex:
-
-Slicing and Joining Layers
-==========================
-
-pad_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: pad_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_cost_layers:
-
-Cost Layers
-===========
-
-cross_entropy
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: cross_entropy
-    :noindex:
-
-cross_entropy_with_selfnorm
----------------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: cross_entropy_with_selfnorm
-    :noindex:
-
-multi_binary_label_cross_entropy
---------------------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: multi_binary_label_cross_entropy
-    :noindex:
-
-mse_cost
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: mse_cost
-    :noindex:
-
-huber_cost
-----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: huber_cost
-    :noindex:
-
-lambda_cost
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: lambda_cost
-    :noindex:
-
-rank_cost
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: rank_cost
-    :noindex:
-
-sum_cost
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: sum_cost
-    :noindex:
-
-crf_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: crf_layer
-    :noindex:
-
-crf_decoding_layer
--------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: crf_decoding_layer
-    :noindex:
-
-ctc_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: ctc_layer
-    :noindex:
-
-warp_ctc_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: warp_ctc_layer
-    :noindex:
-
-nce_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: nce_layer
-    :noindex:
-
-hsigmoid
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: hsigmoid
-    :noindex:
-
-smooth_l1_cost
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: smooth_l1_cost
-    :noindex:
-
-Check Layer 
-============
-
-eos_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: eos_layer
-    :noindex:
diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/networks.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/networks.rst.txt
deleted file mode 100644
index edb53acbf0c31532aa34bda044066fed72eaa426..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/networks.rst.txt
+++ /dev/null
@@ -1,123 +0,0 @@
-========
-Networks
-========
-
-The networks module contains pieces of neural network that combine multiple layers.
-
-NLP
-===
-
-sequence_conv_pool
-------------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: sequence_conv_pool
-    :noindex:
-
-..  _api_trainer_config_helpers_network_text_conv_pool:
-
-text_conv_pool
---------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: text_conv_pool
-    :noindex:
-
-Images
-======
-
-img_conv_bn_pool
-----------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: img_conv_bn_pool
-    :noindex:
-
-img_conv_group
---------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: img_conv_group
-    :noindex:
-
-..  _api_trainer_config_helpers_network_simple_img_conv_pool:
-
-simple_img_conv_pool
---------------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_img_conv_pool
-    :noindex:
-
-vgg_16_network
----------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: vgg_16_network
-    :noindex:
-
-Recurrent
-=========
-
-LSTM
-----
-
-lstmemory_unit
-``````````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: lstmemory_unit
-    :noindex:
-
-lstmemory_group
-```````````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: lstmemory_group
-    :noindex:
-
-simple_lstm
-```````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_lstm
-    :noindex:
-
-bidirectional_lstm
-``````````````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: bidirectional_lstm
-    :noindex:
-
-GRU
----
-
-gru_unit
-````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: gru_unit
-    :noindex:
-
-gru_group
-`````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: gru_group
-    :noindex:
-
-simple_gru
-``````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_gru
-    :noindex:
-
-simple_attention
-----------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_attention
-    :noindex:
-
-Miscs
-=====
-
-dropout_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: dropout_layer
-    :noindex:
-
-outputs
--------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: outputs
-    :noindex:
diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/optimizers.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/optimizers.rst.txt
deleted file mode 100644
index d2f4958c92b8e3b7426945f1af07112ab4071136..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/optimizers.rst.txt
+++ /dev/null
@@ -1,61 +0,0 @@
-..  _api_trainer_config_helpers_optimizers:
-
-==========
-Optimizers
-==========
-
-BaseSGDOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: BaseSGDOptimizer
-    :noindex:
-
-MomentumOptimizer
-=================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: MomentumOptimizer
-    :noindex:
-
-AdamOptimizer
-=============
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdamOptimizer
-    :noindex:
-
-AdamaxOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdamaxOptimizer
-    :noindex:
-
-AdaGradOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdaGradOptimizer
-    :noindex:
-
-DecayedAdaGradOptimizer
-=======================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: DecayedAdaGradOptimizer
-    :noindex:
-
-AdaDeltaOptimizer
-=================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdaDeltaOptimizer
-    :noindex:
-
-RMSPropOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: RMSPropOptimizer
-    :noindex:
-
-..  _api_trainer_config_helpers_optimizers_settings:
-
-settings
-========
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: settings
-    :noindex:
diff --git a/develop/doc/_sources/api/v1/trainer_config_helpers/poolings.rst.txt b/develop/doc/_sources/api/v1/trainer_config_helpers/poolings.rst.txt
deleted file mode 100644
index 66566809d26f59263597b5286c5b27e0bbc9415a..0000000000000000000000000000000000000000
--- a/develop/doc/_sources/api/v1/trainer_config_helpers/poolings.rst.txt
+++ /dev/null
@@ -1,33 +0,0 @@
-========
-Poolings
-========
-
-BasePoolingType
-===============
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: BasePoolingType
-    :noindex:
-
-AvgPooling
-==========
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: AvgPooling
-    :noindex:
-
-MaxPooling
-==========
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: MaxPooling
-    :noindex:
-
-SumPooling
-==========
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: SumPooling
-    :noindex:
-
-SquareRootNPooling
-==================
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: SquareRootNPooling
-    :noindex:
diff --git a/develop/doc/_sources/api/v2/config/evaluators.rst.txt b/develop/doc/_sources/api/v2/config/evaluators.rst.txt
new file mode 100644
index 0000000000000000000000000000000000000000..39db51fa4abc370855ca3f2778b47464f33b6fce
--- /dev/null
+++ b/develop/doc/_sources/api/v2/config/evaluators.rst.txt
@@ -0,0 +1,101 @@
+..  _api_v2:
+
+==========
+Evaluators
+==========
+
+Classification 
+==============
+
+classification_error
+--------------------
+..  automodule:: paddle.v2.evaluator
+    :members: classification_error
+    :noindex:
+
+auc
+---
+..  automodule:: paddle.v2.evaluator
+    :members: auc
+    :noindex:
+
+ctc_error
+---------
+..  automodule:: paddle.v2.evaluator
+    :members: ctc_error
+    :noindex:
+
+chunk
+-----
+..  automodule:: paddle.v2.evaluator
+    :members: chunk
+    :noindex:
+
+precision_recall
+----------------
+..  automodule:: paddle.v2.evaluator
+    :members:  precision_recall
+    :noindex:
+
+Rank
+====
+
+pnpair
+------
+..  automodule:: paddle.v2.evaluator
+    :members:  pnpair
+    :noindex:
+
+Utils
+=====
+
+sum
+---
+..  automodule:: paddle.v2.evaluator
+    :members: sum
+    :noindex:
+
+column_sum
+----------
+..  automodule:: paddle.v2.evaluator
+    :members: column_sum
+    :noindex:
+
+Print
+=====
+
+classification_error_printer
+----------------------------
+..  automodule:: paddle.v2.evaluator
+    :members:  classification_error_printer
+    :noindex:
+
+gradient_printer
+----------------
+..  automodule:: paddle.v2.evaluator
+    :members:  gradient_printer
+    :noindex:
+
+maxid_printer
+-------------
+..  automodule:: paddle.v2.evaluator
+    :members:  maxid_printer
+    :noindex:
+
+maxframe_printer
+----------------
+..  automodule:: paddle.v2.evaluator
+    :members:  maxframe_printer
+    :noindex:
+
+seqtext_printer
+---------------
+..  automodule:: paddle.v2.evaluator
+    :members:  seqtext_printer
+    :noindex:
+
+value_printer
+-------------
+..  automodule:: paddle.v2.evaluator
+    :members:  value_printer
+    :noindex:
diff --git a/develop/doc/_sources/api/v2/config/networks.rst.txt b/develop/doc/_sources/api/v2/config/networks.rst.txt
index 6f209bc95bec7279051118bb857a96515e0371a9..b2a617fff134035c04eeabbbaf6d9cbe2a525f1c 100644
--- a/develop/doc/_sources/api/v2/config/networks.rst.txt
+++ b/develop/doc/_sources/api/v2/config/networks.rst.txt
@@ -44,6 +44,12 @@ simple_img_conv_pool
     :members: simple_img_conv_pool
     :noindex:
 
+small_vgg
+---------
+..  automodule:: paddle.v2.networks
+    :members: small_vgg
+    :noindex:
+
 vgg_16_network
 ---------------
 ..  automodule:: paddle.v2.networks
@@ -101,6 +107,18 @@ simple_gru
     :members: simple_gru
     :noindex:
 
+simple_gru2
+```````````
+..  automodule:: paddle.v2.networks
+    :members: simple_gru2
+    :noindex:
+
+bidirectional_gru
+``````````````````
+..  automodule:: paddle.v2.networks
+    :members: bidirectional_gru
+    :noindex:
+
 simple_attention
 ----------------
 ..  automodule:: paddle.v2.networks
diff --git a/develop/doc/_sources/api/v2/model_configs.rst.txt b/develop/doc/_sources/api/v2/model_configs.rst.txt
index a5fae7e29e56d9e236d489353a5c8967d1954641..992b559cbd87244612521d4c96f84f997d6c4196 100644
--- a/develop/doc/_sources/api/v2/model_configs.rst.txt
+++ b/develop/doc/_sources/api/v2/model_configs.rst.txt
@@ -6,6 +6,7 @@ Model Configuration
 
     config/activation.rst
     config/layer.rst
+    config/evaluators.rst
     config/optimizer.rst
     config/pooling.rst
     config/networks.rst
diff --git a/develop/doc/about/index_en.html b/develop/doc/about/index_en.html
index 70a56744565756a5e92c229abf8a1bbf6aa7e474..7b51d2bad6c8a1221d807683b8d10e18009584f6 100644
--- a/develop/doc/about/index_en.html
+++ b/develop/doc/about/index_en.html
@@ -136,6 +136,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/api/index_en.html b/develop/doc/api/index_en.html
index 1738caea7d0edbecdd781f0bb75c576d78bbdd28..8b12f7178a55f9eb85ec8b89946ebf55811f3061 100644
--- a/develop/doc/api/index_en.html
+++ b/develop/doc/api/index_en.html
@@ -137,6 +137,7 @@
 <li class="toctree-l2"><a class="reference internal" href="v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/api/v1/data_provider/dataprovider_en.html b/develop/doc/api/v1/data_provider/dataprovider_en.html
index ee1b046476565a6e4bc330b101fa5683d1688632..9f7fea10a7a64227e0c1e8583202567878305e65 100644
--- a/develop/doc/api/v1/data_provider/dataprovider_en.html
+++ b/develop/doc/api/v1/data_provider/dataprovider_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/api/v1/data_provider/pydataprovider2_en.html b/develop/doc/api/v1/data_provider/pydataprovider2_en.html
index ed4eacccd7cefb618e2b67af82a25ee6b7171edc..673a29f66f6d1624a09b1bc7b64d1f39708b2c7a 100644
--- a/develop/doc/api/v1/data_provider/pydataprovider2_en.html
+++ b/develop/doc/api/v1/data_provider/pydataprovider2_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/api/v1/index_en.html b/develop/doc/api/v1/index_en.html
index d000bcdc566750d8d0d947264bf8c6f18b348c8f..8ce28b1666a09f5b032846fabaac538f680d8619 100644
--- a/develop/doc/api/v1/index_en.html
+++ b/develop/doc/api/v1/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/networks.html">Networks</a></li>
@@ -189,16 +190,6 @@
 <div class="section" id="model-config-api">
 <span id="api-trainer-config"></span><h2>Model Config API<a class="headerlink" href="#model-config-api" title="Permalink to this headline">¶</a></h2>
 <div class="toctree-wrapper compound">
-<ul>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/optimizers.html">Optimizers</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/data_sources.html">DataSources</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/layers.html">Layers</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/activations.html">Activations</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/poolings.html">Poolings</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/networks.html">Networks</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/evaluators.html">Evaluators</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/attrs.html">Parameter Attributes</a></li>
-</ul>
 </div>
 </div>
 <div class="section" id="applications-api">
diff --git a/develop/doc/api/v1/predict/swig_py_paddle_en.html b/develop/doc/api/v1/predict/swig_py_paddle_en.html
index 02c77b18fd08d47872c0c3e4d71e9e946fd922f9..7af960ddbb5ee2780d223575b25489d9972c396b 100644
--- a/develop/doc/api/v1/predict/swig_py_paddle_en.html
+++ b/develop/doc/api/v1/predict/swig_py_paddle_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/api/v1/trainer_config_helpers/activations.html b/develop/doc/api/v1/trainer_config_helpers/activations.html
deleted file mode 100644
index a8af1617c0fac758c03074bbf026f733b2281bed..0000000000000000000000000000000000000000
--- a/develop/doc/api/v1/trainer_config_helpers/activations.html
+++ /dev/null
@@ -1,373 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Activations &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="../../../genindex.html"/>
-        <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Activations</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="activations">
-<h1>Activations<a class="headerlink" href="#activations" title="Permalink to this headline">¶</a></h1>
-<div class="section" id="baseactivation">
-<h2>BaseActivation<a class="headerlink" href="#baseactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">BaseActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Base</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="absactivation">
-<h2>AbsActivation<a class="headerlink" href="#absactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">AbsActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Abs</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="expactivation">
-<h2>ExpActivation<a class="headerlink" href="#expactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">ExpActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Exp</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="identityactivation">
-<h2>IdentityActivation<a class="headerlink" href="#identityactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">IdentityActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Linear</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="linearactivation">
-<h2>LinearActivation<a class="headerlink" href="#linearactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">LinearActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Linear</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="logactivation">
-<h2>LogActivation<a class="headerlink" href="#logactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">LogActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Log</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="squareactivation">
-<h2>SquareActivation<a class="headerlink" href="#squareactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SquareActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Square</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="sigmoidactivation">
-<h2>SigmoidActivation<a class="headerlink" href="#sigmoidactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SigmoidActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Sigmoid</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="softmaxactivation">
-<h2>SoftmaxActivation<a class="headerlink" href="#softmaxactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SoftmaxActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Softmax</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="sequencesoftmaxactivation">
-<h2>SequenceSoftmaxActivation<a class="headerlink" href="#sequencesoftmaxactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SequenceSoftmaxActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">SequenceSoftmax</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="reluactivation">
-<h2>ReluActivation<a class="headerlink" href="#reluactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">ReluActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Relu</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="breluactivation">
-<h2>BReluActivation<a class="headerlink" href="#breluactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">BReluActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">BRelu</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="softreluactivation">
-<h2>SoftReluActivation<a class="headerlink" href="#softreluactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SoftReluActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">SoftRelu</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="tanhactivation">
-<h2>TanhActivation<a class="headerlink" href="#tanhactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">TanhActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Tanh</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="stanhactivation">
-<h2>STanhActivation<a class="headerlink" href="#stanhactivation" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">STanhActivation</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">STanh</span></code></p>
-</dd></dl>
-
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/api/v1/trainer_config_helpers/attrs.html b/develop/doc/api/v1/trainer_config_helpers/attrs.html
deleted file mode 100644
index 4fec903a1a34e20e68cf704f46a057b42038b82c..0000000000000000000000000000000000000000
--- a/develop/doc/api/v1/trainer_config_helpers/attrs.html
+++ /dev/null
@@ -1,328 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Parameter Attributes &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="../../../genindex.html"/>
-        <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Parameter Attributes</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="module-paddle.trainer_config_helpers.attrs">
-<span id="parameter-attributes"></span><h1>Parameter Attributes<a class="headerlink" href="#module-paddle.trainer_config_helpers.attrs" title="Permalink to this headline">¶</a></h1>
-<dl class="class">
-<dt id="paddle.trainer_config_helpers.attrs.ParameterAttribute">
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ParameterAttribute</code><span class="sig-paren">(</span><em>name=None</em>, <em>is_static=False</em>, <em>initial_std=None</em>, <em>initial_mean=None</em>, <em>initial_max=None</em>, <em>initial_min=None</em>, <em>l1_rate=None</em>, <em>l2_rate=None</em>, <em>learning_rate=None</em>, <em>momentum=None</em>, <em>gradient_clipping_threshold=None</em>, <em>sparse_update=False</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="Permalink to this definition">¶</a></dt>
-<dd><p>Parameter Attributes object. To fine-tuning network training process, user
-can set attribute to control training details, such as l1,l2 rate / learning
-rate / how to init param.</p>
-<p>NOTE: IT IS A HIGH LEVEL USER INTERFACE.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>is_static</strong> (<em>bool</em>) &#8211; True if this parameter will be fixed while training.</li>
-<li><strong>initial_std</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Gauss Random initialization standard deviation.
-None if not using Gauss Random initialize parameter.</li>
-<li><strong>initial_mean</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Gauss Random initialization mean.
-None if not using Gauss Random initialize parameter.</li>
-<li><strong>initial_max</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Uniform initialization max value.</li>
-<li><strong>initial_min</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Uniform initialization min value.</li>
-<li><strong>l1_rate</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; the l1 regularization factor</li>
-<li><strong>l2_rate</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; the l2 regularization factor</li>
-<li><strong>learning_rate</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; The parameter learning rate. None means 1.
-The learning rate when optimize is LEARNING_RATE =
-GLOBAL_LEARNING_RATE * PARAMETER_LEARNING_RATE
-* SCHEDULER_FACTOR.</li>
-<li><strong>momentum</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; The parameter momentum. None means use global value.</li>
-<li><strong>gradient_clipping_threshold</strong> (<em>float</em>) &#8211; gradient clipping threshold. If gradient
-value larger than some value, will be
-clipped.</li>
-<li><strong>sparse_update</strong> (<em>bool</em>) &#8211; Enable sparse update for this parameter. It will
-enable both local and remote sparse update.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-<dl class="method">
-<dt id="paddle.trainer_config_helpers.attrs.ParameterAttribute.set_default_parameter_name">
-<code class="descname">set_default_parameter_name</code><span class="sig-paren">(</span><em>name</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ParameterAttribute.set_default_parameter_name" title="Permalink to this definition">¶</a></dt>
-<dd><p>Set default parameter name. If parameter not set, then will use default
-parameter name.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>name</strong> (<em>basestring</em>) &#8211; default parameter name.</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</dd></dl>
-
-<dl class="class">
-<dt id="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute">
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ExtraLayerAttribute</code><span class="sig-paren">(</span><em>error_clipping_threshold=None</em>, <em>drop_rate=None</em>, <em>device=None</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="Permalink to this definition">¶</a></dt>
-<dd><p>Some high level layer attributes config. You can set all attributes here,
-but some layer doesn&#8217;t support all attributes. If you set an attribute to a
-layer that not support this attribute, paddle will print an error and core.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>error_clipping_threshold</strong> (<em>float</em>) &#8211; Error clipping threshold.</li>
-<li><strong>drop_rate</strong> (<em>float</em>) &#8211; Dropout rate. Dropout will create a mask on layer output.
-The dropout rate is the zero rate of this mask. The
-details of what dropout is please refer to <a class="reference external" href="https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf">here</a>.</li>
-<li><strong>device</strong> (<em>int</em>) &#8211; <p>device ID of layer. device=-1, use CPU. device&gt;=0, use GPU.
-The details allocation in parallel_nn please refer to <a class="reference external" href="http://www.paddlepaddle.org/doc/ui/cmd_argument/use_case.html#case-2-specify-layers-in-different-devices">here</a>.</p>
-</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-<dl class="attribute">
-<dt id="paddle.trainer_config_helpers.attrs.ParamAttr">
-<code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ParamAttr</code><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ParamAttr" title="Permalink to this definition">¶</a></dt>
-<dd><p>alias of <a class="reference internal" href="#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><code class="xref py py-class docutils literal"><span class="pre">ParameterAttribute</span></code></a></p>
-</dd></dl>
-
-<dl class="attribute">
-<dt id="paddle.trainer_config_helpers.attrs.ExtraAttr">
-<code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ExtraAttr</code><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ExtraAttr" title="Permalink to this definition">¶</a></dt>
-<dd><p>alias of <a class="reference internal" href="#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><code class="xref py py-class docutils literal"><span class="pre">ExtraLayerAttribute</span></code></a></p>
-</dd></dl>
-
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/api/v1/trainer_config_helpers/data_sources.html b/develop/doc/api/v1/trainer_config_helpers/data_sources.html
deleted file mode 100644
index f2bcc578e491b02ca51c248fe76424170cdb66f5..0000000000000000000000000000000000000000
--- a/develop/doc/api/v1/trainer_config_helpers/data_sources.html
+++ /dev/null
@@ -1,284 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>DataSources &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="../../../genindex.html"/>
-        <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>DataSources</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="module-paddle.trainer_config_helpers.data_sources">
-<span id="datasources"></span><span id="api-trainer-config-helpers-data-sources"></span><h1>DataSources<a class="headerlink" href="#module-paddle.trainer_config_helpers.data_sources" title="Permalink to this headline">¶</a></h1>
-<p>Data Sources are helpers to define paddle training data or testing data.</p>
-<dl class="function">
-<dt id="paddle.trainer_config_helpers.data_sources.define_py_data_sources2">
-<code class="descclassname">paddle.trainer_config_helpers.data_sources.</code><code class="descname">define_py_data_sources2</code><span class="sig-paren">(</span><em>train_list</em>, <em>test_list</em>, <em>module</em>, <em>obj</em>, <em>args=None</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.data_sources.define_py_data_sources2" title="Permalink to this definition">¶</a></dt>
-<dd><p>Define python Train/Test data sources in one method. If train/test use
-the same Data Provider configuration, module/obj/args contain one argument,
-otherwise contain a list or tuple of arguments. For example:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">define_py_data_sources2</span><span class="p">(</span><span class="n">train_list</span><span class="o">=</span><span class="s2">&quot;train.list&quot;</span><span class="p">,</span>
-                        <span class="n">test_list</span><span class="o">=</span><span class="s2">&quot;test.list&quot;</span><span class="p">,</span>
-                        <span class="n">module</span><span class="o">=</span><span class="s2">&quot;data_provider&quot;</span>
-                        <span class="c1"># if train/test use different configurations,</span>
-                        <span class="c1"># obj=[&quot;process_train&quot;, &quot;process_test&quot;]</span>
-                        <span class="n">obj</span><span class="o">=</span><span class="s2">&quot;process&quot;</span><span class="p">,</span>
-                        <span class="n">args</span><span class="o">=</span><span class="p">{</span><span class="s2">&quot;dictionary&quot;</span><span class="p">:</span> <span class="n">dict_name</span><span class="p">})</span>
-</pre></div>
-</div>
-<p>The related data provider can refer to <a class="reference internal" href="../data_provider/pydataprovider2_en.html#api-pydataprovider2-sequential-model"><span class="std std-ref">DataProvider for the sequential model</span></a> .</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>train_list</strong> (<em>basestring</em>) &#8211; Train list name.</li>
-<li><strong>test_list</strong> (<em>basestring</em>) &#8211; Test list name.</li>
-<li><strong>module</strong> (<em>basestring</em><em> or </em><em>tuple</em><em> or </em><em>list</em>) &#8211; python module name. If train and test is different, then
-pass a tuple or list to this argument.</li>
-<li><strong>obj</strong> (<em>basestring</em><em> or </em><em>tuple</em><em> or </em><em>list</em>) &#8211; python object name. May be a function name if using
-PyDataProviderWrapper. If train and test is different, then pass
-a tuple or list to this argument.</li>
-<li><strong>args</strong> (<em>string</em><em> or </em><em>picklable object</em><em> or </em><em>list</em><em> or </em><em>tuple.</em>) &#8211; The best practice is using dict() to pass arguments into
-DataProvider, and use <code class="code docutils literal"><span class="pre">&#64;init_hook_wrapper</span></code> to receive
-arguments. If train and test is different, then pass a tuple
-or list to this argument.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">None</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">None</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/api/v1/trainer_config_helpers/layers.html b/develop/doc/api/v1/trainer_config_helpers/layers.html
deleted file mode 100644
index 93be72a97e2c63c8fa40dcd32b13f86ab8ef6595..0000000000000000000000000000000000000000
--- a/develop/doc/api/v1/trainer_config_helpers/layers.html
+++ /dev/null
@@ -1,3561 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Layers &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="../../../genindex.html"/>
-        <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Layers</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="layers">
-<span id="api-trainer-config-helpers-layers"></span><h1>Layers<a class="headerlink" href="#layers" title="Permalink to this headline">¶</a></h1>
-<div class="section" id="base">
-<h2>Base<a class="headerlink" href="#base" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="layertype">
-<h3>LayerType<a class="headerlink" href="#layertype" title="Permalink to this headline">¶</a></h3>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">LayerType</code></dt>
-<dd><p>Layer type enumerations.</p>
-<dl class="staticmethod">
-<dt>
-<em class="property">static </em><code class="descname">is_layer_type</code><span class="sig-paren">(</span><em>type_name</em><span class="sig-paren">)</span></dt>
-<dd><p>If type_name is a layer type.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>type_name</strong> (<em>basestring</em>) &#8211; layer type name. Because layer type enumerations are
-strings.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body">True if is a layer_type</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">bool</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</dd></dl>
-
-</div>
-<div class="section" id="layeroutput">
-<h3>LayerOutput<a class="headerlink" href="#layeroutput" title="Permalink to this headline">¶</a></h3>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">LayerOutput</code><span class="sig-paren">(</span><em>name</em>, <em>layer_type</em>, <em>parents=None</em>, <em>activation=None</em>, <em>num_filters=None</em>, <em>img_norm_type=None</em>, <em>size=None</em>, <em>outputs=None</em>, <em>reverse=None</em><span class="sig-paren">)</span></dt>
-<dd><p>LayerOutput is output for layer function. It is used internally by several
-reasons.</p>
-<ul>
-<li><p class="first">Check layer connection make sense.</p>
-<blockquote>
-<div><ul class="simple">
-<li>FC(Softmax) =&gt; Cost(MSE Error) is not good for example.</li>
-</ul>
-</div></blockquote>
-</li>
-<li><p class="first">Tracking layer connection.</p>
-</li>
-<li><p class="first">Pass to layer methods as input.</p>
-</li>
-</ul>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer output name.</li>
-<li><strong>layer_type</strong> (<em>basestring</em>) &#8211; Current Layer Type. One of LayerType enumeration.</li>
-<li><strong>activation</strong> (<em>BaseActivation.</em>) &#8211; Layer Activation.</li>
-<li><strong>parents</strong> (<em>list|tuple|collections.Sequence</em>) &#8211; Layer&#8217;s parents.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-<dl class="method">
-<dt>
-<code class="descname">set_input</code><span class="sig-paren">(</span><em>input</em><span class="sig-paren">)</span></dt>
-<dd><p>Set the input for a memory layer. Can only be used for memory layer</p>
-</dd></dl>
-
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="data-layer">
-<h2>Data layer<a class="headerlink" href="#data-layer" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="api-trainer-config-helpers-layers-data-layer">
-<span id="id1"></span><h3>data_layer<a class="headerlink" href="#api-trainer-config-helpers-layers-data-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">data_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define DataLayer For NeuralNetwork.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">data</span> <span class="o">=</span> <span class="n">data_layer</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s2">&quot;input&quot;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">1000</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Name of this data layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; Size of this data layer.</li>
-<li><strong>height</strong> (<em>int|None</em>) &#8211; Height of this data layer, used for image</li>
-<li><strong>width</strong> (<em>int|None</em>) &#8211; Width of this data layer, used for image</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="fully-connected-layers">
-<h2>Fully Connected Layers<a class="headerlink" href="#fully-connected-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="fc-layer">
-<span id="api-trainer-config-helpers-layers-fc-layer"></span><h3>fc_layer<a class="headerlink" href="#fc-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">fc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Helper for declare fully connected layer.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">fc</span> <span class="o">=</span> <span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-              <span class="n">size</span><span class="o">=</span><span class="mi">1024</span><span class="p">,</span>
-              <span class="n">act</span><span class="o">=</span><span class="n">LinearActivation</span><span class="p">(),</span>
-              <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>which is equal to:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">1024</span><span class="p">)</span> <span class="k">as</span> <span class="n">fc</span><span class="p">:</span>
-    <span class="n">fc</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; The input layer. Could be a list/tuple of input layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The layer dimension.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Default is tanh.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; The Parameter Attribute|list.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|Any</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="selective-fc-layer">
-<h3>selective_fc_layer<a class="headerlink" href="#selective-fc-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">selective_fc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Selectived fully connected layer. Different from fc_layer, the output
-of this layer maybe sparse. It requires an additional input to indicate
-several selected columns for output. If the selected columns is not
-specified, selective_fc_layer acts exactly like fc_layer.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">sel_fc</span> <span class="o">=</span> <span class="n">selective_fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">128</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; The input layer.</li>
-<li><strong>select</strong> (<em>LayerOutput</em>) &#8211; The select layer. The output of select layer should be a
-sparse binary matrix, and treat as the mask of selective fc.
-If is None, acts exactly like fc_layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The layer dimension.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Default is tanh.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; The Parameter Attribute.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|Any</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="conv-layers">
-<h2>Conv Layers<a class="headerlink" href="#conv-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="conv-operator">
-<h3>conv_operator<a class="headerlink" href="#conv-operator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">conv_operator</code><span class="sig-paren">(</span><em>img</em>, <em>filter</em>, <em>filter_size</em>, <em>num_filters</em>, <em>num_channels=None</em>, <em>stride=1</em>, <em>padding=0</em>, <em>filter_size_y=None</em>, <em>stride_y=None</em>, <em>padding_y=None</em>, <em>trans=False</em><span class="sig-paren">)</span></dt>
-<dd><p>Different from img_conv_layer, conv_op is an Operator, which can be used
-in mixed_layer. And conv_op takes two inputs to perform convolution.
-The first input is the image and the second is filter kernel. It only
-support GPU mode.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">op</span> <span class="o">=</span> <span class="n">conv_operator</span><span class="p">(</span><span class="n">img</span><span class="o">=</span><span class="n">input1</span><span class="p">,</span>
-                   <span class="nb">filter</span><span class="o">=</span><span class="n">input2</span><span class="p">,</span>
-                   <span class="n">filter_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
-                   <span class="n">num_filters</span><span class="o">=</span><span class="mi">64</span><span class="p">,</span>
-                   <span class="n">num_channels</span><span class="o">=</span><span class="mi">64</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>img</strong> (<em>LayerOutput</em>) &#8211; input image</li>
-<li><strong>filter</strong> (<em>LayerOutput</em>) &#8211; input filter</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; The x dimension of a filter kernel.</li>
-<li><strong>filter_size_y</strong> (<em>int</em>) &#8211; The y dimension of a filter kernel. Since
-PaddlePaddle now supports rectangular filters,
-the filter&#8217;s shape can be (filter_size, filter_size_y).</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; channel of output data.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; channel of input data.</li>
-<li><strong>stride</strong> (<em>int</em>) &#8211; The x dimension of the stride.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The y dimension of the stride.</li>
-<li><strong>padding</strong> (<em>int</em>) &#8211; The x dimension of padding.</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The y dimension of padding.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A ConvOperator Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">ConvOperator</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="conv-projection">
-<h3>conv_projection<a class="headerlink" href="#conv-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">conv_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Different from img_conv_layer and conv_op, conv_projection is an Projection,
-which can be used in mixed_layer and conat_layer. It use cudnn to implement
-conv and only support GPU mode.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">conv_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input1</span><span class="p">,</span>
-                       <span class="n">filter_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
-                       <span class="n">num_filters</span><span class="o">=</span><span class="mi">64</span><span class="p">,</span>
-                       <span class="n">num_channels</span><span class="o">=</span><span class="mi">64</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; The x dimension of a filter kernel.</li>
-<li><strong>filter_size_y</strong> (<em>int</em>) &#8211; The y dimension of a filter kernel. Since
-PaddlePaddle now supports rectangular filters,
-the filter&#8217;s shape can be (filter_size, filter_size_y).</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; channel of output data.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; channel of input data.</li>
-<li><strong>stride</strong> (<em>int</em>) &#8211; The x dimension of the stride.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The y dimension of the stride.</li>
-<li><strong>padding</strong> (<em>int</em>) &#8211; The x dimension of padding.</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The y dimension of padding.</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; The group number.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Convolution param attribute. None means default attribute</li>
-<li><strong>trans</strong> (<em>boolean</em>) &#8211; whether it is convTrans or conv</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A DotMulProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">DotMulProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="conv-shift-layer">
-<h3>conv_shift_layer<a class="headerlink" href="#conv-shift-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">conv_shift_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>This layer performs cyclic convolution for two input. For example:</dt>
-<dd><ul class="first last simple">
-<li>a[in]: contains M elements.</li>
-<li>b[in]: contains N elements (N should be odd).</li>
-<li>c[out]: contains M elements.</li>
-</ul>
-</dd>
-</dl>
-<div class="math">
-\[c[i] = \sum_{j=-(N-1)/2}^{(N-1)/2}a_{i+j} * b_{j}\]</div>
-<dl class="docutils">
-<dt>In this formular:</dt>
-<dd><ul class="first last simple">
-<li>a&#8217;s index is computed modulo M. When it is negative, then get item from
-the right side (which is the end of array) to the left.</li>
-<li>b&#8217;s index is computed modulo N. When it is negative, then get item from
-the right size (which is the end of array) to the left.</li>
-</ul>
-</dd>
-</dl>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">conv_shift</span> <span class="o">=</span> <span class="n">conv_shift_layer</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; Input layer a.</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input layer b.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="img-conv-layer">
-<h3>img_conv_layer<a class="headerlink" href="#img-conv-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">img_conv_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Convolution layer for image. Paddle can support both square and non-square
-input currently.</p>
-<p>The details of convolution layer, please refer UFLDL&#8217;s <a class="reference external" href="http://ufldl.stanford.edu/tutorial/supervised/FeatureExtractionUsingConvolution/">convolution</a> .</p>
-<p>Convolution Transpose (deconv) layer for image. Paddle can support both square
-and non-square input currently.</p>
-<p>The details of convolution transpose layer,
-please refer to the following explanation and references therein
-&lt;<a class="reference external" href="http://datascience.stackexchange.com/questions/6107/">http://datascience.stackexchange.com/questions/6107/</a>
-what-are-deconvolutional-layers/&gt;`_ .
-The num_channel means input image&#8217;s channel number. It may be 1 or 3 when
-input is raw pixels of image(mono or RGB), or it may be the previous layer&#8217;s
-num_filters * num_group.</p>
-<p>There are several group of filter in PaddlePaddle implementation.
-Each group will process some channel of the inputs. For example, if an input
-num_channel = 256, group = 4, num_filter=32, the PaddlePaddle will create
-32*4 = 128 filters to process inputs. The channels will be split into 4
-pieces. First 256/4 = 64 channels will process by first 32 filters. The
-rest channels will be processed by rest group of filters.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">conv</span> <span class="o">=</span> <span class="n">img_conv_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">data</span><span class="p">,</span> <span class="n">filter_size</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">filter_size_y</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                      <span class="n">num_channels</span><span class="o">=</span><span class="mi">8</span><span class="p">,</span>
-                      <span class="n">num_filters</span><span class="o">=</span><span class="mi">16</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                      <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">,</span>
-                      <span class="n">act</span><span class="o">=</span><span class="n">ReluActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Layer Input.</li>
-<li><strong>filter_size</strong> (<em>int|tuple|list</em>) &#8211; The x dimension of a filter kernel. Or input a tuple for
-two image dimension.</li>
-<li><strong>filter_size_y</strong> (<em>int|None</em>) &#8211; The y dimension of a filter kernel. Since PaddlePaddle
-currently supports rectangular filters, the filter&#8217;s
-shape will be (filter_size, filter_size_y).</li>
-<li><strong>num_filters</strong> &#8211; Each filter group&#8217;s number of filter</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type. Default is tanh</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; Group size of filters.</li>
-<li><strong>stride</strong> (<em>int|tuple|list</em>) &#8211; The x dimension of the stride. Or input a tuple for two image
-dimension.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The y dimension of the stride.</li>
-<li><strong>padding</strong> (<em>int|tuple|list</em>) &#8211; The x dimension of the padding. Or input a tuple for two
-image dimension</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The y dimension of the padding.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Convolution bias attribute. None means default bias.
-False means no bias.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; number of input channels. If None will be set
-automatically from previous output.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Convolution param attribute. None means default attribute</li>
-<li><strong>shared_biases</strong> (<em>bool</em>) &#8211; Is biases will be shared between filters or not.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Layer Extra Attribute.</li>
-<li><strong>trans</strong> (<em>bool</em>) &#8211; true if it is a convTransLayer, false if it is a convLayer</li>
-<li><strong>layer_type</strong> (<em>String</em>) &#8211; specify the layer_type, default is None. If trans=True,
-layer_type has to be &#8220;exconvt&#8221; or &#8220;cudnn_convt&#8221;,
-otherwise layer_type has to be either &#8220;exconv&#8221; or
-&#8220;cudnn_conv&#8221;</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="context-projection">
-<span id="api-trainer-config-helpers-layers-context-projection"></span><h3>context_projection<a class="headerlink" href="#context-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">context_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Context Projection.</p>
-<p>It just simply reorganizes input sequence, combines &#8220;context_len&#8221; sequence
-to one context from context_start. &#8220;context_start&#8221; will be set to
--(context_len - 1) / 2 by default. If context position out of sequence
-length, padding will be filled as zero if padding_attr = False, otherwise
-it is trainable.</p>
-<p>For example, origin sequence is [A B C D E F G], context len is 3, then
-after context projection and not set padding_attr, sequence will
-be [ 0AB ABC BCD CDE DEF EFG FG0 ].</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Sequence.</li>
-<li><strong>context_len</strong> (<em>int</em>) &#8211; context length.</li>
-<li><strong>context_start</strong> (<em>int</em>) &#8211; context start position. Default is
--(context_len - 1)/2</li>
-<li><strong>padding_attr</strong> (<em>bool|ParameterAttribute</em>) &#8211; Padding Parameter Attribute. If false, it means padding
-always be zero. Otherwise Padding is learnable, and
-parameter attribute is set by this parameter.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">Projection</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">Projection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="image-pooling-layer">
-<h2>Image Pooling Layer<a class="headerlink" href="#image-pooling-layer" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="img-pool-layer">
-<h3>img_pool_layer<a class="headerlink" href="#img-pool-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">img_pool_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Image pooling Layer.</p>
-<p>The details of pooling layer, please refer ufldl&#8217;s <a class="reference external" href="http://ufldl.stanford.edu/tutorial/supervised/Pooling/">pooling</a> .</p>
-<ul class="simple">
-<li>ceil_mode=True:</li>
-</ul>
-<div class="math">
-\[w = 1 + int(ceil(input\_width + 2 * padding - pool\_size) / float(stride))
-h = 1 + int(ceil(input\_height + 2 * padding\_y - pool\_size\_y) / float(stride\_y))\]</div>
-<ul class="simple">
-<li>ceil_mode=False:</li>
-</ul>
-<div class="math">
-\[w = 1 + int(floor(input\_width + 2 * padding - pool\_size) / float(stride))
-h = 1 + int(floor(input\_height + 2 * padding\_y - pool\_size\_y) / float(stride\_y))\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">maxpool</span> <span class="o">=</span> <span class="n">img_pool_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">conv</span><span class="p">,</span>
-                         <span class="n">pool_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
-                         <span class="n">pool_size_y</span><span class="o">=</span><span class="mi">5</span><span class="p">,</span>
-                         <span class="n">num_channels</span><span class="o">=</span><span class="mi">8</span><span class="p">,</span>
-                         <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                         <span class="n">stride_y</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span>
-                         <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                         <span class="n">padding_y</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span>
-                         <span class="n">pool_type</span><span class="o">=</span><span class="n">MaxPooling</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>padding</strong> (<em>int</em>) &#8211; pooling padding width.</li>
-<li><strong>padding_y</strong> (<em>int|None</em>) &#8211; pooling padding height. It&#8217;s equal to padding by default.</li>
-<li><strong>name</strong> (<em>basestring.</em>) &#8211; name of pooling layer</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input</li>
-<li><strong>pool_size</strong> (<em>int</em>) &#8211; pooling window width</li>
-<li><strong>pool_size_y</strong> (<em>int|None</em>) &#8211; pooling window height. It&#8217;s eaqual to pool_size by default.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; number of input channel.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType</em>) &#8211; pooling type. MaxPooling or AvgPooling. Default is
-MaxPooling.</li>
-<li><strong>stride</strong> (<em>int</em>) &#8211; stride width of pooling.</li>
-<li><strong>stride_y</strong> (<em>int|None</em>) &#8211; stride height of pooling. It is equal to stride by default.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-<li><strong>ceil_mode</strong> (<em>bool</em>) &#8211; Wether to use ceil mode to calculate output height and with.
-Defalut is True. If set false, Otherwise use floor.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="spp-layer">
-<h3>spp_layer<a class="headerlink" href="#spp-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">spp_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition.
-The details please refer to
-<a class="reference external" href="https://arxiv.org/abs/1406.4729">Kaiming He&#8217;s paper</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">spp</span> <span class="o">=</span> <span class="n">spp_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">data</span><span class="p">,</span>
-                <span class="n">pyramid_height</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span>
-                <span class="n">num_channels</span><span class="o">=</span><span class="mi">16</span><span class="p">,</span>
-                <span class="n">pool_type</span><span class="o">=</span><span class="n">MaxPooling</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; number of input channel.</li>
-<li><strong>pool_type</strong> &#8211; Pooling type. MaxPooling or AveragePooling. Default is MaxPooling.</li>
-<li><strong>pyramid_height</strong> (<em>int</em>) &#8211; pyramid height.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="maxout-layer">
-<h3>maxout_layer<a class="headerlink" href="#maxout-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">maxout_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>A layer to do max out on conv layer output.</dt>
-<dd><ul class="first last simple">
-<li>Input: output of a conv layer.</li>
-<li>Output: feature map size same as input. Channel is (input channel) / groups.</li>
-</ul>
-</dd>
-</dl>
-<p>So groups should be larger than 1, and the num of channels should be able
-to devided by groups.</p>
-<dl class="docutils">
-<dt>Please refer to Paper:</dt>
-<dd><ul class="first last simple">
-<li>Maxout Networks: <a class="reference external" href="http://www.jmlr.org/proceedings/papers/v28/goodfellow13.pdf">http://www.jmlr.org/proceedings/papers/v28/goodfellow13.pdf</a></li>
-<li>Multi-digit Number Recognition from Street View         Imagery using Deep Convolutional Neural Networks:         <a class="reference external" href="https://arxiv.org/pdf/1312.6082v4.pdf">https://arxiv.org/pdf/1312.6082v4.pdf</a></li>
-</ul>
-</dd>
-</dl>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">maxout</span> <span class="o">=</span> <span class="n">maxout_layer</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span>
-                      <span class="n">num_channels</span><span class="o">=</span><span class="mi">128</span><span class="p">,</span>
-                      <span class="n">groups</span><span class="o">=</span><span class="mi">4</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>num_channels</strong> (<em>int|None</em>) &#8211; The channel number of input layer. If None will be set
-automatically from previous output.</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; The group number of input layer.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layer, which can not specify.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="norm-layer">
-<h2>Norm Layer<a class="headerlink" href="#norm-layer" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="img-cmrnorm-layer">
-<h3>img_cmrnorm_layer<a class="headerlink" href="#img-cmrnorm-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">img_cmrnorm_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Response normalization across feature maps.
-The details please refer to
-<a class="reference external" href="http://www.cs.toronto.edu/~fritz/absps/imagenet.pdf">Alex&#8217;s paper</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">norm</span> <span class="o">=</span> <span class="n">img_cmrnorm_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">net</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">5</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; Normalize in number of <span class="math">\(size\)</span> feature maps.</li>
-<li><strong>scale</strong> (<em>float</em>) &#8211; The hyper-parameter.</li>
-<li><strong>power</strong> (<em>float</em>) &#8211; The hyper-parameter.</li>
-<li><strong>num_channels</strong> &#8211; input layer&#8217;s filers number or channels. If
-num_channels is None, it will be set automatically.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="batch-norm-layer">
-<h3>batch_norm_layer<a class="headerlink" href="#batch-norm-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">batch_norm_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Batch Normalization Layer. The notation of this layer as follow.</p>
-<p><span class="math">\(x\)</span> is the input features over a mini-batch.</p>
-<div class="math">
-\[\begin{split}\mu_{\beta} &amp;\gets \frac{1}{m} \sum_{i=1}^{m} x_i \qquad &amp;//\
-\ mini-batch\ mean \\
-\sigma_{\beta}^{2} &amp;\gets \frac{1}{m} \sum_{i=1}^{m}(x_i - \
-\mu_{\beta})^2 \qquad &amp;//\ mini-batch\ variance \\
-\hat{x_i} &amp;\gets \frac{x_i - \mu_\beta} {\sqrt{\
-\sigma_{\beta}^{2} + \epsilon}} \qquad &amp;//\ normalize \\
-y_i &amp;\gets \gamma \hat{x_i} + \beta \qquad &amp;//\ scale\ and\ shift\end{split}\]</div>
-<p>The details of batch normalization please refer to this
-<a class="reference external" href="http://arxiv.org/abs/1502.03167">paper</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">norm</span> <span class="o">=</span> <span class="n">batch_norm_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">net</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="n">ReluActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; batch normalization input. Better be linear activation.
-Because there is an activation inside batch_normalization.</li>
-<li><strong>batch_norm_type</strong> (<em>None|string</em><em>, </em><em>None</em><em> or </em><em>&quot;batch_norm&quot;</em><em> or </em><em>&quot;cudnn_batch_norm&quot;</em>) &#8211; We have batch_norm and cudnn_batch_norm. batch_norm
-supports both CPU and GPU. cudnn_batch_norm requires
-cuDNN version greater or equal to v4 (&gt;=v4). But
-cudnn_batch_norm is faster and needs less memory
-than batch_norm. By default (None), we will
-automaticly select cudnn_batch_norm for GPU and
-batch_norm for CPU. Otherwise, select batch norm
-type based on the specified type. If you use cudnn_batch_norm,
-we suggested you use latest version, such as v5.1.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Better be relu. Because batch
-normalization will normalize input near zero.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; num of image channels or previous layer&#8217;s number of
-filters. None will automatically get from layer&#8217;s
-input.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; <span class="math">\(\beta\)</span>, better be zero when initialize. So the
-initial_std=0, initial_mean=1 is best practice.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; <span class="math">\(\gamma\)</span>, better be one when initialize. So the
-initial_std=0, initial_mean=1 is best practice.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-<li><strong>use_global_stats</strong> (<em>bool|None.</em>) &#8211; whether use moving mean/variance statistics
-during testing peroid. If None or True,
-it will use moving mean/variance statistics during
-testing. If False, it will use the mean
-and variance of current batch of test data for
-testing.</li>
-<li><strong>moving_average_fraction</strong> (<em>float.</em>) &#8211; Factor used in the moving average
-computation, referred to as facotr,
-<span class="math">\(runningMean = newMean*(1-factor)
-+ runningMean*factor\)</span></li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="sum-to-one-norm-layer">
-<h3>sum_to_one_norm_layer<a class="headerlink" href="#sum-to-one-norm-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">sum_to_one_norm_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for sum-to-one normalization,
-which is used in NEURAL TURING MACHINE.</p>
-<div class="math">
-\[out[i] = \frac {in[i]} {\sum_{k=1}^N in[k]}\]</div>
-<p>where <span class="math">\(in\)</span> is a (batchSize x dataDim) input vector,
-and <span class="math">\(out\)</span> is a (batchSize x dataDim) output vector.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">sum_to_one_norm</span> <span class="o">=</span> <span class="n">sum_to_one_norm_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="recurrent-layers">
-<h2>Recurrent Layers<a class="headerlink" href="#recurrent-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="recurrent-layer">
-<h3>recurrent_layer<a class="headerlink" href="#recurrent-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">recurrent_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Simple recurrent unit layer. It is just a fully connect layer through both
-time and neural network.</p>
-<p>For each sequence [start, end] it performs the following computation:</p>
-<div class="math">
-\[\begin{split}out_{i} = act(in_{i})     \      \      \text{for} \ i = start \\
-out_{i} = act(in_{i} + out_{i-1} * W) \ \ \text{for} \ start &lt; i &lt;= end\end{split}\]</div>
-<p>If reversed is true, the order is reversed:</p>
-<div class="math">
-\[\begin{split}out_{i} = act(in_{i})           \    \   \text{for} \ i = end  \\
-out_{i} = act(in_{i} + out_{i+1} * W) \ \ \text{for} \ start &lt;= i &lt; end\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; activation.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; bias attribute.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; parameter attribute.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the layer</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lstmemory">
-<h3>lstmemory<a class="headerlink" href="#lstmemory" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">lstmemory</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Long Short-term Memory Cell.</p>
-<p>The memory cell was implemented as follow equations.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>NOTE: In PaddlePaddle&#8217;s implementation, the multiplications
-<span class="math">\(W_{xi}x_{t}\)</span> , <span class="math">\(W_{xf}x_{t}\)</span>,
-<span class="math">\(W_{xc}x_t\)</span>, <span class="math">\(W_{xo}x_{t}\)</span> are not done in the lstmemory layer,
-so an additional mixed_layer with full_matrix_projection or a fc_layer must
-be included in the configuration file to complete the input-to-hidden
-mappings before lstmemory is called.</p>
-<p>NOTE: This is a low level user interface. You can use network.simple_lstm
-to config a simple plain lstm layer.</p>
-<p>Please refer to <strong>Generating Sequences With Recurrent Neural Networks</strong> for
-more details about LSTM.</p>
-<p><a class="reference external" href="http://arxiv.org/abs/1308.0850">Link</a> goes as below.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The lstmemory layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; is sequence process reversed or not.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; activation type, TanhActivation by default. <span class="math">\(h_t\)</span></li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; gate activation type, SigmoidActivation by default.</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; state activation type, TanhActivation by default.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias attribute. None means default bias. False means no
-bias.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Parameter Attribute.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer attribute</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="grumemory">
-<h3>grumemory<a class="headerlink" href="#grumemory" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">grumemory</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Gate Recurrent Unit Layer.</p>
-<p>The memory cell was implemented as follow equations.</p>
-<p>1. update gate <span class="math">\(z\)</span>: defines how much of the previous memory to
-keep around or the unit updates its activations. The update gate
-is computed by:</p>
-<div class="math">
-\[z_t = \sigma(W_{z}x_{t} + U_{z}h_{t-1} + b_z)\]</div>
-<p>2. reset gate <span class="math">\(r\)</span>: determines how to combine the new input with the
-previous memory. The reset gate is computed similarly to the update gate:</p>
-<div class="math">
-\[r_t = \sigma(W_{r}x_{t} + U_{r}h_{t-1} + b_r)\]</div>
-<p>3. The candidate activation <span class="math">\(\tilde{h_t}\)</span> is computed similarly to
-that of the traditional recurrent unit:</p>
-<div class="math">
-\[{\tilde{h_t}} = tanh(W x_{t} + U (r_{t} \odot h_{t-1}) + b)\]</div>
-<p>4. The hidden activation <span class="math">\(h_t\)</span> of the GRU at time t is a linear
-interpolation between the previous activation <span class="math">\(h_{t-1}\)</span> and the
-candidate activation <span class="math">\(\tilde{h_t}\)</span>:</p>
-<div class="math">
-\[h_t = (1 - z_t) h_{t-1} + z_t {\tilde{h_t}}\]</div>
-<p>NOTE: In PaddlePaddle&#8217;s implementation, the multiplication operations
-<span class="math">\(W_{r}x_{t}\)</span>, <span class="math">\(W_{z}x_{t}\)</span> and <span class="math">\(W x_t\)</span> are not computed in
-gate_recurrent layer. Consequently, an additional mixed_layer with
-full_matrix_projection or a fc_layer must be included before grumemory
-is called.</p>
-<p>More details can be found by referring to <a class="reference external" href="https://arxiv.org/abs/1412.3555">Empirical Evaluation of Gated
-Recurrent Neural Networks on Sequence Modeling.</a></p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">grumemory</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The gru layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; input layer.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; Whether sequence process is reversed or not.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; activation type, TanhActivation by default. This activation
-affects the <span class="math">\({\tilde{h_t}}\)</span>.</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; gate activation type, SigmoidActivation by default.
-This activation affects the <span class="math">\(z_t\)</span> and <span class="math">\(r_t\)</span>. It is the
-<span class="math">\(\sigma\)</span> in the above formula.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias attribute. None means default bias. False means no
-bias.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Parameter Attribute.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer attribute</li>
-<li><strong>size</strong> (<em>None</em>) &#8211; Stub parameter of size, but actually not used. If set this size
-will get a warning.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="recurrent-layer-group">
-<h2>Recurrent Layer Group<a class="headerlink" href="#recurrent-layer-group" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="memory">
-<h3>memory<a class="headerlink" href="#memory" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">memory</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>The memory layers is a layer cross each time step. Reference this output
-as previous time step layer <code class="code docutils literal"><span class="pre">name</span></code> &#8216;s output.</p>
-<p>The default memory is zero in first time step, previous time step&#8217;s
-output in the rest time steps.</p>
-<p>If boot_bias, the first time step value is this bias and
-with activation.</p>
-<p>If boot_with_const_id, then the first time stop is a IndexSlot, the
-Arguments.ids()[0] is this <code class="code docutils literal"><span class="pre">cost_id</span></code>.</p>
-<p>If boot_layer is not null, the memory is just the boot_layer&#8217;s output.
-Set <code class="code docutils literal"><span class="pre">is_seq</span></code> is true boot layer is sequence.</p>
-<p>The same name layer in recurrent group will set memory on each time
-step.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">mem</span> <span class="o">=</span> <span class="n">memory</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;state&#39;</span><span class="p">)</span>
-<span class="n">state</span> <span class="o">=</span> <span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">mem</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;state&#39;</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>If you do not want to specify the name, you can equivalently use set_input()
-to specify the layer needs to be remembered as the following:</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; the name of the layer which this memory remembers.
-If name is None, user should call set_input() to specify the
-name of the layer which this memory remembers.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; size of memory.</li>
-<li><strong>memory_name</strong> (<em>basestring</em>) &#8211; the name of the memory.
-It is ignored when name is provided.</li>
-<li><strong>is_seq</strong> (<em>bool</em>) &#8211; is sequence for boot_layer</li>
-<li><strong>boot_layer</strong> (<em>LayerOutput|None</em>) &#8211; boot layer of memory.</li>
-<li><strong>boot_bias</strong> (<em>ParameterAttribute|None</em>) &#8211; boot layer&#8217;s bias</li>
-<li><strong>boot_bias_active_type</strong> (<em>BaseActivation</em>) &#8211; boot layer&#8217;s active type.</li>
-<li><strong>boot_with_const_id</strong> (<em>int</em>) &#8211; boot layer&#8217;s id.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object which is a memory.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="recurrent-group">
-<h3>recurrent_group<a class="headerlink" href="#recurrent-group" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">recurrent_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Recurrent layer group is an extremely flexible recurrent unit in
-PaddlePaddle. As long as the user defines the calculation done within a
-time step, PaddlePaddle will iterate such a recurrent calculation over
-sequence input. This is extremely usefull for attention based model, or
-Neural Turning Machine like models.</p>
-<p>The basic usage (time steps) is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">step</span><span class="p">(</span><span class="nb">input</span><span class="p">):</span>
-    <span class="n">output</span> <span class="o">=</span> <span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                      <span class="n">size</span><span class="o">=</span><span class="mi">1024</span><span class="p">,</span>
-                      <span class="n">act</span><span class="o">=</span><span class="n">LinearActivation</span><span class="p">(),</span>
-                      <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-    <span class="k">return</span> <span class="n">output</span>
-
-<span class="n">group</span> <span class="o">=</span> <span class="n">recurrent_group</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                        <span class="n">step</span><span class="o">=</span><span class="n">step</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>You can see following configs for further usages:</p>
-<ul class="simple">
-<li>time steps: lstmemory_group, paddle/gserver/tests/sequence_layer_group.conf,                   demo/seqToseq/seqToseq_net.py</li>
-<li>sequence steps: paddle/gserver/tests/sequence_nest_layer_group.conf</li>
-</ul>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>step</strong> (<em>callable</em>) &#8211; <p>recurrent one time step function.The input of this function is
-input of the group. The return of this function will be
-recurrent group&#8217;s return value.</p>
-<p>The recurrent group scatter a sequence into time steps. And
-for each time step, will invoke step function, and return
-a time step result. Then gather each time step of output into
-layer group&#8217;s output.</p>
-</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; recurrent_group&#8217;s name.</li>
-<li><strong>input</strong> (<em>LayerOutput|StaticInput|SubsequenceInput|list|tuple</em>) &#8211; <p>Input links array.</p>
-<p>LayerOutput will be scattered into time steps.
-SubsequenceInput will be scattered into sequence steps.
-StaticInput will be imported to each time step, and doesn&#8217;t change
-through time. It&#8217;s a mechanism to access layer outside step function.</p>
-</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; If reverse is set true, the recurrent unit will process the
-input sequence in a reverse order.</li>
-<li><strong>targetInlink</strong> (<em>LayerOutput|SubsequenceInput</em>) &#8211; <p>the input layer which share info with layer group&#8217;s output</p>
-<p>Param input specifies multiple input layers. For
-SubsequenceInput inputs, config should assign one input
-layer that share info(the number of sentences and the number
-of words in each sentence) with all layer group&#8217;s outputs.
-targetInlink should be one of the layer group&#8217;s input.</p>
-</li>
-<li><strong>is_generating</strong> &#8211; If is generating, none of input type should be LayerOutput;
-else, for training or testing, one of the input type must
-be LayerOutput.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-<p>: type is_generating: bool</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Returns:</th><td class="field-body">LayerOutput object.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Return type:</th><td class="field-body">LayerOutput</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lstm-step-layer">
-<h3>lstm_step_layer<a class="headerlink" href="#lstm-step-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">lstm_step_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>LSTM Step Layer. It used in recurrent_group. The lstm equations are shown
-as follow.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>The input of lstm step is <span class="math">\(Wx_t + Wh_{t-1}\)</span>, and user should use
-<code class="code docutils literal"><span class="pre">mixed_layer</span></code> and <code class="code docutils literal"><span class="pre">full_matrix_projection</span></code> to calculate these
-input vector.</p>
-<p>The state of lstm step is <span class="math">\(c_{t-1}\)</span>. And lstm step layer will do</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t = \sigma(input + W_{ci}c_{t-1} + b_i)\\...\end{aligned}\end{align} \]</div>
-<p>This layer contains two outputs. Default output is <span class="math">\(h_t\)</span>. The other
-output is <span class="math">\(o_t\)</span>, which name is &#8216;state&#8217; and can use
-<code class="code docutils literal"><span class="pre">get_output_layer</span></code> to extract this output.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer&#8217;s name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; Layer&#8217;s size. NOTE: lstm layer&#8217;s size, should be equal as
-<code class="code docutils literal"><span class="pre">input.size/4</span></code>, and should be equal as
-<code class="code docutils literal"><span class="pre">state.size</span></code>.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer. <span class="math">\(Wx_t + Wh_{t-1}\)</span></li>
-<li><strong>state</strong> (<em>LayerOutput</em>) &#8211; State Layer. <span class="math">\(c_{t-1}\)</span></li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type. Default is tanh</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; Gate Activation Type. Default is sigmoid, and should
-be sigmoid only.</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; State Activation Type. Default is sigmoid, and should
-be sigmoid only.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Bias Attribute.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="gru-step-layer">
-<h3>gru_step_layer<a class="headerlink" href="#gru-step-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">gru_step_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; </li>
-<li><strong>output_mem</strong> &#8211; </li>
-<li><strong>size</strong> &#8211; </li>
-<li><strong>act</strong> &#8211; </li>
-<li><strong>name</strong> &#8211; </li>
-<li><strong>gate_act</strong> &#8211; </li>
-<li><strong>bias_attr</strong> &#8211; </li>
-<li><strong>param_attr</strong> &#8211; the parameter_attribute for transforming the output_mem
-from previous step.</li>
-<li><strong>layer_attr</strong> &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="beam-search">
-<h3>beam_search<a class="headerlink" href="#beam-search" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">beam_search</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Beam search is a heuristic search algorithm used in sequence generation.
-It explores a graph by expanding the most promising nodes in a limited set
-to maintain tractability.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">rnn_step</span><span class="p">(</span><span class="nb">input</span><span class="p">):</span>
-    <span class="n">last_time_step_output</span> <span class="o">=</span> <span class="n">memory</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;rnn&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">)</span>
-    <span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;rnn&#39;</span><span class="p">)</span> <span class="k">as</span> <span class="n">simple_rnn</span><span class="p">:</span>
-        <span class="n">simple_rnn</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
-        <span class="n">simple_rnn</span> <span class="o">+=</span> <span class="n">last_time_step_output</span>
-    <span class="k">return</span> <span class="n">simple_rnn</span>
-
-<span class="n">generated_word_embedding</span> <span class="o">=</span> <span class="n">GeneratedInput</span><span class="p">(</span>
-                       <span class="n">size</span><span class="o">=</span><span class="n">target_dictionary_dim</span><span class="p">,</span>
-                       <span class="n">embedding_name</span><span class="o">=</span><span class="s2">&quot;target_language_embedding&quot;</span><span class="p">,</span>
-                       <span class="n">embedding_size</span><span class="o">=</span><span class="n">word_vector_dim</span><span class="p">)</span>
-
-<span class="n">beam_gen</span> <span class="o">=</span> <span class="n">beam_search</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s2">&quot;decoder&quot;</span><span class="p">,</span>
-                       <span class="n">step</span><span class="o">=</span><span class="n">rnn_step</span><span class="p">,</span>
-                       <span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">StaticInput</span><span class="p">(</span><span class="n">encoder_last</span><span class="p">),</span>
-                              <span class="n">generated_word_embedding</span><span class="p">],</span>
-                       <span class="n">bos_id</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span>
-                       <span class="n">eos_id</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                       <span class="n">beam_size</span><span class="o">=</span><span class="mi">5</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>Please see the following demo for more details:</p>
-<ul class="simple">
-<li>machine translation : demo/seqToseq/translation/gen.conf                             demo/seqToseq/seqToseq_net.py</li>
-</ul>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>base string</em>) &#8211; Name of the recurrent unit that generates sequences.</li>
-<li><strong>step</strong> (<em>callable</em>) &#8211; <p>A callable function that defines the calculation in a time
-step, and it is applied to sequences with arbitrary length by
-sharing a same set of weights.</p>
-<p>You can refer to the first parameter of recurrent_group, or
-demo/seqToseq/seqToseq_net.py for more details.</p>
-</li>
-<li><strong>input</strong> (<em>list</em>) &#8211; Input data for the recurrent unit, which should include the
-previously generated words as a GeneratedInput object.</li>
-<li><strong>bos_id</strong> (<em>int</em>) &#8211; Index of the start symbol in the dictionary. The start symbol
-is a special token for NLP task, which indicates the
-beginning of a sequence. In the generation task, the start
-symbol is essential, since it is used to initialize the RNN
-internal state.</li>
-<li><strong>eos_id</strong> (<em>int</em>) &#8211; Index of the end symbol in the dictionary. The end symbol is
-a special token for NLP task, which indicates the end of a
-sequence. The generation process will stop once the end
-symbol is generated, or a pre-defined max iteration number
-is exceeded.</li>
-<li><strong>max_length</strong> (<em>int</em>) &#8211; Max generated sequence length.</li>
-<li><strong>beam_size</strong> (<em>int</em>) &#8211; Beam search for sequence generation is an iterative search
-algorithm. To maintain tractability, every iteration only
-only stores a predetermined number, called the beam_size,
-of the most promising next words. The greater the beam
-size, the fewer candidate words are pruned.</li>
-<li><strong>num_results_per_sample</strong> (<em>int</em>) &#8211; Number of the generated results per input
-sequence. This number must always be less than
-beam size.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">The generated word index.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="get-output-layer">
-<h3>get_output_layer<a class="headerlink" href="#get-output-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">get_output_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Get layer&#8217;s output by name. In PaddlePaddle, a layer might return multiple
-values, but returns one layer&#8217;s output. If the user wants to use another
-output besides the default one, please use get_output_layer first to get
-the output from input.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer&#8217;s name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; get output layer&#8217;s input. And this layer should contains
-multiple outputs.</li>
-<li><strong>arg_name</strong> (<em>basestring</em>) &#8211; Output name from input.</li>
-<li><strong>layer_attr</strong> &#8211; Layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="mixed-layer">
-<h2>Mixed Layer<a class="headerlink" href="#mixed-layer" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="api-trainer-config-helpers-layers-mixed-layer">
-<span id="id2"></span><h3>mixed_layer<a class="headerlink" href="#api-trainer-config-helpers-layers-mixed-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">mixed_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Mixed Layer. A mixed layer will add all inputs together, then activate.
-Each inputs is a projection or operator.</p>
-<p>There are two styles of usages.</p>
-<ol class="arabic simple">
-<li>When not set inputs parameter, use mixed_layer like this:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">)</span> <span class="k">as</span> <span class="n">m</span><span class="p">:</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">)</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">identity_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<ol class="arabic simple" start="2">
-<li>You can also set all inputs when invoke mixed_layer as follows:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">m</span> <span class="o">=</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                <span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">),</span>
-                       <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer2</span><span class="p">)])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; mixed layer name. Can be referenced by other layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; layer size.</li>
-<li><strong>input</strong> &#8211; inputs layer. It is an optional parameter. If set,
-then this function will just return layer&#8217;s name.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; The extra layer config. Default is None.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">MixedLayerType object can add inputs or layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">MixedLayerType</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="embedding-layer">
-<span id="api-trainer-config-helpers-layers-embedding-layer"></span><h3>embedding_layer<a class="headerlink" href="#embedding-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">embedding_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define a embedding Layer.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Name of this embedding layer.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer for this embedding. NOTE: must be Index Data.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The embedding dimension.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None</em>) &#8211; The embedding parameter attribute. See ParameterAttribute
-for details.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra layer Config. Default is None.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="scaling-projection">
-<h3>scaling_projection<a class="headerlink" href="#scaling-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">scaling_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>scaling_projection multiplies the input with a scalar parameter and add to
-the output.</p>
-<div class="math">
-\[out += w * in\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">scaling_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A ScalingProjection object</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">ScalingProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="dotmul-projection">
-<h3>dotmul_projection<a class="headerlink" href="#dotmul-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">dotmul_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>DotMulProjection with a layer as input.
-It performs element-wise multiplication with weight.</p>
-<div class="math">
-\[out.row[i] += in.row[i] .* weight\]</div>
-<p>where <span class="math">\(.*\)</span> means element-wise multiplication.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">dotmul_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A DotMulProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">DotMulProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="dotmul-operator">
-<h3>dotmul_operator<a class="headerlink" href="#dotmul-operator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">dotmul_operator</code><span class="sig-paren">(</span><em>a=None</em>, <em>b=None</em>, <em>scale=1</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>DotMulOperator takes two inputs and performs element-wise multiplication:</p>
-<div class="math">
-\[out.row[i] += scale * (a.row[i] .* b.row[i])\]</div>
-<p>where <span class="math">\(.*\)</span> means element-wise multiplication, and
-scale is a config scalar, its default value is one.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">op</span> <span class="o">=</span> <span class="n">dotmul_operator</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">scale</span><span class="o">=</span><span class="mf">0.5</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; Input layer1</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; Input layer2</li>
-<li><strong>scale</strong> (<em>float</em>) &#8211; config scalar, default value is one.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A DotMulOperator Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">DotMulOperator</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="full-matrix-projection">
-<h3>full_matrix_projection<a class="headerlink" href="#full-matrix-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">full_matrix_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Full Matrix Projection. It performs full matrix multiplication.</p>
-<div class="math">
-\[out.row[i] += in.row[i] * weight\]</div>
-<p>There are two styles of usage.</p>
-<ol class="arabic simple">
-<li>When used in mixed_layer like this, you can only set the input:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span> <span class="k">as</span> <span class="n">m</span><span class="p">:</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<ol class="arabic simple" start="2">
-<li>When used as an independant object like this, you must set the size:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                              <span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                              <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;_proj&#39;</span><span class="p">))</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The parameter size. Means the width of parameter.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A FullMatrixProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">FullMatrixProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="identity-projection">
-<h3>identity_projection<a class="headerlink" href="#identity-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">identity_projection</code><span class="sig-paren">(</span><em>input</em>, <em>offset=None</em><span class="sig-paren">)</span></dt>
-<dd><ol class="arabic simple">
-<li>IdentityProjection if offset=None. It performs:</li>
-</ol>
-<div class="math">
-\[out.row[i] += in.row[i]\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">identity_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>2. IdentityOffsetProjection if offset!=None. It likes IdentityProjection,
-but layer size may be smaller than input size.
-It select dimesions [offset, offset+layer_size) from input:</p>
-<div class="math">
-\[out.row[i] += in.row[i + \textrm{offset}]\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">identity_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                           <span class="n">offset</span><span class="o">=</span><span class="mi">10</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>Note that both of two projections should not have any parameter.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer.</li>
-<li><strong>offset</strong> (<em>int</em>) &#8211; Offset, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A IdentityProjection or IdentityOffsetProjection object</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">IdentityProjection or IdentityOffsetProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="table-projection">
-<h3>table_projection<a class="headerlink" href="#table-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">table_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Table Projection. It selects rows from parameter where row_id
-is in input_ids.</p>
-<div class="math">
-\[out.row[i] += table.row[ids[i]]\]</div>
-<p>where <span class="math">\(out\)</span> is output, <span class="math">\(table\)</span> is parameter, <span class="math">\(ids\)</span> is input_ids,
-and <span class="math">\(i\)</span> is row_id.</p>
-<p>There are two styles of usage.</p>
-<ol class="arabic simple">
-<li>When used in mixed_layer like this, you can only set the input:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span> <span class="k">as</span> <span class="n">m</span><span class="p">:</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">table_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<ol class="arabic simple" start="2">
-<li>When used as an independant object like this, you must set the size:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">table_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                        <span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                        <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;_proj&#39;</span><span class="p">))</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer, which must contains id fields.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The parameter size. Means the width of parameter.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A TableProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">TableProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="trans-full-matrix-projection">
-<h3>trans_full_matrix_projection<a class="headerlink" href="#trans-full-matrix-projection" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">trans_full_matrix_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Different from full_matrix_projection, this projection performs matrix
-multiplication, using transpose of weight.</p>
-<div class="math">
-\[out.row[i] += in.row[i] * w^\mathrm{T}\]</div>
-<p><span class="math">\(w^\mathrm{T}\)</span> means transpose of weight.
-The simply usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">trans_full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                                    <span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                                    <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span>
-                                         <span class="n">name</span><span class="o">=</span><span class="s1">&#39;_proj&#39;</span><span class="p">,</span>
-                                         <span class="n">initial_mean</span><span class="o">=</span><span class="mf">0.0</span><span class="p">,</span>
-                                         <span class="n">initial_std</span><span class="o">=</span><span class="mf">0.01</span><span class="p">))</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The parameter size. Means the width of parameter.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">A TransposedFullMatrixProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">TransposedFullMatrixProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="aggregate-layers">
-<h2>Aggregate Layers<a class="headerlink" href="#aggregate-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="pooling-layer">
-<span id="api-trainer-config-helpers-layers-pooling-layer"></span><h3>pooling_layer<a class="headerlink" href="#pooling-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">pooling_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Pooling layer for sequence inputs, not used for Image.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">seq_pool</span> <span class="o">=</span> <span class="n">pooling_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                         <span class="n">pooling_type</span><span class="o">=</span><span class="n">AvgPooling</span><span class="p">(),</span>
-                         <span class="n">agg_level</span><span class="o">=</span><span class="n">AggregateLevel</span><span class="o">.</span><span class="n">EACH_SEQUENCE</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>agg_level</strong> (<em>AggregateLevel</em>) &#8211; AggregateLevel.EACH_TIMESTEP or
-AggregateLevel.EACH_SEQUENCE</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>pooling_type</strong> (<em>BasePoolingType|None</em>) &#8211; Type of pooling, MaxPooling(default), AvgPooling,
-SumPooling, SquareRootNPooling.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias parameter attribute. False if no bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; The Extra Attributes for layer, such as dropout.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="last-seq">
-<span id="api-trainer-config-helpers-layers-last-seq"></span><h3>last_seq<a class="headerlink" href="#last-seq" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">last_seq</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Get Last Timestamp Activation of a sequence.</p>
-<p>If stride &gt; 0, this layer slides a window whose size is determined by stride,
-and return the last value of the window as the output. Thus, a long sequence
-will be shorten. Note that for sequence with sub-sequence, the default value
-of stride is -1.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">seq</span> <span class="o">=</span> <span class="n">last_seq</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>agg_level</strong> &#8211; Aggregated level</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>stride</strong> (<em>Int</em>) &#8211; window size.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="first-seq">
-<span id="api-trainer-config-helpers-layers-first-seq"></span><h3>first_seq<a class="headerlink" href="#first-seq" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">first_seq</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Get First Timestamp Activation of a sequence.</p>
-<p>If stride &gt; 0, this layer slides a window whose size is determined by stride,
-and return the first value of the window as the output. Thus, a long sequence
-will be shorten. Note that for sequence with sub-sequence, the default value
-of stride is -1.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">seq</span> <span class="o">=</span> <span class="n">first_seq</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>agg_level</strong> &#8211; aggregation level</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>stride</strong> (<em>Int</em>) &#8211; window size.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="concat-layer">
-<h3>concat_layer<a class="headerlink" href="#concat-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">concat_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Concat all input vector into one huge vector.
-Inputs can be list of LayerOutput or list of projection.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">concat</span> <span class="o">=</span> <span class="n">concat_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>list|tuple|collections.Sequence</em>) &#8211; input layers or projections</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="seq-concat-layer">
-<h3>seq_concat_layer<a class="headerlink" href="#seq-concat-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">seq_concat_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Concat sequence a with sequence b.</p>
-<dl class="docutils">
-<dt>Inputs:</dt>
-<dd><ul class="first last simple">
-<li>a = [a1, a2, ..., an]</li>
-<li>b = [b1, b2, ..., bn]</li>
-<li>Note that the length of a and b should be the same.</li>
-</ul>
-</dd>
-</dl>
-<p>Output: [a1, b1, a2, b2, ..., an, bn]</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">concat</span> <span class="o">=</span> <span class="n">seq_concat_layer</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; input sequence layer</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input sequence layer</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="reshaping-layers">
-<h2>Reshaping Layers<a class="headerlink" href="#reshaping-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="block-expand-layer">
-<h3>block_expand_layer<a class="headerlink" href="#block-expand-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">block_expand_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>Expand feature map to minibatch matrix.</dt>
-<dd><ul class="first last simple">
-<li>matrix width is: block_y * block_x * num_channels</li>
-<li>matirx height is: outputH * outputW</li>
-</ul>
-</dd>
-</dl>
-<div class="math">
-\[ \begin{align}\begin{aligned}outputH = 1 + (2 * padding_y + imgSizeH - block_y + stride_y - 1) / stride_y\\outputW = 1 + (2 * padding_x + imgSizeW - block_x + stride_x - 1) / stride_x\end{aligned}\end{align} \]</div>
-<p>The expand method is the same with ExpandConvLayer, but saved the transposed
-value. After expanding, output.sequenceStartPositions will store timeline.
-The number of time steps are outputH * outputW and the dimension of each
-time step is block_y * block_x * num_channels. This layer can be used after
-convolution neural network, and before recurrent neural network.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">block_expand</span> <span class="o">=</span> <span class="n">block_expand_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                                  <span class="n">num_channels</span><span class="o">=</span><span class="mi">128</span><span class="p">,</span>
-                                  <span class="n">stride_x</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                                  <span class="n">stride_y</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                                  <span class="n">block_x</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                                  <span class="n">block_x</span><span class="o">=</span><span class="mi">3</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>num_channels</strong> (<em>int|None</em>) &#8211; The channel number of input layer.</li>
-<li><strong>block_x</strong> (<em>int</em>) &#8211; The width of sub block.</li>
-<li><strong>block_y</strong> (<em>int</em>) &#8211; The width of sub block.</li>
-<li><strong>stride_x</strong> (<em>int</em>) &#8211; The stride size in horizontal direction.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The stride size in vertical direction.</li>
-<li><strong>padding_x</strong> (<em>int</em>) &#8211; The padding size in horizontal direction.</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The padding size in vertical direction.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layer, which can not specify.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="expand-layer">
-<span id="api-trainer-config-helpers-layers-expand-layer"></span><h3>expand_layer<a class="headerlink" href="#expand-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">expand_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for &#8220;Expand Dense data or (sequence data where the length of each
-sequence is one) to sequence data.&#8221;</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">expand</span> <span class="o">=</span> <span class="n">expand_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span>
-                      <span class="n">expand_as</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span>
-                      <span class="n">expand_level</span><span class="o">=</span><span class="n">ExpandLevel</span><span class="o">.</span><span class="n">FROM_TIMESTEP</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer</li>
-<li><strong>expand_as</strong> (<em>LayerOutput</em>) &#8211; Expand as this layer&#8217;s sequence info.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias attribute. None means default bias. False means no
-bias.</li>
-<li><strong>expand_level</strong> (<em>ExpandLevel</em>) &#8211; whether input layer is timestep(default) or sequence.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="repeat-layer">
-<h3>repeat_layer<a class="headerlink" href="#repeat-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">repeat_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for repeating the input for num_repeats times. This is equivalent
-to apply concat_layer() with num_repeats same input.</p>
-<div class="math">
-\[y  = [x, x, \cdots, x]\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">expand</span> <span class="o">=</span> <span class="n">repeat_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span> <span class="n">num_repeats</span><span class="o">=</span><span class="mi">4</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer</li>
-<li><strong>num_repeats</strong> (<em>int</em>) &#8211; Repeat the input so many times</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="rotate-layer">
-<h3>rotate_layer<a class="headerlink" href="#rotate-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">rotate_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for rotating 90 degrees (clock-wise) for each feature channel,
-usually used when the input sample is some image or feature map.</p>
-<div class="math">
-\[y(j,i,:) = x(M-i-1,j,:)\]</div>
-<p>where <span class="math">\(x\)</span> is (M x N x C) input, and <span class="math">\(y\)</span> is (N x M x C) output.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">rot</span> <span class="o">=</span> <span class="n">rotate_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                   <span class="n">height</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                   <span class="n">width</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>height</strong> (<em>int</em>) &#8211; The height of the sample matrix</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="seq-reshape-layer">
-<h3>seq_reshape_layer<a class="headerlink" href="#seq-reshape-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">seq_reshape_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for reshaping the sequence. Assume the input sequence has T instances,
-the dimension of each instance is M, and the input reshape_size is N, then the
-output sequence has T*M/N instances, the dimension of each instance is N.</p>
-<p>Note that T*M/N must be an integer.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">reshape</span> <span class="o">=</span> <span class="n">seq_reshape_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span> <span class="n">reshape_size</span><span class="o">=</span><span class="mi">4</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>reshape_size</strong> (<em>int</em>) &#8211; the size of reshaped sequence.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="math-layers">
-<h2>Math Layers<a class="headerlink" href="#math-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="addto-layer">
-<h3>addto_layer<a class="headerlink" href="#addto-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">addto_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>AddtoLayer.</p>
-<div class="math">
-\[y = f(\sum_{i} x_i + b)\]</div>
-<p>where <span class="math">\(y\)</span> is output, <span class="math">\(x\)</span> is input, <span class="math">\(b\)</span> is bias,
-and <span class="math">\(f\)</span> is activation function.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">addto</span> <span class="o">=</span> <span class="n">addto_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">],</span>
-                    <span class="n">act</span><span class="o">=</span><span class="n">ReluActivation</span><span class="p">(),</span>
-                    <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>This layer just simply add all input layers together, then activate the sum
-inputs. Each input of this layer should be the same size, which is also the
-output size of this layer.</p>
-<p>There is no weight matrix for each input, because it just a simple add
-operation. If you want a complicated operation before add, please use
-mixed_layer.</p>
-<p>It is a very good way to set dropout outside the layers. Since not all
-PaddlePaddle layer support dropout, you can add an add_to layer, set
-dropout here.
-Please refer to dropout_layer for details.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; Input layers. It could be a LayerOutput or list/tuple of
-LayerOutput.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type, default is tanh.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|bool</em>) &#8211; Bias attribute. If False, means no bias. None is default
-bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="linear-comb-layer">
-<h3>linear_comb_layer<a class="headerlink" href="#linear-comb-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">linear_comb_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>A layer for weighted sum of vectors takes two inputs.</dt>
-<dd><ul class="first last simple">
-<li><dl class="first docutils">
-<dt>Input: size of weights is M</dt>
-<dd>size of vectors is M*N</dd>
-</dl>
-</li>
-<li>Output: a vector of size=N</li>
-</ul>
-</dd>
-</dl>
-<div class="math">
-\[z(i) = \sum_{j=0}^{M-1} x(j) y(i+Nj)\]</div>
-<p>where <span class="math">\(0 \le i \le N-1\)</span></p>
-<p>Or in the matrix notation:</p>
-<div class="math">
-\[z = x^\mathrm{T} Y\]</div>
-<dl class="docutils">
-<dt>In this formular:</dt>
-<dd><ul class="first last simple">
-<li><span class="math">\(x\)</span>: weights</li>
-<li><span class="math">\(y\)</span>: vectors.</li>
-<li><span class="math">\(z\)</span>: the output.</li>
-</ul>
-</dd>
-</dl>
-<p>Note that the above computation is for one sample. Multiple samples are
-processed in one batch.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">linear_comb</span> <span class="o">=</span> <span class="n">linear_comb_layer</span><span class="p">(</span><span class="n">weights</span><span class="o">=</span><span class="n">weight</span><span class="p">,</span> <span class="n">vectors</span><span class="o">=</span><span class="n">vectors</span><span class="p">,</span>
-                                <span class="n">size</span><span class="o">=</span><span class="n">elem_dim</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>weights</strong> (<em>LayerOutput</em>) &#8211; The weight layer.</li>
-<li><strong>vectors</strong> (<em>LayerOutput</em>) &#8211; The vector layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; the dimension of this layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="interpolation-layer">
-<h3>interpolation_layer<a class="headerlink" href="#interpolation-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">interpolation_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer is for linear interpolation with two inputs,
-which is used in NEURAL TURING MACHINE.</p>
-<div class="math">
-\[y.row[i] = w[i] * x_1.row[i] + (1 - w[i]) * x_2.row[i]\]</div>
-<p>where <span class="math">\(x_1\)</span> and <span class="math">\(x_2\)</span> are two (batchSize x dataDim) inputs,
-<span class="math">\(w\)</span> is (batchSize x 1) weight vector, and <span class="math">\(y\)</span> is
-(batchSize x dataDim) output.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">interpolation</span> <span class="o">=</span> <span class="n">interpolation_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">],</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer3</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>list|tuple</em>) &#8211; Input layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="bilinear-interp-layer">
-<h3>bilinear_interp_layer<a class="headerlink" href="#bilinear-interp-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">bilinear_interp_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer is to implement bilinear interpolation on conv layer output.</p>
-<p>Please refer to Wikipedia: <a class="reference external" href="https://en.wikipedia.org/wiki/Bilinear_interpolation">https://en.wikipedia.org/wiki/Bilinear_interpolation</a></p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">bilinear</span> <span class="o">=</span> <span class="n">bilinear_interp_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">out_size_x</span><span class="o">=</span><span class="mi">64</span><span class="p">,</span> <span class="n">out_size_y</span><span class="o">=</span><span class="mi">64</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; A input layer.</li>
-<li><strong>out_size_x</strong> (<em>int|None</em>) &#8211; bilinear interpolation output width.</li>
-<li><strong>out_size_y</strong> (<em>int|None</em>) &#8211; bilinear interpolation output height.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The layer&#8217;s name, which cna not be specified.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="power-layer">
-<h3>power_layer<a class="headerlink" href="#power-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">power_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer applies a power function to a vector element-wise,
-which is used in NEURAL TURING MACHINE.</p>
-<div class="math">
-\[y = x^w\]</div>
-<p>where <span class="math">\(x\)</span> is a input vector, <span class="math">\(w\)</span> is scalar weight,
-and <span class="math">\(y\)</span> is a output vector.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">power</span> <span class="o">=</span> <span class="n">power_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="scaling-layer">
-<h3>scaling_layer<a class="headerlink" href="#scaling-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">scaling_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for multiplying input vector by weight scalar.</p>
-<div class="math">
-\[y  = w x\]</div>
-<p>where <span class="math">\(x\)</span> is size=dataDim input, <span class="math">\(w\)</span> is size=1 weight,
-and <span class="math">\(y\)</span> is size=dataDim output.</p>
-<p>Note that the above computation is for one sample. Multiple samples are
-processed in one batch.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">scale</span> <span class="o">=</span> <span class="n">scaling_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="slope-intercept-layer">
-<h3>slope_intercept_layer<a class="headerlink" href="#slope-intercept-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">slope_intercept_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer for applying a slope and an intercept to the input
-element-wise. There is no activation and weight.</p>
-<div class="math">
-\[y = slope * x + intercept\]</div>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">scale</span> <span class="o">=</span> <span class="n">slope_intercept_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">slope</span><span class="o">=-</span><span class="mf">1.0</span><span class="p">,</span> <span class="n">intercept</span><span class="o">=</span><span class="mf">1.0</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>slope</strong> (<em>float.</em>) &#8211; the scale factor.</li>
-<li><strong>intercept</strong> (<em>float.</em>) &#8211; the offset.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="tensor-layer">
-<h3>tensor_layer<a class="headerlink" href="#tensor-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">tensor_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer performs tensor operation for two input.
-For example, each sample:</p>
-<div class="math">
-\[y_{i} = a * W_{i} * {b^\mathrm{T}}, i=0,1,...,K-1\]</div>
-<dl class="docutils">
-<dt>In this formular:</dt>
-<dd><ul class="first last simple">
-<li><span class="math">\(a\)</span>: the first input contains M elements.</li>
-<li><span class="math">\(b\)</span>: the second input contains N elements.</li>
-<li><span class="math">\(y_{i}\)</span>: the i-th element of y.</li>
-<li><span class="math">\(W_{i}\)</span>: the i-th learned weight, shape if [M, N]</li>
-<li><span class="math">\(b^\mathrm{T}\)</span>: the transpose of <span class="math">\(b_{2}\)</span>.</li>
-</ul>
-</dd>
-</dl>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">tensor</span> <span class="o">=</span> <span class="n">tensor_layer</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">1000</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; Input layer a.</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input layer b.</li>
-<li><strong>size</strong> (<em>int.</em>) &#8211; the layer dimension.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Default is tanh.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; The Parameter Attribute.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|Any</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="cos-sim">
-<span id="api-trainer-config-helpers-layers-cos-sim"></span><h3>cos_sim<a class="headerlink" href="#cos-sim" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">cos_sim</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Cosine Similarity Layer. The cosine similarity equation is here.</p>
-<div class="math">
-\[similarity = cos(\theta) = {\mathbf{a} \cdot \mathbf{b}
-\over \|\mathbf{a}\| \|\mathbf{b}\|}\]</div>
-<p>The size of a is M, size of b is M*N,
-Similarity will be calculated N times by step M. The output size is
-N. The scale will be multiplied to similarity.</p>
-<p>Note that the above computation is for one sample. Multiple samples are
-processed in one batch.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cos</span> <span class="o">=</span> <span class="n">cos_sim</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">3</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; input layer a</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input layer b</li>
-<li><strong>scale</strong> (<em>float</em>) &#8211; scale for cosine value. default is 5.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; layer size. NOTE size_a * size should equal size_b.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="trans-layer">
-<h3>trans_layer<a class="headerlink" href="#trans-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">trans_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for transposing a minibatch matrix.</p>
-<div class="math">
-\[y = x^\mathrm{T}\]</div>
-<p>where <span class="math">\(x\)</span> is (M x N) input, and <span class="math">\(y\)</span> is (N x M) output.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">trans</span> <span class="o">=</span> <span class="n">trans_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="sampling-layers">
-<h2>Sampling Layers<a class="headerlink" href="#sampling-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="maxid-layer">
-<h3>maxid_layer<a class="headerlink" href="#maxid-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">maxid_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for finding the id which has the maximal value for each sample.
-The result is stored in output.ids.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">maxid</span> <span class="o">=</span> <span class="n">maxid_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="sampling-id-layer">
-<h3>sampling_id_layer<a class="headerlink" href="#sampling-id-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">sampling_id_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for sampling id from multinomial distribution from the input layer.
-Sampling one id for one sample.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">samping_id</span> <span class="o">=</span> <span class="n">sampling_id_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="slicing-and-joining-layers">
-<h2>Slicing and Joining Layers<a class="headerlink" href="#slicing-and-joining-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="pad-layer">
-<h3>pad_layer<a class="headerlink" href="#pad-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">pad_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This operation pads zeros to the input data according to pad_c,pad_h
-and pad_w. pad_c, pad_h, pad_w specifies the which dimension and size
-of padding. And the input data shape is NCHW.</p>
-<p>For example, pad_c=[2,3] means padding 2 zeros before the
-input data and 3 zeros after the input data in channel dimension.
-pad_h means padding zeros in height dimension. pad_w means padding zeros
-in width dimension.</p>
-<p>For example,</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">input</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">)</span>  <span class="o">=</span> <span class="p">[</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">1</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">],</span> <span class="p">[</span><span class="mi">3</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">5</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">6</span><span class="p">,</span><span class="mi">7</span><span class="p">]]</span> <span class="p">],</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">4</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">7</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">3</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">9</span><span class="p">],</span> <span class="p">[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">]]</span> <span class="p">]</span>
-                  <span class="p">]</span>
-
-<span class="n">pad_c</span><span class="o">=</span><span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">1</span><span class="p">],</span> <span class="n">pad_h</span><span class="o">=</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="n">pad_w</span><span class="o">=</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]</span>
-
-<span class="n">output</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">)</span> <span class="o">=</span> <span class="p">[</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">1</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">],</span> <span class="p">[</span><span class="mi">3</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">5</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">6</span><span class="p">,</span><span class="mi">7</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]]</span> <span class="p">],</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">4</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">7</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">3</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">9</span><span class="p">],</span> <span class="p">[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]]</span> <span class="p">]</span>
-                  <span class="p">]</span>
-</pre></div>
-</div>
-<p>The simply usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">pad</span> <span class="o">=</span> <span class="n">pad_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">ipt</span><span class="p">,</span>
-                <span class="n">pad_c</span><span class="o">=</span><span class="p">[</span><span class="mi">4</span><span class="p">,</span><span class="mi">4</span><span class="p">],</span>
-                <span class="n">pad_h</span><span class="o">=</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span>
-                <span class="n">pad_w</span><span class="o">=</span><span class="p">[</span><span class="mi">2</span><span class="p">,</span><span class="mi">2</span><span class="p">])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input.</li>
-<li><strong>pad_c</strong> (<em>list|None</em>) &#8211; padding size in channel dimension.</li>
-<li><strong>pad_h</strong> (<em>list|None</em>) &#8211; padding size in height dimension.</li>
-<li><strong>pad_w</strong> (<em>list|None</em>) &#8211; padding size in width dimension.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="cost-layers">
-<span id="api-trainer-config-helpers-layers-cost-layers"></span><h2>Cost Layers<a class="headerlink" href="#cost-layers" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="cross-entropy">
-<h3>cross_entropy<a class="headerlink" href="#cross-entropy" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">cross_entropy</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for multi class entropy.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">cross_entropy</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                     <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float.</em>) &#8211; The cost is multiplied with coeff.
-The coefficient affects the gradient in the backward.</li>
-<li><strong>weight</strong> (<em>LayerOutout</em>) &#8211; The cost of each sample is multiplied with each weight.
-The weight should be a layer with size=1. Note that gradient
-will not be calculated for weight.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="cross-entropy-with-selfnorm">
-<h3>cross_entropy_with_selfnorm<a class="headerlink" href="#cross-entropy-with-selfnorm" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">cross_entropy_with_selfnorm</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for multi class entropy with selfnorm.
-Input should be a vector of positive numbers, without normalization.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">cross_entropy_with_selfnorm</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                                   <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float.</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>softmax_selfnorm_alpha</strong> (<em>float.</em>) &#8211; The scale factor affects the cost.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="multi-binary-label-cross-entropy">
-<h3>multi_binary_label_cross_entropy<a class="headerlink" href="#multi-binary-label-cross-entropy" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">multi_binary_label_cross_entropy</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for multi binary label cross entropy.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">multi_binary_label_cross_entropy</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                                        <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="mse-cost">
-<h3>mse_cost<a class="headerlink" href="#mse-cost" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">mse_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><blockquote>
-<div><p>mean squared error cost:</p>
-<div class="math">
-\[\]</div>
-</div></blockquote>
-<p>rac{1}{N}sum_{i=1}^N(t_i-y_i)^2</p>
-<blockquote>
-<div><table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">param name:</th><td class="field-body">layer name.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type name:</th><td class="field-body">basestring</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">param input:</th><td class="field-body">Network prediction.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type input:</th><td class="field-body">LayerOutput</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">param label:</th><td class="field-body">Data label.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type label:</th><td class="field-body">LayerOutput</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">param weight:</th><td class="field-body">The weight affects the cost, namely the scale of cost.
-It is an optional argument.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type weight:</th><td class="field-body">LayerOutput</td>
-</tr>
-<tr class="field-odd field"><th class="field-name" colspan="2">param layer_attr:</th></tr>
-<tr class="field-odd field"><td>&#160;</td><td class="field-body">layer&#8217;s extra attribute.</td>
-</tr>
-<tr class="field-even field"><th class="field-name" colspan="2">type layer_attr:</th></tr>
-<tr class="field-even field"><td>&#160;</td><td class="field-body">ExtraLayerAttribute</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">return:</th><td class="field-body">LayerOutput object.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">rtype:</th><td class="field-body">LayerOutput</td>
-</tr>
-</tbody>
-</table>
-</div></blockquote>
-</dd></dl>
-
-</div>
-<div class="section" id="huber-cost">
-<h3>huber_cost<a class="headerlink" href="#huber-cost" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">huber_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for huber loss.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">huber_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                  <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float.</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lambda-cost">
-<h3>lambda_cost<a class="headerlink" href="#lambda-cost" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">lambda_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>lambdaCost for lambdaRank LTR approach.</p>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">lambda_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                   <span class="n">score</span><span class="o">=</span><span class="n">score</span><span class="p">,</span>
-                   <span class="n">NDCG_num</span><span class="o">=</span><span class="mi">8</span><span class="p">,</span>
-                   <span class="n">max_sort_size</span><span class="o">=-</span><span class="mi">1</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Samples of the same query should be loaded as sequence.</li>
-<li><strong>score</strong> &#8211; The 2nd input. Score of each sample.</li>
-<li><strong>NDCG_num</strong> (<em>int</em>) &#8211; The size of NDCG (Normalized Discounted Cumulative Gain),
-e.g., 5 for NDCG&#64;5. It must be less than for equal to the
-minimum size of lists.</li>
-<li><strong>max_sort_size</strong> (<em>int</em>) &#8211; The size of partial sorting in calculating gradient.
-If max_sort_size = -1, then for each list, the
-algorithm will sort the entire list to get gradient.
-In other cases, max_sort_size must be greater than or
-equal to NDCG_num. And if max_sort_size is greater
-than the size of a list, the algorithm will sort the
-entire list of get gradient.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="rank-cost">
-<h3>rank_cost<a class="headerlink" href="#rank-cost" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">rank_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A cost Layer for learning to rank using gradient descent. Details can refer
-to <a class="reference external" href="http://research.microsoft.com/en-us/um/people/cburges/papers/ICML_ranking.pdf">papers</a>.
-This layer contains at least three inputs. The weight is an optional
-argument, which affects the cost.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}C_{i,j} &amp; = -\tilde{P_{ij}} * o_{i,j} + log(1 + e^{o_{i,j}})\\o_{i,j} &amp; =  o_i - o_j\\\tilde{P_{i,j}} &amp; = \{0, 0.5, 1\} \ or \ \{0, 1\}\end{aligned}\end{align} \]</div>
-<dl class="docutils">
-<dt>In this formula:</dt>
-<dd><ul class="first last simple">
-<li><span class="math">\(C_{i,j}\)</span> is the cross entropy cost.</li>
-<li><span class="math">\(\tilde{P_{i,j}}\)</span> is the label. 1 means positive order
-and 0 means reverse order.</li>
-<li><span class="math">\(o_i\)</span> and <span class="math">\(o_j\)</span>: the left output and right output.
-Their dimension is one.</li>
-</ul>
-</dd>
-</dl>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">rank_cost</span><span class="p">(</span><span class="n">left</span><span class="o">=</span><span class="n">out_left</span><span class="p">,</span>
-                 <span class="n">right</span><span class="o">=</span><span class="n">out_right</span><span class="p">,</span>
-                 <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>left</strong> (<em>LayerOutput</em>) &#8211; The first input, the size of this layer is 1.</li>
-<li><strong>right</strong> (<em>LayerOutput</em>) &#8211; The right input, the size of this layer is 1.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label is 1 or 0, means positive order and reverse order.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; The weight affects the cost, namely the scale of cost.
-It is an optional argument.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="sum-cost">
-<h3>sum_cost<a class="headerlink" href="#sum-cost" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">sum_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer which calculate the sum of the input as loss</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">sum_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="crf-layer">
-<h3>crf_layer<a class="headerlink" href="#crf-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">crf_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for calculating the cost of sequential conditional random
-field model.</p>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">crf</span> <span class="o">=</span> <span class="n">crf_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">,</span>
-                <span class="n">size</span><span class="o">=</span><span class="n">label_dim</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The first input layer is the feature.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; The second input layer is label.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The category number.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; The third layer is &#8220;weight&#8221; of each sample, which is an
-optional argument.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter attribute. None means default attribute</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="crf-decoding-layer">
-<h3>crf_decoding_layer<a class="headerlink" href="#crf-decoding-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">crf_decoding_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for calculating the decoding sequence of sequential conditional
-random field model. The decoding sequence is stored in output.ids.
-If a second input is provided, it is treated as the ground-truth label, and
-this layer will also calculate error. output.value[i] is 1 for incorrect
-decoding or 0 for correct decoding.</p>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">crf_decoding</span> <span class="o">=</span> <span class="n">crf_decoding_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                                  <span class="n">size</span><span class="o">=</span><span class="n">label_dim</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The first input layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; size of this layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em><em> or </em><em>None</em>) &#8211; None or ground-truth label.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter attribute. None means default attribute</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="ctc-layer">
-<h3>ctc_layer<a class="headerlink" href="#ctc-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">ctc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Connectionist Temporal Classification (CTC) is designed for temporal
-classication task. That is, for sequence labeling problems where the
-alignment between the inputs and the target labels is unknown.</p>
-<p>More details can be found by referring to <a class="reference external" href="http://machinelearning.wustl.edu/mlpapers/paper_files/icml2006_GravesFGS06.pdf">Connectionist Temporal
-Classification: Labelling Unsegmented Sequence Data with Recurrent
-Neural Networks</a></p>
-<div class="admonition note">
-<p class="first admonition-title">Note</p>
-<p class="last">Considering the &#8216;blank&#8217; label needed by CTC, you need to use
-(num_classes + 1) as the input size. num_classes is the category number.
-And the &#8216;blank&#8217; is the last category index. So the size of &#8216;input&#8217; layer, such as
-fc_layer with softmax activation, should be num_classes + 1. The size of ctc_layer
-should also be num_classes + 1.</p>
-</div>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">ctc</span> <span class="o">=</span> <span class="n">ctc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">,</span>
-                <span class="n">size</span><span class="o">=</span><span class="mi">9055</span><span class="p">,</span>
-                <span class="n">norm_by_times</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; The data layer of label with variable length.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; category numbers + 1.</li>
-<li><strong>name</strong> (<em>basestring|None</em>) &#8211; The name of this layer</li>
-<li><strong>norm_by_times</strong> (<em>bool</em>) &#8211; Whether to normalization by times. False by default.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="warp-ctc-layer">
-<h3>warp_ctc_layer<a class="headerlink" href="#warp-ctc-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">warp_ctc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer intergrating the open-source <cite>warp-ctc
-&lt;https://github.com/baidu-research/warp-ctc&gt;</cite> library, which is used in
-<cite>Deep Speech 2: End-toEnd Speech Recognition in English and Mandarin
-&lt;https://arxiv.org/pdf/1512.02595v1.pdf&gt;</cite>, to compute Connectionist Temporal
-Classification (CTC) loss.</p>
-<p>More details of CTC can be found by referring to <a class="reference external" href="http://machinelearning.wustl.edu/mlpapers/paper_files/icml2006_GravesFGS06.pdf">Connectionist Temporal
-Classification: Labelling Unsegmented Sequence Data with Recurrent
-Neural Networks</a></p>
-<div class="admonition note">
-<p class="first admonition-title">Note</p>
-<ul class="last simple">
-<li>Let num_classes represent the category number. Considering the &#8216;blank&#8217;
-label needed by CTC, you need to use (num_classes + 1) as the input
-size. Thus, the size of both warp_ctc_layer and &#8216;input&#8217; layer should
-be set to num_classes + 1.</li>
-<li>You can set &#8216;blank&#8217; to any value ranged in [0, num_classes], which
-should be consistent as that used in your labels.</li>
-<li>As a native &#8216;softmax&#8217; activation is interated to the warp-ctc library,
-&#8216;linear&#8217; activation is expected instead in the &#8216;input&#8217; layer.</li>
-</ul>
-</div>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">ctc</span> <span class="o">=</span> <span class="n">warp_ctc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                     <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">,</span>
-                     <span class="n">size</span><span class="o">=</span><span class="mi">1001</span><span class="p">,</span>
-                     <span class="n">blank</span><span class="o">=</span><span class="mi">1000</span><span class="p">,</span>
-                     <span class="n">norm_by_times</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; The data layer of label with variable length.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; category numbers + 1.</li>
-<li><strong>name</strong> (<em>basestring|None</em>) &#8211; The name of this layer, which can not specify.</li>
-<li><strong>blank</strong> (<em>int</em>) &#8211; the &#8216;blank&#8217; label used in ctc</li>
-<li><strong>norm_by_times</strong> (<em>bool</em>) &#8211; Whether to normalization by times. False by default.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="nce-layer">
-<h3>nce_layer<a class="headerlink" href="#nce-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">nce_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Noise-contrastive estimation.
-Implements the method in the following paper:
-A fast and simple algorithm for training neural probabilistic language models.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">nce_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer3</span><span class="p">,</span>
-                 <span class="n">num_classes</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">neg_distribution</span><span class="o">=</span><span class="p">[</span><span class="mf">0.1</span><span class="p">,</span><span class="mf">0.3</span><span class="p">,</span><span class="mf">0.6</span><span class="p">])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple|collections.Sequence</em>) &#8211; input layers. It could be a LayerOutput of list/tuple of LayerOutput.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; label layer</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; weight layer, can be None(default)</li>
-<li><strong>num_classes</strong> (<em>int</em>) &#8211; number of classes.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation, default is Sigmoid.</li>
-<li><strong>num_neg_samples</strong> (<em>int</em>) &#8211; number of negative samples. Default is 10.</li>
-<li><strong>neg_distribution</strong> (<em>list|tuple|collections.Sequence|None</em>) &#8211; The distribution for generating the random negative labels.
-A uniform distribution will be used if not provided.
-If not None, its length must be equal to num_classes.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias parameter attribute. True if no bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="hsigmoid">
-<h3>hsigmoid<a class="headerlink" href="#hsigmoid" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">hsigmoid</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Organize the classes into a binary tree. At each node, a sigmoid function
-is used to calculate the probability of belonging to the right branch.
-This idea is from &#8220;F. Morin, Y. Bengio (AISTATS 05):
-Hierarchical Probabilistic Neural Network Language Model.&#8221;</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">hsigmoid</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">],</span>
-                <span class="n">label</span><span class="o">=</span><span class="n">data_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; Input layers. It could be a LayerOutput or list/tuple of
-LayerOutput.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label layer.</li>
-<li><strong>num_classes</strong> (<em>int|None</em>) &#8211; number of classes.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Bias attribute. None means default bias.
-False means no bias.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None</em>) &#8211; Parameter Attribute. None means default parameter.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="smooth-l1-cost">
-<h3>smooth_l1_cost<a class="headerlink" href="#smooth-l1-cost" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">smooth_l1_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This is a L1 loss but more smooth. It requires that the
-size of input and label are equal. The formula is as follows,</p>
-<div class="math">
-\[L = \sum_{i} smooth_{L1}(input_i - label_i)\]</div>
-<p>in which</p>
-<div class="math">
-\[\begin{split}smooth_{L1}(x) = \begin{cases} 0.5x^2&amp; \text{if}  \ |x| &lt; 1 \\ |x|-0.5&amp; \text{otherwise} \end{cases}\end{split}\]</div>
-<p>More details can be found by referring to <a class="reference external" href="https://arxiv.org/pdf/1504.08083v2.pdf">Fast R-CNN</a></p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">smooth_l1_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                      <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="check-layer">
-<h2>Check Layer<a class="headerlink" href="#check-layer" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="eos-layer">
-<h3>eos_layer<a class="headerlink" href="#eos-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">eos_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for checking EOS for each sample:
-- output_id = (input_id == conf.eos_id)</p>
-<p>The result is stored in output_.ids.
-It is used by recurrent layer group.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">eos</span> <span class="o">=</span> <span class="n">eos_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span> <span class="n">eos_id</span><span class="o">=</span><span class="nb">id</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>eos_id</strong> (<em>int</em>) &#8211; end id of sequence</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/api/v1/trainer_config_helpers/networks.html b/develop/doc/api/v1/trainer_config_helpers/networks.html
deleted file mode 100644
index 9769b3b4bf8364610f1a032536c14b7135e9a152..0000000000000000000000000000000000000000
--- a/develop/doc/api/v1/trainer_config_helpers/networks.html
+++ /dev/null
@@ -1,949 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Networks &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="../../../genindex.html"/>
-        <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Networks</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="networks">
-<h1>Networks<a class="headerlink" href="#networks" title="Permalink to this headline">¶</a></h1>
-<p>The networks module contains pieces of neural network that combine multiple layers.</p>
-<div class="section" id="nlp">
-<h2>NLP<a class="headerlink" href="#nlp" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="sequence-conv-pool">
-<h3>sequence_conv_pool<a class="headerlink" href="#sequence-conv-pool" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">sequence_conv_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Text convolution pooling layers helper.</p>
-<p>Text input =&gt; Context Projection =&gt; FC Layer =&gt; Pooling =&gt; Output.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of output layer(pooling layer name)</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; name of input layer</li>
-<li><strong>context_len</strong> (<em>int</em>) &#8211; context projection length. See
-context_projection&#8217;s document.</li>
-<li><strong>hidden_size</strong> (<em>int</em>) &#8211; FC Layer size.</li>
-<li><strong>context_start</strong> (<em>int</em><em> or </em><em>None</em>) &#8211; context projection length. See
-context_projection&#8217;s context_start.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType.</em>) &#8211; pooling layer type. See pooling_layer&#8217;s document.</li>
-<li><strong>context_proj_layer_name</strong> (<em>basestring</em>) &#8211; context projection layer name.
-None if user don&#8217;t care.</li>
-<li><strong>context_proj_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; context projection parameter attribute.
-None if user don&#8217;t care.</li>
-<li><strong>fc_layer_name</strong> (<em>basestring</em>) &#8211; fc layer name. None if user don&#8217;t care.</li>
-<li><strong>fc_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc layer parameter attribute. None if user don&#8217;t care.</li>
-<li><strong>fc_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc bias parameter attribute. False if no bias,
-None if user don&#8217;t care.</li>
-<li><strong>fc_act</strong> (<em>BaseActivation</em>) &#8211; fc layer activation type. None means tanh</li>
-<li><strong>pool_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; pooling layer bias attr. None if don&#8217;t care.
-False if no bias.</li>
-<li><strong>fc_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; fc layer extra attribute.</li>
-<li><strong>context_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; context projection layer extra attribute.</li>
-<li><strong>pool_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; pooling layer extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">output layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="text-conv-pool">
-<span id="api-trainer-config-helpers-network-text-conv-pool"></span><h3>text_conv_pool<a class="headerlink" href="#text-conv-pool" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">text_conv_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Text convolution pooling layers helper.</p>
-<p>Text input =&gt; Context Projection =&gt; FC Layer =&gt; Pooling =&gt; Output.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of output layer(pooling layer name)</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; name of input layer</li>
-<li><strong>context_len</strong> (<em>int</em>) &#8211; context projection length. See
-context_projection&#8217;s document.</li>
-<li><strong>hidden_size</strong> (<em>int</em>) &#8211; FC Layer size.</li>
-<li><strong>context_start</strong> (<em>int</em><em> or </em><em>None</em>) &#8211; context projection length. See
-context_projection&#8217;s context_start.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType.</em>) &#8211; pooling layer type. See pooling_layer&#8217;s document.</li>
-<li><strong>context_proj_layer_name</strong> (<em>basestring</em>) &#8211; context projection layer name.
-None if user don&#8217;t care.</li>
-<li><strong>context_proj_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; context projection parameter attribute.
-None if user don&#8217;t care.</li>
-<li><strong>fc_layer_name</strong> (<em>basestring</em>) &#8211; fc layer name. None if user don&#8217;t care.</li>
-<li><strong>fc_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc layer parameter attribute. None if user don&#8217;t care.</li>
-<li><strong>fc_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc bias parameter attribute. False if no bias,
-None if user don&#8217;t care.</li>
-<li><strong>fc_act</strong> (<em>BaseActivation</em>) &#8211; fc layer activation type. None means tanh</li>
-<li><strong>pool_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; pooling layer bias attr. None if don&#8217;t care.
-False if no bias.</li>
-<li><strong>fc_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; fc layer extra attribute.</li>
-<li><strong>context_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; context projection layer extra attribute.</li>
-<li><strong>pool_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; pooling layer extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">output layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="images">
-<h2>Images<a class="headerlink" href="#images" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="img-conv-bn-pool">
-<h3>img_conv_bn_pool<a class="headerlink" href="#img-conv-bn-pool" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">img_conv_bn_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Convolution, batch normalization, pooling group.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; group name</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document</li>
-<li><strong>pool_size</strong> (<em>int</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; see batch_norm_layer&#8217;s document.</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document</li>
-<li><strong>conv_stride</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_padding</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>num_channel</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>shared_bias</strong> (<em>bool</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_layer_attr</strong> (<em>ExtraLayerOutput</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>bn_param_attr</strong> (<em>ParameterAttribute.</em>) &#8211; see batch_norm_layer&#8217;s document.</li>
-<li><strong>bn_bias_attr</strong> &#8211; see batch_norm_layer&#8217;s document.</li>
-<li><strong>bn_layer_attr</strong> &#8211; ParameterAttribute.</li>
-<li><strong>pool_stride</strong> (<em>int</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>pool_padding</strong> (<em>int</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>pool_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; see img_pool_layer&#8217;s document.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">Layer groups output</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="img-conv-group">
-<h3>img_conv_group<a class="headerlink" href="#img-conv-group" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">img_conv_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Image Convolution Group, Used for vgg net.</p>
-<p>TODO(yuyang18): Complete docs</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>conv_batchnorm_drop_rate</strong> &#8211; </li>
-<li><strong>input</strong> &#8211; </li>
-<li><strong>conv_num_filter</strong> &#8211; </li>
-<li><strong>pool_size</strong> &#8211; </li>
-<li><strong>num_channels</strong> &#8211; </li>
-<li><strong>conv_padding</strong> &#8211; </li>
-<li><strong>conv_filter_size</strong> &#8211; </li>
-<li><strong>conv_act</strong> &#8211; </li>
-<li><strong>conv_with_batchnorm</strong> &#8211; </li>
-<li><strong>pool_stride</strong> &#8211; </li>
-<li><strong>pool_type</strong> &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first last"></p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="simple-img-conv-pool">
-<span id="api-trainer-config-helpers-network-simple-img-conv-pool"></span><h3>simple_img_conv_pool<a class="headerlink" href="#simple-img-conv-pool" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_img_conv_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Simple image convolution and pooling group.</p>
-<p>Input =&gt; conv =&gt; pooling</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; group name</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>pool_size</strong> (<em>int</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>conv_stride</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>conv_padding</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer for details</li>
-<li><strong>num_channel</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer for details</li>
-<li><strong>shared_bias</strong> (<em>bool</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>conv_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; see img_conv_layer for details</li>
-<li><strong>pool_stride</strong> (<em>int</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>pool_padding</strong> (<em>int</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>pool_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; see img_pool_layer for details</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">Layer&#8217;s output</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="vgg-16-network">
-<h3>vgg_16_network<a class="headerlink" href="#vgg-16-network" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">vgg_16_network</code><span class="sig-paren">(</span><em>input_image</em>, <em>num_channels</em>, <em>num_classes=1000</em><span class="sig-paren">)</span></dt>
-<dd><p>Same model from <a class="reference external" href="https://gist.github.com/ksimonyan/211839e770f7b538e2d8">https://gist.github.com/ksimonyan/211839e770f7b538e2d8</a></p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>num_classes</strong> &#8211; </li>
-<li><strong>input_image</strong> (<em>LayerOutput</em>) &#8211; </li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first last"></p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="recurrent">
-<h2>Recurrent<a class="headerlink" href="#recurrent" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="lstm">
-<h3>LSTM<a class="headerlink" href="#lstm" title="Permalink to this headline">¶</a></h3>
-<div class="section" id="lstmemory-unit">
-<h4>lstmemory_unit<a class="headerlink" href="#lstmemory-unit" title="Permalink to this headline">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">lstmemory_unit</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define calculations that a LSTM unit performs in a single time step.
-This function itself is not a recurrent layer, so that it can not be
-directly applied to sequence input. This function is always used in
-recurrent_group (see layers.py for more details) to implement attention
-mechanism.</p>
-<p>Please refer to  <strong>Generating Sequences With Recurrent Neural Networks</strong>
-for more details about LSTM. The link goes as follows:
-.. _Link: <a class="reference external" href="https://arxiv.org/abs/1308.0850">https://arxiv.org/abs/1308.0850</a></p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">lstm_step</span> <span class="o">=</span> <span class="n">lstmemory_unit</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span>
-                           <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                           <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">(),</span>
-                           <span class="n">gate_act</span><span class="o">=</span><span class="n">SigmoidActivation</span><span class="p">(),</span>
-                           <span class="n">state_act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; lstmemory unit name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstmemory unit size.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; lstm final activiation type</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; lstm gate activiation type</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; lstm state activiation type.</li>
-<li><strong>mixed_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of mixed layer.
-False means no bias, None means default bias.</li>
-<li><strong>lstm_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of lstm layer.
-False means no bias, None means default bias.</li>
-<li><strong>mixed_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; mixed layer&#8217;s extra attribute.</li>
-<li><strong>lstm_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; lstm layer&#8217;s extra attribute.</li>
-<li><strong>get_output_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; get output layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">lstmemory unit name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lstmemory-group">
-<h4>lstmemory_group<a class="headerlink" href="#lstmemory-group" title="Permalink to this headline">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">lstmemory_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>lstm_group is a recurrent layer group version of Long Short Term Memory. It
-does exactly the same calculation as the lstmemory layer (see lstmemory in
-layers.py for the maths) does. A promising benefit is that LSTM memory
-cell states, or hidden states in every time step are accessible to the
-user. This is especially useful in attention model. If you do not need to
-access the internal states of the lstm, but merely use its outputs,
-it is recommended to use the lstmemory, which is relatively faster than
-lstmemory_group.</p>
-<p>NOTE: In PaddlePaddle&#8217;s implementation, the following input-to-hidden
-multiplications:
-<span class="math">\(W_{xi}x_{t}\)</span> , <span class="math">\(W_{xf}x_{t}\)</span>,
-<span class="math">\(W_{xc}x_t\)</span>, <span class="math">\(W_{xo}x_{t}\)</span> are not done in lstmemory_unit to
-speed up the calculations. Consequently, an additional mixed_layer with
-full_matrix_projection must be included before lstmemory_unit is called.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">lstm_step</span> <span class="o">=</span> <span class="n">lstmemory_group</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span>
-                            <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                            <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">(),</span>
-                            <span class="n">gate_act</span><span class="o">=</span><span class="n">SigmoidActivation</span><span class="p">(),</span>
-                            <span class="n">state_act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; lstmemory group name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstmemory group size.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; is lstm reversed</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; lstm final activiation type</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; lstm gate activiation type</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; lstm state activiation type.</li>
-<li><strong>mixed_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of mixed layer.
-False means no bias, None means default bias.</li>
-<li><strong>lstm_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of lstm layer.
-False means no bias, None means default bias.</li>
-<li><strong>mixed_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; mixed layer&#8217;s extra attribute.</li>
-<li><strong>lstm_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; lstm layer&#8217;s extra attribute.</li>
-<li><strong>get_output_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; get output layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">the lstmemory group.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="simple-lstm">
-<h4>simple_lstm<a class="headerlink" href="#simple-lstm" title="Permalink to this headline">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_lstm</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Simple LSTM Cell.</p>
-<p>It just combine a mixed layer with fully_matrix_projection and a lstmemory
-layer. The simple lstm cell was implemented as follow equations.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>Please refer <strong>Generating Sequences With Recurrent Neural Networks</strong> if you
-want to know what lstm is. <a class="reference external" href="http://arxiv.org/abs/1308.0850">Link</a> is here.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; lstm layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstm layer size.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
-<li><strong>mat_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; mixed layer&#8217;s matrix projection parameter attribute.</li>
-<li><strong>bias_param_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute. False means no bias, None
-means default bias.</li>
-<li><strong>inner_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; lstm cell parameter attribute.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; lstm final activiation type</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; lstm gate activiation type</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; lstm state activiation type.</li>
-<li><strong>mixed_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; mixed layer&#8217;s extra attribute.</li>
-<li><strong>lstm_cell_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; lstm layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">lstm layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="bidirectional-lstm">
-<h4>bidirectional_lstm<a class="headerlink" href="#bidirectional-lstm" title="Permalink to this headline">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">bidirectional_lstm</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A bidirectional_lstm is a recurrent unit that iterates over the input
-sequence both in forward and bardward orders, and then concatenate two
-outputs form a final output. However, concatenation of two outputs
-is not the only way to form the final output, you can also, for example,
-just add them together.</p>
-<p>Please refer to  <strong>Neural Machine Translation by Jointly Learning to Align
-and Translate</strong> for more details about the bidirectional lstm.
-The link goes as follows:
-.. _Link: <a class="reference external" href="https://arxiv.org/pdf/1409.0473v3.pdf">https://arxiv.org/pdf/1409.0473v3.pdf</a></p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">bi_lstm</span> <span class="o">=</span> <span class="n">bidirectional_lstm</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">input1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; bidirectional lstm layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstm layer size.</li>
-<li><strong>return_seq</strong> (<em>bool</em>) &#8211; If set False, outputs of the last time step are
-concatenated and returned.
-If set True, the entire output sequences that are
-processed in forward and backward directions are
-concatenated and returned.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">LayerOutput object accroding to the return_seq.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="gru">
-<h3>GRU<a class="headerlink" href="#gru" title="Permalink to this headline">¶</a></h3>
-<div class="section" id="gru-unit">
-<h4>gru_unit<a class="headerlink" href="#gru-unit" title="Permalink to this headline">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">gru_unit</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define calculations that a gated recurrent unit performs in a single time
-step. This function itself is not a recurrent layer, so that it can not be
-directly applied to sequence input. This function is almost always used in
-the recurrent_group (see layers.py for more details) to implement attention
-mechanism.</p>
-<p>Please see grumemory in layers.py for the details about the maths.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; type of the activation</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; type of the gate activation</li>
-<li><strong>gru_layer_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">the gru output layer.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="gru-group">
-<h4>gru_group<a class="headerlink" href="#gru-group" title="Permalink to this headline">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">gru_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>gru_group is a recurrent layer group version of Gated Recurrent Unit. It
-does exactly the same calculation as the grumemory layer does. A promising
-benefit is that gru hidden states are accessible to the user. This is
-especially useful in attention model. If you do not need to access
-any internal state, but merely use the outputs of a GRU, it is recommended
-to use the grumemory, which is relatively faster.</p>
-<p>Please see grumemory in layers.py for more detail about the maths.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">gur_group</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span>
-                <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">(),</span>
-                <span class="n">gate_act</span><span class="o">=</span><span class="n">SigmoidActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; type of the activiation</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; type of the gate activiation</li>
-<li><strong>gru_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias. False means no bias, None means default bias.</li>
-<li><strong>gru_layer_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">the gru group.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="simple-gru">
-<h4>simple_gru<a class="headerlink" href="#simple-gru" title="Permalink to this headline">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_gru</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>You maybe see gru_step_layer, grumemory in layers.py, gru_unit, gru_group,
-simple_gru in network.py. The reason why there are so many interfaces is
-that we have two ways to implement recurrent neural network. One way is to
-use one complete layer to implement rnn (including simple rnn, gru and lstm)
-with multiple time steps, such as recurrent_layer, lstmemory, grumemory. But,
-the multiplication operation <span class="math">\(W x_t\)</span> is not computed in these layers.
-See details in their interfaces in layers.py.
-The other implementation is to use an recurrent group which can ensemble a
-series of layers to compute rnn step by step. This way is flexible for
-attenion mechanism or other complex connections.</p>
-<ul class="simple">
-<li>gru_step_layer: only compute rnn by one step. It needs an memory as input
-and can be used in recurrent group.</li>
-<li>gru_unit: a wrapper of gru_step_layer with memory.</li>
-<li>gru_group: a GRU cell implemented by a combination of multiple layers in
-recurrent group.
-But <span class="math">\(W x_t\)</span> is not done in group.</li>
-<li>gru_memory: a GRU cell implemented by one layer, which does same calculation
-with gru_group and is faster than gru_group.</li>
-<li>simple_gru: a complete GRU implementation inlcuding <span class="math">\(W x_t\)</span> and
-gru_group. <span class="math">\(W\)</span> contains <span class="math">\(W_r\)</span>, <span class="math">\(W_z\)</span> and <span class="math">\(W\)</span>, see
-formula in grumemory.</li>
-</ul>
-<p>The computational speed is that, grumemory is relatively better than
-gru_group, and gru_group is relatively better than simple_gru.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">simple_gru</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; type of the activiation</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; type of the gate activiation</li>
-<li><strong>gru_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias. False means no bias, None means default bias.</li>
-<li><strong>gru_layer_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">the gru group.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="simple-attention">
-<h3>simple_attention<a class="headerlink" href="#simple-attention" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_attention</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Calculate and then return a context vector by attention machanism.
-Size of the context vector equals to size of the encoded_sequence.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}a(s_{i-1},h_{j}) &amp; = v_{a}f(W_{a}s_{t-1} + U_{a}h_{j})\\e_{i,j} &amp; = a(s_{i-1}, h_{j})\\a_{i,j} &amp; = \frac{exp(e_{i,j})}{\sum_{k=1}^{T_x}{exp(e_{i,k})}}\\c_{i} &amp; = \sum_{j=1}^{T_{x}}a_{i,j}h_{j}\end{aligned}\end{align} \]</div>
-<p>where <span class="math">\(h_{j}\)</span> is the jth element of encoded_sequence,
-<span class="math">\(U_{a}h_{j}\)</span> is the jth element of encoded_proj
-<span class="math">\(s_{i-1}\)</span> is decoder_state
-<span class="math">\(f\)</span> is weight_act, and is set to tanh by default.</p>
-<p>Please refer to <strong>Neural Machine Translation by Jointly Learning to
-Align and Translate</strong> for more details. The link is as follows:
-<a class="reference external" href="https://arxiv.org/abs/1409.0473">https://arxiv.org/abs/1409.0473</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">context</span> <span class="o">=</span> <span class="n">simple_attention</span><span class="p">(</span><span class="n">encoded_sequence</span><span class="o">=</span><span class="n">enc_seq</span><span class="p">,</span>
-                           <span class="n">encoded_proj</span><span class="o">=</span><span class="n">enc_proj</span><span class="p">,</span>
-                           <span class="n">decoder_state</span><span class="o">=</span><span class="n">decoder_prev</span><span class="p">,)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the attention model.</li>
-<li><strong>softmax_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; parameter attribute of sequence softmax
-that is used to produce attention weight</li>
-<li><strong>weight_act</strong> (<em>Activation</em>) &#8211; activation of the attention model</li>
-<li><strong>encoded_sequence</strong> (<em>LayerOutput</em>) &#8211; output of the encoder</li>
-<li><strong>encoded_proj</strong> (<em>LayerOutput</em>) &#8211; attention weight is computed by a feed forward neural
-network which has two inputs : decoder&#8217;s hidden state
-of previous time step and encoder&#8217;s output.
-encoded_proj is output of the feed-forward network for
-encoder&#8217;s output. Here we pre-compute it outside
-simple_attention for speed consideration.</li>
-<li><strong>decoder_state</strong> (<em>LayerOutput</em>) &#8211; hidden state of decoder in previous time step</li>
-<li><strong>transform_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; parameter attribute of the feed-forward
-network that takes decoder_state as inputs to
-compute attention weight.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first last">a context vector</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="miscs">
-<h2>Miscs<a class="headerlink" href="#miscs" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="dropout-layer">
-<h3>dropout_layer<a class="headerlink" href="#dropout-layer" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">dropout_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>&#64;TODO(yuyang18): Add comments.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> &#8211; </li>
-<li><strong>input</strong> &#8211; </li>
-<li><strong>dropout_rate</strong> &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first last"></p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="outputs">
-<h3>outputs<a class="headerlink" href="#outputs" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">outputs</code><span class="sig-paren">(</span><em>layers</em>, <em>*args</em><span class="sig-paren">)</span></dt>
-<dd><p>Declare the outputs of network. If user have not defined the inputs of
-network, this method will calculate the input order by dfs travel.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>layers</strong> (<em>list|tuple|LayerOutput</em>) &#8211; Output layers.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/api/v1/trainer_config_helpers/optimizers.html b/develop/doc/api/v1/trainer_config_helpers/optimizers.html
deleted file mode 100644
index c8f15d6a847f12fa581c6861b80a01af66015c9f..0000000000000000000000000000000000000000
--- a/develop/doc/api/v1/trainer_config_helpers/optimizers.html
+++ /dev/null
@@ -1,474 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Optimizers &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="../../../genindex.html"/>
-        <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Optimizers</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="optimizers">
-<span id="api-trainer-config-helpers-optimizers"></span><h1>Optimizers<a class="headerlink" href="#optimizers" title="Permalink to this headline">¶</a></h1>
-<div class="section" id="basesgdoptimizer">
-<h2>BaseSGDOptimizer<a class="headerlink" href="#basesgdoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">BaseSGDOptimizer</code></dt>
-<dd><p>SGD Optimizer.</p>
-<p>SGD is an optimization method, trying to find a neural network that
-minimize the &#8220;cost/error&#8221; of it by iteration. In paddle&#8217;s implementation
-SGD Optimizer is synchronized, which means all gradients will be wait to
-calculate and reduced into one gradient, then do optimize operation.</p>
-<p>The neural network consider the learning problem of minimizing an objective
-function, that has the form of a sum</p>
-<div class="math">
-\[Q(w) = \sum_{i}^{n} Q_i(w)\]</div>
-<p>The value of function Q sometimes is the cost of neural network (Mean
-Square Error between prediction and label for example). The function Q is
-parametrised by w, the weight/bias of neural network. And weights is what to
-be learned. The i is the i-th observation in (trainning) data.</p>
-<p>So, the SGD method will optimize the weight by</p>
-<div class="math">
-\[w = w - \eta \nabla Q(w) = w - \eta \sum_{i}^{n} \nabla Q_i(w)\]</div>
-<p>where <span class="math">\(\eta\)</span> is learning rate. And <span class="math">\(n\)</span> is batch size.</p>
-</dd></dl>
-
-</div>
-<div class="section" id="momentumoptimizer">
-<h2>MomentumOptimizer<a class="headerlink" href="#momentumoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">MomentumOptimizer</code><span class="sig-paren">(</span><em>momentum=None</em>, <em>sparse=False</em><span class="sig-paren">)</span></dt>
-<dd><p>MomentumOptimizer.</p>
-<p>When sparse=True, the update scheme:</p>
-<div class="math">
-\[\begin{split}\alpha_t &amp;= \alpha_{t-1} / k \\
-\beta_t &amp;= \beta_{t-1} / (1 + \lambda \gamma_t) \\
-u_t &amp;= u_{t-1} - \alpha_t \gamma_t g_t \\
-v_t &amp;= v_{t-1} + \tau_{t-1} \alpha_t \gamma_t g_t \\
-\tau_t &amp;= \tau_{t-1} + \beta_t / \alpha_t\end{split}\]</div>
-<p>where <span class="math">\(k\)</span> is momentum, <span class="math">\(\lambda\)</span> is decay rate,
-<span class="math">\(\gamma_t\)</span> is learning rate at the t&#8217;th step.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>sparse</strong> (<em>bool</em>) &#8211; with sparse support or not.</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adamoptimizer">
-<h2>AdamOptimizer<a class="headerlink" href="#adamoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdamOptimizer</code><span class="sig-paren">(</span><em>beta1=0.9</em>, <em>beta2=0.999</em>, <em>epsilon=1e-08</em><span class="sig-paren">)</span></dt>
-<dd><p>Adam optimizer.
-The details of please refer <a class="reference external" href="https://arxiv.org/abs/1412.6980">Adam: A Method for Stochastic Optimization</a></p>
-<div class="math">
-\[\begin{split}m(w, t) &amp; = \beta_1 m(w, t-1) + (1 - \beta_1) \nabla Q_i(w) \\
-v(w, t) &amp; = \beta_2 v(w, t-1) + (1 - \beta_2)(\nabla Q_i(w)) ^2 \\
-w &amp; = w - \frac{\eta}{\sqrt{v(w,t) + \epsilon}}\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>beta1</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_1\)</span> in equation.</li>
-<li><strong>beta2</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_2\)</span> in equation.</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; the <span class="math">\(\epsilon\)</span> in equation. It is used to prevent
-divided by zero.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adamaxoptimizer">
-<h2>AdamaxOptimizer<a class="headerlink" href="#adamaxoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdamaxOptimizer</code><span class="sig-paren">(</span><em>beta1</em>, <em>beta2</em><span class="sig-paren">)</span></dt>
-<dd><p>Adamax optimizer.</p>
-<p>The details of please refer this <a class="reference external" href="https://arxiv.org/abs/1412.6980">Adam: A Method for Stochastic Optimization</a></p>
-<div class="math">
-\[\begin{split}m_t &amp; = \beta_1 * m_{t-1} + (1-\beta_1)* \nabla Q_i(w) \\
-u_t &amp; = max(\beta_2*u_{t-1}, abs(\nabla Q_i(w))) \\
-w_t &amp; = w_{t-1} - (\eta/(1-\beta_1^t))*m_t/u_t\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>beta1</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_1\)</span> in the equation.</li>
-<li><strong>beta2</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_2\)</span> in the equation.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adagradoptimizer">
-<h2>AdaGradOptimizer<a class="headerlink" href="#adagradoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdaGradOptimizer</code></dt>
-<dd><p>Adagrad(for ADAptive GRAdient algorithm) optimizer.</p>
-<p>For details please refer this <a class="reference external" href="http://www.magicbroom.info/Papers/DuchiHaSi10.pdf">Adaptive Subgradient Methods for
-Online Learning and Stochastic Optimization</a>.</p>
-<div class="math">
-\[\begin{split}G &amp;= \sum_{\tau=1}^{t} g_{\tau} g_{\tau}^T \\
-w &amp; = w - \eta diag(G)^{-\frac{1}{2}} \circ g\end{split}\]</div>
-</dd></dl>
-
-</div>
-<div class="section" id="decayedadagradoptimizer">
-<h2>DecayedAdaGradOptimizer<a class="headerlink" href="#decayedadagradoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">DecayedAdaGradOptimizer</code><span class="sig-paren">(</span><em>rho=0.95</em>, <em>epsilon=1e-06</em><span class="sig-paren">)</span></dt>
-<dd><p>AdaGrad method with decayed sum gradients. The equations of this method
-show as follow.</p>
-<div class="math">
-\[\begin{split}E(g_t^2) &amp;= \rho * E(g_{t-1}^2) + (1-\rho) * g^2 \\
-learning\_rate &amp;= 1/sqrt( ( E(g_t^2) + \epsilon )\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>rho</strong> (<em>float</em>) &#8211; The <span class="math">\(\rho\)</span> parameter in that equation</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; The <span class="math">\(\epsilon\)</span> parameter in that equation.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adadeltaoptimizer">
-<h2>AdaDeltaOptimizer<a class="headerlink" href="#adadeltaoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdaDeltaOptimizer</code><span class="sig-paren">(</span><em>rho=0.95</em>, <em>epsilon=1e-06</em><span class="sig-paren">)</span></dt>
-<dd><p>AdaDelta method. The details of adadelta please refer to this
-<a class="reference external" href="http://www.matthewzeiler.com/pubs/googleTR2012/googleTR2012.pdf">ADADELTA: AN ADAPTIVE LEARNING RATE METHOD</a>.</p>
-<div class="math">
-\[\begin{split}E(g_t^2) &amp;= \rho * E(g_{t-1}^2) + (1-\rho) * g^2 \\
-learning\_rate &amp;= sqrt( ( E(dx_{t-1}^2) + \epsilon ) / ( \
-                  E(g_t^2) + \epsilon ) ) \\
-E(dx_t^2) &amp;= \rho * E(dx_{t-1}^2) + (1-\rho) * (-g*learning\_rate)^2\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>rho</strong> (<em>float</em>) &#8211; <span class="math">\(\rho\)</span> in equation</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; <span class="math">\(\rho\)</span> in equation</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="rmspropoptimizer">
-<h2>RMSPropOptimizer<a class="headerlink" href="#rmspropoptimizer" title="Permalink to this headline">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">RMSPropOptimizer</code><span class="sig-paren">(</span><em>rho=0.95</em>, <em>epsilon=1e-06</em><span class="sig-paren">)</span></dt>
-<dd><p>RMSProp(for Root Mean Square Propagation) optimizer. For details please
-refer this <a class="reference external" href="http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf">slide</a>.</p>
-<p>The equations of this method as follows:</p>
-<div class="math">
-\[\begin{split}v(w, t) &amp; = \rho v(w, t-1) + (1 - \rho)(\nabla Q_{i}(w))^2 \\
-w &amp; = w - \frac{\eta} {\sqrt{v(w,t) + \epsilon}} \nabla Q_{i}(w)\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>rho</strong> (<em>float</em>) &#8211; the <span class="math">\(\rho\)</span> in the equation. The forgetting factor.</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; the <span class="math">\(\epsilon\)</span> in the equation.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="settings">
-<span id="api-trainer-config-helpers-optimizers-settings"></span><h2>settings<a class="headerlink" href="#settings" title="Permalink to this headline">¶</a></h2>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">settings</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Set the optimization method, learning rate, batch size, and other training
-settings. The currently supported algorithms are SGD and Async-SGD.</p>
-<div class="admonition warning">
-<p class="first admonition-title">Warning</p>
-<p class="last">Note that the &#8216;batch_size&#8217; in PaddlePaddle is not equal to global
-training batch size. It represents the single training process&#8217;s batch
-size. If you use N processes to train one model, for example use three
-GPU machines, the global batch size is N*&#8217;batch_size&#8217;.</p>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>batch_size</strong> (<em>int</em>) &#8211; batch size for one training process.</li>
-<li><strong>learning_rate</strong> (<em>float</em>) &#8211; learning rate for SGD</li>
-<li><strong>learning_method</strong> (<em>BaseSGDOptimizer</em>) &#8211; The extension optimization algorithms of gradient
-descent, such as momentum, adagrad, rmsprop, etc.
-Note that it should be instance with base type
-BaseSGDOptimizer.</li>
-<li><strong>regularization</strong> (<em>BaseRegularization</em>) &#8211; The regularization method.</li>
-<li><strong>is_async</strong> (<em>bool</em>) &#8211; Is Async-SGD or not. Default value is False.</li>
-<li><strong>model_average</strong> (<em>ModelAverage</em>) &#8211; Model Average Settings.</li>
-<li><strong>gradient_clipping_threshold</strong> (<em>float</em>) &#8211; gradient clipping threshold. If gradient
-value larger than some value, will be
-clipped.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/api/v1/trainer_config_helpers/poolings.html b/develop/doc/api/v1/trainer_config_helpers/poolings.html
deleted file mode 100644
index 7fcbc1dd72a5c9f6b71dea720a3e0c8515e9a96a..0000000000000000000000000000000000000000
--- a/develop/doc/api/v1/trainer_config_helpers/poolings.html
+++ /dev/null
@@ -1,283 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Poolings &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="../../../genindex.html"/>
-        <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Poolings</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="poolings">
-<h1>Poolings<a class="headerlink" href="#poolings" title="Permalink to this headline">¶</a></h1>
-<div class="section" id="basepoolingtype">
-<h2>BasePoolingType<a class="headerlink" href="#basepoolingtype" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">BasePoolingType</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">BasePool</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="avgpooling">
-<h2>AvgPooling<a class="headerlink" href="#avgpooling" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">AvgPooling</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Avg</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="maxpooling">
-<h2>MaxPooling<a class="headerlink" href="#maxpooling" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">MaxPooling</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Max</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="sumpooling">
-<h2>SumPooling<a class="headerlink" href="#sumpooling" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">SumPooling</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">Sum</span></code></p>
-</dd></dl>
-
-</div>
-<div class="section" id="squarerootnpooling">
-<h2>SquareRootNPooling<a class="headerlink" href="#squarerootnpooling" title="Permalink to this headline">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">SquareRootNPooling</code></dt>
-<dd><p>alias of <code class="xref py py-class docutils literal"><span class="pre">SquareRootN</span></code></p>
-</dd></dl>
-
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/api/v2/config/activation.html b/develop/doc/api/v2/config/activation.html
index 570a39e2f32b7450d2b834ff64fe460d7911ef39..53eae0a0d1d71675b6240ba9f15b39dc19083678 100644
--- a/develop/doc/api/v2/config/activation.html
+++ b/develop/doc/api/v2/config/activation.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">Model Configuration</a><ul class="current">
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
diff --git a/develop/doc/api/v2/config/attr.html b/develop/doc/api/v2/config/attr.html
index ff2b03bb8233f443b0b29ba2c919bf4272bd3bb7..a15623989ec5c8e1e218a6db76b9db9b7e3df645 100644
--- a/develop/doc/api/v2/config/attr.html
+++ b/develop/doc/api/v2/config/attr.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">Model Configuration</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
diff --git a/develop/doc/api/v1/trainer_config_helpers/evaluators.html b/develop/doc/api/v2/config/evaluators.html
similarity index 68%
rename from develop/doc/api/v1/trainer_config_helpers/evaluators.html
rename to develop/doc/api/v2/config/evaluators.html
index 2411ba063cc5b07271d2a3375a9224deef322e8c..7e084cd1d29b9b36101513078af3360b10a216f0 100644
--- a/develop/doc/api/v1/trainer_config_helpers/evaluators.html
+++ b/develop/doc/api/v2/config/evaluators.html
@@ -32,7 +32,10 @@
         <link rel="index" title="Index"
               href="../../../genindex.html"/>
         <link rel="search" title="Search" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/> 
+    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/>
+        <link rel="up" title="Model Configuration" href="../model_configs.html"/>
+        <link rel="next" title="Optimizer" href="optimizer.html"/>
+        <link rel="prev" title="Layers" href="layer.html"/> 
 
   <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
   <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
@@ -80,10 +83,10 @@
       </div>
       <div class="doc-module">
         
-        <ul>
+        <ul class="current">
 <li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
 <li class="toctree-l1"><a class="reference internal" href="../../../howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a></li>
+<li class="toctree-l1 current"><a class="reference internal" href="../../index_en.html">API</a></li>
 <li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
 </ul>
 
@@ -105,7 +108,7 @@
     <nav class="doc-menu-vertical" role="navigation">
         
           
-          <ul>
+          <ul class="current">
 <li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a><ul>
 <li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
@@ -131,18 +134,19 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
 </ul>
 </li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
+<li class="toctree-l1 current"><a class="reference internal" href="../../index_en.html">API</a><ul class="current">
+<li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">Model Configuration</a><ul class="current">
+<li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
+<li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3 current"><a class="current reference internal" href="#">Evaluators</a></li>
+<li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
+<li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
+<li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
+<li class="toctree-l3"><a class="reference internal" href="attr.html">Parameter Attribute</a></li>
 </ul>
 </li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">Training and Inference</a></li>
+<li class="toctree-l2"><a class="reference internal" href="../data.html">Data Reader Interface and DataSets</a></li>
+<li class="toctree-l2"><a class="reference internal" href="../run_logic.html">Training and Inference</a></li>
 </ul>
 </li>
 <li class="toctree-l1"><a class="reference internal" href="../../../about/index_en.html">ABOUT</a></li>
@@ -166,6 +170,10 @@
 <div role="navigation" aria-label="breadcrumbs navigation">
   <ul class="wy-breadcrumbs">
       
+        <li><a href="../../index_en.html">API</a> > </li>
+      
+        <li><a href="../model_configs.html">Model Configuration</a> > </li>
+      
     <li>Evaluators</li>
   </ul>
 </div>
@@ -176,54 +184,14 @@
            <div itemprop="articleBody">
             
   <div class="section" id="evaluators">
-<span id="api-trainer-config-helpers-evaluators"></span><h1>Evaluators<a class="headerlink" href="#evaluators" title="Permalink to this headline">¶</a></h1>
-<div class="section" id="base">
-<h2>Base<a class="headerlink" href="#base" title="Permalink to this headline">¶</a></h2>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">evaluator_base</code><span class="sig-paren">(</span><em>input</em>, <em>type</em>, <em>label=None</em>, <em>weight=None</em>, <em>name=None</em>, <em>chunk_scheme=None</em>, <em>num_chunk_types=None</em>, <em>classification_threshold=None</em>, <em>positive_label=None</em>, <em>dict_file=None</em>, <em>result_file=None</em>, <em>num_results=None</em>, <em>delimited=None</em>, <em>top_k=None</em>, <em>excluded_chunk_types=None</em><span class="sig-paren">)</span></dt>
-<dd><p>Evaluator will evaluate the network status while training/testing.</p>
-<p>User can use evaluator by classify/regression job. For example.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">classify</span><span class="p">(</span><span class="n">prediction</span><span class="p">,</span> <span class="n">output</span><span class="p">,</span> <span class="n">evaluator</span><span class="o">=</span><span class="n">classification_error_evaluator</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>And user could define evaluator separately as follow.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">classification_error_evaluator</span><span class="p">(</span><span class="s2">&quot;ErrorRate&quot;</span><span class="p">,</span> <span class="n">prediction</span><span class="p">,</span> <span class="n">label</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>The evaluator often contains a name parameter. It will also be printed when
-evaluating network. The printed information may look like the following.</p>
-<div class="highlight-text"><div class="highlight"><pre><span></span>Batch=200 samples=20000 AvgCost=0.679655 CurrentCost=0.662179 Eval:
-classification_error_evaluator=0.4486
-CurrentEval: ErrorRate=0.3964
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>list|LayerOutput</em>) &#8211; Input layers, a object of LayerOutput or a list of
-LayerOutput.</li>
-<li><strong>label</strong> (<em>LayerOutput|None</em>) &#8211; An input layer containing the ground truth label.</li>
-<li><strong>weight</strong> (<em>LayerOutput.</em>) &#8211; An input layer which is a weight for each sample.
-Each evaluator may calculate differently to use this weight.</li>
-<li><strong>top_k</strong> (<em>int</em>) &#8211; number k in top-k error rate</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
+<span id="api-v2"></span><h1>Evaluators<a class="headerlink" href="#evaluators" title="Permalink to this headline">¶</a></h1>
 <div class="section" id="classification">
 <h2>Classification<a class="headerlink" href="#classification" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="classification-error-evaluator">
-<h3>classification_error_evaluator<a class="headerlink" href="#classification-error-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="classification-error">
+<h3>classification_error<a class="headerlink" href="#classification-error" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">classification_error_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">classification_error</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Classification Error Evaluator. It will print error rate for classification.</p>
 <p>The classification error is:</p>
 <div class="math">
@@ -238,9 +206,9 @@ Each evaluator may calculate differently to use this weight.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
 <li><strong>name</strong> (<em>basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
 <li><strong>label</strong> (<em>basestring</em>) &#8211; Label layer name.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. And will just multiply to NumOfWrongPredicts
 and NumOfAllSamples. So, the elements of weight are all one,
 then means not set weight. The larger weight it is, the more
@@ -258,11 +226,11 @@ important this sample is.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="auc-evaluator">
-<h3>auc_evaluator<a class="headerlink" href="#auc-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="auc">
+<h3>auc<a class="headerlink" href="#auc" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">auc_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">auc</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Auc Evaluator which adapts to binary classification.</p>
 <p>The simple usage:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">auc_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">label</span><span class="p">)</span>
@@ -274,9 +242,9 @@ important this sample is.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
 <li><strong>label</strong> (<em>None|basestring</em>) &#8211; Label layer name.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1].</li>
 </ul>
 </td>
@@ -286,11 +254,11 @@ important this sample is.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="ctc-error-evaluator">
-<h3>ctc_error_evaluator<a class="headerlink" href="#ctc-error-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="ctc-error">
+<h3>ctc_error<a class="headerlink" href="#ctc-error" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">ctc_error_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">ctc_error</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This evaluator is to calculate sequence-to-sequence edit distance.</p>
 <p>The simple usage is :</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">ctc_error_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="n">lbl</span><span class="p">)</span>
@@ -302,9 +270,9 @@ important this sample is.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer. Should be the same as the input for ctc_layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; input label, which is a data_layer. Should be the same as the
-label for ctc_layer</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer. Should be the same as the input for ctc.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; input label, which is a data. Should be the same as the
+label for ctc</li>
 </ul>
 </td>
 </tr>
@@ -313,11 +281,11 @@ label for ctc_layer</li>
 </dd></dl>
 
 </div>
-<div class="section" id="chunk-evaluator">
-<h3>chunk_evaluator<a class="headerlink" href="#chunk-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="chunk">
+<h3>chunk<a class="headerlink" href="#chunk" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">chunk_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">chunk</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Chunk evaluator is used to evaluate segment labelling accuracy for a
 sequence. It calculates the chunk detection F1 score.</p>
 <p>A chunk is correctly detected if its beginning, end and type are correct.
@@ -348,8 +316,8 @@ The tag type for each of the scheme is shown as follows:</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layers.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; An input layer containing the ground truth label.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; The input layers.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; An input layer containing the ground truth label.</li>
 <li><strong>chunk_scheme</strong> (<em>basestring</em>) &#8211; The labelling schemes support 4 types. It is one of
 &#8220;IOB&#8221;, &#8220;IOE&#8221;, &#8220;IOBES&#8221;, &#8220;plain&#8221;. It is required.</li>
 <li><strong>num_chunk_types</strong> &#8211; number of chunk types other than &#8220;other&#8221;</li>
@@ -363,11 +331,11 @@ The tag type for each of the scheme is shown as follows:</p>
 </dd></dl>
 
 </div>
-<div class="section" id="precision-recall-evaluator">
-<h3>precision_recall_evaluator<a class="headerlink" href="#precision-recall-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="precision-recall">
+<h3>precision_recall<a class="headerlink" href="#precision-recall" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">precision_recall_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">precision_recall</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>An Evaluator to calculate precision and recall, F1-score.
 It is adapt to the task with multiple labels.</p>
 <ul class="simple">
@@ -386,10 +354,10 @@ F1-score of this label.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label layer name.</li>
-<li><strong>positive_label</strong> (<em>LayerOutput.</em>) &#8211; The input label layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Label layer name.</li>
+<li><strong>positive_label</strong> (<em>paddle.v2.config_base.Layer.</em>) &#8211; The input label layer.</li>
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. (TODO, explaination)</li>
 </ul>
 </td>
@@ -402,11 +370,11 @@ F1-score of this label.</li>
 </div>
 <div class="section" id="rank">
 <h2>Rank<a class="headerlink" href="#rank" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="pnpair-evaluator">
-<h3>pnpair_evaluator<a class="headerlink" href="#pnpair-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="pnpair">
+<h3>pnpair<a class="headerlink" href="#pnpair" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">pnpair_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">pnpair</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Positive-negative pair rate Evaluator which adapts to rank task like
 learning to rank. This evaluator must contain at least three layers.</p>
 <p>The simple usage:</p>
@@ -419,10 +387,10 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label layer name.</li>
-<li><strong>info</strong> (<em>LayerOutput</em>) &#8211; Label layer name. (TODO, explaination)</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Label layer name.</li>
+<li><strong>info</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Label layer name. (TODO, explaination)</li>
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. (TODO, explaination)</li>
 </ul>
 </td>
@@ -435,11 +403,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </div>
 <div class="section" id="utils">
 <h2>Utils<a class="headerlink" href="#utils" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="sum-evaluator">
-<h3>sum_evaluator<a class="headerlink" href="#sum-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="sum">
+<h3>sum<a class="headerlink" href="#sum" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">sum_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">sum</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>An Evaluator to sum the result of input.</p>
 <p>The simple usage:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">sum_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
@@ -451,8 +419,8 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name.</li>
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. (TODO, explaination)</li>
 </ul>
 </td>
@@ -462,11 +430,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="column-sum-evaluator">
-<h3>column_sum_evaluator<a class="headerlink" href="#column-sum-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="column-sum">
+<h3>column_sum<a class="headerlink" href="#column-sum" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">column_sum_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">column_sum</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to sum the last column of input.</p>
 <p>The simple usage is:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">column_sum_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">label</span><span class="p">)</span>
@@ -478,7 +446,7 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name.</li>
 </ul>
 </td>
 </tr>
@@ -490,11 +458,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </div>
 <div class="section" id="print">
 <h2>Print<a class="headerlink" href="#print" title="Permalink to this headline">¶</a></h2>
-<div class="section" id="classification-error-printer-evaluator">
-<h3>classification_error_printer_evaluator<a class="headerlink" href="#classification-error-printer-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="classification-error-printer">
+<h3>classification_error_printer<a class="headerlink" href="#classification-error-printer" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">classification_error_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">classification_error_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the classification error of each sample.</p>
 <p>The simple usage is:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">classification_error_printer_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
@@ -505,8 +473,8 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Input label layer.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input layer.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input label layer.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -516,11 +484,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="gradient-printer-evaluator">
-<h3>gradient_printer_evaluator<a class="headerlink" href="#gradient-printer-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="gradient-printer">
+<h3>gradient_printer<a class="headerlink" href="#gradient-printer" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">gradient_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">gradient_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the gradient of input layers. It contains
 one or more input layers.</p>
 <p>The simple usage is:</p>
@@ -532,7 +500,7 @@ one or more input layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; One or more input layers.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; One or more input layers.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -542,11 +510,11 @@ one or more input layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="maxid-printer-evaluator">
-<h3>maxid_printer_evaluator<a class="headerlink" href="#maxid-printer-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="maxid-printer">
+<h3>maxid_printer<a class="headerlink" href="#maxid-printer" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">maxid_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">maxid_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print maximum top k values and their indexes
 of each row of input layers. It contains one or more input layers.
 k is specified by num_results.</p>
@@ -559,7 +527,7 @@ k is specified by num_results.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; Input Layer name.</li>
 <li><strong>num_results</strong> (<em>int.</em>) &#8211; This number is used to specify the top k numbers.
 It is 1 by default.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
@@ -571,11 +539,11 @@ It is 1 by default.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="maxframe-printer-evaluator">
-<h3>maxframe_printer_evaluator<a class="headerlink" href="#maxframe-printer-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="maxframe-printer">
+<h3>maxframe_printer<a class="headerlink" href="#maxframe-printer" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">maxframe_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">maxframe_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the top k frames of each input layers.
 The input layers should contain sequences info or sequences type.
 k is specified by num_results.
@@ -593,7 +561,7 @@ It contains one or more input layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; Input Layer name.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -603,11 +571,11 @@ It contains one or more input layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="seqtext-printer-evaluator">
-<h3>seqtext_printer_evaluator<a class="headerlink" href="#seqtext-printer-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="seqtext-printer">
+<h3>seqtext_printer<a class="headerlink" href="#seqtext-printer" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">seqtext_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">seqtext_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Sequence text printer will print text according to index matrix and a
 dictionary. There can be multiple input to this layer:</p>
 <p>1. If there is no id_input, the input must be a matrix containing
@@ -639,7 +607,7 @@ the sequence of indices;</p>
 <p>Typically SequenceTextPrinter layer takes output of maxid or RecurrentGroup
 with maxid (when generating) as an input.</p>
 <p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">seqtext_printer_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">maxid_layer</span><span class="p">,</span>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">seqtext_printer_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">maxid</span><span class="p">,</span>
                                  <span class="n">id_input</span><span class="o">=</span><span class="n">sample_id</span><span class="p">,</span>
                                  <span class="n">dict_file</span><span class="o">=</span><span class="n">dict_file</span><span class="p">,</span>
                                  <span class="n">result_file</span><span class="o">=</span><span class="n">result_file</span><span class="p">)</span>
@@ -650,9 +618,9 @@ with maxid (when generating) as an input.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; Input Layer name.</li>
 <li><strong>result_file</strong> (<em>basestring</em>) &#8211; Path of the file to store the generated results.</li>
-<li><strong>id_input</strong> (<em>LayerOutput</em>) &#8211; Index of the input sequence, and the specified index will
+<li><strong>id_input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Index of the input sequence, and the specified index will
 be prited in the gereated results. This an optional
 parameter.</li>
 <li><strong>dict_file</strong> (<em>basestring</em>) &#8211; Path of dictionary. This is an optional parameter.
@@ -677,11 +645,11 @@ Default is True. No space is added if set to False.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="value-printer-evaluator">
-<h3>value_printer_evaluator<a class="headerlink" href="#value-printer-evaluator" title="Permalink to this headline">¶</a></h3>
-<dl class="function">
+<div class="section" id="value-printer">
+<h3>value_printer<a class="headerlink" href="#value-printer" title="Permalink to this headline">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">value_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">value_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the values of input layers. It contains
 one or more input layers.</p>
 <p>The simple usage is:</p>
@@ -693,7 +661,7 @@ one or more input layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; One or more input layers.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; One or more input layers.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -711,6 +679,15 @@ one or more input layers.</p>
           </div>
           <footer>
   
+    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
+      
+        <a href="optimizer.html" class="btn btn-neutral float-right" title="Optimizer" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
+      
+      
+        <a href="layer.html" class="btn btn-neutral" title="Layers" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
+      
+    </div>
+  
 
   <hr/>
 
diff --git a/develop/doc/api/v2/config/layer.html b/develop/doc/api/v2/config/layer.html
index 0fb2ef5fe3010fbed0af43362131921f77a102fb..55bc8c45ce9cf5683347be911006de5bd8064c9f 100644
--- a/develop/doc/api/v2/config/layer.html
+++ b/develop/doc/api/v2/config/layer.html
@@ -34,7 +34,7 @@
         <link rel="search" title="Search" href="../../../search.html"/>
     <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/>
         <link rel="up" title="Model Configuration" href="../model_configs.html"/>
-        <link rel="next" title="Optimizer" href="optimizer.html"/>
+        <link rel="next" title="Evaluators" href="evaluators.html"/>
         <link rel="prev" title="Activation" href="activation.html"/> 
 
   <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
@@ -138,6 +138,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">Model Configuration</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
@@ -1453,10 +1454,10 @@ Each inputs is a projection or operator.</p>
 <li><strong>input</strong> &#8211; inputs layer. It is an optional parameter. If set,
 then this function will just return layer&#8217;s name.</li>
 <li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="../../v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
+<li><strong>bias_attr</strong> (<em>ParameterAttribute</em><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
 something not type of ParameterAttribute. None will get a
 default Bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="../../v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; The extra layer config. Default is None.</li>
+<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute</em>) &#8211; The extra layer config. Default is None.</li>
 </ul>
 </td>
 </tr>
@@ -3470,7 +3471,7 @@ It is used by recurrent layer group.</p>
   
     <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
       
-        <a href="optimizer.html" class="btn btn-neutral float-right" title="Optimizer" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
+        <a href="evaluators.html" class="btn btn-neutral float-right" title="Evaluators" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
       
       
         <a href="activation.html" class="btn btn-neutral" title="Activation" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
diff --git a/develop/doc/api/v2/config/networks.html b/develop/doc/api/v2/config/networks.html
index 3965aa50da5fb9f48bf8b099de329047b99f0a14..69e2b9c5c491c8fe8a40395fd609829e68e33d2e 100644
--- a/develop/doc/api/v2/config/networks.html
+++ b/develop/doc/api/v2/config/networks.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">Model Configuration</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Networks</a></li>
@@ -409,6 +410,9 @@ False if no bias.</li>
 </table>
 </dd></dl>
 
+</div>
+<div class="section" id="small-vgg">
+<h3>small_vgg<a class="headerlink" href="#small-vgg" title="Permalink to this headline">¶</a></h3>
 </div>
 <div class="section" id="vgg-16-network">
 <h3>vgg_16_network<a class="headerlink" href="#vgg-16-network" title="Permalink to this headline">¶</a></h3>
@@ -791,6 +795,86 @@ gru_group, and gru_group is relatively better than simple_gru.</p>
 </table>
 </dd></dl>
 
+</div>
+<div class="section" id="simple-gru2">
+<h4>simple_gru2<a class="headerlink" href="#simple-gru2" title="Permalink to this headline">¶</a></h4>
+<dl class="class">
+<dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.networks.</code><code class="descname">simple_gru2</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<dd><p>simple_gru2 is the same with simple_gru, but using grumemory instead
+Please see grumemory in layers.py for more detail about the maths.
+simple_gru2 is faster than simple_gru.</p>
+<p>The example usage is:</p>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">simple_gru2</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">)</span>
+</pre></div>
+</div>
+<table class="docutils field-list" frame="void" rules="none">
+<col class="field-name" />
+<col class="field-body" />
+<tbody valign="top">
+<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; input layer name.</li>
+<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
+<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
+<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
+<li><strong>act</strong> (<em>paddle.v2.Activation.Base</em>) &#8211; type of the activiation</li>
+<li><strong>gate_act</strong> (<em>paddle.v2.Activation.Base</em>) &#8211; type of the gate activiation</li>
+<li><strong>gru_bias_attr</strong> (<em>paddle.v2.attr.ParameterAttribute|False</em>) &#8211; bias. False means no bias, None means default bias.</li>
+<li><strong>gru_attr</strong> (<em>paddle.v2.attr.ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
+</ul>
+</td>
+</tr>
+<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">the gru group.</p>
+</td>
+</tr>
+<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">paddle.v2.config_base.Layer</p>
+</td>
+</tr>
+</tbody>
+</table>
+</dd></dl>
+
+</div>
+<div class="section" id="bidirectional-gru">
+<h4>bidirectional_gru<a class="headerlink" href="#bidirectional-gru" title="Permalink to this headline">¶</a></h4>
+<dl class="class">
+<dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.networks.</code><code class="descname">bidirectional_gru</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<dd><p>A bidirectional_gru is a recurrent unit that iterates over the input
+sequence both in forward and bardward orders, and then concatenate two
+outputs to form a final output. However, concatenation of two outputs
+is not the only way to form the final output, you can also, for example,
+just add them together.</p>
+<p>The example usage is:</p>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">bi_gru</span> <span class="o">=</span> <span class="n">bidirectional_gru</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">input1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">)</span>
+</pre></div>
+</div>
+<table class="docutils field-list" frame="void" rules="none">
+<col class="field-name" />
+<col class="field-body" />
+<tbody valign="top">
+<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first simple">
+<li><strong>name</strong> (<em>basestring</em>) &#8211; bidirectional gru layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; input layer.</li>
+<li><strong>size</strong> (<em>int</em>) &#8211; gru layer size.</li>
+<li><strong>return_seq</strong> (<em>bool</em>) &#8211; If set False, outputs of the last time step are
+concatenated and returned.
+If set True, the entire output sequences that are
+processed in forward and backward directions are
+concatenated and returned.</li>
+</ul>
+</td>
+</tr>
+<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">paddle.v2.config_base.Layer object.</p>
+</td>
+</tr>
+<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body"><p class="first last">paddle.v2.config_base.Layer</p>
+</td>
+</tr>
+</tbody>
+</table>
+</dd></dl>
+
 </div>
 </div>
 <div class="section" id="simple-attention">
diff --git a/develop/doc/api/v2/config/optimizer.html b/develop/doc/api/v2/config/optimizer.html
index 01071e7410085be37ca272ced1e34a0a9f78c986..440c1fc64b0fa2a736e37ff18c9a702f6d52ba25 100644
--- a/develop/doc/api/v2/config/optimizer.html
+++ b/develop/doc/api/v2/config/optimizer.html
@@ -35,7 +35,7 @@
     <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/>
         <link rel="up" title="Model Configuration" href="../model_configs.html"/>
         <link rel="next" title="Pooling" href="pooling.html"/>
-        <link rel="prev" title="Layers" href="layer.html"/> 
+        <link rel="prev" title="Evaluators" href="evaluators.html"/> 
 
   <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
   <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
@@ -138,6 +138,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">Model Configuration</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
@@ -367,7 +368,7 @@ w &amp; = w - \frac{\eta} {\sqrt{v(w,t) + \epsilon}} \nabla Q_{i}(w)\end{split}\
         <a href="pooling.html" class="btn btn-neutral float-right" title="Pooling" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
       
       
-        <a href="layer.html" class="btn btn-neutral" title="Layers" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
+        <a href="evaluators.html" class="btn btn-neutral" title="Evaluators" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
       
     </div>
   
diff --git a/develop/doc/api/v2/config/pooling.html b/develop/doc/api/v2/config/pooling.html
index 7871dfcc7d6bc1f90a5fde9b08f1866efbb45f3d..53fb82325c36ff8257a3ed9b998b741828532711 100644
--- a/develop/doc/api/v2/config/pooling.html
+++ b/develop/doc/api/v2/config/pooling.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">Model Configuration</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
diff --git a/develop/doc/api/v2/data.html b/develop/doc/api/v2/data.html
index 65d1a7c10c7dd976c78067decb82e58dc3ec2e47..d423c9493340a369d355ba8902f5fa4560213bcc 100644
--- a/develop/doc/api/v2/data.html
+++ b/develop/doc/api/v2/data.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/networks.html">Networks</a></li>
diff --git a/develop/doc/api/v2/model_configs.html b/develop/doc/api/v2/model_configs.html
index a6ea1f333f669e180101adbafad74aba7c58a76a..6e3d3285ca1de6d7912ad1a84212180792a4626e 100644
--- a/develop/doc/api/v2/model_configs.html
+++ b/develop/doc/api/v2/model_configs.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2 current"><a class="current reference internal" href="#">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/networks.html">Networks</a></li>
@@ -186,6 +187,7 @@
 <ul>
 <li class="toctree-l1"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l1"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/networks.html">Networks</a></li>
diff --git a/develop/doc/api/v2/run_logic.html b/develop/doc/api/v2/run_logic.html
index cd2522ec132765888bc3c5381487dbc4f813553f..c02345235c1aa43c05aca18c9e4c1a8b249169d3 100644
--- a/develop/doc/api/v2/run_logic.html
+++ b/develop/doc/api/v2/run_logic.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/networks.html">Networks</a></li>
@@ -494,8 +495,9 @@ index that reader returns.</li>
 <dd><p>Infer a neural network by given neural network output and parameters.  The
 user should pass either a batch of input data or reader method.</p>
 <p>Example usages:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">result</span> <span class="o">=</span> <span class="n">paddle</span><span class="o">.</span><span class="n">infer</span><span class="p">(</span><span class="n">prediction</span><span class="p">,</span> <span class="n">parameters</span><span class="p">,</span> <span class="nb">input</span><span class="o">=</span><span class="n">SomeData</span><span class="p">,</span>
-                      <span class="n">batch_size</span><span class="o">=</span><span class="mi">32</span><span class="p">)</span>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">result</span> <span class="o">=</span> <span class="n">paddle</span><span class="o">.</span><span class="n">infer</span><span class="p">(</span><span class="n">outptut_layer</span><span class="o">=</span><span class="n">prediction</span><span class="p">,</span>
+                      <span class="n">parameters</span><span class="o">=</span><span class="n">parameters</span><span class="p">,</span>
+                      <span class="nb">input</span><span class="o">=</span><span class="n">SomeData</span><span class="p">)</span>
 <span class="k">print</span> <span class="n">result</span>
 </pre></div>
 </div>
diff --git a/develop/doc/design/api.html b/develop/doc/design/api.html
index 4c6aba58305d4a6c3ea4fab84374f0728a67283d..816a89ec12e97e9566b6976eba6afaf3eff8da1c 100644
--- a/develop/doc/design/api.html
+++ b/develop/doc/design/api.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/design/cluster_train/README.html b/develop/doc/design/cluster_train/README.html
index 9f91cdc3c2d0e8b54e74479a60e7436ad9ccad85..75790248d9949e679514c8171ace9f33da943b84 100644
--- a/develop/doc/design/cluster_train/README.html
+++ b/develop/doc/design/cluster_train/README.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/design/cluster_train/checkpointing.html b/develop/doc/design/cluster_train/checkpointing.html
index 6b321d31e77c2621ebaaffd6fc970d5d88213529..1e4ee832338c81861e67ad13c21998ba01e31468 100644
--- a/develop/doc/design/cluster_train/checkpointing.html
+++ b/develop/doc/design/cluster_train/checkpointing.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/design/cluster_train/data_dispatch.html b/develop/doc/design/cluster_train/data_dispatch.html
index b5e248794398b8f24aac2d3c8f48ac30bb8f4b79..882c009b3f019922d6dc2736ca8ce64377c7952e 100644
--- a/develop/doc/design/cluster_train/data_dispatch.html
+++ b/develop/doc/design/cluster_train/data_dispatch.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/design/multi_language_interface/00.why_plain_c.html b/develop/doc/design/multi_language_interface/00.why_plain_c.html
index cf45c569b1c10d958ae549c145b3e5b5983628a8..ce47a58d7da33dae4a8a0a63ba35ec89951d4c7c 100644
--- a/develop/doc/design/multi_language_interface/00.why_plain_c.html
+++ b/develop/doc/design/multi_language_interface/00.why_plain_c.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/design/multi_language_interface/01.inference_implementation.html b/develop/doc/design/multi_language_interface/01.inference_implementation.html
index 43447a3730608e913dcc9397ce741a08902561ad..8fd2672e3ec2561bfc660c651edb44c6326dd18c 100644
--- a/develop/doc/design/multi_language_interface/01.inference_implementation.html
+++ b/develop/doc/design/multi_language_interface/01.inference_implementation.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/design/reader/README.html b/develop/doc/design/reader/README.html
index 4de366bd53ddbb2d646443eb4747b219070d324e..285c19a8e4329266c39c483ee5a15a111a6c8fbc 100644
--- a/develop/doc/design/reader/README.html
+++ b/develop/doc/design/reader/README.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/design/releasing_process.html b/develop/doc/design/releasing_process.html
index 902df2cb15109c725759e02d7b19ad2ea614cb5a..2844f232688baf70ed50c41827d8181b1c7474e8 100644
--- a/develop/doc/design/releasing_process.html
+++ b/develop/doc/design/releasing_process.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/genindex.html b/develop/doc/genindex.html
index ee5993c9cbdcd98fa7d49a811d8745561c0b3b13..02bf19465f8166b03173d87239f157b58746c76e 100644
--- a/develop/doc/genindex.html
+++ b/develop/doc/genindex.html
@@ -136,6 +136,7 @@
 <li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
@@ -180,58 +181,17 @@
 <h1 id="index">Index</h1>
 
 <div class="genindex-jumpbox">
- <a href="#D"><strong>D</strong></a>
- | <a href="#E"><strong>E</strong></a>
- | <a href="#P"><strong>P</strong></a>
- | <a href="#S"><strong>S</strong></a>
+ <a href="#P"><strong>P</strong></a>
  
 </div>
-<h2 id="D">D</h2>
-<table style="width: 100%" class="indextable genindextable"><tr>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/data_sources.html#paddle.trainer_config_helpers.data_sources.define_py_data_sources2">define_py_data_sources2() (in module paddle.trainer_config_helpers.data_sources)</a>
-</li>
-  </ul></td>
-</tr></table>
-
-<h2 id="E">E</h2>
-<table style="width: 100%" class="indextable genindextable"><tr>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ExtraAttr">ExtraAttr (in module paddle.trainer_config_helpers.attrs)</a>
-</li>
-  </ul></td>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute">ExtraLayerAttribute (class in paddle.trainer_config_helpers.attrs)</a>
-</li>
-  </ul></td>
-</tr></table>
-
 <h2 id="P">P</h2>
 <table style="width: 100%" class="indextable genindextable"><tr>
   <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#module-paddle.trainer_config_helpers.attrs">paddle.trainer_config_helpers.attrs (module)</a>
-</li>
-      <li><a href="api/v1/trainer_config_helpers/data_sources.html#module-paddle.trainer_config_helpers.data_sources">paddle.trainer_config_helpers.data_sources (module)</a>
-</li>
-  </ul></td>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParamAttr">ParamAttr (in module paddle.trainer_config_helpers.attrs)</a>
-</li>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute">ParameterAttribute (class in paddle.trainer_config_helpers.attrs)</a>
-</li>
       <li><a href="api/v1/data_provider/pydataprovider2_en.html#paddle.trainer.PyDataProvider2.provider">provider() (in module paddle.trainer.PyDataProvider2)</a>
 </li>
   </ul></td>
 </tr></table>
 
-<h2 id="S">S</h2>
-<table style="width: 100%" class="indextable genindextable"><tr>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute.set_default_parameter_name">set_default_parameter_name() (paddle.trainer_config_helpers.attrs.ParameterAttribute method)</a>
-</li>
-  </ul></td>
-</tr></table>
-
 
 
            </div>
diff --git a/develop/doc/getstarted/basic_usage/index_en.html b/develop/doc/getstarted/basic_usage/index_en.html
index 974ef3ce8a27aeac57aa745201e2b0750ad79cab..5dbd3ed684afe9dd5b653089a3c67587b8dca115 100644
--- a/develop/doc/getstarted/basic_usage/index_en.html
+++ b/develop/doc/getstarted/basic_usage/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/getstarted/build_and_install/build_from_source_en.html b/develop/doc/getstarted/build_and_install/build_from_source_en.html
index 7b8c78f7377a7cd2fe111ac352460349b250637b..798e9df10324c19df0619dea0544849223637c2c 100644
--- a/develop/doc/getstarted/build_and_install/build_from_source_en.html
+++ b/develop/doc/getstarted/build_and_install/build_from_source_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/getstarted/build_and_install/docker_install_en.html b/develop/doc/getstarted/build_and_install/docker_install_en.html
index 31fef38e3f5e7ad94607732ea6065caed3d507bf..c81084520a1690f1d7f611c6f376763c4b817f5b 100644
--- a/develop/doc/getstarted/build_and_install/docker_install_en.html
+++ b/develop/doc/getstarted/build_and_install/docker_install_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/getstarted/build_and_install/index_en.html b/develop/doc/getstarted/build_and_install/index_en.html
index 97f9451697b8edf5f48138f5138bb1896e611e93..e9b54a9520df927dc58a32560ed1d8e36a94f9d8 100644
--- a/develop/doc/getstarted/build_and_install/index_en.html
+++ b/develop/doc/getstarted/build_and_install/index_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/getstarted/build_and_install/ubuntu_install_en.html b/develop/doc/getstarted/build_and_install/ubuntu_install_en.html
index 6a641b85aeb07dbf4b1f993d0781b40465da09f1..691bad263475888e359648e1c991b1e3c0d0e09e 100644
--- a/develop/doc/getstarted/build_and_install/ubuntu_install_en.html
+++ b/develop/doc/getstarted/build_and_install/ubuntu_install_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/getstarted/index_en.html b/develop/doc/getstarted/index_en.html
index 95aabc4acd9a02149341e35f0a6fc0f84078eb12..c4a1be6c2b09c0daa6f1dd2bf14e0ae39e121aa4 100644
--- a/develop/doc/getstarted/index_en.html
+++ b/develop/doc/getstarted/index_en.html
@@ -137,6 +137,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/deep_model/rnn/index_en.html b/develop/doc/howto/deep_model/rnn/index_en.html
index 6c3b311ca5c6f0b37a907658ed2d30322c08fb06..51e523af5d6315922f3a061f0d75a10eb02a6dec 100644
--- a/develop/doc/howto/deep_model/rnn/index_en.html
+++ b/develop/doc/howto/deep_model/rnn/index_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/deep_model/rnn/rnn_config_en.html b/develop/doc/howto/deep_model/rnn/rnn_config_en.html
index fa9f70dff2622ad7278979e4178cb09ca2dd87d1..91b60fc4981f1d6d48e9c69b7d694e248bf998b0 100644
--- a/develop/doc/howto/deep_model/rnn/rnn_config_en.html
+++ b/develop/doc/howto/deep_model/rnn/rnn_config_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
@@ -247,7 +248,7 @@ Its <strong>output function</strong> simply takes <span class="math">\(x_t\)</sp
 <p>We will use the sequence to sequence model with attention as an example to demonstrate how you can configure complex recurrent neural network models. An illustration of the sequence to sequence model with attention is shown in the following figure.</p>
 <img alt="../../../_images/encoder-decoder-attention-model.png" class="align-center" src="../../../_images/encoder-decoder-attention-model.png" />
 <p>In this model, the source sequence <span class="math">\(S = \{s_1, \dots, s_T\}\)</span> is encoded with a bidirectional gated recurrent neural networks. The hidden states of the bidirectional gated recurrent neural network <span class="math">\(H_S = \{H_1, \dots, H_T\}\)</span> is called <em>encoder vector</em> The decoder is a gated recurrent neural network. When decoding each token <span class="math">\(y_t\)</span>, the gated recurrent neural network generates a set of weights <span class="math">\(W_S^t = \{W_1^t, \dots, W_T^t\}\)</span>, which are used to compute a weighted sum of the encoder vector. The weighted sum of the encoder vector is utilized to condition the generation of the token <span class="math">\(y_t\)</span>.</p>
-<p>The encoder part of the model is listed below. It calls <code class="code docutils literal"><span class="pre">grumemory</span></code> to represent gated recurrent neural network. It is the recommended way of using recurrent neural network if the network architecture is simple, because it is faster than <code class="code docutils literal"><span class="pre">recurrent_group</span></code>. We have implemented most of the commonly used recurrent neural network architectures, you can refer to <a class="reference internal" href="../../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers"><span class="std std-ref">Layers</span></a> for more details.</p>
+<p>The encoder part of the model is listed below. It calls <code class="code docutils literal"><span class="pre">grumemory</span></code> to represent gated recurrent neural network. It is the recommended way of using recurrent neural network if the network architecture is simple, because it is faster than <code class="code docutils literal"><span class="pre">recurrent_group</span></code>. We have implemented most of the commonly used recurrent neural network architectures, you can refer to <span class="xref std std-ref">api_trainer_config_helpers_layers</span> for more details.</p>
 <p>We also project the encoder vector to <code class="code docutils literal"><span class="pre">decoder_size</span></code> dimensional space, get the first instance of the backward recurrent network, and project it to <code class="code docutils literal"><span class="pre">decoder_size</span></code> dimensional space:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="c1"># Define the data layer of the source sentence.</span>
 <span class="n">src_word_id</span> <span class="o">=</span> <span class="n">data_layer</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;source_language_word&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="n">source_dict_dim</span><span class="p">)</span>
diff --git a/develop/doc/howto/dev/contribute_to_paddle_en.html b/develop/doc/howto/dev/contribute_to_paddle_en.html
index 4c2dcae5a29fb65c5c69f6f39cf722621a7ba23d..c26b184ef22f37b21cd8c82b04eb480cf67a7b17 100644
--- a/develop/doc/howto/dev/contribute_to_paddle_en.html
+++ b/develop/doc/howto/dev/contribute_to_paddle_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/dev/new_layer_en.html b/develop/doc/howto/dev/new_layer_en.html
index c56fe7bfdcb3ca74482e202b248c8164f0e14465..67a42328afc2b4130d52ddb4de1c2e46abc006a0 100644
--- a/develop/doc/howto/dev/new_layer_en.html
+++ b/develop/doc/howto/dev/new_layer_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/index_en.html b/develop/doc/howto/index_en.html
index 11f58305f734fa9cc30e1ce8585542af4298aae2..32f19524b1d0b0f07c38a33b9c3f8d126b887072 100644
--- a/develop/doc/howto/index_en.html
+++ b/develop/doc/howto/index_en.html
@@ -137,6 +137,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/optimization/gpu_profiling_en.html b/develop/doc/howto/optimization/gpu_profiling_en.html
index 913cfe90b030ed9ce49762650fc3ecf0c4349306..002fa6c751c0b8d8df64c1daeb06cbeb1e211f01 100644
--- a/develop/doc/howto/optimization/gpu_profiling_en.html
+++ b/develop/doc/howto/optimization/gpu_profiling_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/cluster/cluster_train_en.html b/develop/doc/howto/usage/cluster/cluster_train_en.html
index e4153c2e3c8cfc4639369703d11ced3da31717c2..1b13461b021a4a69863e26e1c28044ca66e0eea2 100644
--- a/develop/doc/howto/usage/cluster/cluster_train_en.html
+++ b/develop/doc/howto/usage/cluster/cluster_train_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/cmd_parameter/arguments_en.html b/develop/doc/howto/usage/cmd_parameter/arguments_en.html
index 5235f793b003493f668f59a8044e017f909f85e3..3cce1414b4808d5e9fb73555a339629b8dd8f675 100644
--- a/develop/doc/howto/usage/cmd_parameter/arguments_en.html
+++ b/develop/doc/howto/usage/cmd_parameter/arguments_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/cmd_parameter/detail_introduction_en.html b/develop/doc/howto/usage/cmd_parameter/detail_introduction_en.html
index 037ae4ce951bd9be494ec54b63c059a5ef4ac636..817ca01fe86a60103ab00398a5f1d4fa30e7499e 100644
--- a/develop/doc/howto/usage/cmd_parameter/detail_introduction_en.html
+++ b/develop/doc/howto/usage/cmd_parameter/detail_introduction_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/cmd_parameter/index_en.html b/develop/doc/howto/usage/cmd_parameter/index_en.html
index 1b2c692b8a296b8fd8df514994a16447cdd7d874..a453a06041bbd43316711624251ea90e5b3504e2 100644
--- a/develop/doc/howto/usage/cmd_parameter/index_en.html
+++ b/develop/doc/howto/usage/cmd_parameter/index_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/cmd_parameter/use_case_en.html b/develop/doc/howto/usage/cmd_parameter/use_case_en.html
index 490771bb9c2fc2b50e475e3f4a6d44b535e97c01..c21c12ee1a25815fc8e9434966db4b777082da62 100644
--- a/develop/doc/howto/usage/cmd_parameter/use_case_en.html
+++ b/develop/doc/howto/usage/cmd_parameter/use_case_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/k8s/k8s_aws_en.html b/develop/doc/howto/usage/k8s/k8s_aws_en.html
index 4ec021b26875891cb020059cdc54c5bee6621e37..cedcfdd5a8f0472bb98b1e0398db19477103895e 100644
--- a/develop/doc/howto/usage/k8s/k8s_aws_en.html
+++ b/develop/doc/howto/usage/k8s/k8s_aws_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/k8s/k8s_en.html b/develop/doc/howto/usage/k8s/k8s_en.html
index d23d440e9aca1646e0d7d5ac5603cef8e148419a..288570e6b902715f42c5de30c87d602f696c029f 100644
--- a/develop/doc/howto/usage/k8s/k8s_en.html
+++ b/develop/doc/howto/usage/k8s/k8s_en.html
@@ -138,6 +138,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/k8s/src/k8s_data/README.html b/develop/doc/howto/usage/k8s/src/k8s_data/README.html
index 48f563142949bb8e773f50cdd3a546fb5f9a3181..83751775c9bd43774356946a6e01e3f63c484c4d 100644
--- a/develop/doc/howto/usage/k8s/src/k8s_data/README.html
+++ b/develop/doc/howto/usage/k8s/src/k8s_data/README.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/howto/usage/k8s/src/k8s_train/README.html b/develop/doc/howto/usage/k8s/src/k8s_train/README.html
index 9725be05449bbd86927db0d08ecd2440a9e82366..161180d2000dd94090fe82703610e55a3ee018f2 100644
--- a/develop/doc/howto/usage/k8s/src/k8s_train/README.html
+++ b/develop/doc/howto/usage/k8s/src/k8s_train/README.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/index_en.html b/develop/doc/index_en.html
index 4c6d8694402296e9dad2fb3c72000c4621df48ca..ebbb30c36ef2c2daa19f40177dca39e24916290e 100644
--- a/develop/doc/index_en.html
+++ b/develop/doc/index_en.html
@@ -136,6 +136,7 @@
 <li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/objects.inv b/develop/doc/objects.inv
index 2bb1f11ae905ce22445b262d12a30b8efa170648..b7e4ab120149808608be42db016131a123d4b400 100644
Binary files a/develop/doc/objects.inv and b/develop/doc/objects.inv differ
diff --git a/develop/doc/py-modindex.html b/develop/doc/py-modindex.html
deleted file mode 100644
index 40422411bb9b6fc275484a9459da3716312d541e..0000000000000000000000000000000000000000
--- a/develop/doc/py-modindex.html
+++ /dev/null
@@ -1,266 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Python Module Index &mdash; PaddlePaddle  documentation</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="Index"
-              href="genindex.html"/>
-        <link rel="search" title="Search" href="search.html"/>
-    <link rel="top" title="PaddlePaddle  documentation" href="index.html"/>
- 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-
-
-  
-  <script src="_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="getstarted/index_en.html">GET STARTED</a></li>
-<li class="toctree-l1"><a class="reference internal" href="howto/index_en.html">HOW TO</a></li>
-<li class="toctree-l1"><a class="reference internal" href="api/index_en.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="getstarted/index_en.html">GET STARTED</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="getstarted/build_and_install/index_en.html">Install and Build</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="getstarted/build_and_install/docker_install_en.html">PaddlePaddle in Docker Containers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="getstarted/build_and_install/ubuntu_install_en.html">Debian Package installation guide</a></li>
-<li class="toctree-l3"><a class="reference internal" href="getstarted/build_and_install/build_from_source_en.html">Installing from Sources</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="howto/index_en.html">HOW TO</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="howto/usage/cmd_parameter/use_case_en.html">Use Case</a></li>
-<li class="toctree-l3"><a class="reference internal" href="howto/usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
-<li class="toctree-l3"><a class="reference internal" href="howto/usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/cluster/cluster_train_en.html">Run Distributed Training</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/dev/new_layer_en.html">Write New Layers</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/dev/contribute_to_paddle_en.html">Contribute Code</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/deep_model/rnn/index_en.html">RNN Models</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="api/index_en.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">Model Configuration</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="api/v2/data.html">Data Reader Interface and DataSets</a></li>
-<li class="toctree-l2"><a class="reference internal" href="api/v2/run_logic.html">Training and Inference</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="about/index_en.html">ABOUT</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
-
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Python Module Index</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-
-   <h1>Python Module Index</h1>
-
-   <div class="modindex-jumpbox">
-   <a href="#cap-p"><strong>p</strong></a>
-   </div>
-
-   <table class="indextable modindextable">
-     <tr class="pcap"><td></td><td>&#160;</td><td></td></tr>
-     <tr class="cap" id="cap-p"><td></td><td>
-       <strong>p</strong></td><td></td></tr>
-     <tr>
-       <td><img src="_static/minus.png" class="toggler"
-              id="toggle-1" style="display: none" alt="-" /></td>
-       <td>
-       <code class="xref">paddle</code></td><td>
-       <em></em></td></tr>
-     <tr class="cg-1">
-       <td></td>
-       <td>&#160;&#160;&#160;
-       <a href="api/v1/trainer_config_helpers/attrs.html#module-paddle.trainer_config_helpers.attrs"><code class="xref">paddle.trainer_config_helpers.attrs</code></a></td><td>
-       <em></em></td></tr>
-     <tr class="cg-1">
-       <td></td>
-       <td>&#160;&#160;&#160;
-       <a href="api/v1/trainer_config_helpers/data_sources.html#module-paddle.trainer_config_helpers.data_sources"><code class="xref">paddle.trainer_config_helpers.data_sources</code></a></td><td>
-       <em></em></td></tr>
-   </table>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'./',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="_static/jquery.js"></script>
-      <script type="text/javascript" src="_static/underscore.js"></script>
-      <script type="text/javascript" src="_static/doctools.js"></script>
-      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc/search.html b/develop/doc/search.html
index 09f082299f10449f0c952e3a54debe593cb8153f..c23994e0898d5194b12f0bf20d2d54660f3b081d 100644
--- a/develop/doc/search.html
+++ b/develop/doc/search.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/searchindex.js b/develop/doc/searchindex.js
index 68b67e9352c0253387b3348153c42cca03903445..719c682252ca2f15e139659c56cb103521e19f02 100644
--- a/develop/doc/searchindex.js
+++ b/develop/doc/searchindex.js
@@ -1 +1 @@
-Search.setIndex({docnames:["about/index_en","api/index_en","api/v1/data_provider/dataprovider_en","api/v1/data_provider/pydataprovider2_en","api/v1/index_en","api/v1/predict/swig_py_paddle_en","api/v1/trainer_config_helpers/activations","api/v1/trainer_config_helpers/attrs","api/v1/trainer_config_helpers/data_sources","api/v1/trainer_config_helpers/evaluators","api/v1/trainer_config_helpers/layers","api/v1/trainer_config_helpers/networks","api/v1/trainer_config_helpers/optimizers","api/v1/trainer_config_helpers/poolings","api/v2/config/activation","api/v2/config/attr","api/v2/config/layer","api/v2/config/networks","api/v2/config/optimizer","api/v2/config/pooling","api/v2/data","api/v2/model_configs","api/v2/run_logic","design/api","design/cluster_train/README","design/cluster_train/checkpointing","design/cluster_train/data_dispatch","design/multi_language_interface/00.why_plain_c","design/multi_language_interface/01.inference_implementation","design/reader/README","design/releasing_process","getstarted/basic_usage/index_en","getstarted/build_and_install/build_from_source_en","getstarted/build_and_install/docker_install_en","getstarted/build_and_install/index_en","getstarted/build_and_install/ubuntu_install_en","getstarted/index_en","howto/deep_model/rnn/index_en","howto/deep_model/rnn/rnn_config_en","howto/dev/contribute_to_paddle_en","howto/dev/new_layer_en","howto/index_en","howto/optimization/gpu_profiling_en","howto/usage/cluster/cluster_train_en","howto/usage/cmd_parameter/arguments_en","howto/usage/cmd_parameter/detail_introduction_en","howto/usage/cmd_parameter/index_en","howto/usage/cmd_parameter/use_case_en","howto/usage/k8s/k8s_aws_en","howto/usage/k8s/k8s_en","howto/usage/k8s/src/k8s_data/README","howto/usage/k8s/src/k8s_train/README","index_en","tutorials/embedding_model/index_en","tutorials/gan/index_en","tutorials/image_classification/index_en","tutorials/imagenet_model/resnet_model_en","tutorials/index_en","tutorials/quick_start/index_en","tutorials/rec/ml_dataset_en","tutorials/rec/ml_regression_en","tutorials/semantic_role_labeling/index_en","tutorials/sentiment_analysis/index_en","tutorials/text_generation/index_en"],envversion:50,filenames:["about/index_en.rst","api/index_en.rst","api/v1/data_provider/dataprovider_en.rst","api/v1/data_provider/pydataprovider2_en.rst","api/v1/index_en.rst","api/v1/predict/swig_py_paddle_en.rst","api/v1/trainer_config_helpers/activations.rst","api/v1/trainer_config_helpers/attrs.rst","api/v1/trainer_config_helpers/data_sources.rst","api/v1/trainer_config_helpers/evaluators.rst","api/v1/trainer_config_helpers/layers.rst","api/v1/trainer_config_helpers/networks.rst","api/v1/trainer_config_helpers/optimizers.rst","api/v1/trainer_config_helpers/poolings.rst","api/v2/config/activation.rst","api/v2/config/attr.rst","api/v2/config/layer.rst","api/v2/config/networks.rst","api/v2/config/optimizer.rst","api/v2/config/pooling.rst","api/v2/data.rst","api/v2/model_configs.rst","api/v2/run_logic.rst","design/api.md","design/cluster_train/README.md","design/cluster_train/checkpointing.md","design/cluster_train/data_dispatch.md","design/multi_language_interface/00.why_plain_c.md","design/multi_language_interface/01.inference_implementation.md","design/reader/README.md","design/releasing_process.md","getstarted/basic_usage/index_en.rst","getstarted/build_and_install/build_from_source_en.md","getstarted/build_and_install/docker_install_en.rst","getstarted/build_and_install/index_en.rst","getstarted/build_and_install/ubuntu_install_en.rst","getstarted/index_en.rst","howto/deep_model/rnn/index_en.rst","howto/deep_model/rnn/rnn_config_en.rst","howto/dev/contribute_to_paddle_en.md","howto/dev/new_layer_en.rst","howto/index_en.rst","howto/optimization/gpu_profiling_en.rst","howto/usage/cluster/cluster_train_en.md","howto/usage/cmd_parameter/arguments_en.md","howto/usage/cmd_parameter/detail_introduction_en.md","howto/usage/cmd_parameter/index_en.rst","howto/usage/cmd_parameter/use_case_en.md","howto/usage/k8s/k8s_aws_en.md","howto/usage/k8s/k8s_en.md","howto/usage/k8s/src/k8s_data/README.md","howto/usage/k8s/src/k8s_train/README.md","index_en.rst","tutorials/embedding_model/index_en.md","tutorials/gan/index_en.md","tutorials/image_classification/index_en.md","tutorials/imagenet_model/resnet_model_en.md","tutorials/index_en.md","tutorials/quick_start/index_en.md","tutorials/rec/ml_dataset_en.md","tutorials/rec/ml_regression_en.rst","tutorials/semantic_role_labeling/index_en.md","tutorials/sentiment_analysis/index_en.md","tutorials/text_generation/index_en.md"],objects:{"paddle.trainer.PyDataProvider2":{provider:[3,0,1,""]},"paddle.trainer_config_helpers":{attrs:[7,1,0,"-"],data_sources:[8,1,0,"-"]},"paddle.trainer_config_helpers.attrs":{ExtraAttr:[7,2,1,""],ExtraLayerAttribute:[7,3,1,""],ParamAttr:[7,2,1,""],ParameterAttribute:[7,3,1,""]},"paddle.trainer_config_helpers.attrs.ParameterAttribute":{set_default_parameter_name:[7,4,1,""]},"paddle.trainer_config_helpers.data_sources":{define_py_data_sources2:[8,0,1,""]}},objnames:{"0":["py","function","Python function"],"1":["py","module","Python module"],"2":["py","attribute","Python attribute"],"3":["py","class","Python class"],"4":["py","method","Python method"]},objtypes:{"0":"py:function","1":"py:module","2":"py:attribute","3":"py:class","4":"py:method"},terms:{"0000x":58,"00186201e":5,"00m":42,"02595v1":[10,16],"03m":42,"0424m":42,"0473v3":[11,17],"055ee37d":48,"05d":55,"0630u":42,"06u":42,"0810u":42,"08823112e":5,"0957m":42,"0ab":[10,16],"0rc1":30,"0rc2":[30,33],"0th":63,"10007_10":62,"10014_7":62,"100gb":42,"100gi":48,"10m":42,"1150u":42,"11\u5b9e\u73b0\u4e86c":28,"11e6":49,"12194102e":5,"124n":42,"13m":49,"1490u":42,"15501715e":5,"1550u":42,"15mb":58,"1636k":63,"16mb":58,"16u":42,"173m":56,"173n":42,"1770u":42,"18ad":48,"18e457ce3d362ff5f3febf8e7f85ffec852f70f3b629add10aed84f930a68750":49,"197u":42,"1gb":42,"1st":[53,56,62,63],"202mb":63,"210u":42,"211839e770f7b538e2d8":[11,17],"215n":42,"228u":42,"234m":56,"2520u":42,"252kb":58,"25639710e":5,"25k":58,"2680u":42,"27787406e":5,"279n":42,"27m":42,"285m":42,"2863m":42,"28m":42,"28x28":3,"2977m":42,"2cbf7385":48,"2nd":[10,16,62,63],"302n":42,"30u":42,"32777140e":5,"328n":42,"32u":42,"32x32":[20,55],"331n":42,"3320u":42,"36540484e":5,"365e":48,"36u":42,"3710m":42,"3768m":42,"387u":42,"38u":42,"3920u":42,"39u":42,"3rd":[60,62,63],"4035m":42,"4090u":42,"4096mb":45,"4279m":42,"43630644e":5,"43u":42,"448a5b355b84":49,"4560u":42,"4563m":42,"45u":42,"4650u":42,"4726m":42,"473m":49,"48565123e":5,"48684503e":5,"49316648e":5,"4gb":45,"50bd":48,"50gi":48,"51111044e":5,"514u":42,"525n":42,"526u":42,"53018653e":5,"536u":42,"5460u":42,"5470u":42,"54u":42,"55g":63,"5690m":42,"573u":42,"578n":42,"5798m":42,"586u":42,"58s":49,"5969m":42,"6080u":42,"6082v4":[10,16],"6140u":42,"6305m":42,"639u":42,"655u":42,"6780u":42,"6810u":42,"682u":42,"6970u":42,"6ce9":48,"6node":43,"6th":63,"704u":42,"70634608e":5,"7090u":42,"72296313e":5,"72u":42,"73u":42,"75u":42,"760u":42,"767u":42,"783n":42,"784u":42,"78m":42,"7eamaa":20,"7kb":49,"8250u":42,"8300u":42,"830n":42,"849m":42,"85625684e":5,"861u":42,"864k":63,"8661m":42,"892m":42,"901n":42,"90u":42,"918u":42,"9247m":42,"924n":42,"9261m":42,"93137714e":5,"9330m":42,"94u":42,"9530m":42,"96644767e":5,"983m":42,"988u":42,"997u":42,"99982715e":5,"99m":56,"99u":42,"9f18":49,"\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u6a21\u578b\u7531\u5927\u91cf\u7684\u53c2\u6570\u7ec4\u6210":25,"\u4e00\u822c\u4e0d\u5141\u8bb8\u518d\u4ece":30,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u67d0\u4e00\u7c7b\u6587\u4ef6\u7684\u5b9e\u73b0\u65b9\u5f0f":28,"\u4e0d\u4f7f\u7528\u9759\u6001\u5e93":27,"\u4e0d\u4f7f\u7528c":27,"\u4e0d\u4f7f\u7528swig":27,"\u4e0d\u540c\u7248\u672c\u7684\u7f16\u8bd1\u5668\u4e4b\u95f4":27,"\u4e0d\u540c\u8bed\u8a00\u7684\u63a5\u53e3\u9002\u5e94\u4e0d\u540c\u8bed\u8a00\u7684\u7279\u6027":27,"\u4e0d\u5728":28,"\u4e0d\u5d4c\u5165\u5176\u4ed6\u8bed\u8a00\u89e3\u91ca\u5668":27,"\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":27,"\u4e0d\u663e\u793a\u7684\u5199\u6bcf\u4e2a\u7c7b\u5177\u4f53\u5305\u542b\u4ec0\u4e48":27,"\u4e0e\u529f\u80fd\u5206\u652f\u4e0d\u540c\u7684\u662f":30,"\u4e0e\u53ef\u80fd\u6709\u7684":30,"\u4e14\u589e\u52a0\u4e00\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00":27,"\u4e14\u8c03\u7528\u65f6\u4e0d\u80fd\u629b\u51fa\u5f02\u5e38\u6216\u51fa\u73b0\u8fd0\u884c\u65f6\u9519\u8bef":28,"\u4e14c99\u652f\u6301bool\u7c7b\u578b\u548c\u5b9a\u957f\u6574\u6570":27,"\u4e14c99\u76f8\u5bf9\u4e8ec11\u4f7f\u7528\u66f4\u52a0\u5e7f\u6cdb":27,"\u4e2a\u6027\u5316\u63a8\u8350":30,"\u4e2d":[27,28],"\u4e2d\u5199\u5165json\u5185\u5bb9":25,"\u4e2d\u5b8c\u5168\u4e00\u81f4":27,"\u4e2d\u5b9e\u73b0\u7684\u7ed3\u6784\u4f53":28,"\u4e3a\u4e86\u66b4\u9732\u7684\u63a5\u53e3\u5c3d\u91cf\u7b80\u5355":28,"\u4e4b\u5916\u7684\u6240\u6709\u5934\u6587\u4ef6":28,"\u4e5f\u4e0d\u4f7f\u7528\u5176\u4ed6\u52a8\u6001\u5e93":27,"\u4e5f\u4e0d\u5e94\u8be5\u62a5\u9519":28,"\u4e5f\u4e0d\u751f\u6210":28,"\u4e66\u5199":27,"\u4eba\u8138\u8bc6\u522b":26,"\u4ec5\u4ec5\u4f7f\u7528":27,"\u4ece":30,"\u4ece\u78c1\u76d8\u6587\u4ef6\u4e2d\u52a0\u8f7duuid\u6587\u4ef6\u540d\u7684\u68c0\u67e5\u70b9\u5feb\u7167\u6587\u4ef6":25,"\u4eceetcd\u4e2d\u8bfb\u53d6\u8282\u70b9":25,"\u4ed6\u4e3b\u8981\u5305\u542b\u4e86\u5b9e\u9645\u66b4\u9732\u7684\u7c7b\u578b\u7ed3\u6784":28,"\u4ed6\u662f\u5c06":28,"\u4ed6\u7684\u76ee\u6807\u662f\u4f7f\u7528c":27,"\u4ee3\u7801\u751f\u6210\u7684\u7b26\u53f7\u53ef\u80fd\u4e0d\u4e00\u81f4":27,"\u4ee3\u8868\u8fd9\u4e2ashard\u7684\u6700\u5927index":26,"\u4ee3\u8868shard\u7684index":26,"\u4ee5\u4e0a\u4ee3\u7801\u7684reader\u8f93\u51fa\u7684data":26,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u4e0b\u751f\u6210100\u4e2a\u6587\u4ef6":26,"\u4ee5\u4e0b":26,"\u4ee5\u53canumpi":26,"\u4efb\u610f\u65f6\u523b\u53ea\u53ef\u80fd\u540c\u65f6\u6709\u4e00\u53f0\u670d\u52a1\u5668\u6545\u969c":25,"\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7248\u672cpython\u5728\u4e00\u4e2a\u8fdb\u7a0b\u91cc\u7684bug":27,"\u4f1a\u76f4\u63a5\u62a5\u9519\u9000\u51fa":27,"\u4f1a\u88abpickle\u5e8f\u5217\u5316\u6210\u5b57\u7b26\u4e32":26,"\u4f20\u5165":26,"\u4f46":28,"\u4f46\u4e0d\u66b4\u9732":28,"\u4f46\u5e76\u6ca1\u6709\u7ecf\u8fc7\u56de\u5f52\u6d4b\u8bd5":30,"\u4f46\u6240\u6709fork\u7684\u7248\u672c\u5e93\u7684\u6240\u6709\u5206\u652f\u90fd\u76f8\u5f53\u4e8e\u7279\u6027\u5206\u652f":30,"\u4f46\u662f\u53c8\u8fc7\u4e8e\u7410\u788e":28,"\u4f46\u662f\u89e3\u91ca\u6027\u8bed\u8a00":27,"\u4f5c\u4e3a\u7c7b\u53e5\u67c4":27,"\u4f7f\u7528":[28,30],"\u4f7f\u7528\u4e0b\u9762\u547d\u4ee4":26,"\u4f7f\u7528\u52a8\u6001\u5e93":27,"\u4f7f\u7528\u540c\u6837\u7684\u8bad\u7ec3\u6570\u636eblock":25,"\u4f7f\u7528\u667a\u80fd\u6307\u9488\u7684\u539f\u56e0\u662f":28,"\u4f7f\u7528\u76f8\u5bf9\u8def\u5f84\u7684\u5f15\u7528\u65b9\u5f0f":28,"\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u5b8c\u6210\u5bf9\u65b0\u6570\u636e\u7684\u9884\u6d4b":25,"\u4f7f\u7528\u9759\u6001\u5e93\u548c\u52a8\u6001\u5e93\u96be\u5ea6\u5dee\u4e0d\u591a":27,"\u4f7f\u7528c":28,"\u4f7f\u7528c99\u505a\u63a5\u53e3":27,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c11\u7684\u539f\u56e0\u662f":27,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c89":27,"\u4f7f\u7528regress":30,"\u4f7f\u7528swig\u53ea\u652f\u6301cpython\u89e3\u91ca\u5668":27,"\u4f7f\u7528swig\u9700\u8981\u591a\u8bed\u8a00\u7ed1\u5b9a\u7684\u5f00\u53d1\u4eba\u5458\u719f\u7ec3\u638c\u63e1swig\u914d\u7f6e":27,"\u4f7f\u7528void":27,"\u4f8b\u5982":[27,28,30],"\u4f8b\u5982\u5bf9\u4e8ejava\u6216\u8005python":27,"\u4f8b\u5982\u5bf9\u4e8ejava\u6765\u8bf4":27,"\u4f8b\u5982\u5bf9\u4e8epython":27,"\u4f8b\u5982c":27,"\u4f8b\u5982java\u4e0epython\u7684\u9519\u8bef\u5904\u7406\u662f\u76f4\u63a5\u6254\u51fa\u6765except":27,"\u4f8b\u5982python\u53ef\u4ee5\u4f7f\u7528":27,"\u4f8b\u5982python\u7684":27,"\u4f9d\u6b21\u7c7b\u63a8":30,"\u4fbf\u662f\u5c06\u9759\u6001\u5e93\u52a0\u5165jvm\u4e2d":27,"\u4fee\u590d\u6240\u6709bug\u540e":30,"\u4fee\u590ddocker\u7f16\u8bd1\u955c\u50cf\u95ee\u9898":30,"\u4fee\u590dubuntu":30,"\u505a\u5982\u4e0b\u51e0\u4e2a\u64cd\u4f5c":30,"\u505a\u63a5\u53e3":27,"\u505c\u6b62\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":25,"\u5148\u5b9e\u73b0\u6a21\u578b\u63a8\u65ad\u7684api":28,"\u5176\u4e2d":[27,30],"\u5176\u4ed6\u51fd\u6570\u5747\u8fd4\u56de":28,"\u5176\u4ed6\u7528\u6237\u7684fork\u7248\u672c\u5e93\u5e76\u4e0d\u9700\u8981\u4e25\u683c\u9075\u5b88":30,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":28,"\u5177\u4f53\u539f\u56e0\u53c2\u8003":28,"\u5177\u4f53\u8bf7\u53c2\u8003":28,"\u5185\u90e8\u9a71\u52a8python\u89e3\u91ca\u5668\u8fdb\u884c\u6a21\u578b\u914d\u7f6e\u89e3\u6790\u548c\u6570\u636e\u8bfb\u53d6":27,"\u518d\u5728\u6bcf\u4e00\u4e2aapi\u4e2d\u81ea\u5df1\u68c0\u67e5\u7c7b\u578b":27,"\u518d\u57fa\u4e8e":30,"\u5199\u4ee3\u7801":27,"\u5199\u5165\u5feb\u7167\u6570\u636e":25,"\u51fd\u6570\u5373\u53ef\u5b8c\u6210\u8f6c\u6362":26,"\u51fd\u6570\u540d\u4e3a":28,"\u51fd\u6570\u547d\u540d":27,"\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":25,"\u5206\u652f":30,"\u5206\u652f\u4e00\u65e6\u5efa\u7acb":30,"\u5206\u652f\u4e2d":30,"\u5206\u652f\u4e3a\u5f00\u53d1":30,"\u5206\u652f\u4e3a\u6bcf\u4e00\u6b21release\u65f6\u5efa\u7acb\u7684\u4e34\u65f6\u5206\u652f":30,"\u5206\u652f\u4e3a\u7a33\u5b9a":30,"\u5206\u652f\u529f\u80fd\u7684\u5c01\u95ed":30,"\u5206\u652f\u5408\u5165":30,"\u5206\u652f\u5408\u5165master\u5206\u652f":30,"\u5206\u652f\u540c\u6b65\u4e3b\u7248\u672c\u5e93\u7684":30,"\u5206\u652f\u540d\u4e3a":30,"\u5206\u652f\u5b58\u5728\u7684\u65f6\u5019":30,"\u5206\u652f\u6d3e\u751f\u51fa\u65b0\u7684\u5206\u652f":30,"\u5206\u652f\u7684\u7248\u672c\u90fd\u662f\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5\u548c\u56de\u5f52\u6d4b\u8bd5\u7684\u7248\u672c":30,"\u5206\u652f\u7684\u7248\u672c\u90fd\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5":30,"\u5206\u7247":25,"\u5219\u4f7f\u7528\u542f\u52a8\u53c2\u6570\u5b9a\u4e49\u7684\u521d\u59cb\u5316\u65b9\u6cd5\u521d\u59cb\u5316\u53c2\u6570":25,"\u5219\u5ffd\u7565":25,"\u5219\u628a\u53e6\u4e00\u4e2a\u6162\u901f\u7684kill\u6389":25,"\u5219\u76f4\u63a5\u5f15\u5165\u53e6\u4e00\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5219\u9700\u8981\u56de\u6eda\u5230\u4e0a\u4e00\u4e2a\u68c0\u67e5\u70b9":25,"\u5220\u9664\u78c1\u76d8\u76ee\u5f55\u4e2d\u4e0d\u662f\u5f53\u524duuid\u7684\u5feb\u7167\u6587\u4ef6":25,"\u5230":25,"\u529f\u80fd\u7684\u6b63\u786e\u6027\u5305\u62ec\u9a8c\u8bc1paddle\u76ee\u524d\u7684":30,"\u52a8\u6001\u5e93":27,"\u5305\u542b\u4e86\u67d0\u79cd\u7c7b\u578b\u7684\u7c7b\u578b\u5b9a\u4e49\u548c\u66b4\u9732\u7684\u5168\u90e8\u51fd\u6570":28,"\u5305\u62ec":26,"\u5305\u62ec\u6743\u91cdw\u548c\u504f\u7f6eb":25,"\u534f\u540c\u5b8c\u6210releas":30,"\u5355\u4e2a\u503c":26,"\u5355\u70b9\u6545\u969c":25,"\u5373":28,"\u5373\u4f7f\u7528":28,"\u5373\u4f7f\u7528\u6237\u76f4\u63a5\u5f15\u7528\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5373\u4f7fc":28,"\u5373\u4f8b\u5982":28,"\u5373\u4fbfpaddl":28,"\u5373\u5b8c\u6210\u67d0\u4e00\u4e2a\u4efb\u52a1\u7684\u6700\u5c11\u51fd\u6570":28,"\u5373\u66b4\u9732":28,"\u5373\u8fd9\u4e2a\u52a8\u6001\u5e93\u662f\u4e0d\u4f9d\u8d56\u4e8e\u5176\u4ed6\u4efb\u4f55\u6587\u4ef6\u7684":27,"\u53c2\u6570":27,"\u53c2\u8003":27,"\u53d1\u5e03\u5230dockerhub":30,"\u53d1\u5e03\u5230github":30,"\u53ea\u5bf9\u7279\u6b8a\u5728\u7ebf\u7cfb\u7edf\u8003\u8651\u4e24\u53f0\u4ee5\u4e0a\u540c\u65f6\u6545\u969c\u7684\u5bb9\u707e":25,"\u53ea\u66b4\u9732\u6982\u5ff5\u7684\u63a5\u53e3":28,"\u53ea\u80fd\u8c03\u7528paddle\u7684\u52a8\u6001\u5e93":27,"\u53ea\u9700\u8981\u6062\u590d\u8fd9\u53f0\u8282\u70b9":25,"\u53ef\u4ee5\u51cf\u5c0f\u7cfb\u7edf\u590d\u6742\u6027":25,"\u53ef\u4ee5\u5728\u4efb\u4f55\u673a\u5668\u4e0a\u6267\u884c\u7684":27,"\u53ef\u4ee5\u628a\u672c\u5730\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5b58\u50a8\u96c6\u7fa4\u4e2d":26,"\u53ef\u4ee5\u6709\u6548\u7684\u907f\u514dparamet":25,"\u53ef\u4ee5\u7528\u4ee5\u4e0b\u6307\u4ee4":26,"\u53ef\u4ee5\u7ee7\u7eed\u5728\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f\u63d0\u4ea4\u4ee3\u7801":30,"\u53ef\u4ee5\u901a\u8fc7\u9636\u6bb5\u6027\u7684\u4fdd\u5b58\u6bcf\u4e2aparamet":25,"\u53ef\u4ee5\u901a\u8fc7volume\u6216persistentvolume\u6302\u8f7d\u5b58\u50a8\u7a7a\u95f4\u5230\u6bcf\u4e2a\u5bb9\u5668\u4e2d":26,"\u53ef\u80fd\u4f1a\u9020\u6210\u7f51\u7edc\u62e5\u585e":25,"\u540c\u65f6\u518d\u5c06":30,"\u540c\u65f6\u63d0\u8d77":30,"\u540d\u5b57\u4fee\u9970":27,"\u5411\u6307\u5b9a\u7684\u76ee\u5f55\u4e2d\u4e00\u4e2a\u65b0\u7684\u6587\u4ef6":25,"\u5411paddle\u7684\u4e3b\u7248\u672c\u5e93\u63d0\u4ea4":30,"\u5426\u5219\u5f97\u628apaddle\u9759\u6001\u5e93\u94fe\u63a5\u5230\u89e3\u91ca\u5668\u91cc":27,"\u542f\u52a8\u4e00\u4e2a\u65b0\u7684\u7ebf\u7a0b\u5f00\u59cb\u4fdd\u5b58\u68c0\u67e5\u70b9":25,"\u548c":[26,27,28,30],"\u548c\u79bb\u7ebf\u6570\u636e\u7684\u65b9\u5f0f":26,"\u54ea\u4e2atrainer\u5148\u5b8c\u6210block\u7684\u8bad\u7ec3":25,"\u56e0\u4e3a\u8fd9\u6837\u505a\u4e5f\u6ca1\u6cd5\u4fdd\u8bc1\u6d88\u9664\u968f\u673a\u6027":25,"\u56e0\u4e3aswig\u5728\u7b2c\u4e09\u65b9\u8bed\u8a00\u4e2d\u66b4\u9732\u7684\u51fd\u6570\u540d":27,"\u56fe\u50cf\u5206\u7c7b":30,"\u5728":[28,30],"\u5728\u4e00\u4e2a\u4e0d\u53ef\u4e2d\u65ad\u5e76\u7f3a\u5c11\u5907\u4efd\u7684\u8bad\u7ec3\u4efb\u52a1\u4e2d":25,"\u5728\u4e0a\u56fe\u4e2d\u663e\u793a\u4e86\u5728\u4e00\u4e2a\u5b9e\u9645\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u5e94\u7528":26,"\u5728\u51fa\u73b0\u5355\u70b9\u6545\u969c\u65f6":25,"\u5728\u5b9e\u73b0\u8fc7\u7a0b\u4e2d":28,"\u5728\u5f02\u6784\u96c6\u7fa4\u4e2d":25,"\u5728\u5f15\u5165\u5176\u4ed6\u7c7b\u578b\u7684\u5934\u6587\u4ef6\u65f6":28,"\u5728\u5feb\u7167\u5199\u5165\u5b8c\u6210\u540e":25,"\u5728\u60a8\u7684\u5b9e\u9645\u73af\u5883\u4e2d":25,"\u5728\u6570\u636e\u96c6\u53ef\u4ee5\u88ab\u8bad\u7ec3\u4e4b\u524d":26,"\u5728\u6837\u4f8b\u4e2d":28,"\u5728\u7528\u6237\u4f7f\u7528c":28,"\u5728\u7ebf\u6a21\u578b\u9884\u6d4b\u670d\u52a1":26,"\u5728\u8bc4\u5ba1\u8fc7\u7a0b\u4e2d":30,"\u5728\u8fd9\u4e2a":30,"\u5728\u8fd9\u4e2a\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":27,"\u5728\u8fd9\u4e2a\u9636\u6bb5\u7684\u4ee3\u7801\u6b63\u5728\u7ecf\u5386\u56de\u5f52\u6d4b\u8bd5":30,"\u5728\u8fd9\u4e9b\u5934\u6587\u4ef6\u4e2d":28,"\u5728\u8fd9\u4e9b\u6587\u4ef6\u4e2d":28,"\u5728\u96c6\u7fa4\u8fd0\u884cmapreduc":26,"\u5728c":27,"\u5728c\u7684\u5934\u6587\u4ef6":27,"\u5728cephfs\u5b58\u50a8\u7cfb\u7edf\u4e2d\u7684\u516c\u5f00\u76ee\u5f55":26,"\u5728kubernetes\u4e0a\u8fd0\u884c\u7684\u4e0d\u540c\u7684\u8ba1\u7b97\u6846\u67b6":26,"\u5728paddle\u4e4b\u4e0a\u8fd0\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u8f93\u51fa\u7684\u6a21\u578b\u4f1a\u63d0\u4f9b\u7ed9\u5728\u7ebf\u4eba\u8138\u8bc6\u522b\u7684\u5e94\u7528\u4f7f\u7528":26,"\u5728paramet":25,"\u5747\u4f1a\u88ab\u5b89\u88c5\u5230includ":28,"\u5747\u662f\u5728":28,"\u591a\u4e2a\u503c":26,"\u591a\u4e2aparamet":25,"\u5927\u591a\u6570\u8bed\u8a00\u90fd\u652f\u6301\u4f7f\u7528c\u8bed\u8a00api":27,"\u5982\u56fe\u4e2dtrainer":25,"\u5982\u679c\u4e0a\u9762\u4e24\u6b65\u51fa\u73b0\u9519\u8bef":25,"\u5982\u679c\u4f7f\u7528swig\u6211\u4eec\u9700\u8981\u5c06\u5728interface\u6587\u4ef6\u91cc":27,"\u5982\u679c\u5931\u8d25":30,"\u5982\u679c\u5b58\u5728\u67d0\u4e9btrainer\u6267\u884c\u901f\u5ea6\u8fc7\u6162\u4f1a\u5f71\u54cd\u6574\u4f53\u96c6\u7fa4\u7684\u901f\u5ea6":25,"\u5982\u679c\u5df2\u7ecf\u6b63\u5728\u6267\u884c\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":25,"\u5982\u679c\u662f\u5176\u5b83\u7c7b\u578b":26,"\u5982\u679c\u6709bugfix\u7684\u884c\u4e3a":30,"\u5982\u679c\u67d0\u4e00\u4e2a\u7c7b\u578b\u9700\u8981\u5f15\u7528\u53e6\u4e00\u4e2a\u7c7b\u578b":28,"\u5982\u679c\u67d0\u4e00\u4e2apaddl":28,"\u5982\u679c\u67d0\u4e00\u4e2apaddle\u6982\u5ff5\u5fc5\u987b\u8981\u66b4\u9732":28,"\u5982\u679c\u6ee1\u8db3\u6761\u4ef6":25,"\u5982\u679c\u7528\u6237\u8981\u628apaddle\u7684\u9759\u6001\u5e93":27,"\u5982\u679c\u8c03\u7528\u9759\u6001\u5e93\u53ea\u80fd\u5c06\u9759\u6001\u5e93\u4e0e\u89e3\u91ca\u5668\u94fe\u63a5":27,"\u5982\u679cparamet":25,"\u5b57\u7b26\u4e32":26,"\u5b58\u50a8":26,"\u5b66\u4e60\u6210\u672c\u9ad8":27,"\u5b83\u4eec\u7684\u6587\u4ef6\u540d\u662f":26,"\u5b89\u88c5\u540e\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":28,"\u5b8c\u6210\u6570\u636e\u7684\u9884\u5904\u7406":26,"\u5b9e\u73b0\u7b80\u5355":27,"\u5bf9\u4e8e\u4e0d\u540c\u8bed\u8a00":27,"\u5bf9\u4e8e\u540c\u4e00\u6bb5c":27,"\u5bf9\u4e8e\u591a\u8bed\u8a00\u63a5\u53e3":27,"\u5bf9\u4e8e\u5927\u591a\u6570\u8bed\u8a00":27,"\u5bf9\u4e8e\u6bcf\u79cd\u7c7b\u578b":28,"\u5bf9\u4e8e\u6bcf\u79cdc":28,"\u5bf9\u6bd4":27,"\u5bf9\u8f93\u5165\u53c2\u6570\u7684\u5b89\u5168\u6027\u8fdb\u884c\u4e86\u5fc5\u8981\u7684\u5224\u65ad":28,"\u5bf9\u8fd9\u4e2a\u7248\u672c\u7684\u63d0\u4ea4":30,"\u5bfc\u51fa\u8fd9\u4e9b\u63a5\u53e3":28,"\u5c06":30,"\u5c06\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u62c6\u5206\u6210\u591a\u4efd":25,"\u5c06\u5927\u91cf\u7684":27,"\u5c06\u65b0\u5206\u652f\u7684\u7248\u672c\u6253\u4e0atag":30,"\u5c06master\u5206\u652f\u7684\u5408\u5165commit\u6253\u4e0atag":30,"\u5c31\u9700\u8981\u5bf9\u8fd9\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00\u589e\u52a0\u4e00\u4e9b\u5b9a\u4e49":27,"\u5e73\u5747\u6545\u969c\u4fee\u590d\u65f6\u95f4":25,"\u5e73\u5747\u6545\u969c\u7387":25,"\u5e76\u4e14\u4f7f\u7528":28,"\u5e76\u4e14\u53ef\u4ee5\u88ab\u63d0\u4ea4\u7684job\u76f4\u63a5\u4f7f\u7528":26,"\u5e76\u4e14\u5728\u5e38\u89c1\u7684\u5e73\u53f0\u4e0a":27,"\u5e76\u4e14\u8ba9\u63a5\u53e3\u8131\u79bb\u5b9e\u73b0\u7ec6\u8282":27,"\u5e76\u5220\u9664":30,"\u5e76\u5220\u9664\u66f4\u65e9\u7684\u5feb\u7167":25,"\u5e76\u52a0\u8f7d\u5176\u4e2d\u7684\u53c2\u6570":25,"\u5e76\u5728\u96c6\u7fa4\u4e2d\u8fd0\u884c\u591a\u4e2a\u5206\u5e03\u5f0f\u6570\u636e\u5904\u7406\u4efb\u52a1":26,"\u5e76\u5c06c":28,"\u5e76\u628a\u5feb\u7167\u4fdd\u5b58\u5230\u8fd9\u4e2a\u76ee\u5f55\u4e0b":25,"\u5e76\u6ca1\u6709paddle\u7279\u522b\u9700\u8981\u7684\u7279\u6027":27,"\u5e76\u88ab\u5b58\u50a8\u5728\u8bf8\u5982hadoop":26,"\u5e76\u9002\u5e94github\u7684\u7279\u6027\u505a\u4e86\u4e00\u4e9b\u533a\u522b":30,"\u5efa\u8bae":30,"\u5f00\u53d1\u4e86\u6a21\u578b\u9884\u6d4b\u7684\u6837\u4f8b\u4ee3\u7801":28,"\u5f00\u53d1\u8005\u4fee\u6539\u81ea\u5df1\u7684\u4ee3\u7801":30,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4e2d":30,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4f7f\u7528":30,"\u5f00\u59cb\u63d0\u4f9b\u670d\u52a1":25,"\u5f15\u5165\u4e86\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5f53\u529f\u80fd\u5206\u652f\u5f00\u53d1\u5b8c\u6bd5\u540e":30,"\u5f53\u7528\u6237\u4f7f\u7528\u5b8c\u8fd9\u4e2a\u53c2\u6570\u540e":28,"\u5f88\u96be\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":27,"\u5f97\u4f7f\u7528":27,"\u5fc5\u8981":28,"\u60c5\u611f\u5206\u6790":30,"\u6211\u4eec\u4e5f\u53ef\u4ee5\u786e\u5b9a\u6bcf\u4e00\u4e2a\u53c2\u6570\u7684\u7c7b\u578b":28,"\u6211\u4eec\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":27,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u8f6c\u6362\u65b9\u5f0f":26,"\u6211\u4eec\u6700\u7ec8\u7684\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165python\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u8bed\u8a00\u7684\u89e3\u91ca\u5668":27,"\u6211\u4eec\u90fd\u63d0\u4f9bpython\u7684\u8f6c\u6362\u5e93":26,"\u6216\u8005":[27,28],"\u6216\u8005\u5c06\u8fd9\u53f0\u8282\u70b9\u8fc1\u79fb\u5230\u53e6\u4e00\u4e2a\u8282\u70b9\u5e76\u542f\u52a8\u5373\u53ef\u6062\u590d\u8bad\u7ec3\u4efb\u52a1":25,"\u6216\u8005\u7528tuple\u8868\u793a\u7684\u591a\u4e2a\u503c":26,"\u6216\u8005\u7531\u5b83\u4eec\u7ec4\u6210\u7684list":26,"\u6240\u4ee5\u5728\u5199\u5165\u5feb\u7167\u7684\u8fc7\u7a0b\u4e2d":25,"\u6240\u6709\u4e0e\u7c7b\u578b\u76f8\u5173\u7684\u51fd\u6570":28,"\u6240\u6709\u7684\u63a5\u53e3\u5747\u4e3ac\u63a5\u53e3":28,"\u6240\u6709\u7c7b\u578b\u540d\u4e3a":28,"\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":27,"\u6253\u5f00\u8fd9\u4e2a\u7f16\u8bd1\u9009\u9879":28,"\u628a":26,"\u628a\u4e4b\u524d\u793a\u4f8b\u4e2d\u8f6c\u6362\u5b8c\u6bd5\u7684random":26,"\u6307\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u4e4b\u540e\u5f97\u5230\u7684\u6240\u6709\u53c2\u6570":25,"\u63a5\u53e3":[27,28],"\u63a5\u53e3\u5c42\u505a\u8fc7\u591a\u5c01\u88c5":28,"\u63a5\u53e3\u662f":26,"\u63d0\u4f9b\u7ed9\u7528\u6237\u672c\u5730\u8f6c\u6362\u7684\u5e93":26,"\u63d0\u4f9b\u7ed9paddle\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e":26,"\u6570\u636e\u8bfb\u53d6\u5747\u4ea4\u7531\u5176\u4ed6\u8bed\u8a00\u5b8c\u6210":27,"\u6570\u636e\u9884\u5904\u7406\u4efb\u52a1":26,"\u6587\u4ef6":27,"\u6587\u4ef6\u5185\u5bb9\u4e3a":27,"\u6587\u4ef6\u540d\u4e3a\u6b64uuid":25,"\u6587\u4ef6\u9700\u8981\u9884\u5148\u88ab\u8f6c\u6362\u6210paddlepaddle\u96c6\u7fa4\u5185\u90e8\u7684\u5b58\u50a8\u683c\u5f0f":26,"\u65b0\u624b\u5165\u95e8\u7ae0\u8282":30,"\u65b9\u4fbf\u6d4b\u8bd5\u4eba\u5458\u6d4b\u8bd5paddle\u7684\u884c\u4e3a":30,"\u65e0\u6cd5\u505a\u5230\u5bf9\u4e8e\u5404\u79cd\u8bed\u8a00\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u7684\u9002\u914d":27,"\u65e0\u8bba\u5728\u672c\u5730\u8fd8\u662f\u5728\u4e91\u7aef":26,"\u65e0\u8bba\u662f\u5728\u672c\u5730\u6216\u662f\u4e91\u7aef\u8f6c\u6362":26,"\u65f6":25,"\u662f\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3\u7684\u4ee3\u7801\u751f\u6210\u5668":27,"\u662f\u4e00\u4e2a\u7c7b\u578b\u7684\u6807\u5fd7":28,"\u662f\u4e0d\u5e38\u89c1\u7684\u505a\u6cd5":27,"\u662f\u5404\u4e2a\u5b9e\u73b0\u4e2d\u5171\u4eab\u7684\u5934\u6587\u4ef6":28,"\u662f\u56e0\u4e3ac99\u652f\u6301":27,"\u662f\u6307":28,"\u662f\u7528\u6237\u4f7f\u7528c":28,"\u662fc":28,"\u6682\u65f6\u4e0d\u8003\u8651\u591a\u4e2aparamet":25,"\u66b4\u9732\u8fd9\u4e2a\u6982\u5ff5\u5fc5\u8981\u51fd\u6570":28,"\u6700\u540e\u5220\u9664":30,"\u6700\u5e38\u89c1\u7684\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662fexcept":27,"\u6709\u6807\u51c6\u7684":27,"\u6709\u7684\u65f6\u5019":27,"\u672c\u5217\u8868\u8bf4\u660epaddle\u53d1\u7248\u4e4b\u524d\u9700\u8981\u6d4b\u8bd5\u7684\u529f\u80fd\u70b9":30,"\u672c\u6587\u6863\u63cf\u8ff0paddl":28,"\u673a\u5668\u7ffb\u8bd1":30,"\u6765\u4fdd\u8bc1\u8bad\u7ec3\u8fc7\u7a0b\u53ef\u4ee5\u4ece\u4e2d\u95f4\u72b6\u6001\u91cd\u65b0\u542f\u52a8":25,"\u6765\u786e\u4fdd\u628a":27,"\u6765\u8868\u793apaddle\u5185\u90e8\u7c7b":27,"\u6765\u8fdb\u884c\u8ba8\u8bba":28,"\u6807\u51c6\u8868\u793apaddle\u7248\u672c\u53f7":30,"\u68c0\u67e5\u70b9\u4fdd\u5b58\u7a0b\u5e8f\u6d41\u7a0b":25,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\u901a\u8fc7\u5b9a\u671f\u5411\u78c1\u76d8\u4e0a\u4fdd\u5b58\u4e00\u4efd\u5b58\u50a8\u5728paramet":25,"\u6a21\u578b\u6570\u636e\u68c0\u67e5\u70b9\u7684\u5b9e\u73b0":25,"\u6a21\u578b\u914d\u7f6e\u89e3\u6790":27,"\u6b64\u65f6master\u5c06\u8d1f\u8d23\u542f\u52a8\u4e00\u4e2a\u65b0\u7684train":25,"\u6bcf\u4e00\u4e2a":30,"\u6bcf\u4e00\u4e2a\u6587\u4ef6\u662f\u6570\u636e\u96c6\u7684\u4e00\u4e2ashard":26,"\u6bcf\u4e2a\u503c\u7684\u7c7b\u578b\u53ef\u4ee5\u662f\u6574\u5f62":26,"\u6bcf\u4e2adata":26,"\u6bcf\u4e2aparamet":25,"\u6bcf\u4e2ashard\u5206\u522b\u5b58\u50a8\u5728\u5176\u4e2d\u4e00\u53f0paramet":25,"\u6bcf\u6b21\u8f93\u51fa\u4e00\u4e2adata":26,"\u6bcf\u969410\u5206\u949f":25,"\u6bd4\u5982":26,"\u6bd4\u5982\u6bcf\u969410\u5206\u949f\u6700\u65b0\u7684\u5feb\u7167":25,"\u6bd4\u5982\u6d41\u5f0f\u6570\u636e\u5904\u7406":26,"\u6bd4\u5982imagenet\u8fd9\u4e2a\u6570\u636e\u96c6\u53ef\u80fd\u88ab\u5206\u62101000\u4e2ashard":26,"\u6bd4\u5982mnist":26,"\u6ce8":25,"\u6d4b\u8bd5docker\u955c\u50cf":30,"\u6d6e\u70b9\u578b\u6570\u636e":26,"\u7136\u540e\u5728etcd\u7684":25,"\u7248\u672c\u5206\u652f":30,"\u7248\u672c\u53f7":30,"\u7248\u672c\u53f7rc":30,"\u7248\u672cfork\u51fa\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f":30,"\u73b0\u9636\u6bb5paddle\u6709\u4e00\u4e2a\u95ee\u9898\u662f":27,"\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u8bad\u7ec3\u6570\u636e\u96c6\u901a\u5e38\u4f53\u79ef\u5f88\u5927":26,"\u751f\u4ea7\u73af\u5883\u7684\u65e5\u5fd7\u6570\u636e\u4f1a\u901a\u8fc7\u5b9e\u65f6\u6d41\u7684\u65b9\u5f0f":26,"\u751f\u6210\u5404\u79cd\u8bed\u8a00\u7684\u7ed1\u5b9a\u4ee3\u7801":27,"\u751f\u6210\u6587\u6863":27,"\u751f\u6210\u7684":26,"\u751f\u6210\u7ed9\u5b9asstable\u6587\u4ef6\u5bf9\u5e94\u7684data":26,"\u751f\u6210api\u6587\u6863":27,"\u7528\u4e8e\u4e5f\u53ef\u4ee5\u4e0a\u4f20label":26,"\u7528\u6237\u53ef\u4ee5\u4e0a\u4f20\u81ea\u5df1\u7684\u6570\u636e\u96c6":26,"\u7528\u6237\u53ef\u4ee5\u5b89\u5168\u7684\u91ca\u653e\u67d0\u4e2ac":28,"\u7528\u6237\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u8fd9\u4e2a\u52a8\u6001\u5e93\u6765\u5f15\u5165paddl":28,"\u7528\u6237\u53ef\u4ee5\u7f16\u5199\u7a0b\u5e8f\u5b8c\u6210\u8f6c\u6362":26,"\u7528\u6237\u901a\u8fc7c":28,"\u7531\u4e8e\u5bf9parameters\u7684\u66f4\u65b0\u9700\u8981\u83b7\u53d6parameters\u5185\u5b58\u7684":25,"\u7531\u4e8e\u96c6\u7fa4\u4e2d\u540c\u65f6\u5b58\u5728\u4e24\u53f0\u673a\u5668\u6545\u969c\u7684\u6982\u7387\u6781\u4f4e":25,"\u7531\u4e8ec":27,"\u7684\u547d\u540d\u98ce\u683c\u5e76\u4e0d\u80fd\u9002\u5e94\u5176\u4ed6\u7b2c\u4e09\u65b9\u8bed\u8a00":27,"\u7684\u5934\u6587\u4ef6":27,"\u7684\u63a5\u53e3\u6837\u5f0f":27,"\u7684\u6570\u636e\u6d41\u56fe":26,"\u7684\u6e90\u7801\u91cc\u4f7f\u7528\u4e86":27,"\u7684\u89c4\u8303":27,"\u76ee\u524d\u53ea\u8003\u8651\u52a8\u6001\u6269\u5bb9trainer\u6570\u91cf":25,"\u76ee\u524d\u5d4c\u5165python\u89e3\u91ca\u5668":27,"\u76ee\u524dpaddle\u7684\u8fdb\u7a0b\u6a21\u578b\u662fc":27,"\u76ee\u5f55\u4e0b":28,"\u76f4\u63a5\u4f7f\u7528c\u8bed\u8a00\u7684":27,"\u76f4\u63a5\u5220\u9664\u8fd9\u4e2a\u53c2\u6570\u5373\u53ef":28,"\u76f4\u63a5\u5bfc\u51fa\u5230c\u7684\u63a5\u53e3\u6bd4\u8f83\u56f0\u96be":27,"\u793e\u533a\u53c2\u4e0e\u56f0\u96be":27,"\u793e\u533a\u8d21\u732e\u4ee3\u7801\u5b66\u4e60\u6210\u672c\u9ad8":27,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u7684\u53c2\u6570":25,"\u79bb\u7ebf\u6279\u5904\u7406":26,"\u7b2c\u4e00\u4e2atag\u4e3a":30,"\u7b2c\u4e09\u6b65\u5b8c\u6210\u540e":30,"\u7b2c\u4e8c\u4e2a\u4e3a":30,"\u7b49":28,"\u7b49\u5168\u90e8\u9759\u6001\u5e93\u4e2d\u7684\u76ee\u6807\u6587\u4ef6\u5168\u90e8\u6253\u5305\u540e\u4ea7\u751f\u7684\u6587\u4ef6":28,"\u7b49\u6587\u4ef6":28,"\u7c7b\u4f3c":28,"\u7c7b\u540d\u548cc":27,"\u7c7b\u578b":27,"\u7ea2\u697c\u68a6":53,"\u7ed3\u8bba":27,"\u7f16\u8bd1\u5668\u6ca1\u6709":27,"\u7f16\u8bd1\u578b\u8bed\u8a00":27,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684docker\u53d1\u884c\u955c\u50cf":30,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684ubuntu":30,"\u7f16\u8bd1c":28,"\u7f16\u8bd1master\u5206\u652f\u7684docker\u53d1\u884c\u955c\u50cf":30,"\u7f16\u8bd1ubuntu\u7684deb\u5305":30,"\u800c\u4e0d\u5fc5\u5728\u610fpaddl":28,"\u800c\u4e0d\u652f\u6301pypy\u89e3\u91ca\u5668":27,"\u800c\u4e0d\u66b4\u9732\u6982\u5ff5\u7684\u5b9e\u73b0":28,"\u800c\u51fa\u73b0\u9636\u6bb5\u6027\u7684\u8fd0\u884c\u505c\u6ede":25,"\u800c\u5728cpp\u91cc\u9762\u5b9e\u73b0\u8fd9\u4e2ac\u7684\u63a5\u53e3":27,"\u800c\u591a\u8bed\u8a00\u63a5\u53e3\u9700\u8981\u76f4\u63a5\u8bfb\u53d6\u751f\u6210\u7684\u4e8c\u8fdb\u5236":27,"\u800c\u5bf9\u4e8egolang":27,"\u800c\u5bf9\u4e8egolang\u9519\u8bef\u5904\u7406\u5e94\u8be5\u4f7f\u7528\u8fd4\u56de\u503c":27,"\u800c\u662f\u76f4\u63a5\u4fee\u6539paddl":28,"\u800cswig\u53ea\u80fd\u7b80\u5355\u7684\u66b4\u9732c":27,"\u81ea\u52a8\u6302\u8f7d\u5206\u5e03\u5f0f\u5b58\u50a8\u76ee\u5f55":25,"\u81f3\u4e8e\u4e3a\u4ec0\u4e48\u9700\u8981c":28,"\u826f\u597d\u7684\u6587\u6863":27,"\u83b7\u53d6\u6700\u65b0\u7684\u68c0\u67e5\u70b9\u7684\u6587\u4ef6uuid":25,"\u867d\u7136\u4e0d\u9f13\u52b1\u8fd9\u6837":28,"\u89e3\u91ca\u578b\u8bed\u8a00\u53ea\u80fd\u8c03\u7528\u52a8\u6001\u5e93":27,"\u89e3\u91ca\u6027\u8bed\u8a00\u5b9e\u9645\u8fd0\u884c\u7684\u4e8c\u8fdb\u5236\u662f\u89e3\u91ca\u5668\u672c\u8eab":27,"\u8ba1\u7b97\u8fd9\u4e2a\u6587\u4ef6\u7684md5":25,"\u8ba9paddle\u6838\u5fc3\u4e2d":28,"\u8bad\u7ec3\u4efb\u52a1\u7684\u8fd0\u884c\u53ef\u80fd\u4f1a\u5360\u6ee1trainer\u548cparamet":25,"\u8bad\u7ec3\u548c\u7eaf\u4f7f\u7528":30,"\u8bad\u7ec3\u6a21\u578b\u6b63\u786e\u6027":30,"\u8bb0\u5f55\u4e0b\u6240\u6709\u5931\u8d25\u7684\u4f8b\u5b50":30,"\u8bbe\u7f6e":28,"\u8bc6\u522b\u6570\u5b57":30,"\u8bcd\u5411\u91cf":30,"\u8bed\u610f\u89d2\u8272\u6807\u6ce8":30,"\u8bf4\u660e":25,"\u8bf7\u53c2\u8003":28,"\u8f6c\u6362\u751f\u6210\u7684\u6587\u4ef6\u540d\u4f1a\u662f\u4ee5\u4e0b\u683c\u5f0f":26,"\u8fbe\u5230\u5bb9\u707e\u7684\u76ee\u7684":25,"\u8fd4\u56de\u7b2c\u4e8c\u6b65":30,"\u8fd9\u4e00\u5c42\u8fdb\u884c\u5c01\u88c5":28,"\u8fd9\u4e00\u6982\u5ff5\u4e0d\u518d\u7410\u788e":28,"\u8fd9\u4e09\u4e2a\u5206\u652f":30,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u7684\u8fde\u63a5\u53c2\u6570\u4e0epaddle\u7684\u5176\u4ed6\u4e8c\u8fdb\u5236":28,"\u8fd9\u4e2a\u53c2\u6570\u4e5f\u4e0d\u4f1a\u4e00\u5e76\u5220\u9664":28,"\u8fd9\u4e2a\u5934\u6587\u4ef6\u4e0d\u5047\u8bbe\u5176\u4ed6\u6587\u4ef6\u7684\u5f15\u7528\u987a\u5e8f":28,"\u8fd9\u4e2a\u63a5\u53e3\u9700\u8981\u505a\u5230":27,"\u8fd9\u4e2a\u6587\u4ef6\u5177\u6709\u72ec\u7279\u7684\u8bed\u6cd5":27,"\u8fd9\u4e2a\u76ee\u5f55\u4e2d\u9664\u4e86":28,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u4e2d\u7684\u53e6\u4e00\u4e2a\u9879\u76ee\u662f":28,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u5305\u542b\u4e24\u4e2a\u9879\u76ee":28,"\u8fd9\u4e2a\u9759\u6001\u5e93\u5305\u542b\u4e86paddle\u7684\u5168\u90e8\u7b26\u53f7":28,"\u8fd9\u4e2ainstance\u53ef\u4ee5\u662f\u5355\u4e2a\u503c":26,"\u8fd9\u4e9b\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1\u901a\u5e38\u4f1a\u628a\u6570\u636e\u5207\u5272\u6210\u591a\u4e2a\u5206\u7247\u5206\u5e03\u5f0f\u7684\u5b58\u50a8\u5728\u591a\u4e2a\u8282\u70b9\u4e4b\u4e0a":26,"\u8fd9\u5bf9\u4e8e\u901a\u5e38\u7684java\u7684\u5f00\u53d1\u8005\u6765\u8bf4":27,"\u8fd9\u662f\u56e0\u4e3a":27,"\u8fd9\u6837":28,"\u8fd9\u6837\u4fdd\u8bc1":30,"\u8fd9\u6837\u5c31\u53ef\u4ee5\u5728\u4e91\u7aef\u6267\u884c\u591a\u79cd\u6570\u636e\u7c7b\u8ba1\u7b97\u4efb\u52a1":26,"\u8fd9\u90fd\u9700\u8981\u8fd9\u4e2a\u63a5\u53e3\u6309\u7167\u7ea6\u5b9a\u4fd7\u6210\u7684\u89c4\u5219\u6765\u6ce8\u91ca\u5b8c\u5907":27,"\u8fd9\u91cc\u9700\u8981\u7528\u6237\u989d\u5916\u6ce8\u610f":25,"\u8fdb\u800c\u8fdb\u884c\u4ee3\u7801\u8bc4\u5ba1":30,"\u9009\u62e9cephfs\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u670d\u52a1":26,"\u901a\u5e38":28,"\u901a\u5e38\u6307\u5c06\u4e00\u4e2a\u6574\u4f53\u62c6\u5206\u6210\u591a\u4efd\u7684\u5176\u4e2d\u7684\u4e00\u4efd":25,"\u901a\u8fc7\u6a21\u578b\u63a8\u65adapi\u7684\u5b9e\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u6837\u4f8b":28,"\u9075\u5faa\u4ee5\u4e0b\u6d41\u7a0b":30,"\u90a3\u4e48":28,"\u90fd\u662f\u4e94\u4f4d\u7684\u6570\u5b57":26,"\u90fd\u662fabi\u8c03\u7528\u6807\u51c6\u7684":27,"\u91ca\u653e\u5bf9paramters\u5185\u5b58\u7684\u9501\u5b9a":25,"\u91cc\u6240\u6709\u7684\u7b26\u53f7\u90fd\u5199\u5165\u81ea\u5df1\u7684\u7a0b\u5e8f\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cc":27,"\u91cd\u547d\u540d\u6210":27,"\u94fe\u63a5\u5230\u81ea\u5df1\u7684\u7a0b\u5e8f\u91cc":27,"\u9519\u8bef\u5904\u7406":27,"\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662f\u8fd4\u56de\u503c":27,"\u9519\u8bef\u5904\u7406\u7684\u65b9\u5f0f\u4e5f\u4e0d\u5c3d\u76f8\u540c":27,"\u9664\u6784\u9020\u67d0\u79cd\u7c7b\u578b\u7684\u51fd\u6570":28,"\u9700\u8981\u4fdd\u5b58\u4e00\u4e9b\u9884\u7f6e\u7684\u516c\u5f00\u6570\u636e\u96c6":26,"\u9700\u8981\u5728cmake\u7684\u65f6\u5019":28,"\u9700\u8981\u5c06bugfix\u7684\u5206\u652f\u540c\u65f6merge\u5230":30,"\u9700\u8981\u5f15\u7528":28,"\u9700\u8981\u6709\u7a33\u5b9a\u7684\u5bfc\u51fa\u7b26\u53f7":27,"\u9700\u8981\u6ce8\u610f\u7684\u662f":30,"\u9700\u8981\u88ab\u66b4\u9732\u5230\u5176\u4ed6\u8bed\u8a00":28,"\ufb01xed":63,"abstract":[40,45],"api\u4e2d\u4f7f\u7528":27,"api\u5bfc\u51fa\u7684\u52a8\u6001\u5e93":28,"api\u5bfc\u51fa\u7684\u9759\u6001\u5e93":28,"api\u63a5\u53d7\u7684\u7c7b\u578b\u5168\u662f":28,"api\u63a5\u53e3\u7684\u53c2\u6570\u8f6c\u53d1\u7ed9":28,"api\u65f6":28,"api\u65f6\u6240\u552f\u4e00\u9700\u8981\u5f15\u5165\u7684\u5934\u6587\u4ef6":28,"api\u662f\u591a\u8bed\u8a00api\u7684\u57fa\u7840\u90e8\u5206":28,"api\u66b4\u9732\u7684\u7c7b\u578b":28,"api\u751f\u6210\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u4f1a\u88ab\u5b89\u88c5\u5230":28,"api\u7684\u5b9e\u4f8b":28,"api\u7684\u5b9e\u73b0\u7ec6\u8282":28,"api\u7684\u63a5\u53e3":28,"api\u7684\u65f6\u5019\u63a8\u8350paddle\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":28,"api\u7684\u7f16\u8bd1\u9009\u9879\u9ed8\u8ba4\u5173\u95ed":28,"api\u76ee\u5f55\u7ed3\u6784\u5982\u4e0a\u56fe\u8868\u6240\u793a":28,"api\u83b7\u5f97\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u5b9e\u4f8b":28,"block\u6784\u6210\u4e00\u4e2amodel":25,"book\u4e2d\u6240\u6709\u7ae0\u8282\u529f\u80fd\u7684\u6b63\u786e\u6027":30,"boolean":[10,16,27],"break":58,"bugfix\u5206\u652f\u4e5f\u662f\u5728\u5f00\u53d1\u8005\u81ea\u5df1\u7684fork\u7248\u672c\u5e93\u7ef4\u62a4":30,"bugfix\u5206\u652f\u9700\u8981\u5206\u522b\u7ed9\u4e3b\u7248\u672c\u5e93\u7684":30,"c99\u662f\u76ee\u524dc\u6700\u5e7f\u6cdb\u7684\u4f7f\u7528\u6807\u51c6":27,"c\u6709\u6807\u51c6\u7684abi":27,"c\u8bed\u8a00\u662f\u6709\u5bfc\u51fa\u7b26\u53f7\u7684\u6807\u51c6\u7684":27,"case":[10,16,28,29,31,38,39,40,42,46,48,54,58],"char":60,"class":[5,7,10,12,14,15,16,17,18,19,20,22,23,27,44,55,62],"const":40,"core\u4e2d\u7684\u6a21\u578b\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u53c2\u6570":28,"core\u4e2d\u8fd9\u4e00\u7c7b\u578b\u63a5\u53e3\u7684\u667a\u80fd\u6307\u9488":28,"core\u662f\u5426\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u5b9e\u4f8b":28,"core\u6982\u5ff5":28,"data\u5230\u5206\u5e03\u5f0f\u5b58\u50a8\u8865\u5145\u8bad\u7ec3\u6570\u636e":26,"deb\u5305":30,"deb\u5305\u7f16\u8bd1\u95ee\u9898":30,"default":[3,7,9,10,11,12,15,16,17,19,20,22,23,33,43,45,47,48,49,58,60,62,63],"export":[32,55],"final":[11,17,31,32,40,60,62],"float":[3,7,9,10,12,15,16,18,20,31,40,42,47,53,56,60],"function":[3,5,8,10,11,12,16,17,18,20,23,24,29,31,38,40,42,43,45,54,55,58,61,62,63],"golang\u53ef\u4ee5\u4f7f\u7528":27,"golang\u7684":27,"h\u5e76\u4e0d\u56f0\u96be":27,"imagenet\u6570\u636e\u96c6\u7b49":26,"images\u6570\u636e\u96c6\u4e0a\u4f20\u5230\u4e91\u7aef\u7684":26,"import":[3,5,9,10,16,23,31,38,42,48,53,54,55,56,58,60,62,63],"instance\u4e0e\u751f\u6210\u6570\u636e\u96c6\u65f6":26,"instance\u5305\u6db5\u4e24\u4e2a\u503c":26,"instance\u662f\u4e00\u6a21\u4e00\u6837\u7684":26,"int":[3,7,9,10,11,12,15,16,17,20,27,28,29,40,47,58,60,61],"interface\u6587\u4ef6\u7684\u5199\u6cd5\u975e\u5e38":27,"job\u5b8c\u6210\u8f6c\u6362":26,"list\u4f5c\u4e3a\u68c0\u67e5\u5217\u8868":30,"long":[2,10,11,16,17,20,33,42,61,62],"model\u505a\u5206\u652f\u7ba1\u7406":30,"ndarray\u7c7b\u578b\u7684\u503c\u548c\u6574\u578b\u7684\u503c":26,"new":[3,10,16,20,24,29,39,41,48,49,54,58,61,62],"note\u7684\u4e66\u5199":30,"null":[10,16,40,45,60],"paddle\u4e00\u4e2a\u52a8\u6001\u5e93\u53ef\u4ee5\u5728\u4efb\u4f55linux\u7cfb\u7edf\u4e0a\u8fd0\u884c":27,"paddle\u4f7f\u7528git":30,"paddle\u5185\u5d4c\u7684python\u89e3\u91ca\u5668\u548c\u5916\u90e8\u4f7f\u7528\u7684python\u5982\u679c\u7248\u672c\u4e0d\u540c":27,"paddle\u5185\u90e8\u7684\u7c7b\u4e3ac":27,"paddle\u5f00\u53d1\u8fc7\u7a0b\u4f7f\u7528":30,"paddle\u6bcf\u6b21\u53d1\u65b0\u7684\u7248\u672c":30,"paddle\u6bcf\u6b21\u53d1\u7248\u672c\u9996\u5148\u8981\u4fdd\u8bc1paddl":30,"paddle\u7684\u4e3b\u7248\u672c\u5e93\u9075\u5faa":30,"paddle\u7684\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0\u5305\u62ec\u4e00\u4e0b\u51e0\u4e2a\u65b9\u9762":27,"paddle\u7684\u7c7b\u578b\u5168\u90e8\u9000\u5316\u6210":28,"paddle\u7684\u94fe\u63a5\u65b9\u5f0f\u6bd4\u8f83\u590d\u6742":27,"paddle\u7684c":28,"paddle\u8bad\u7ec3\u4efb\u52a1":26,"paddle\u8def\u5f84\u4e0b":28,"paddle\u9700\u8981\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3":27,"paddle\u9700\u8981\u66b4\u9732\u7684api\u5f88\u591a":28,"paddle\u9759\u6001\u5e93\u94fe\u63a5\u590d\u6742":27,"paddle_\u7c7b\u578b\u540d":28,"paddle_\u7c7b\u578b\u540d_\u51fd\u6570\u540d":28,"paddlepaddle\u63d0\u4f9b\u4e13\u7528\u7684":26,"patch\u53f7":30,"patch\u53f7\u52a0\u4e00":30,"public":[20,40,43,48,49,62],"reader\u7684\u4f7f\u7528\u65b9\u5f0f\u90fd\u662f\u4e00\u81f4\u7684":26,"reader\u8f93\u51fa\u7684data":26,"release\u9875\u9762":30,"return":[3,8,9,10,11,16,17,19,20,22,23,26,31,38,40,48,54,56,58,59,60,63],"s3\u4e4b\u7c7b\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u4e4b\u4e0a":26,"server\u4e4b\u4e0a":25,"server\u4e4b\u95f4\u7684\u7f51\u7edc\u5e26\u5bbd":25,"server\u4f1a\u6682\u505c\u53c2\u6570\u66f4\u65b0\u5e76\u7b49\u5f85":25,"server\u4f1a\u83b7\u53d6parameters\u5185\u5b58\u7684":25,"server\u5185\u5b58\u4e2d\u7684\u6a21\u578b\u6570\u636e\u7684\u5b8c\u6574\u955c\u50cf":25,"server\u540c\u6b65\u7684\u4fdd\u5b58\u4e00\u4e2a\u7279\u5b9a\u65f6\u95f4\u70b9\u7684\u5168\u5c40\u68c0\u67e5\u70b9":25,"server\u5728\u96c6\u7fa4\u4e2d\u542f\u52a8\u540e":25,"server\u6545\u969c\u540e\u88abkubernetes\u91cd\u65b0\u542f\u52a8":25,"server\u6b64\u65f6\u8fd8\u9700\u8981\u901a\u8fc7\u7f51\u7edc\u8bbf\u95ee\u5206\u5e03\u5f0f\u5b58\u50a8\u4ee5\u4fdd\u5b58\u5feb\u7167":25,"server\u751f\u6210\u4e00\u4e2auuid":25,"server\u7684\u5355\u70b9\u6216\u591a\u70b9\u540c\u65f6\u6545\u969c":25,"server\u7684\u6570\u636e\u5feb\u7167":25,"server\u7684\u68c0\u67e5\u70b9\u5404\u81ea\u72ec\u7acb\u4fdd\u5b58":25,"server\u7b2c\u4e00\u6b21\u542f\u52a8\u6216\u4efb\u610f\u65f6\u95f4paramet":25,"short":[10,11,16,17,31,60,61,62],"static":[10,28,48],"super":40,"swig\u652f\u6301\u7684\u8bed\u8a00\u6216\u8005\u89e3\u91ca\u5668\u6709\u5c40\u9650":27,"swig\u66b4\u9732\u7684\u63a5\u53e3\u4fdd\u7559\u4e86c":27,"swig\u751f\u6210\u7684\u4ee3\u7801\u4e0d\u80fd\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":27,"swig\u76f4\u63a5\u8bfb\u53d6c":27,"swig\u9700\u8981\u5199\u4e00\u4e2ainterface\u6587\u4ef6":27,"switch":[28,48,62],"tag\u4e3a":30,"throw":48,"true":[3,7,9,10,11,12,15,16,17,19,20,22,23,29,31,38,40,45,47,48,56,60,61,62,63],"try":[12,18,24,29,42,54,60],"type\u5b57\u6bb5\u5747\u4e0d\u5c3d\u76f8\u540c":28,"ubuntu\u5b89\u88c5\u5305\u7684\u529f\u80fd\u6b63\u786e\u6027":30,"void":[27,28,40],"while":[2,3,7,9,15,20,29,33,38,45,54,58,62,63],AGE:[48,49],AND:60,ARE:60,AWS:[26,41,50,51],Abs:6,Age:59,And:[3,9,10,12,16,18,20,29,35,39,47,48,49,53,56,60,62,63],But:[3,10,11,16,17],EOS:[10,16],For:[2,3,8,9,10,12,16,18,20,23,29,31,32,33,38,40,42,43,44,45,47,53,55,56,58,62,63],Going:62,Has:3,IDs:[20,58],Ids:58,Into:48,Its:[3,38,48,60],Not:[23,24,43],ONE:3,One:[9,10,11,17,22,38,40,45,54,58,62,63],QoS:49,THE:3,TLS:[23,48],That:[10,16,20,29,33,45,47],The:[2,3,5,7,8,9,10,11,12,14,15,16,17,18,20,22,23,24,28,29,31,32,33,34,35,38,39,40,42,43,45,47,48,49,53,54,55,56,58,59,60,61,62,63],Their:[3,10,16,24],Then:[5,10,16,32,33,38,39,40,42,48,49,53,55,60,61,62],There:[9,10,16,20,22,23,24,31,33,35,42,48,54,55,56,57,58,60,63],These:[43,47,55,61],USE:60,USING:60,Use:[3,23,29,33,40,42,45,46,48,60],Used:[11,17],Useful:3,Using:[24,49,62],VPS:48,WITH:39,Will:[20,22],With:[3,10,11,16,17,31,54,61],Yes:33,___fc_layer_0__:48,__init__:40,__list_to_map__:60,__main__:56,__meta__:60,__name__:56,__rnn_step__:38,_error:54,_link:[11,17],_proj:[10,16],_res2_1_branch1_bn:56,_source_language_embed:[38,53],_target_language_embed:[38,53],aaaaa:26,aaaaaaaaaaaaa:48,abc:[10,16],abl:[10,16,23,54,62],about:[5,10,11,16,17,31,42,44,45,48,52,61,62,63],abov:[3,5,10,16,23,24,31,33,42,48,49,54,56,58,61],abs:[11,17,54],absolut:[2,43],academ:59,acceler:[25,47],accept:[3,5,20,23,29,58,61],acceptor:61,access:[2,10,11,16,17,23,38,63],accessmod:48,accident:59,accomplish:33,accord:[2,3,9,10,16,38,39,43,44,45,47],accordingli:[5,40],accordingto:61,accrod:[11,17],accumul:24,accuraci:[9,40,58,59,62],achiev:[42,55],ack:45,acl:62,aclimdb:62,across:[10,16],act:[10,11,16,17,31,38,58],act_typ:58,action:[48,59],activ:[0,4,5,10,11,16,17,21,31,32,40,45,58,62],activi:[11,17],actual:[3,10,16,31],adadelta:[12,58],adagrad:[12,58],adam:[12,23,58,62,63],adamax:[12,58],adamoptim:[53,58,62,63],adapt:[9,12,18,31,62,63],add:[3,10,11,16,17,20,31,32,39,40,42,47,58,60],add_input:40,add_test:40,add_to:[10,16],add_unittest_without_exec:40,addbia:40,added:[3,9,40],adding:56,addit:[10,11,16,17,33,58],address:[24,33,42,45],addrow:40,addtion:43,addto:10,addtolay:[10,16],adject:62,adjust:31,admin:59,adopt:61,advanc:[38,42,45],advantag:62,adventur:59,adverb:62,adversari:29,advic:42,affect:[10,16],afi:3,aforement:43,after:[10,16,20,32,35,38,40,43,45,47,48,49,54,55,56,58,60,61,62,63],again:[23,24,42],against:48,age:[20,60],agg_level:[10,16],aggreg:48,aggregatelevel:[10,16],aid:42,aim:[62,63],aircraft:63,airplan:55,aistat:[10,16],alex:[10,16,62],alexnet_pass1:47,alexnet_pass2:47,algorithm:[10,12,16,18,31,38,53,55,62,63],alia:[6,7,13,14,15],align:[10,11,16,17,20,63],all:[0,3,7,9,10,12,15,16,18,23,24,28,31,33,38,39,40,42,43,44,45,47,48,49,53,54,56,58,59,60,61,62,63],alloc:[7,15,40,47],allow:[23,33,39,40,42,45,48,58],allow_only_one_model_on_one_gpu:[44,45,47],almost:[11,17,31,43,53],along:62,alreadi:[24,33,42,43,45,48,49,62],alreali:[44,63],also:[2,3,9,10,11,16,17,20,23,29,32,33,38,40,42,43,49,54,55,56,58,61,62],although:31,alwai:[5,10,11,16,17,22,29,31,45,48,63],amaz:55,amazon:[48,49,58,62],amazonaw:48,amazonec2fullaccess:48,amazonelasticfilesystemfullaccess:48,amazonroute53domainsfullaccess:48,amazonroute53fullaccess:48,amazons3fullaccess:48,amazonvpcfullaccess:48,ambigu:[29,61],amd64:48,amend:39,american:55,among:[48,62],amount:[42,62],analysi:[31,42,57,61],analyz:[58,62],andd:48,ani:[2,3,10,11,16,17,20,23,24,29,38,39,42,48,58,60,63],anim:59,annot:61,annual:61,anoth:[3,10,16,23,33,45,48,61,62],ans:48,answer:[31,48,61],anyth:[20,29,39,48,61],api:[20,22,23,30,32,40,42,48,52,54,58,60,62],apiserv:48,apivers:[48,49],apo:63,appar:63,appear:61,append:[3,22,29,38,40,43,60],append_gradient_machin:22,appleclang:32,appleyard:42,appli:[0,10,11,16,17,38,40,55,58],applic:[33,42,48,49,62],appreci:[39,62],approach:[10,16],apt:[32,35,55],arbitrari:[10,16],architectur:[53,61,62,63],architecur:62,archiv:[20,27,28],arg:[3,8,9,10,11,12,16,17,20,31,44,54,55,56,58,60,61,62],arg_nam:[10,16],argu:61,argument:[3,5,8,10,16,20,38,40,45,46,53,54,55,56,60,61,62,63],argv:56,arn:48,around:[3,10,16,48],arrai:[5,10,16,20,22,29,31,56],art:[31,61],articl:[43,49],artifact:48,artifici:54,artist:59,arxiv:[10,11,16,17,54,62],ask:24,aspect:62,assign:[10,16,45,48],associ:[61,62,63],assum:[10,16,38,47,53],assur:2,astyp:[29,54],asyc:24,async:[12,24,44],async_count:45,async_lagged_grad_discard_ratio:45,async_lagged_ratio_default:[44,45],async_lagged_ratio_min:[44,45],asynchron:[24,45],atla:32,atlas_root:32,attenion:[11,17],attent:[10,11,16,17,33,63],attitud:62,attr:[7,11,15,16,17],attribut:[3,4,10,11,16,17,21,40,53,61],auc:[9,44],aucvalidationlay:45,authent:48,author:[48,56],authorized_kei:43,autmot:39,auto:[27,40,42,57,60],autom:[48,63],automak:32,automat:[10,16,23,32,38,40,43,44,45,48,60,61,63],automaticli:[10,16],automobil:55,avail:[24,32,48],availabel:32,averag:[9,10,12,16,19,45,56,58,60,61,62,63],average_test_period:[44,45,61],average_window:62,averagepool:[10,16],avg:[13,42,58],avgcost:[9,58,60,62,63],avgpool:[10,16,58],avoid:[24,42],avx:[32,33,35],await:49,awar:[23,24,33,48],aws_account_id:48,awsaccountid:48,awskeymanagementservicepowerus:48,b2t:53,b363:49,b8561f5c79193550d64fa47418a9e67ebdd71546186e840f88de5026b8097465:49,ba5f:48,back:[3,24,33],backward:[10,11,14,16,17,38,40,45,47],backward_first:38,backwardactiv:40,bag:[58,62],baidu:[0,10,16,31,35,39,49,53],baik:53,balanc:[45,48,54],balasubramanyan:62,bank:61,bardward:[11,17],bare:49,barrier:45,barrierstatset:42,base:[6,12,16,17,19,20,23,31,35,38,39,40,42,43,45,48,53,54,58,60,62,63],baseactiv:[10,11,16],baseev:22,basematrix:40,basenam:9,basepool:13,basepoolingtyp:[10,11,16,17],baseregular:12,basestr:[7,8,9,10,11,15,16,17,19,22,60],bash:[33,48,49],bashrc:32,basic:[3,10,16,22,33,39,40,58,59,62],batch:[3,9,10,11,12,16,17,18,20,22,23,24,26,40,43,45,48,49,54,55,56,58,60,61,62,63],batch_0:56,batch_id:22,batch_norm:[10,17],batch_norm_lay:11,batch_norm_typ:[10,16],batch_read:[26,29],batch_siz:[3,12,20,22,31,43,53,54,55,58,60,62,63],batchsiz:[10,16,40],bbbbb:26,bcd:[10,16],bcebo:20,beam:[10,16,38,45,61,63],beam_gen:[10,16,38],beam_search:[22,38],beam_siz:[10,16,38,44,45,47],beamsiz:63,becaus:[5,10,16,20,23,24,29,38,39,40,47,48,55,58,61],becom:[39,42],been:[3,32,39,55,58,61,62,63],befor:[5,10,11,16,17,24,29,33,39,43,48,55,60,62,63],begin:[5,9,10,16,40],beginiter:[22,23],beginn:38,beginpass:[22,23],begintrain:23,behavior:42,being:[24,29,54],belong:[10,16,63],below:[3,10,16,20,24,29,38,40,42,43,48,54,55,58,60],benefit:[11,17],bengio:[10,16],bertolami:62,besid:[2,10,16,20,63],best:[8,10,16,32,33,45,58,60,62,63],best_model_path:61,besteffort:49,beta1:[12,18],beta2:[12,18],beta:56,better:[10,11,16,17,31,43,48,54,60],between:[10,12,16,18,24,28,31,39,48,54,58,59,62,63],bgr:56,bi_lstm:[11,17],bia:[10,11,12,16,17,18,38,40,56],bias:[10,16,40],bias_attr:[10,11,16,17,31,38],bias_param_attr:[11,17],biases_:40,biasparameter_:40,biassiz:40,bidi:49,bidirect:[11,17,38,61,63],bidirectional_lstm_net:62,big:42,bigger:24,biggest:62,bilinear:[10,16],bilinear_interpol:[10,16],bilinearfwdbwd:42,bin:[32,33,43,48,49,60],binari:[3,9,10,16,20,42,48,53,58,62],bird:55,bison:32,bit:58,bitext:63,bla:32,blank:[10,16,48],block:[10,16,25,31,40,42,45,56,62],block_expand:10,block_i:[10,16],block_x:[10,16],blog:62,bn_attr:17,bn_bias_attr:[11,17],bn_layer_attr:11,bn_param_attr:[11,17],bollen:62,book:20,bool:[3,7,9,10,11,12,15,16,17,19,20,40,45,47,58,60,62],boot:[10,16,38],boot_bia:[10,16],boot_bias_active_typ:[10,16],boot_lay:[10,16,38],boot_with_const_id:[10,16],bootstrap:32,bos_id:[10,16,38],both:[0,7,10,11,14,15,16,17,23,24,33,38,40,42,48,54,56,58],bottleneck:[42,56],bottom:62,bow:[26,58,62],box:42,branch:[10,16,23,30,39],breadth:[45,63],brelu:6,brendan:62,brew:32,briefli:42,broadcast:24,brows:33,browser:[33,48],bryan:62,bucket_nam:48,buf_siz:20,buffer:[3,20,29,45],buffered_read:29,bug:48,bui:62,build:[0,20,31,33,36,45,48,50,51,53,55,56,58,60,62,63],build_dict:20,built:[0,32,33,54,61],bunch:[42,58],bunk:62,button:[39,48],c11:27,c99:28,c99e:48,cach:[58,60,61],cache_pass_in_mem:[3,58,60,61],cachetyp:[3,58,60,61],calc_batch_s:[3,61],calcul:[3,9,10,11,12,16,17,18,24,38,40,42,45,47,54,60],call:[3,10,11,16,17,23,31,33,38,40,42,45,48,55,56,58,62,63],callabl:[3,10,16,20],callback:40,caller:48,caltech:55,can:[2,3,5,7,8,9,10,11,15,16,17,20,23,24,29,31,32,33,35,38,39,40,42,43,44,45,47,48,49,53,54,55,56,58,60,61,62,63],can_over_batch_s:[3,61],candid:[10,16],cannot:40,caoi:63,capabl:[32,62],capac:48,capi:27,capi_prvi:28,caption:[31,63],captur:[31,43],card:43,care:[11,17,29,44,45,59],carefulli:[43,45,56],cat:[33,55,56,62],categor:61,categori:[10,16,20,24,58,62],categorig:20,categoryfil:49,caus:24,caution:[48,49],ccb2_pc30:63,cde:[10,16],cdn:20,ceil:[10,16],ceil_mod:[10,16],cell:[10,11,16,17,62],center:3,ceph:[26,49],certain:[2,44,61],certif:[23,48],cffi:27,cfg:49,cgo:27,chain:[20,40],challeng:24,chanc:[23,40,58],chang:[10,16,20,29,31,33,38,39,40,42,45,48,58,62],channel:[10,16,42,43,56],channl:[43,56],char_bas:60,charact:[58,60],character:31,characterist:[47,55],check:[3,20,31,32,33,39,45,47,48,59],check_align:20,check_eq:40,check_fail_continu:3,check_l:40,check_sparse_distribution_batch:[44,45],check_sparse_distribution_in_pserv:[44,45],check_sparse_distribution_ratio:[44,45],check_sparse_distribution_unbalance_degre:[44,45],checkgrad:45,checkgrad_ep:45,checkout:39,children:59,china:33,chines:57,chmod:[32,48],choic:[33,59],choos:[45,58,60],chosen:[2,59,63],chunk:[9,54,61],chunk_schem:9,chunktyp:9,cifar:[54,55],cifar_vgg_model:55,claim:48,claimnam:48,clang:[27,32,33,39],class1:62,class2:62,class_dim:62,classfic:[56,62],classfiic:55,classic:[10,16,31],classif:[3,5,10,16,47,56,57,58,62,63],classifc:62,classifi:[9,54,55,56,58,62],classification_cost:[55,58],classification_error_evalu:[54,58,62,63],classification_threshold:9,claster:48,clean:[5,60],cleric:59,cli:48,click:[39,42,48],client:39,clip:[7,12,15,45,58,62],clock:[10,16],clone:[32,33],close:[3,29],closer:31,cloud:24,cls:58,cludform:48,cluster:[23,24,44,45,49,58,63],cluster_train:43,cm469:48,cmake3:32,cmake:[28,32,40,42],cmakelist:40,cmatrix:[27,28],cmd:49,cna:[10,16],cname:48,cnn:[10,16,49,56,58],code:[0,3,5,20,23,29,31,32,33,34,38,40,41,42,43,48,49,54,58,59],coeff:[10,16],coeffici:[10,16],collabor:24,collect:[10,16,20,22,31,59],collectbia:40,colleg:59,color:[55,56],colour:20,column:[9,10,16,29,40,53,63],colunm:63,com:[10,11,16,17,20,32,33,35,39,48,49,56],combin:[10,11,16,17,20,22,54,60,62],come:62,comedi:59,comma:[45,53],command:[2,5,31,32,33,35,39,40,41,42,43,48,49,50,51,53,54,55,56,60,61,62],commandlin:[42,62],commenc:58,comment:[11,17,39,58,62],commnun:43,common:[38,40,44],common_util:[43,60],commonli:[38,42,47],commun:[0,24,40,43,48],compani:62,compar:[40,54,58],compat:3,compet:62,competit:54,compil:[32,33,39,40],complet:[0,5,10,11,16,17,20,22,24,40,48,49,58],complex:[2,3,11,17,29,38,42,58],complic:[10,16],compon:40,compos:[20,23,54,61],composenotalign:20,comput:[10,11,16,17,23,24,31,32,33,38,40,42,47,48,58,60,61,62],computation:38,conat:16,conat_lay:10,concat:[10,63],concat_lay:38,concaten:[11,17],concept:[3,23,33,38],concern:23,concurr:24,concurrentremoteparameterupdat:45,condit:[10,16,38,43,49,63],conduct:42,conf:[5,10,16,43,53,54,56,63],conf_paddle_gradient_num:48,conf_paddle_n:48,conf_paddle_port:48,conf_paddle_ports_num:48,conf_paddle_ports_num_spars:48,confid:62,config:[3,7,10,11,15,16,17,31,40,43,44,45,48,49,53,54,55,56,58,62,63],config_:45,config_arg:[44,45,47,56,58,61,62],config_bas:[16,17,22],config_fil:61,config_gener:[43,60],config_lay:40,config_pars:[5,40],configur:[1,2,3,5,8,10,16,31,39,40,42,45,53,55,56,62,63],conflict:39,confront:63,congest:45,conll05st:61,conll:[20,61],connect:[2,11,17,31,40,48,49,54,55,56,58,60,62],connectionist:[10,16,62],connor:62,consequ:[10,11,16,17],consid:[9,10,12,16,18,32,33,42,47,55],consider:[3,11,17],consist:[10,16,20,29,55,56,58,61,63],consol:[42,48],constant:40,construct:[3,5,23,38,60],construct_featur:60,constructor:40,consum:[24,62],contact:24,contain:[3,8,9,10,11,16,17,19,20,22,23,34,35,38,39,43,48,55,56,58,59,62,63],containerport:48,contemporan:62,content:[49,61,62],context:[10,11,16,17,20,38,53,58,60,61,62,63],context_attr:[11,17],context_len:[10,11,16,17,58,60],context_proj_layer_nam:11,context_proj_nam:17,context_proj_param_attr:[11,17],context_project:[11,17,60],context_start:[10,11,16,17,58],contibut:39,contin:48,continu:[3,24,35,45],contrast:[10,16,63],contribut:[0,34,41,62],contributor:0,control:[7,15,33,45,48,49,63],conv:[11,17],conv_act:[11,17],conv_attr:17,conv_batchnorm_drop_r:[11,17],conv_bias_attr:[11,17],conv_filter_s:[11,17],conv_layer_attr:11,conv_num_filt:[11,17],conv_op:[10,16],conv_pad:[11,17],conv_param_attr:[11,17],conv_shift:10,conv_strid:[11,17],conv_with_batchnorm:[11,17],conveni:[23,43],converg:[43,54,62],convert:[3,5,20,26,29,38,53,55,56,58,60],convlay:[10,16],convolut:[10,11,16,17,54,56,60],convoper:[10,16],convtran:[10,16],convtranslay:[10,16],cool:[3,39],coordin:24,copi:[22,23,33,48,54,60],copy_shared_paramet:54,copytonumpymat:54,core:[3,7,15,28,45,63],coreo:48,corespond:61,corpora:63,corpu:[20,61],correct:[3,9,10,16,40,48],correctli:[9,20,40,54],correl:[31,55,62],correspoind:23,correspond:[3,5,23,31,38,40,55,59,61,62,63],corss_entropi:23,cos:[10,16],cos_sim:60,cosin:[10,16,60],cost:[5,12,18,22,23,31,45,54,58,60,62,63],cost_id:[10,16],could:[3,5,9,10,16,20,22,23,29,33,42,43,48,58,60],count:[24,29,42,45,47,49,53,60,61,62,63],counter:24,coupl:31,coverag:32,coveral:32,coveralls_uploadpackag:32,cpickl:[56,60],cpp:[27,28,39,40,42,58,60,63],cpu:[2,3,7,10,15,16,30,32,35,42,45,49,54,61,62,63],cpuinfo:33,cpusparsematrix:28,craftsman:59,crash:[24,42,43,45],crazi:43,creat:[5,7,10,15,16,20,22,23,24,31,32,33,40,43,45,53,54,55,63],create_bias_paramet:40,create_input_paramet:40,createargu:54,createfromconfigproto:[5,54],createstack:48,creation:48,creationd:48,creator:[20,26],credit:54,crf:[10,61],crf_decod:10,crime:59,critic:62,crop:56,crop_siz:56,cross:[10,16,58,61],cross_entropi:[16,23,54],cross_entropy_with_selfnorm:16,csc:40,cslm:63,csr:40,csv:59,ctc:10,ctc_layer:9,ctest:33,ctrl:[43,60],ctx:61,ctx_0:61,ctx_0_slot:61,ctx_n1:61,ctx_n1_slot:61,ctx_n2:61,ctx_n2_slot:61,ctx_p1:61,ctx_p1_slot:61,ctx_p2:61,ctx_p2_slot:61,cub:55,cuda:[32,33,35,42,43,45],cuda_dir:[44,45],cudaconfigurecal:42,cudadevicegetattribut:42,cudaeventcr:42,cudaeventcreatewithflag:42,cudafre:42,cudagetdevic:42,cudagetdevicecount:42,cudagetdeviceproperti:42,cudagetlasterror:42,cudahostalloc:42,cudalaunch:42,cudamalloc:42,cudamemcpi:42,cudaprofilerstart:42,cudaprofilerstop:42,cudaruntimegetvers:42,cudasetdevic:42,cudasetupargu:42,cudastreamcr:42,cudastreamcreatewithflag:42,cudastreamsynchron:42,cudeviceget:42,cudevicegetattribut:42,cudevicegetcount:42,cudevicegetnam:42,cudevicetotalmem:42,cudnn:[10,16,19,32,35,45],cudnn_batch_norm:[10,16],cudnn_conv:[10,16],cudnn_conv_workspace_limit_in_mb:[44,45],cudnn_convt:[10,16],cudnn_dir:[44,45],cudrivergetvers:42,cuinit:42,cumul:[10,16],curl:[32,48],current:[3,10,12,16,24,31,33,38,39,40,43,45,48,58,62,63],current_word:38,currentcost:[9,58,60,62,63],currentev:[9,58,60,62,63],curv:[23,55,61],custom:[2,3,23,40,48,59,62],custom_batch_read:29,cutoff:20,cycl:24,cyclic:[10,16],cython:27,d3e0:48,dai:63,daili:62,dalla:3,dan:61,danger:3,darwin:48,dat:[20,43,60],data:[1,2,3,5,8,11,12,17,18,22,23,26,32,33,40,42,43,44,45,47,50,56,59],data_batch_gen:54,data_dir:[53,55,62,63],data_feed:20,data_fil:31,data_initialz:58,data_lay:[3,9,31,38,54,55,58,60,61],data_nam:20,data_provid:8,data_read:[20,29],data_reader_creator_random_imag:29,data_sourc:[8,54],data_typ:[16,20],databas:[20,62],datadim:[10,16],datalay:[10,16],dataprovid:[2,8,31,38,43,60,61],dataprovider_bow:58,dataprovider_emb:58,dataproviderconvert:5,datasci:[10,16],dataset:[1,3,26,29,31,45,53,55,56,58,61,62],datasourc:[4,60],date:61,db_lstm:61,dcgan:54,dcmake_install_prefix:32,dead:24,deal:[39,54],deb:[34,35],debian:[33,34],debug:3,decai:[12,18,55],decid:[23,29],declar:[10,11,16,60],decod:[10,11,16,17,38,61,63],decoder_boot:38,decoder_group_nam:38,decoder_input:38,decoder_mem:38,decoder_prev:[11,17],decoder_s:38,decoder_st:[11,17,38],deconv:[10,16],deconvolut:[10,16],decor:[3,20,40],decreas:31,decrypt:48,deep:[0,10,16,31,33,36,42,54,55,56,58,61],deeper:[31,33,56],deer:55,def:[3,10,16,20,23,26,29,31,38,40,54,56,58,60,61],defalut:[10,16,45,47],default_devic:47,default_valu:47,defferenct:3,defin:[2,3,8,9,10,11,16,17,20,23,24,29,31,38,40,43,45,53,54,55,60,61],define_py_data_sources2:[3,8,31,55,56,58,60],defini:63,definit:[3,20,24,31,33,53,58,62],degre:[10,16],del:60,delai:45,delar:58,delet:24,deletestack:48,delimit:[9,59,60],demand:24,demo:[10,16,20,38,43,49,50,53,54,55,56,57,58,59,60,61,62,63],demograph:59,demolish:49,demonstr:[31,38,54,60],denot:[47,58,59,61],dens:[3,10,16,20,40,48,58,60],dense_vector:[3,5,16,20,31,60],dense_vector_sequ:20,depend:[24,31,33,35,43,47,55,59],deploi:[43,47],deploy:[43,48],deriv:[14,23],descent:[10,12,16,24],describ:[23,31,40,48,49,54,58,61],describestack:48,describestackev:48,describestackresourc:48,descript:[5,32,38,46,48,55,60],deseri:22,design:[3,10,16,20,27,62],desir:[24,48,49,53],destructor:40,detail:[3,5,7,10,11,12,15,16,17,18,38,39,40,42,43,46,47,48,49,53,54,56,58,60,62,63],detect:9,determin:[3,10,16,20,40,54],dev:[32,33,55,60,63],devel:32,develop:[0,30,32,39,44,45,63],deverlop:45,deviat:[7,15],devic:[7,15,45,63],deviceid:47,devid:[10,16,45],dez:62,dfs:11,diagnos:43,diagram:56,dict:[3,8,20,22,58,60,62,63],dict_dim:62,dict_fil:[9,38,58,61],dict_nam:8,dict_siz:20,dictionai:58,dictionari:[3,8,9,10,16,20,22,23,38,47,56,58,60,61,62,63],dictsiz:63,did:3,differ:[3,8,9,10,16,24,31,33,38,39,40,43,45,48,49,53,55,56,58,62,63],difficult:31,dig:[33,42,48],digit:[3,10,16],dim:[20,40,53,56,58,62],dimens:[10,14,16,19,20,40,47,53,58,60,62],dimension:[3,31,38,40,54,58],dimenst:53,dimes:[10,16],din:60,dir:[43,56,58,60,61,62,63],dirctori:33,direct:[10,11,16,17,33,56,61],directli:[2,3,11,17,31,33,43,49,62],directori:[2,26,32,33,39,42,43,45,49,55,56,58,60,61,62,63],diretcoti:56,dis_conf:54,dis_train:54,dis_training_machin:54,disabl:3,discard:[20,24,45],discount:[10,16],discov:[24,61],discoveri:48,discrep:42,discrimin:54,discriminator_train:54,discuss:23,disk:49,dispatch:[43,45],disput:63,dist_train:23,distanc:9,distibut:53,distinguish:[43,54,63],distribut:[10,16,32,41,49,50,51,54,58,61],distribute_test:[44,45],distributedli:40,disucss:23,divid:[12,18,44,55,63],diy_beam_search_prob_so:[44,45],dmkl_root:32,dns:48,do_forward_backward:29,doc:[5,11,17,20,32,33,43],docker:[30,34,48,50,51],docker_build:23,docker_push:23,dockerfil:33,dockerhub:33,doctor:59,document:[3,5,11,17,32,39,47,55,58,60,61,62],documentari:[3,59],doe:[3,5,11,17,24,29,31,35,38,40,42,58,60,61],doesn:[7,10,15,16,20,23,29,39,42,49,63],dog:[55,56],doing:42,domain:48,don:[11,17,23,29,31,48,62],done:[10,11,16,17,24,38,42,48,54,62],dopenblas_root:32,dot:[45,56,63],dot_period:[45,47,54,55,60,62,63],dotmuloper:[10,16],dotmulproject:[10,16],doubl:[3,32,45],down:[42,58],download:[20,24,33,35,54,55,58,61,62],download_cifar:55,downsampl:55,doxygen:[32,39],dpkg:35,drama:59,drop:3,drop_rat:[7,15],dropout:[7,10,15,16,40,58],dropout_lay:10,dropout_r:[11,17],drwxr:49,dtoh:42,dtype:[5,31,56],dubai:63,due:[59,60],duplic:59,durat:42,dure:[2,3,10,16,24,31,39,40,44,45,48,58,60,61,63],durn:3,dwith_c_api:28,dwith_doc:32,dwith_profil:42,dwith_python:28,dwith_swig_pi:28,dwith_tim:42,dynam:[2,3,28,29,32,42,45],dynamic_cast:40,each:[2,3,5,9,10,16,19,20,22,24,29,31,33,38,39,40,43,45,47,48,53,55,56,58,59,60,61,62,63],each_feature_vector:14,each_meta:60,each_pixel_str:3,each_sequ:[10,16],each_time_step_output:14,each_timestep:[10,16],each_word:3,eaqual:[10,16],eas:[20,29,56],easi:[0,29,33,40,43,58],easier:[23,29,40],easili:[23,29,31],echo:[33,60,62],edit:[9,33,48],editor:[33,39],edu:[20,48,49,55],educ:59,eeoi3ezpr86c:48,effect:[3,45,48],effici:[0,2,3,38,40],efg:[10,16],efs:48,efs_dns_nam:48,efsvol:48,eight:61,either:[10,16,20,22,23,42,58,60],elb:48,elbapis:48,elec:58,electron:[49,58],elem_dim:[10,16],element:[3,5,9,10,11,16,17,20,22,29,58,62,63],elif:[23,60],elimin:61,els:[10,16,23,33,40,56,58,60],emac:[33,39],emb:[49,58],embed:[10,23,38,57,60,62],embedd:61,embedding_lay:[38,58,60],embedding_nam:[10,16,38],embedding_s:[10,16,38],emphas:42,empir:[10,16],emplace_back:40,emploi:[38,59],empti:[9,20,24,31],emul:63,enabl:[3,7,15,42,43,45,48],enable_grad_shar:[44,45],enable_parallel_vector:45,enc_proj:[11,17,38],enc_seq:[11,17],enc_vec:38,encod:[11,17,38,63],encoded_proj:[11,17,38],encoded_sequ:[11,17,38],encoded_vector:38,encoder_last:[10,16],encoder_proj:38,encoder_s:38,encrypt:48,encrypt_decrypt:48,end:[3,9,10,16,29,31,38,45,53,61,62,63],end_pass:23,enditer:[22,23],endpass:[22,23],endpoint:48,endtrain:23,engin:[0,42,59],english:[3,10,16,63],enjoi:33,enough:31,ensembl:[11,17],ensur:[3,24,40],enter:[33,59],entir:[10,11,16,17,62],entri:[20,40,48,59],entropi:[10,16,58,61],enumer:[10,14,58,60],enumerate_data_types_of_data_lay:20,env:[39,48],environ:[23,32,33,35,42,43,44,45,48,49,54,55,60],eol:39,eos:10,eos_id:[10,16,38],epel:32,epoch:59,epsilon:[12,18],equal:[10,11,12,16,17,24,45],equat:[10,11,12,16,17,18,33],equilibrium:54,equip:[32,38],equival:[10,16,23],error:[7,9,10,12,15,16,18,23,31,35,40,43,45,48,55,56,58,59,60,62,63],error_clipping_threshold:[7,15],errorr:9,especi:[3,11,17,61],essenc:23,essenti:[10,16,23,32,61,63],estat:31,estim:[10,16,23],eta:49,etc:[12,20,29,33,43,44,47,48,62,63],etcd:24,eth0:[43,48],ethternet:43,eval:[9,58,60,62,63],eval_bleu:63,evalu:[2,4,10,16,22,42,43,58,62,63],evaluate_pass:62,evaluator_bas:9,evalut:[31,63],even:[23,29,42,45,62],evenli:48,event:49,event_handl:[22,23],everi:[2,3,9,10,11,16,17,20,23,24,38,39,40,45,58,61,62,63],everyth:[31,39],exactli:[3,9,10,11,16,17,33,48,61],exampl:[2,3,8,9,10,11,12,16,17,18,20,22,29,31,32,33,38,40,42,43,44,45,47,48,49,55,56,57,58,62,63],exceed:[10,16],except:[3,20,47,53,60,62],excluded_chunk_typ:9,exconv:[10,16],exconvt:[10,16],exdb:20,exec:[33,45],execut:[24,40,42,48,59,61,62],exist:[23,24,29,40,45,48,59,62],exit:[45,49],exp:6,expand:[10,40,61,62,63],expand_a:[10,16],expand_level:[10,16],expandconvlay:[10,16],expandlevel:[10,16],expect:[10,16,42,62],expens:63,experi:47,expir:24,explain:[3,9,24,43,54,62],explan:[10,16,58,63],explanatori:[31,33],explicit:40,explicitli:[3,23],exploit:55,explor:[10,16],exponenti:14,expos:[33,48],express:[23,48,62],extend:[0,60],extens:[12,59,60,63],extent:28,extern:[3,27,28],extra:[10,11,15,16,17,31],extra_input:16,extra_lay:22,extraattr:[7,15,47],extraattribut:[16,17],extraattributenon:16,extract:[10,16,48,55,61,62],extract_fea_c:56,extract_fea_pi:56,extract_para:53,extralayerattribut:[7,10,11,15,16],extralayeroutput:11,extrapaddl:17,extrem:[10,16,42],extremli:2,f120da72:49,f7e3:48,fa0wx:49,fabric:43,facotr:[10,16],fact:56,factor:[7,10,12,15,16,18],factori:27,fail:[3,24,45,47,49,55],failur:24,fake:54,fake_imag:29,fals:[3,7,9,10,11,12,15,16,17,18,20,29,31,38,40,45,47,49,53,58,60,61,62,63],false_label:29,false_read:29,famili:63,familiar:[3,31],fanscin:3,fantasi:59,fantast:58,far:0,farmer:59,fascinatingli:2,fast:[10,16,39,42],faster:[10,11,16,17,24,38,42,62],favorit:39,favourit:33,fbd1f2bb71f4:49,fc1:[40,47],fc2:47,fc3:47,fc4:47,fc8a365:48,fc8a:48,fc_act:[11,17],fc_attr:[11,17],fc_bias_attr:[11,17],fc_layer:[16,31,40,47,58,60],fc_layer_nam:11,fc_mat:22,fc_name:17,fc_param_attr:[11,17],fclayer:40,fdata:61,fea:56,fea_output:56,feat:62,featur:[3,10,14,16,20,39,45,55,58,62,63],feature_map:60,feed:[11,17,20,22,23,31,62],feedback:0,feeder:20,feedforward:55,femal:59,fernan:62,festiv:3,fetch:[20,24,38,40],few:[3,24,29,33],fewer:[10,16],fg0:[10,16],field:[10,16,22,42,48],figur:[23,38,40,42,53,54,55,56,61,62,63],file1:63,file2:63,file:[2,3,5,9,10,16,20,22,23,24,26,28,29,31,32,33,38,39,40,43,45,53,55,56,61,62,63],file_list:3,file_nam:[3,31,56,58,61],filenam:[3,26,60],filer:[10,16],filesystem:[33,48],fill:[10,16,24,48,58],film:59,filter:[10,16,56],filter_s:[10,11,16,17],filter_size_i:[10,16],finali:43,find:[10,12,16,18,24,33,42,55,62,63],fine:[7,15,60],fingerprint:48,finish:[3,24,33,43,48,49,55],finit:40,first:[3,10,16,20,23,24,31,33,35,38,39,40,42,45,47,48,53,54,55,56,58,60,61,62,63],first_seq:38,firstn:20,firstseen:49,fit:[2,20,39],five:[42,58],fix:[3,7,15,27,63],flag:[20,45,54,55,61],flexiabl:29,flexibl:[0,2,10,11,16,17,23,38],flight:63,float32:[5,20,29,31,54,56],floor:[10,16],flow:[30,39],fly:[31,58],fnt03:48,focu:[3,42],folder:[26,32,48,55,62,63],follow:[2,3,9,10,11,12,16,17,18,20,23,24,29,32,33,35,38,39,40,42,43,47,48,49,50,51,53,54,55,56,58,59,60,61,62,63],fool:54,forbid:23,force_load:27,forecast:62,forget:[12,18,23,62],form:[2,3,11,12,17,18,42,61],format:[2,3,9,31,39,40,45,48,53,55,59,60,62],former:[23,63],formula:[10,11,16,17],formular:[10,16],forward:[11,14,17,38,39,40,47,54,61,62],forwardactiv:40,forwardtest:5,found:[3,5,10,16,32,38,54,55,58,62],four:[3,35,53,56,58,60,61,62],frame:9,framework:[23,40,56,58,62],free:[20,63],french:63,frequenc:[20,42,53,58,62],frequent:[29,43,63],frog:55,from:[0,3,5,10,11,16,17,20,22,24,26,29,31,33,38,39,40,42,43,45,47,48,49,53,54,55,56,58,59,60,61,62,63],from_timestep:[10,16],fromfil:[29,31,56],fulfil:42,full:[10,16,24,33,38,40],full_matrix_project:[11,17,38],fulli:[31,39,40,42,54,55,56,58,60,62],fullmatrixproject:[10,16],fully_matrix_project:[11,17],fullyconnect:53,fullyconnectedlay:40,func:20,fundament:31,further:[10,16],fusion:60,gain:[10,16],game:54,gamma:56,gan:23,gan_train:54,gap:45,gate:[10,11,16,17,62],gate_act:[10,11,16,17],gate_recurr:[10,16],gather:[10,16,40,60],gauss:[7,15],gaussian:54,gcc:[27,32,33],gdebi:35,gen:[10,16,63],gen_conf:[54,63],gen_data:63,gen_result:63,gen_train:54,gen_training_machin:54,gen_trans_fil:38,gender:[20,59,60],gener:[2,3,5,9,10,11,16,17,20,22,23,24,26,29,31,32,33,42,43,45,47,48,53,56,57,58,60,62],generated_word_embed:[10,16],generatedinput:[10,16,38],generator_conf:54,generator_machin:54,generator_train:54,genert:3,genr:[59,60],gereat:9,get:[3,10,11,16,17,20,22,31,32,35,38,40,42,43,48,52,55,56,58,60,61,62],get_batch_s:61,get_best_pass:62,get_config_arg:[47,58,60,62],get_data:[49,58,61],get_dict:20,get_embed:20,get_imdb:62,get_input_lay:40,get_mnist_data:54,get_model:56,get_movie_title_dict:20,get_nois:54,get_output_attr:17,get_output_layer_attr:11,get_shap:22,get_training_loss:54,get_word_dict:20,getbatchs:40,getenv:23,getinput:40,getinputgrad:40,getinputvalu:40,getoutputgrad:40,getoutputvalu:40,getparameterptr:40,getsiz:40,getslotvalu:54,gettempl:48,gettranspos:40,getw:40,getweight:40,getwgrad:40,gfortran:32,gildea:61,gist:[11,17],git:[30,32,33,39],github:[10,11,16,17,32,33,35,56],give:[3,24,31,33,40,42,48,58],given:[3,20,22,29,40,45,54,58,61,62,63],global:[3,7,12,15,23,24,42,45,48,60,62],global_learning_r:[7,15],globalstat:42,globalstatinfo:42,globe:3,goal:[42,61],godoc:27,goe:[10,11,16,17,24,31],going:[58,62],good:[10,16,29,42,62,63],goodfellow13:[10,16],googl:23,googleapi:48,gpg2:48,gpg:48,gpu:[2,3,7,10,12,15,16,19,30,32,35,41,43,54,55,56,60,61,62,63],gpu_id:[45,47,54],gpugpu_id:44,grab:[24,62],grad:[45,59],grad_share_block_num:[44,45],gradient:[7,9,10,12,15,16,18,22,24,45,58,62],gradient_clipping_threshold:[7,12,15,58,62],gradient_machin:[22,28],gradientmachin:[5,22,28,54,60,63],gradual:[31,42],grai:55,gram:[53,62],grant:48,graph:[10,16,22,24,53],graphviz:56,grave:62,grayscal:3,greater:[10,16],grep:[33,62],groudtruth:38,ground:[9,10,16,58,63],group:[11,17,62],group_id:60,group_input:38,grouplen:[20,59],gru:[10,16,38,58,63],gru_attr:17,gru_bias_attr:[11,17],gru_decod:38,gru_decoder_with_attent:38,gru_encoder_decod:[53,63],gru_layer_attr:11,gru_memori:[11,17],gru_siz:58,gru_step:[17,38],gru_step_lay:[11,38],grumemori:[11,17,38],gserver:[10,16,40],gsizex:42,guarante:40,guess:[31,62],gui:42,guid:[34,38,39,40,42,48,49,53,55,62,63],guidenc:31,gur_group:[11,17],gzip:49,hack:[34,43],hadoop:23,half:48,hand:[59,60,62],handl:[23,29,43,60,62],handler:22,handwrit:[3,62],hard:[48,58],hardwar:[33,42],has:[3,5,10,11,12,16,17,18,20,23,24,33,38,40,42,48,49,53,55,58,59,60,61,62,63],has_kei:22,have:[2,3,5,9,10,11,16,17,20,23,24,29,31,32,33,38,39,40,42,43,45,47,48,53,55,58,59,60,62,63],hdf:[2,26],head:[39,53,62],header:[28,31,40,53,56,60],health:59,heavi:43,height:[10,16,20,27,29,40,55],held:24,hello:23,help:[3,5,39,43],helper:[8,10,11,16,17,40],here:[3,5,7,10,11,15,16,17,20,23,24,29,31,32,38,43,44,47,48,49,53,55,56,57,58,59,60,61,62,63],heurist:[10,16,45,63],hidden:[10,11,16,17,38,48,58,60,62],hidden_s:[11,17,60],hierarch:[10,16,38],high:[7,15,40,54],higher:2,highest:[20,63],highli:[2,3,20,38,47,60,62],him:23,hint:31,histor:62,hl_get_sync_flag:40,hold:[23,24,48],home:[26,43,48,49],homemak:59,hook:[3,60,61],hope:0,horizont:[10,16,56],horror:59,hors:55,horst:62,host:[32,33,43,48,49],hostnam:[43,48],hostpath:49,hostport:48,hot:60,hour:63,hous:[3,20,31,53],how:[2,3,7,10,15,16,23,24,31,38,43,45,48,49,52,55,56,58,60],howev:[3,11,17,29,31,38,39,44,45,48,62,63],hpp:27,html:[20,33,55],htod:42,http:[10,11,16,17,20,32,33,35,39,48,49,54,55,56,63],huber:[10,16],huge:[10,16,39],huina:62,human:63,hyper:[10,16,40],hyperplan:20,i0601:60,i0706:63,i0719:63,i1117:42,iamfullaccess:48,iamusersshkei:48,ib0:43,ics:20,icwsm:62,id_input:[9,38],idea:[10,16,29],ident:[31,33,48,59],identifi:[38,40],identityoffsetproject:[10,16],identityproject:[10,16],ids:[9,10,16,40,58,60],idx:40,ieee:62,ignor:[3,9,10,16,45,53],ijcnlp:62,illustr:[3,24,38,40,42,58],ilsvrc:56,imag:[3,19,20,23,29,31,34,47,48,50,51,54,56,57,63],image_a:29,image_b:29,image_classif:55,image_fil:29,image_lay:29,image_list_provid:56,image_nam:23,image_path:29,image_provid:55,image_reader_cr:29,image_s:56,imagenet:[26,57],imagepullpolici:48,imageri:[10,16],images_reader_cr:29,imdb:59,imdber:62,img:[3,10,16,55],img_conv:17,img_conv_lay:11,img_featur:3,img_norm_typ:10,img_pool:17,img_pool_lay:11,img_siz:55,imgsiz:42,imgsizei:42,imgsizex:42,immedi:48,immutable_paramet:23,implement:[3,10,11,12,16,17,18,20,27,28,38,58,61],importerror:60,improv:[0,42,48,62,63],inbound:48,includ:[2,3,10,11,16,17,20,23,27,28,32,33,38,40,42,45,48,49,53,58,59,61,63],inconsist:59,incorrect:[10,16],increas:[24,45,63],increment:45,incupd:40,inde:[20,29,33],independ:[10,16,58],index:[3,9,10,16,19,20,22,24,38,43,48,60],indexslot:[10,16,61],indic:[3,9,10,16,31,43,48,61],individu:[24,31,48],industri:24,infer:[1,23,24,27,32],infiniband:43,info:[9,10,16,20,40,43],infom:39,inform:[5,9,20,40,42,45,48,59,60,61,62,63],infrastructur:[48,54],ingor:45,ininst:23,init:[7,15,40,47,48,54,58,60,61],init_hook:[58,60,61],init_hook_wrapp:8,init_model_path:[44,45,47,53,58,61],initi:[3,5,7,10,15,16,20,38,40,45,53,54,58,61],initial_max:[7,15],initial_mean:[7,10,15,16],initial_min:[7,15],initial_std:[7,10,15,16],initpaddl:[5,54],inlcud:[11,17],inlin:48,inner:40,inner_param_attr:[11,17],input1:[10,11,16,17],input2:[10,16],input:[3,5,9,10,11,14,16,17,19,20,22,29,31,38,40,47,53,54,55,56,58,60,61,62,63],input_data:40,input_data_target:40,input_featur:14,input_fil:[31,61],input_hassub_sequence_data:40,input_id:[10,16],input_imag:[11,17,55],input_index:40,input_label:40,input_lay:[10,40],input_nam:23,input_sequence_data:40,input_sequence_label:40,input_sparse_float_value_data:40,input_sparse_non_value_data:40,input_t:40,input_typ:[31,38,58,60],inputdef:40,inputlayers_:40,inputtyp:[3,20],insid:[9,10,16,24,29,33,48],inspir:53,instal:[33,36,39,43,49,55,56,60,61,62],instanc:[10,12,16,24,26,38,40,42,45,61],instance_ip:48,instanti:24,instead:[10,16,19,29,33,39,43,58,63],instruct:[33,35,42,58],int32:45,integ:[3,9,10,16,20,27,38,40,58,62],integer_valu:[3,20,58],integer_value_sequ:[3,20,38,58,61],integr:[32,61],intend:0,inter:[10,16,43],interact:[33,48],intercept:[10,16],interest:[42,62],interfac:[1,5,7,10,11,15,16,17,43,48,55,60,62],interg:58,intergr:[10,16],intermedi:61,intern:[10,11,16,17,20,22,48],internet:[24,62],interpol:10,interpret:[3,9,32,42],interv:62,intrins:32,introduc:[3,24,49,60,62],introduct:[4,54],invalid:29,invari:55,invok:[3,10,16,22,42,48,60],involv:54,iob:9,ioe:9,ips:48,ipt:[10,16,38],ipython:23,is_async:12,is_discriminator_train:54,is_gener:[10,16,53,54,63],is_generator_train:54,is_kei:60,is_layer_typ:10,is_loc:22,is_predict:[58,60,62],is_seq:[10,16,38,60],is_sequ:60,is_stat:[7,15],is_test:[56,61,62],is_train:3,isn:42,isol:33,isspars:40,issu:[32,33,42],item:[10,16,20,22,29],iter:[10,11,12,16,17,18,20,22,23,24,29,55,61,62],its:[3,9,10,11,16,17,23,24,40,42,45,48,53,54,55,58,62,63],itself:[11,17,24],java:27,jeremi:42,jie:[61,62],jmlr:[10,16],job:[5,9,20,44,45,47,56,58,60,61,62,63],job_dispatch_packag:43,job_id:20,job_mod:53,job_nam:48,job_namespac:48,job_path:48,job_workspac:43,jobpath:48,jobport0:48,jobport1:48,jobport2:48,jobport3:48,johan:62,join:24,joint:[53,63],jointli:[11,17,63],journal:[61,62],journei:33,jpeg:55,jpg:56,json:[43,48,49,60],jth:[11,17],judg:63,jupyt:33,just:[3,9,10,11,14,16,17,20,31,39,43,47,48,53,55,60,61,62],jx4xr:48,jypyt:23,k8s_data:48,k8s_job:23,k8s_token:23,k8s_train:48,k8s_user:23,kafka:26,kaim:[10,16],kaimingh:56,kebilinearinterpbw:42,kebilinearinterpfw:42,keep:[3,10,16,24],kei:[3,20,22,24,42,43,60,62],kernel:[10,16,42,58],key1:45,key2:45,key_pair_nam:48,keyid:48,keymetadata:48,keypair:48,keyserv:48,keystat:48,keyusag:48,keyword:3,kill:[24,48],kind:[2,3,23,24,31,48,49,54,58,60],kingsburi:61,kms:48,know:[3,11,17,23,31,40,42,48,60],knowledg:62,known:[54,62,63],kriz:[20,55],ksimonyan:[11,17],kube_cluster_tl:23,kube_ctrl_start_job:23,kube_list_containers_in_job_and_return_current_containers_rank:23,kubeconfig:48,kubectl:49,kuberent:[24,48],kubernet:[23,24,41,43,50,51],kubernetes_service_host:23,kwarg:[3,9,10,11,12,16,17,18,20,58,60,61],l1_rate:[7,15],l2_rate:[7,15],l2regular:[55,58,62],label:[3,5,9,10,12,16,18,20,22,29,31,38,49,54,55,56,57,58,60,62],label_dict:61,label_dim:[10,16,58],label_fil:[29,61],label_lay:[10,29],label_list:61,label_path:29,label_slot:61,labeledbow:62,labl:62,lag:45,lake:3,lambdacost:[10,16],lambdarank:[10,16],languag:[10,16,20,47,53,61,62,63],larg:[19,20,61,62,63],larger:[3,7,9,10,12,15,16,43],last:[9,10,11,16,17,31,38,43,45,58,62,63],last_time_step_output:[10,16],lastseen:49,late:62,latenc:[43,48],later:[32,39,48,58],latest:[10,16,24,33,39,49,62],latter:63,launch:[45,48,62],launcher:23,lawyer:59,layer1:[10,11,16,17],layer2:[10,16],layer3:[10,16],layer:[4,5,7,9,11,15,17,19,20,21,22,29,31,38,41,44,45,53,54,55,56,58,60,61,62],layer_0:40,layer_attr:[10,16,38,47],layer_num:[47,56],layer_s:[10,16],layer_typ:[10,16],layerbas:40,layerconfig:40,layergradutil:40,layermap:40,layeroutout:[10,16],layeroutput:[9,11,16,60],lbl:[9,55],ld_library_path:[32,35,43],lead:42,learn:[0,7,9,10,11,12,15,16,17,18,20,23,29,31,33,36,38,40,42,55,56,58,61,62,63],learnabl:[10,16],learning_method:[12,31,53,55,58,60,62,63],learning_r:[7,12,15,31,53,55,58,60,62,63],leas:24,least:[9,10,16,24,32,59],leav:[3,48],lecun:20,left:[10,16,31,56],leman:63,len:[3,10,16,38,40,58,60,61],length:[10,11,16,17,20,38,45,49,62,63],less:[10,16,23,43,63],less_than:23,let02:49,let:[5,10,16,23,31,48,60],level:[7,10,15,16,43,45,54,60,62,63],lib64:[32,43,45],lib:28,libari:28,libcudnn:32,libjpeg:55,libpaddl:[27,28],libpaddle_capi:28,libpaddle_gserv:28,libpaddle_math:28,libpython:32,librari:[10,16,28,32,33,43,45,60],licens:61,life:24,like:[3,9,10,16,20,24,29,31,32,38,42,43,44,47,48,53,56,58,60,62,63],limit:[10,16,20,42,45],line:[2,3,5,9,20,31,39,41,42,43,47,48,53,55,56,60,61,62,63],linear:[6,10,16],linear_comb:10,linearactiv:[10,16,31],linguist:61,link:[10,11,16,17,32,48,58,62],linux:[32,33,35,48,63],lipeng:53,lipton:62,list:[2,3,8,9,10,11,16,20,22,23,31,33,38,40,43,45,47,48,55,56,58,60,61,62,63],listen:45,literatur:62,littl:[2,3,45,58,62],lium:63,live:[24,33],liwicki:62,load:[2,3,5,10,16,23,24,31,45,48,56,60,61,62,63],load_featur:56,load_feature_c:56,load_feature_pi:56,load_missing_parameter_strategi:[44,45,47,53,61],load_uniform_data:54,loadparamet:5,loadsave_parameters_in_pserv:[44,45],local:[7,15,24,32,33,39,43,44,45,49,55,62],localhost:33,locat:[38,40,58,61],lock:24,log:[3,6,39,40,43,45,48,49,55,60,61,62,63],log_barrier_abstract:45,log_barrier_lowest_nod:[44,45],log_barrier_show_log:[44,45],log_clip:[44,45],log_error_clip:[44,45],log_period:[45,47,49,54,55,58,60,61,62,63],log_period_serv:[44,45],logarithm:14,logger:3,logic:[3,43],login:33,longer:63,look:[3,9,31,43,44,48,49,54,58],lookup:58,loop:29,loss:[10,16,40,54,58,62,63],lot:44,low:[10,16],lower:43,lowest:45,lpaddle_capi_shar:28,lpaddle_capi_whol:28,lst:60,lstm:[10,16,38,49,58],lstm_attr:17,lstm_bias_attr:[11,17],lstm_cell_attr:[11,17],lstm_group:[11,17],lstm_layer_attr:11,lstm_size:58,lstm_step:[11,17],lstmemori:[11,17,38],lstmemory_group:[10,16],ltr:[10,16],lucki:31,mac:[28,32,33],machan:[11,17],machin:[10,11,12,16,17,20,22,31,39,40,44,45,47,48,49,58,60,62,63],made:[3,24,31,38,59],mai:[3,8,9,10,16,29,33,39,42,48,59],main:[3,5,39,48,55,61,62],mainli:45,maintain:[10,16,48],major:[33,39,54,56,62,63],make:[3,10,23,24,29,32,33,39,40,42,43,48,55,58,60,62],male:59,malloc:40,manag:[24,39,43],manageri:59,mandarin:[10,16],mani:[0,10,11,16,17,31,33,45,58,59,60,62],mannal:43,manual:39,manufactur:63,mao:62,map:[3,10,16,20,22,23,45,55,56,60],map_read:20,mapreduc:23,marcu:62,mark:[3,38,61],mark_slot:61,market:[31,59,62],martha:61,mask:[7,10,15,16],master:[23,30,39,45,62],mat:[27,28],mat_param_attr:[11,17],match:42,math:[11,17,27,40,42],matirx:[10,16],matplotlib:55,matric:[5,38,40],matrix:[9,10,11,16,17,20,22,27,28,38,40,44,47,56,61],matrixptr:40,matrixtyp:28,matter:3,max:[3,7,10,13,15,16,20,42,45,47,55,58,60],max_id:[16,22,58],max_job_id:20,max_length:[10,16,38],max_movie_id:20,max_sort_s:[10,16],max_user_id:20,maxid:[9,10,58],maxid_lay:[9,58],maxim:[10,16,63],maximum:[9,20,38,42,45,58,61,62],maxinum:19,maxout:10,maxpool:[10,16],mayb:[10,11,16,17,55],md5:[20,25],mean:[3,7,9,10,11,12,15,16,17,18,19,20,22,29,31,38,42,43,45,47,48,53,54,55,56,58,60,61,62,63],mean_img_s:55,mean_meta:56,mean_meta_224:56,mean_valu:56,measur:[31,42],mechan:[10,11,16,17,38,48,62],media:62,meet:61,mem:[10,16],member:23,memcpi:42,memor:62,memori:[2,3,11,17,38,40,42,45,47,49,58,61,62,63],memory_nam:[10,16],memory_threshold_on_load_data:45,mere:[11,17],merg:[39,45,53,63],mergedict:[53,63],messag:[31,45,49,60,62,63],meta:[43,55,56,58],meta_config:[43,60],meta_fil:60,meta_gener:[43,60],meta_path:55,meta_to_head:60,metadata:[48,49],metaplotlib:23,method:[3,8,10,11,12,16,18,22,33,40,42,45,47,58,60,62,63],might:[10,16,33,40,48],mileag:42,million:[20,47,59],min:[7,15,42,47,48,60],min_pool_s:3,mind:43,mini:[3,10,16,20,22,24],mini_batch:29,minibatch:[10,16],minibatch_data:20,minim:[3,12,18,31,45],minimum:[10,16],minimun:45,minst:3,minut:[24,48,63],mirror:33,miss:[45,53,61],mit:48,mix:[11,17,38,61],mixed_attr:17,mixed_bias_attr:[11,17],mixed_lay:[11,16,38,61],mixed_layer_attr:11,mixedlayertyp:[10,16],mkdir:[32,33,48],mkl:32,mkl_path:32,mkl_root:32,ml_data:[43,60],mnist:[3,5,26,29],mnist_provid:3,mnist_random_image_batch_read:29,mnist_train:[3,29],mnist_train_batch_read:29,mod:61,modal:61,mode:[10,16,45,54,55,56,60,62,63],model:[1,2,5,8,10,11,12,16,17,20,24,25,39,40,41,45,48,60,61,62],model_averag:12,model_config:[5,54],model_list:[45,47,61,62],model_output:62,model_path:47,model_zoo:[53,56],modelaverag:12,modifi:[5,38,39,40,43,48],modul:[2,3,5,8,11,17,20,22,31,32,55,56,58,60,61],modulo:[10,16],momentum:[7,12,15,31,58],momentumoptim:[31,55],mon:49,monitor:[58,62],mono:[10,16],month:[58,63],mood:62,more:[2,3,5,9,10,11,16,17,20,23,24,29,31,33,38,40,42,43,47,49,55,58,61,62,63],morin:[10,16],mose:[62,63],moses_bleu:63,mosesdecod:62,most:[3,5,10,16,20,23,29,31,38,40,42,44,60,61,62,63],mostli:[55,59],mount:[33,48,49],mountpath:[48,49],move:[10,16,24,42,48,60,62],movement:[42,62],movi:[3,20,62],movie_categori:20,movie_featur:60,movie_head:60,movie_id:60,movie_info:20,movie_meta:60,movie_nam:60,movie_review:20,movieid:59,movieinfo:20,movielen:57,moving_average_fract:[10,16],mpi:43,mse:10,mse_cost:[31,60],much:[10,16,24,29,42],mul:40,mulit:43,multi:[10,16,40,44,45,56,63],multi_binary_label_cross_entropi:16,multi_crop:56,multinomi:[10,16],multipl:[9,10,11,16,17,20,23,33,38,40,45,47,48,54,58,60,62],multipli:[9,10,16,40,55],multithread:3,music:59,must:[3,9,10,11,14,16,17,29,32,33,38,39,40,43,45,47,48,63],my_cluster_nam:48,my_cool_stuff_branch:39,my_external_dns_nam:48,mypaddl:49,mysteri:59,name:[3,7,8,9,10,11,15,16,17,19,20,22,23,24,26,28,31,33,38,40,42,43,45,47,49,50,51,53,54,55,56,58,60,62,63],name_prefix:26,namespac:[27,33,40,49],nano:39,nativ:[10,16],natur:[47,61,62],nchw:[10,16],ndarrai:[22,26],ndarri:22,ndcg:[10,16],ndcg_num:[10,16],nearest:58,necessari:[3,10,16,32,40,43,58,62],necessarili:40,need:[3,10,11,16,17,20,23,31,32,33,35,38,39,40,43,44,45,47,48,49,54,55,56,58,60,61,62,63],neg:[3,9,10,16,58,61,62],neg_distribut:[10,16],negat:61,neighbor:58,nest:[3,20],net:[10,11,16,17],net_conf:62,net_diagram:56,network:[2,3,4,5,7,9,10,12,15,16,18,20,21,22,23,24,29,31,33,40,42,43,45,53,62,63],network_config:47,networkadministr:48,neural:[3,5,10,11,12,16,17,18,20,22,23,24,31,42,45,53,54,56,62,63],neuralnetwork:[10,16],neuron:[5,40,58,62],never:[20,29,48,49],newest:39,newtork:62,next:[10,16,20,24,38,40,42,45,48,49,61,62,63],nfs4:48,nfs:48,nfsver:48,nginx:33,nic:[43,44,45],nine:[20,61],nlp:[3,10,16],nltk:20,nmt:63,nnz:40,no_cach:3,no_sequ:[3,60],noah:62,noavx:[33,35],node:[10,16,40,43,45,48,49,62,63],node_0:48,node_1:48,node_2:48,nodefil:43,noir:59,nois:[10,16,24,54],noise_dim:54,non:[10,16,24,40,45,48],none:[2,3,5,7,8,9,10,11,12,15,16,17,18,19,20,22,23,31,38,56,58],nonlinear:40,norm:54,norm_by_tim:[10,16],normal:[3,5,10,11,16,17,20,35,38,40,43,45,49,53,54,56],normzal:56,north:55,notat:[10,16],note:[3,5,7,10,11,12,15,16,17,19,22,23,29,32,42,45,47,48,53,55,60,62],notebook:33,noth:[14,22,45],notic:[38,40],novel:62,now:[0,3,10,16,24,31,33,39,45,48,54,60,61],np_arrai:20,nproc:32,ntst1213:63,ntst14:63,nullptr:40,num:[10,16,43,45,58,61,62,63],num_channel:[10,11,16,17,55],num_chunk_typ:9,num_class:[10,11,16,17,55],num_filt:[10,11,16,17],num_gradient_serv:[44,45],num_group:[10,16],num_neg_sampl:[10,16],num_parameter_serv:23,num_pass:[22,31,44,45,47,49,58,60,61,62,63],num_repeat:[10,16],num_result:9,num_results_per_sampl:[10,16],num_shard:26,number:[3,9,10,16,20,24,26,29,31,40,43,45,48,53,55,56,58,61,62,63],numchunktyp:9,numdevices_:47,numlogicaldevices_:47,numofallsampl:9,numofwrongpredict:9,numpi:[20,22,26,29,31,32,54,56],numsampl:42,numtagtyp:9,nvcc:33,nvidia:[32,33,42,45],obj:[3,8,31,55,56,58,60],object:[3,5,7,8,9,10,11,12,15,16,17,18,20,22,23,27,42,54,55,56,58,61],observ:[12,18,31,40,42,63],obtain:[58,61,62],occup:[59,60],occur:[20,22,39],oct:49,odd:[10,16],off:[28,33],offer:[5,61],offici:[33,48,55],offlin:[24,26],offset:[10,16,60],often:[9,43,58,63],ograd:40,old:[33,39,45],omit:58,on_init:3,on_travisexclud:32,onc:[3,10,16,24,33,39,40,48,58],one:[3,8,9,10,11,12,14,16,17,18,19,20,23,24,29,31,33,39,40,43,45,47,48,49,53,54,55,56,58,60,61,62,63],one_host_dens:60,one_hot_dens:60,onli:[2,3,5,9,10,11,16,17,19,20,22,23,31,32,38,39,40,42,44,45,47,48,49,53,56,58,59,62,63],onlin:[12,18,24,26,29],onto:48,open:[0,3,10,16,23,29,31,33,48,56,58,60,61],openbla:32,openblas_path:32,openblas_root:32,oper:[10,11,12,16,17,18,33,38,40,42,45,48,53,55,60],opinion:62,opt:[23,32],optim:[3,4,7,15,21,22,31,40,42,62],option:[3,9,10,16,23,31,39,40,43,47],order:[3,10,11,16,17,20,29,40,45,48,49,54,56,58,62,63],ordinari:62,oregon:48,org:[10,11,16,17,20,32,33,54],organ:[10,16,55,62,63],origin:[0,2,3,10,16,20,39,54,61,63],other:[3,9,10,11,12,16,17,20,24,32,33,35,38,39,47,48,49,53,54,55,56,58,59,60,61,62,63],otherchunktyp:9,otherwis:[2,8,10,16,20,23,24,29,38,43,47,60,63],our:[23,33,38,40,48,49,53,55,58,61,62,63],out:[10,16,22,23,31,38,42,45,48,49,55,62],out_dir:48,out_left:[10,16],out_mem:38,out_right:[10,16],out_size_i:[10,16],out_size_x:[10,16],outlin:46,outperform:61,output:[5,7,9,10,14,15,16,17,19,20,22,23,26,29,31,38,40,42,45,47,49,53,54,55,56,58,60,61,62,63],output_:[10,16,40],output_dir:56,output_fil:61,output_id:[10,16],output_lay:[22,56],output_max_index:19,output_mem:[10,16,38],output_path:26,outputh:[10,16],outputw:[10,16],outsid:[3,10,11,16,17],outter_kwarg:3,outv:40,over:[2,10,11,16,17,23,39,40,42,58,61,62],overcom:62,overhead:42,overlap:40,overrid:[24,40],owe:0,own:[33,39,43,48],pacakg:35,pack:33,packag:[3,20,33,34,48],pad:[10,38,58],pad_c:[10,16],pad_h:[10,16],pad_w:[10,16],paddepaddl:2,padding_attr:[10,16],padding_i:[10,16],padding_x:[10,16],paddl:[3,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,22,23,24,26,27,28,31,32,33,34,35,39,40,41,42,43,45,47,48,54,55,58,60,61,62,63],paddle_error:[27,28],paddle_matrix:[27,28],paddle_matrix_cr:28,paddle_matrix_get_shap:27,paddle_matrix_shap:27,paddle_n:43,paddle_output:49,paddle_port:43,paddle_ports_num:43,paddle_ports_num_for_spars:43,paddle_pserver2:43,paddle_root:53,paddle_source_root:53,paddle_train:[28,30,43],paddledev:[48,49],paddlepaddl:[0,2,3,5,10,11,12,16,17,20,24,29,31,32,35,38,39,40,41,42,43,50,51,56,58,60,61,62],paddlepadl:3,paddlpaddl:0,paddpepaddl:3,page:[39,48,60],pai:33,pair:[9,61],palceholder_just_ignore_the_embed:53,palmer:61,paper:[10,16,53,54,56,61,62,63],paraconvert:53,paragraph:62,parallel:[42,45,47,48,49,63],parallel_nn:[7,15,44,45],param:[7,10,15,16,60],param_attr:[10,11,16,17,31,38],paramattr:[7,10,15,16,31,38],paramet:[2,3,4,5,8,9,10,11,12,16,17,18,19,20,21,25,29,31,40,41,47,54,55,58,60,61,62,63],parameter_attribut:[10,16],parameter_block_s:[44,45],parameter_block_size_for_spars:[44,45],parameter_learning_r:[7,15],parameter_nam:[22,23],parameter_serv:23,parameterattribut:[7,10,11,15,16,17],parametermap:40,parameters_:40,parameterset:23,parametris:[12,18],paramt:[48,53],paramutil:60,paraphras:63,paraphrase_data:53,paraphrase_model:53,paraspars:40,parent:[10,40],pars:[5,20,47,48,54,60,61],parse_config:[5,54],parser:60,part:[3,31,38,39,40,42,54,58,60,61,62,63],parti:[42,60],partial:[10,16,54],participl:53,particular:42,partit:[24,26,48],pass:[3,8,10,16,20,22,24,29,31,39,40,42,43,45,48,49,54,55,58,60,61,62,63],pass_id:22,pass_idx:29,pass_test:54,passtyp:40,password:43,past:[23,33,48],path:[2,3,9,20,22,24,29,31,32,38,43,45,47,48,49,53,55,56,58,61,62,63],pattern:[20,24,27,31,48,60,62],paul:61,paus:24,pave:63,pdf:[10,11,16,17],pem:[23,48],pend:24,penn:61,per:[10,16,20,24,29,45,55,58],perfom:[45,47],perform:[2,10,11,16,17,31,38,39,40,41,43,44,54,55,58,62,63],period:[2,24,45,58,60,61,62,63],perl:[62,63],permiss:48,peroid:[10,16],persist:48,persistentvolum:48,persistentvolumeclaim:48,person:23,perspect:42,perturb:40,pfs:26,pgp:48,phase:31,photo:55,pick:[3,48],pickl:60,picklabl:8,pictur:58,piec:[10,11,16,17,31],pillow:55,pip:[32,39,43,55,60],pipe:59,pipelin:61,pixel:[3,10,16,20],pixels_float:3,pixels_str:3,place:[2,3,24,40,42,43,56,63],placehold:[31,53],plai:[61,62],plain:[2,9,10,16,22,28],plan:[24,40],platform:[0,31,48],pleas:[3,5,7,10,11,12,15,16,17,18,23,24,29,32,33,34,38,39,40,48,53,55,58,60,61],plot:[23,55],plotcurv:55,png:[55,56],pnpairvalidationlay:45,pnpairvalidationpredict_fil:44,pod:[48,49],pod_nam:48,point:[24,31,42],polar:[20,62],polici:48,polit:62,poll:62,poo:55,pool3:40,pool:[3,4,11,17,21,55,58,60],pool_attr:[11,17],pool_bias_attr:[11,17],pool_layer_attr:11,pool_pad:[11,17],pool_siz:[3,10,11,16,17],pool_size_i:[10,16],pool_strid:[11,17],pool_typ:[10,11,16,17],pooling_lay:[11,58,60],pooling_typ:[10,16,58],poolingtyp:19,popular:[31,56],port:[33,43,44,45,48,49],port_num:44,ports_num:45,ports_num_for_spars:[44,45,47],pos:[60,62],pose:24,posit:[3,9,10,16,20,58,61,62,63],positive_label:9,possibl:[23,39,42,54],post1:32,potenti:42,power:[10,58,63],practic:[8,10,16,31,38,40],pre:[3,10,11,16,17,20,23,33,48,49,53,55,61,62,63],pre_dictandmodel:53,precis:[9,32],pred:[58,61],predefin:62,predetermin:[10,16,45,63],predic:[20,61],predicate_dict:61,predicate_dict_fil:61,predicate_slot:61,predict:[3,4,9,10,12,16,18,22,31,38,43,45,53,58,63],predict_fil:45,predict_output_dir:[44,45,58],predict_sampl:5,predicted_label_id:58,predictor:60,predin:55,prefer:62,prefetch:40,prefix:[24,26,48],pregrad:40,preinstal:32,premodel:53,prepar:[5,50,58],preprcess:62,preprocess:[20,38,43,49,62],prerequisit:32,present:[23,56,61,63],pretti:31,prev_batch_st:[44,45],prevent:[2,12,18,23,24],previou:[10,11,16,17,24,40,45,48,61,63],previous:[10,16,49,56],price:[20,31],primarili:62,principl:23,print:[7,15,22,23,31,38,45,53,58,60,61,62,63],printallstatu:42,printer:9,printstatu:42,prite:9,privat:28,privileg:48,prob:[9,22,54],probabilist:[10,16,53],probability_of_label_0:58,probability_of_label_1:58,probabl:[9,10,16,22,38,39,56,58,61],problem:[5,10,12,16,18,23,58,61,62],proc:33,proc_from_raw_data:58,proce:[20,24,29,48],procedur:[53,61,63],proceed:[10,16,61],process:[2,3,5,7,8,10,11,12,15,16,17,23,26,31,33,38,43,45,47,48,49,53,55,56,58,60,61,62,63],process_pr:58,process_test:8,process_train:8,processdata:[55,56],processor:42,prod:33,produc:[11,17,20,24,29,33,56,58],product:[0,33,40,48,58,62],productgraph:49,profil:32,proflier:42,program:[2,20,23,26,29,33,42,43,45],programm:59,progress:[24,45],proivid:3,proj:[10,16],project:[10,11,16,17,28,32,38,40,60],promis:[10,11,16,17],prompt:39,prone:23,prop:61,propag:[12,18,45,47],properli:58,properti:[3,45],propos:63,proposit:61,protect:40,proto:19,protobuf:32,protocol:45,prove:58,proven:63,provid:[0,8,10,16,20,23,31,33,38,42,43,48,53,54,55,56,59,62],providermemory_threshold_on_load_data:44,provis:48,provod:3,prune:[10,16],ps_desir:24,pserver:[43,44,45,48],pserver_id:25,pserver_num_thread:[44,45],pserverstart_pserv:44,pseudo:23,psize:40,ptr:28,pull:[30,33,53,63],punctuat:62,purchas:58,purpos:[0,24,42],push_back:40,put:[24,33,40,43,49,58],pvc:48,pwd:33,py_paddl:[5,20,54],pydataprovid:[2,3,58],pydataprovider2:[4,5,31,38,58,60,62],pydataproviderwrapp:8,pyramid:[10,16],pyramid_height:[10,16],python:[2,3,4,8,22,23,27,30,31,32,39,43,53,54,55,61,62,63],pythonpath:55,pzo:62,qualifi:32,qualiti:58,queri:[10,16,48,63],question:[10,16,23,48,61],quick:[45,49,57,63],quick_start:[48,49,50,58],quick_start_data:49,quickli:31,quickstart:49,quit:42,quot:59,rac:[10,16],rais:20,ramnath:62,ran:42,rand:[42,45,47,54,61],random:[3,7,10,15,16,20,26,29,31,45,54,55,61],random_imag:26,randomli:62,randomnumberse:44,rang:[3,10,16,20,26,29,45,47,55,59,61],rank:[10,16,23,48,56,58],rare:3,rate:[7,9,12,15,18,20,40,43,55,58,60,62,63],rather:[5,48,62],ratio:45,raw:[10,16,31,58,62],raw_meta:60,rdma:[32,45],rdma_tcp:[44,45],reach:[24,42,61],read:[2,3,20,22,23,24,26,29,31,38,43,48,56,58,60],read_from_realistic_imag:23,read_from_rng:23,read_lock:25,read_mnist_imag:23,read_ranking_model_data:23,reader:[1,22,26,63],reader_cr:26,reader_creator_bool:29,reader_creator_random_imag:[20,29],reader_creator_random_image_and_label:[20,29],readi:[24,31,48,49,55],readm:[28,59,60,62],readonesamplefromfil:3,readwritemani:48,real:[3,29,31,54],realist:23,reason:[10,11,17,23,24,33,49],rebas:39,recal:9,receiv:[8,24],recent:63,reciev:45,recogn:55,recognit:[3,10,16,56,62],recommand:3,recommend:[2,11,17,23,33,38,40,43,45,60],recommonmark:32,recompil:42,record:[48,60,61],recordio:23,recov:[24,31,54],rectangular:[10,16],recurr:[61,62],recurrent_group:[11,17,38],recurrent_lay:11,recurrentgradientmachin:28,recurrentgroup:9,recurrentlay:45,recv:48,reduc:[12,18,43,45,47],refer:[2,5,7,8,10,11,12,15,16,17,18,24,38,40,43,49,53,55,58,60,63],referenc:[10,16],regard:61,regardless:63,regex:60,region:[42,61],regist:[40,42],register_gpu_profil:42,register_lay:40,register_timer_info:42,registri:[33,49],regress:[9,57],regular:[7,12,15,40,48,55,58,62],rel:[2,11,17,43],relat:[3,8,24,33,35,49,60,62],relationship:[20,31,54],releas:[30,32,33,35,48,59,61],relev:[61,63],reli:32,reliabl:24,relu:[6,10,16,40],reluactiv:10,remain:58,rememb:[10,16],remot:[7,15,39,40,43,45,47,48],remoteparameterupdat:45,remov:[20,43,45,62],renam:63,reorgan:[10,16],repeat:10,replac:[33,62],repo:39,report:[42,43],repositori:39,repres:[3,5,10,12,16,20,38,40,48,55,58,59],represent:[58,62],reproduc:63,request:[24,30,48,49,53,63],requir:[2,9,10,16,23,24,40,43,48,49,54,55,58,60],requrest:39,res5_3_branch2c_bn:56,res5_3_branch2c_conv:56,res:61,research:[10,16,20,55,59,62],resembl:62,reserv:3,reserveoutput:40,reset:[10,16,24],reshap:29,reshape_s:[10,16],residu:56,resnet:57,resnet_101:56,resnet_152:56,resnet_50:56,resolv:[39,49],resourc:[33,48],respect:[3,31,38,40,45,55,56,61,63],respons:[10,16,48,49],rest:[3,10,16,31],restart:[24,48,49],restartpolici:[48,49],restrict:45,resu:29,result:[5,9,10,14,16,22,38,42,45,48,55,56,58,60,61,62],result_fil:[9,38],ret_val:60,retir:59,retran:48,retriev:[40,49],return_seq:[11,17],reuqest:30,reus:[29,40],reveal:23,revers:[10,11,16,17,38,61,62],review:[20,39,49,58,62],reviews_electronics_5:49,revis:58,rewrit:63,rgb:[10,16],rgen:62,rho:[12,18],rich:31,right:[3,10,16,56],rmsprop:[12,58],rmspropoptim:60,rnn:[10,11,16,17,41,44,58,62],rnn_bias_attr:38,rnn_layer_attr:38,rnn_out:38,rnn_step:[10,16],rnn_use_batch:[44,45],rnnlm:20,robot:55,role:[20,23,38,48,57,62],roman:62,romanc:59,root:[12,18,19,43,48,49],root_dir:43,rot:[10,16],rotat:10,roughli:[3,54],routin:60,routledg:62,row:[5,9,10,16,20,40,56],row_id:[10,16],rsize:48,rtype:[10,16,60],rule:[40,48],run:[23,24,33,39,40,41,42,45,48,50,51,53,55,56,58,60,62,63],runinitfunct:42,runtim:[2,3,32,33,43],s_fusion:60,s_id:60,s_param:54,s_recurrent_group:38,sacrif:2,sai:[31,45,47],sake:40,sale:59,same:[3,5,8,9,10,11,16,17,23,38,43,47,48,53,58,60,61,62,63],samping_id:[10,16],sampl:[3,5,9,20,43,45,47,53,54,56,58,60,61,62,63],sample_dim:54,sample_id:9,sample_num:9,santiago:62,satisfi:[43,48,58],save:[3,10,16,20,24,26,31,45,47,48,49,55,56,58,60,61,62,63],save_dir:[31,45,47,49,54,55,58,60,61,62,63],save_only_on:[44,45],saving_period:[44,45],saving_period_by_batch:[44,45,47,58],saw:3,scalabl:0,scalar:[3,10,16],scale:[0,10,14,56,59,60],scalingproject:[10,16],scatter:[10,16],scenario:[31,44],scene:44,schdule:48,schedul:[48,54],scheduler_factor:[7,15],schema:53,scheme:[9,12,61],schmidhub:62,schwenk:63,sci:59,scienc:62,scientist:[0,59],score:[9,10,16,60,62,63],screen:60,scrip:58,script:[5,20,43,48,55,56,58,61,62,63],seaplane_s_000978:55,search:[10,16,24,32,38,45,61,63],seat:63,second:[3,10,16,20,23,29,31,33,39,43,53,56,58,59,60,62],secret:48,section:[3,38,40,43,48,58],sed:62,see:[3,5,10,11,16,17,23,24,31,33,39,42,48,53,54,56,58,60,62,63],seed:[42,45],segment:9,segmentor:53,sel_fc:[10,16],select:[10,16,39,48,59,63],selectiv:[10,16],selector:49,self:[31,40,59,62],selfnorm:[10,16],semant:[20,23,30,38,57,62],semat:23,sen_len:61,send:[24,45,48],sens:10,sent:[23,49],sent_id:38,sentenc:[3,10,16,20,38,58,61,62,63],sentiment:[3,31,57,58,61],sentiment_data:62,sentiment_net:62,sentimental_provid:3,separ:[3,9,45,53,58,59,60,61,63],seq:[10,16,60],seq_pool:[10,16],seq_text_print:9,seq_to_seq_data:[53,63],seq_typ:[3,20,60],seqtext_printer_evalu:38,seqtoseq:[10,16,38,53,63],seqtoseq_net:[10,16,38,53,63],sequel:3,sequenc:[3,9,10,11,14,16,17,19,20,40,53,58,60,61,62,63],sequence_conv_pool:58,sequence_layer_group:[10,16],sequence_nest_layer_group:[10,16],sequencesoftmax:6,sequencestartposit:[10,16],sequencetextprint:9,sequencetyp:3,sequenti:[8,10,16,38,58,61],seri:[11,17,62],serial:[3,22],serv:[33,42,48,54],server:[23,33,40,43,44],serverless:24,servic:[33,59],session:[33,42],set:[2,3,5,7,9,10,11,15,16,17,20,22,23,24,31,32,33,35,38,40,41,42,43,44,45,47,48,49,53,55,56,58,59,60,61,62,63],set_active_typ:40,set_default_parameter_nam:[7,15],set_drop_r:40,set_input:[10,16],set_siz:40,set_typ:40,setp:48,settup:40,setup:[3,33,40,58],sever:[3,10,16,43,47,48,57,58,60,61,62,63],sgd:[12,18,22,23,24,43,54,62,63],sgdasync_count:44,shallow:61,shape:[10,16,22,56],shard:[24,25,26,48],share:[10,16,28,32,33,42,45,49,61],shared_bia:[11,17],shared_bias:[10,16],shared_ptr:[27,28],shell:[48,56],shift:56,ship:55,shold:62,shop:62,shorten:[10,16],shorter:56,should:[3,5,9,10,12,16,20,22,23,29,31,35,38,39,43,48,55,58,60,61,62,63],should_be_fals:23,should_be_tru:23,should_shuffl:[3,61],shouldn:39,show:[5,12,18,20,24,31,39,45,48,49,53,56,58,60,61,62,63],show_check_sparse_distribution_log:[44,45],show_layer_stat:[44,45],show_parameter_stats_period:[44,45,47,49,58,61,62,63],shown:[3,9,10,16,23,38,40,42,48,54,55,56,58,60,62,63],shrink:40,shuf:60,shuffl:[3,20,60,62],sid:48,side:[10,16,22,56],sig:48,sigint:43,sigmoid:[6,10,16,17,40],sigmoidactiv:[10,11],sign:48,signal:43,signatur:48,signific:42,similar:[10,16,29,48,58,60],similarli:[10,16,61],simpl:[2,3,9,10,11,14,16,17,20,22,32,33,39,42,45,58,60,61,62],simple_attent:38,simple_gru:58,simple_lstm:[10,16,58],simple_rnn:[10,16,38],simplest:48,simpli:[2,10,16,23,32,33,38,39,42,53,56,60,62,63],simplifi:[23,40,49],simultan:48,sinc:[10,16,24,29,31,33,42,48,54,58,59,63],sincer:[39,62],singl:[3,9,11,12,17,20,24,33,40,43,49,56,58,61,63],site:48,six:[53,61,63],size:[3,9,10,11,12,16,17,18,20,24,26,29,31,38,40,43,45,54,55,56,58,59,60,61,62,63],size_a:[10,16],size_b:[10,16],size_t:40,sizeof:53,skill:63,skip:[29,31,43,48,56],slide:[10,12,16,18,20,24],slightli:55,slope:[10,16],slot:[60,61],slot_dim:60,slot_nam:60,slottyp:60,slow:[3,42],small:[3,20,40,43,45,55,63],small_messag:[44,45],small_vgg:55,smaller:[10,16,24],smith:62,smooth:[10,16],snap:49,snapshot:[25,48],snippet:[38,40,42,48,58],social:62,sock_recv_buf_s:[44,45],sock_send_buf_s:[44,45],socket:45,softmax:[6,10,11,16,17,23,38,40,53,58,61,62],softmax_param_attr:[11,17],softmax_selfnorm_alpha:[10,16],softmaxactiv:[38,58],softrelu:6,softwar:[33,42],solv:[23,61],solver:63,some:[3,7,10,12,15,16,20,22,23,31,32,39,40,42,44,45,47,48,54,58,59,60,61,62,63],some_c_api_funct:28,some_inst:28,some_python_class:27,somecppclass:27,somedata:22,somegotyp:27,someth:[3,10,16],sometim:[12,18,29,42,62],soon:24,sophist:[31,40,43],sort:[10,16,20,45,48,60,62,63],sourc:[0,8,10,16,20,28,29,31,33,38,39,48,49,53,58,60,63],source_dict_dim:38,source_language_word:38,space:[9,33,38,42],space_seperated_tokens_from_dictionary_according_to_seq:9,space_seperated_tokens_from_dictionary_according_to_sub_seq:9,spars:[3,7,10,12,15,16,18,20,40,43,45,48,58],sparse_binary_vector:[3,20,58],sparse_binary_vector_sequ:20,sparse_float_vector:3,sparse_non_value_slot:20,sparse_upd:[7,15],sparse_value_slot:20,sparse_vector:20,sparse_vector_sequ:20,sparseparam:40,sparseprefetchrowcpumatrix:40,spatial:[10,16,55],speak:[38,63],spec:[48,49],specfii:45,speci:55,special:[10,16,32,53,58,63],specif:[2,47,55,58,60],specifi:[2,3,9,10,16,20,23,31,32,38,40,45,48,54,55,56,58,59,60,62,63],speech:[10,16],speed:[11,17,33],spefici:56,sphinx:[27,32,33],sphinx_rtd_them:32,split:[3,10,16,43,47,48,53,56,58,61],split_count:48,spp:10,sql:2,squar:[6,10,12,16,18,19,31],squarerootn:13,squarerootnpool:[10,16],squash:63,srand:45,src:63,src_backward:38,src_dict:38,src_embed:38,src_forward:38,src_id:38,src_root:5,src_word_id:38,srl:[20,61],ssd:16,ssh:[33,43,48,49],sshd:33,ssl:32,sstabl:[23,26],stabl:[30,48],stack:[31,48,58,61],stacked_lstm_net:62,stacked_num:62,stackexchang:[10,16],stage:43,stake:63,stale:[24,39],stamp:42,standard:[7,15,53,55,61,62,63],stanford:[20,49],stanh:6,star:59,start:[10,16,22,24,31,33,38,39,42,43,45,52,53,57,60,63],start_pass:[44,45],start_pserv:45,startup:[24,48],stat:[32,42,45,61,62,63],state:[10,11,16,17,24,31,38,45,49,54,61,63],state_act:[10,11,16,17],statement:[40,48],staticinput:[10,16,38],statist:[10,16,45,58,61,62,63],statset:42,statu:[9,39,42,48,49],status:49,std:[27,28,40,45],stderr:43,stdout:43,step:[5,10,11,12,16,17,19,24,33,38,40,42,43,48,49,58,60,61,62,63],still:56,stmt1482205552000:48,stmt1482205746000:48,stochast:[12,18,24],stock:62,stop:[10,16,33,43,45,49,60],storag:[48,49,55],store:[9,10,16,20,40,43,45,48,49,53,55,56,58,60,61,62,63],str:[22,47],straight:39,strategi:[3,19,24,45,61],street:[10,16,61],strength:54,strict:29,stride:[10,16],stride_i:[10,16],stride_x:[10,16],string:[2,3,8,9,10,16,40,45,48,62],strip:[58,60,61],struct:28,structur:[20,43,48,53,55,58,60,61,62,63],sts:48,stub:[10,16],student:59,stuff:39,stun:3,style:[3,10,16,32,39],sub:[9,10,16,20,23,38,40,55,58,63],sub_sequ:3,subgradi:[12,18],submit:[39,44,45,48],subnet0:48,subnet:[23,48],subobjectpath:49,subsequenceinput:[10,16],subset:[40,63],substanti:56,substitut:63,succe:62,succeed:49,success:[48,49,56,61],successfulcr:49,successfuli:62,successfulli:[56,60,62],successor:[45,63],sucessfulli:63,sudo:[32,35,48,55],suffic:[29,31],suffici:45,suffix:63,suggest:[10,16,42],suitabl:[39,45,55],sum:[9,10,12,13,16,18,25,38,40],sum_:[10,16],sum_to_one_norm:10,summar:[58,62],sumpool:[10,16],support:[7,9,10,12,15,16,19,20,24,29,32,33,35,38,40,42,45,48,61],suppos:[31,40,58],sure:[39,40,48,55,62],survei:62,swap_channel:56,swig:[5,27,28,32],swig_paddl:[5,20,54],symbol:[10,16,28],sync:[24,39,45,54],syncflag:40,synchron:[12,18,24,43,45,48],syntact:61,syntax:[29,60],synthect:31,synthes:54,synthet:31,sys:56,system:[24,32,33,43,49,58,61,62,63],t2b:53,t_i:[10,16],tab:[33,58],tabl:[3,10,16,56,58,63],tableproject:[10,16],tag:[9,20,33,38],tagtyp:9,take:[3,5,9,10,11,16,17,23,38,40,42,48,49,54,61,63],taken:[3,61],tanh:[6,10,11,16,17,40],tanhactiv:[10,11,38],taobao:62,tar:[32,48],tarbal:48,target:[10,16,20,22,38,53,58,63],target_dict_dim:38,target_dictionary_dim:[10,16],target_language_embed:[10,16],target_language_word:38,targetinlink:[10,16],task:[3,9,10,16,31,38,47,53,56,61,62,63],tconf:62,tcp:[45,48],teach:58,tear:42,technic:24,technician:59,techniqu:[38,40],tee:[49,55,60,61,62,63],tell:[24,33,42,60],tellig:62,templat:[49,61],tempor:[10,16,58,61],tensor:10,term:[10,11,16,17,24,61,62],termin:49,terminolog:31,tese:2,tesh:61,test100:20,test10:20,test:[2,3,8,9,10,16,20,22,23,28,29,30,32,33,35,39,42,43,44,53,55,56,58,59,63],test_all_data_in_one_period:[49,55,60,61,62],test_data:63,test_fcgrad:40,test_gpuprofil:42,test_layergrad:40,test_list:[3,8,31,55,58],test_part_000:62,test_pass:[44,45,47,63],test_period:[44,45,47],test_ratio:60,test_wait:[44,45],testa:23,testb:23,testbilinearfwdbwd:42,testconfig:40,tester:[60,63],testfcgrad:40,testfclay:40,testlayergrad:40,testmodel_list:44,testq:23,testresult:22,testsave_dir:44,testutil:40,text:[2,3,9,11,17,20,23,33,38,48,53,57,58,60,62],text_conv:58,text_conv_pool:60,text_fil:[20,62],tflop:42,tgz:[20,32],than:[3,5,7,9,10,11,12,15,16,17,24,32,33,38,40,43,48,56,61,62,63],thank:[0,53,63],thei:[3,23,24,31,33,38,40,42,43,44,48,56,62],them:[2,3,11,17,23,24,29,31,33,38,42,44,45,48,55,56,58,60,62,63],theori:42,therefor:32,therein:[10,16],therun:56,thi:[2,3,7,8,9,10,11,12,15,16,17,18,20,22,23,24,29,31,32,33,35,38,39,40,42,43,45,47,48,49,53,54,55,56,58,59,60,61,62,63],thing:[3,31,38,39,42,60,61],think:23,third:[10,16,24,42,56,62],those:[24,56,61],thought:42,thread:[40,42,45,47,60,61,62,63],thread_local_rand_use_global_se:[44,45],threadid:47,threadloc:42,three:[3,9,10,12,16,24,29,31,38,45,54,56,62,63],threshold:[7,9,12,15,24,45,62],thriller:59,through:[5,10,16,24,38,40,42,43,53,54,55,62,63],throughout:58,throughput:42,thu:[3,10,16,31,40,48,63],tier:49,tight:32,time:[3,10,11,16,17,19,20,23,24,29,31,38,42,45,47,49,58,59,61,62,63],timelin:[10,16,42],timeo:48,timeout:24,timer:32,timestamp:[10,16,25,59],timestep:[3,10,16],titil:60,titl:[20,39,59,60],tmall:62,todo:[9,11,17,20,24],toend:[10,16],togeth:[3,10,11,16,17,20,22,38],token:[9,10,16,23,38,53,62,63],too:[20,33,35],tool:[33,38,39,48,62],toolchain:32,toolkit:[32,35],top:[9,56,61],top_k:9,topolog:[20,23,24],topolopi:22,toronto:[20,55],total:[9,22,24,29,42,43,49,53,63],total_pass:29,touch:62,tourism:62,tourist:63,toward:31,tra:63,track:[10,24],tractabl:[10,16],tradesman:59,tradit:[10,16],trail:20,train100:20,train10:20,train:[1,2,3,5,7,8,9,10,12,15,16,18,20,26,38,40,41,42,44,50,51,56],train_conf:[53,63],train_config_dir:48,train_data:63,train_id:48,train_list:[3,8,31,55,56,58],train_part_000:62,trainabl:[10,16],traindot_period:44,trainer:[3,5,23,25,26,31,40,43,45,47,54,58,61,62,63],trainer_config:[2,3,31,43,48,49,58,60,62],trainer_config_help:[3,6,7,8,9,10,11,12,13,31,40,55,58,60],trainer_count:[44,45,47,48,49,60,61,62,63],trainer_id:[45,48],trainerintern:[58,60,63],training_machin:54,trainingtest_period:44,trainonedatabatch:54,tran:[10,40,45],trane:3,transact:[24,62],transfer:[2,3],transform:[10,16,38,40,54,55,58,61],transform_param_attr:[11,17],translat:[10,11,16,17,31,53,60,62,63],transpar:43,transport:45,transpos:[10,16,40,54],transposedfullmatrixproject:[10,16],travel:[3,11],travi:[32,39],treat:[10,16,38],tree:[10,16,39,45,63],trg:63,trg_dict:38,trg_dict_path:38,trg_embed:38,trg_id:38,trg_ids_next:38,triain:2,tricki:27,trivial:3,trn:58,truck:55,true_imag:29,true_label:29,true_read:29,truth:[9,10,16,58,63],tst:58,tune:[7,15,41,58,60,63],tuninglog_barrier_abstract:44,tupl:[3,8,10,11,16,20,22,29],ture:[10,16],turn:[10,16,29,54],tutori:[33,38,39,40,42,43,48,49,50,51,56,58],tweet:62,twelv:63,twitter:62,two:[2,3,10,11,16,17,23,29,31,33,38,42,43,47,48,53,54,55,56,58,60,61,62,63],txt:[3,40,43,48,58,60,62],type:[3,8,9,10,11,12,16,17,19,20,22,23,24,27,28,29,31,33,38,40,45,47,48,49,55,56,58,60,61],type_nam:[10,60],typedef:[27,28],typic:[5,9,33,42,62],typo_freq:20,ubuntu:[30,35],ubyt:29,uci:20,ufldl:[10,16],uid:49,uint64:27,uint64_t:27,unbalanc:45,unbound:38,unconstrain:62,under:[31,32,33,48,59,62],underli:31,understand:[33,42,53,55,62],understudi:63,undeterminist:42,unemploi:59,unexist:61,uniform:[7,10,15,16,20,26,29,45,54],uniqu:[23,24,39,45,48],unique_ptr:40,unit:[10,11,16,17,31,32,33,38,39,61],unittest:28,unittestcheckgrad_ep:44,univ:63,unix:43,unk:[53,63],unk_idx:[58,61],unknown:[10,16],unlabel:62,unlik:[61,62,63],unseg:[10,16],unsup:62,unsupbow:62,until:[24,43,48,61],unus:60,unzip:60,updat:[7,10,12,15,16,24,32,40,43,45,47,62],update_equ:22,updatecallback:40,updatestack:48,upload:24,upon:[0,61],upstream:39,uri:48,url:[20,35,62],urls_neg:62,urls_po:62,urls_unsup:62,usag:[2,3,9,10,11,16,17,20,22,31,42,53,54,60],use:[0,2,3,5,7,8,9,10,11,12,15,16,17,19,20,22,23,24,31,32,33,34,35,38,39,40,42,43,45,47,48,49,53,54,55,56,58,59,60,61,62,63],use_global_stat:[10,16],use_gpu:[44,45,47,49,54,55,56,58,60,61,62,63],use_jpeg:55,use_old_updat:[44,45],use_seq:[31,60],use_seq_or_not:60,used:[2,3,5,9,10,11,12,16,17,18,19,20,22,23,24,29,31,34,35,38,40,42,43,44,45,47,48,53,55,56,58,60,61,62,63],useful:[2,3,10,11,16,17,38,40,47,58,61,62],usegpu:[40,54],useless:43,user:[2,3,7,9,10,11,15,16,17,20,22,23,29,31,33,39,43,44,45,48,56,58,61],user_featur:60,user_head:60,user_id:60,user_info:20,user_meta:60,user_nam:60,userid:59,userinfo:20,usernam:39,uses:[3,24,38,39,40,45,48,55,56,58,60,63],using:[2,3,5,7,8,10,15,16,20,23,24,29,31,33,38,39,40,42,45,47,48,49,53,54,55,56,58,61,62],usr:[32,43,45,48],usrdict:53,usrmodel:53,usual:[10,16,20,22,31,32,42,45,47,48,62],utf:53,util:[5,32,38,40,42,55,60,62],uuid:25,v28:[10,16],valid:[29,48,56,62],valu:[3,5,7,9,10,12,15,16,18,19,20,22,24,31,38,40,45,47,48,54,55,56,61,62],value1:45,value2:45,value_rang:20,vanilla:38,vanish:62,vari:[42,48],variabl:[3,10,16,20,23,31,32,35,40,43,48,49,62],varianc:[10,16,56],variant:33,vast:39,vector:[3,10,11,16,17,20,23,38,40,53,58,60,62,63],vectorenable_parallel_vector:44,verb:[20,61],veri:[3,10,16,19,38,42,55,58,62],verifi:[39,40],versa:32,version:[10,11,16,17,30,32,33,35,40,42,43,44,45,48,49,53,55,59,61,62,63],versu:23,vertic:[10,16,56],vgg:[11,17,55],vgg_16_cifar:55,via:[24,29,32,42,43,48,58],vice:32,view:[10,16],vim:39,virtual:33,virtualenv:60,visibl:33,vision:55,visipedia:55,visual:[10,16,33,42],viterbi:61,voc_dim:58,vocab:62,volum:[33,49],volumemount:[48,49],volumn:48,voluntarili:59,vutbr:20,wai:[3,10,11,16,17,23,31,33,38,40,43,47,60,61,63],wait:[12,18,24,45],walk:[5,54],wall:61,want:[3,10,11,16,17,23,29,31,32,33,40,45,47,53,56,58,60,61,62],war:59,warn:[10,16],warp:[10,16,42],watch:24,wbia:[48,56],web:33,websit:[55,58,61,62],wei:[61,62],weight:[9,10,11,12,16,17,18,38,40,45,47,55,56],weight_act:[11,17],weightlist:40,weights_:40,weights_t:40,welcom:[60,62],well:[33,40,45,48,55,58],west:48,western:59,wether:[10,16],what:[7,10,11,12,15,16,17,18,31,43,58,60],wheel:32,when:[2,3,7,9,10,12,15,16,20,22,24,35,38,39,40,42,45,47,48,49,53,54,55,61,62,63],whenev:60,where:[3,10,11,12,16,17,18,23,24,31,38,40,42,43,45,47,53,56,61,63],whether:[9,10,11,16,17,29,40,45,54,55,60,62,63],which:[0,2,3,5,9,10,11,12,16,17,18,20,23,24,26,29,31,35,38,40,42,43,45,47,48,54,55,56,58,59,60,61,62,63],whichev:54,whl:32,who:[53,56,59],whole:[3,9,20,27,28,48,49,58,59,60,63],whole_cont:60,whose:[3,10,16,20,24,38,60,61],why:[11,17,28],wide:61,width:[9,10,16,20,27,29,40,55,63],wiki:[10,16],wikipedia:[10,16,20],wilder:3,window:[10,16,19,20,33,62],wise:[10,16],with_avx:33,with_avxcompil:32,with_coveragecompil:32,with_doccompil:32,with_doubl:40,with_doublecompil:32,with_dsocompil:32,with_gpu:33,with_gpucompil:32,with_profil:42,with_profilercompil:32,with_pythoncompil:32,with_rdmacompil:32,with_style_checkcompil:32,with_swig_pycompil:32,with_test:33,with_testingcompil:32,with_tim:42,with_timercompil:32,within:[10,16,31],without:[9,10,16,24,29,43,62],wmt14:63,wmt14_data:63,wmt14_model:63,wmt:63,wmt_shrinked_data:20,woboq:33,won:[42,56],wonder:3,word:[3,9,10,16,20,38,47,57,60,61,62,63],word_dict:[58,61],word_dim:58,word_id:3,word_idx:20,word_slot:61,word_vector:58,word_vector_dim:[10,16,38,53],words_freq_sort:20,work:[3,5,20,23,24,29,32,38,39,40,42,43,45,48,49,58,60],worker:48,workercount:48,workflow:[33,39,48],workspac:[33,45,60],worri:31,wors:54,would:[22,24,29,33,43,48,54,58,61],wrap:61,wrapper:[11,17,42],writ:60,write:[3,20,23,24,29,33,38,39,41,43,48,55,60,61,63],write_lock:25,writelin:31,writer:[23,59],written:[60,62],wrong:[3,29],wsize:48,wsj:61,www:[10,16,20,55,63],x64:32,xarg:40,xgbe0:45,xgbe1:45,xiaojun:62,xrang:[29,31,40],xxbow:62,xxx:[23,56,63],xxxx:25,xxxxxxxxx:48,xxxxxxxxxx:48,xxxxxxxxxxxxx:48,xxxxxxxxxxxxxxxxxxx:48,xzf:32,y_i:[10,16],y_predict:31,yaml:[48,60],yann:20,year:59,yeild:[22,55],yield:[3,20,23,26,29,31,38,58,60,61,62],you:[2,3,5,7,10,11,12,15,16,17,31,32,33,35,38,39,40,42,43,45,47,48,53,54,55,56,58,60,61,62,63],your:[3,10,16,23,32,33,40,42,43,47,48,58,62],your_access_key_id:48,your_secrete_access_kei:48,your_source_root:28,yum:32,yuyang18:[11,17,20],zachari:62,zeng:62,zero:[3,7,10,12,15,16,18,20,24,40,45,48,58],zhidao:53,zhou:[61,62],zip:[20,59],zone:48,zxvf:48},titles:["ABOUT","API","Introduction","PyDataProvider2","API","Python Prediction","Activations","Parameter Attributes","DataSources","Evaluators","Layers","Networks","Optimizers","Poolings","Activation","Parameter Attribute","Layers","Networks","Optimizer","Pooling","Data Reader Interface and DataSets","Model Configuration","Training and Inference","PaddlePaddle Design Doc","Design Doc: Distributed Training","\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\uff08Checkpointing\uff09","\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Python Data Reader Design Doc","Paddle\u53d1\u884c\u89c4\u8303","Simple Linear Regression","Installing from Sources","PaddlePaddle in Docker Containers","Install and Build","Debian Package installation guide","GET STARTED","RNN Models","RNN Configuration","Contribute Code","Write New Layers","HOW TO","Tune GPU Performance","Run Distributed Training","Argument Outline","Detail Description","Set Command-line Parameters","Use Case","Distributed PaddlePaddle Training on AWS with Kubernetes","Paddle On Kubernetes","&lt;no title&gt;","&lt;no title&gt;","PaddlePaddle Documentation","Chinese Word Embedding Model Tutorial","Generative Adversarial Networks (GAN)","Image Classification Tutorial","Model Zoo - ImageNet","TUTORIALS","Quick Start","MovieLens Dataset","Regression MovieLens Ratting","Semantic Role labeling Tutorial","Sentiment Analysis Tutorial","Text generation Tutorial"],titleterms:{"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":26,"\u4e0d\u4f7f\u7528":27,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":27,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":27,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":27,"\u4ec5\u4ec5\u4f7f\u7528void":27,"\u4ece\u5feb\u7167\u6062\u590d":25,"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":27,"\u4f7f\u7528\u8f6c\u6362\u5e93":26,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":28,"\u5206\u652f\u89c4\u8303":30,"\u52a0\u901f\u6267\u884c":25,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":27,"\u52a8\u6001\u6269\u5bb9":25,"\u539f\u56e0":27,"\u539f\u56e0\u5217\u8868":27,"\u57fa\u672c\u8981\u6c42":27,"\u5b9e\u73b0":27,"\u5b9e\u73b0\u65b9\u5f0f":28,"\u5bfc\u51fac":27,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":25,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":27,"\u63a8\u6d4b\u6267\u884c":25,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":26,"\u6587\u4ef6\u9884\u5904\u7406":26,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":28,"\u672f\u8bed":25,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":25,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":28,"\u6d41\u7a0b\u4ecb\u7ecd":26,"\u76ee\u5f55\u7ed3\u6784":28,"\u793a\u4f8b\u7a0b\u5e8f":26,"\u7b26\u53f7":27,"\u7c7b":27,"\u7f16\u8bd1\u9009\u9879":28,"\u7f29\u5bb9":25,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":27,"\u80cc\u666f":27,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8":26,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":26,"\u8f6c\u6362\u5e93":26,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":27,"\u8fdb\u884c\u8bad\u7ec3":26,"book\u4e2d\u6240\u6709\u7ae0\u8282":30,"case":47,"class":40,"function":53,"new":40,"paddle\u52a8\u6001\u5e93\u4e2d":27,"paddle\u53d1\u884c\u89c4\u8303":30,"paddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":30,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":27,"return":29,AWS:48,Abs:14,DNS:48,EFS:48,For:49,KMS:48,Use:[47,49],Using:[33,39],With:33,about:0,absactiv:6,access:48,account:48,activ:[6,14],adadelta:18,adadeltaoptim:12,adagrad:18,adagradoptim:12,adam:18,adamax:18,adamaxoptim:12,adamoptim:12,add:48,address:48,addto:16,addto_lay:10,adversari:54,aggreg:[10,16],algorithm:[24,58],analysi:62,api:[1,4,28,33],appendix:58,applic:4,approach:42,architectur:[38,58],argument:[29,44,47,58],asset:48,associ:48,async:45,attent:38,attribut:[7,15],auc_evalu:9,avg:19,avgpool:13,aws:48,background:31,base:[9,10],baseactiv:6,basepool:19,basepoolingtyp:13,basesgdoptim:12,batch:29,batch_norm:16,batch_norm_lay:10,batch_siz:29,beam_search:[10,16],between:23,bidirect:62,bidirectional_lstm:[11,17],bilinear_interp:16,bilinear_interp_lay:10,bleu:63,block_expand:16,block_expand_lay:10,book:33,brelu:14,breluactiv:6,bucket:48,build:[32,34,49],built:42,cach:3,capi:28,capi_priv:28,cento:32,check:[10,16,40,43],checkpoint:[24,25],chines:53,choos:48,chunk_evalu:9,cifar:20,classif:[9,55],classification_error_evalu:9,classification_error_printer_evalu:9,clone:39,cloudform:48,cluster:[43,47,48],code:39,column_sum_evalu:9,command:[46,47,58,63],commit:[39,49],common:45,commun:45,compos:29,concat:16,concat_lay:10,concept:48,config:[4,47,60,61],configur:[21,38,41,43,48,58,60],conll05:20,connect:[10,16],contain:[33,49],content:[28,42,48],context_project:[10,16],contribut:39,conv:[10,16],conv_oper:[10,16],conv_project:[10,16],conv_shift:16,conv_shift_lay:10,convolut:[55,58],core:48,cos_sim:[10,16],cost:[10,16],cpu:[33,47],creat:[29,39,48,49],creator:29,credenti:48,credit:0,crf:16,crf_decod:16,crf_decoding_lay:10,crf_layer:10,cross_channel_norm:16,cross_entropi:10,cross_entropy_cost:16,cross_entropy_with_selfnorm:10,cross_entropy_with_selfnorm_cost:16,ctc:16,ctc_error_evalu:9,ctc_layer:10,cudnnavg:19,cudnnmax:19,custom:29,dat:59,data:[10,16,20,24,29,31,38,48,49,53,54,55,58,60,61,62,63],data_lay:10,datafeed:20,dataprovid:[3,4,45],dataset:[20,24,59,60,63],datasourc:8,datatyp:20,date:39,debian:35,decayedadagrad:18,decayedadagradoptim:12,decor:29,defin:[48,58,62,63],delet:48,delv:55,demo:48,depend:32,deriv:40,descript:[45,54,59,61],design:[23,24,29],destroi:48,detail:[45,55],develop:[33,41],devic:47,dictionari:[29,53],differ:47,directori:48,dispatch:24,distribut:[23,24,43,45,48],doc:[23,24,29],docker:[33,49],document:[33,52],dotmul_oper:[10,16],dotmul_project:[10,16],down:48,download:[32,48,49,53,56,60,63],dropout_lay:[11,17],dylib:28,dynam:24,ec2:48,elast:48,embed:[16,53,58],embedding_lay:10,entri:29,eos:16,eos_lay:10,equat:40,evalu:[9,31,60],evalutaion:63,event:[22,23],exampl:[23,28,53,54],exercis:55,exp:14,expactiv:6,expand:16,expand_lay:10,extern:48,extract:[53,56,60,63],fault:24,fc_layer:10,featur:[56,59,60,61],field:60,file:[48,49,58,59,60],find:48,first_seq:[10,16],fork:39,format:[24,58],from:[23,32,34],full_matrix_project:[10,16],fulli:[10,16],gan:54,gate:38,gener:[38,54,63],get:[36,49],get_output:16,get_output_lay:10,github:39,gpu:[33,42,45,47],gradient:40,gradient_printer_evalu:9,group:[10,16,48],gru:[11,17,45],gru_group:[11,17],gru_step:16,gru_step_lay:10,gru_unit:[11,17],grumemori:[10,16],guid:35,hand:42,handler:[23,27],hook:39,how:[29,41,42],hsigmoid:[10,16],huber_cost:[10,16],iam:48,ident:14,identity_project:[10,16],identityactiv:6,imag:[10,11,16,17,33,49,55],imagenet:56,imdb:[20,62],img_cmrnorm:16,img_cmrnorm_lay:10,img_conv:16,img_conv_bn_pool:[11,17],img_conv_group:[11,17],img_conv_lay:10,img_pool:16,img_pool_lay:10,imikolov:20,implement:[29,40,54],infer:[22,58],info:56,ingredi:23,init_hook:3,initi:[47,48],input_typ:3,inspect:48,instal:[32,34,35,48,58],instanc:48,integr:48,interfac:[20,24,29,56],interpol:16,interpolation_lay:10,introduct:[2,53,56,62,63],isn:29,job:[24,43,48,49],join:[10,16],keep:39,kei:48,kill:43,kube:48,kubectl:48,kubernet:[48,49],label:61,lambda_cost:[10,16],last_seq:[10,16],lastest:39,launch:43,layer:[10,16,23,40,47],layeroutput:10,layertyp:10,libpaddle_capi_shar:28,libpaddle_capi_whol:28,line:[46,58],linear:[14,31],linear_comb:16,linear_comb_lay:10,linearactiv:6,list:[25,29],local:[47,48],log:[14,58],logactiv:6,logist:58,lstm:[11,17,45,61,62],lstm_step:16,lstm_step_lay:10,lstmemori:[10,16],lstmemory_group:[11,17],lstmemory_unit:[11,17],map:29,master:24,math:[10,16],matrix:45,max:19,maxframe_printer_evalu:9,maxid:16,maxid_lay:10,maxid_printer_evalu:9,maxout:16,maxout_lay:10,maxpool:13,memori:[10,16],meta:60,mini:29,minibatch:20,misc:[11,17],mix:[10,16,47],mixed_lay:10,mnist:[20,54],model:[3,4,21,23,31,33,37,38,43,47,53,54,55,56,57,58,63],modifi:49,momentum:18,momentumoptim:12,movi:[59,60],movielen:[20,59,60],mse_cost:[10,16],multi_binary_label_cross_entropi:10,multi_binary_label_cross_entropy_cost:16,multipl:29,name:48,nce:16,nce_lay:10,need:[29,42],network:[11,17,38,47,54,55,56,58,60,61],neural:[38,55,58,60,61],neuralnetwork:31,nlp:[11,17,45],non:3,norm:[10,16],nvprof:42,nvvp:42,object:[24,60],observ:[53,56],onli:[29,33],optim:[12,18,24,41,58],option:[32,53],outlin:44,output:[11,43,48],overview:58,packag:35,pad:16,pad_lay:10,paddl:[29,30,49],paddlepaddl:[23,33,34,48,52,53,63],pair:48,parallel_nn:47,paramet:[7,15,22,23,24,45,46,48,53,56],paraphras:53,pass:47,perform:[42,45],pnpair_evalu:9,point:48,pool:[10,13,16,19],pooling_lay:10,power:16,power_lay:10,pre:39,precision_recall_evalu:9,predict:[5,55,56,60,61,62],prefetch:29,prepar:[31,38,43,48,53,54,55,60,62,63],preprocess:[53,55,58,60,63],prerequisit:43,pretrain:[53,63],print:9,privat:48,problem:31,process:24,profil:42,provid:[3,29,58,60,61],pull:39,push:39,pydataprovider2:3,python:[5,29,33,40,56,58,60],queue:24,quick:58,randomnumb:45,rank:9,rank_cost:[10,16],rat:60,rate:59,reader:[20,23,29],recoveri:24,recurr:[10,11,16,17,38,58],recurrent_group:[10,16],recurrent_lay:10,refer:[3,42,61,62],region:48,regress:[31,58,60],relu:14,reluactiv:6,render:48,repeat:16,repeat_lay:10,request:39,requir:[32,39],reshap:[10,16],resnet:56,result:[43,49,63],revis:[39,53],rmsprop:18,rmspropoptim:12,rnn:[37,38,45],role:61,rotat:16,rotate_lay:10,route53:48,run:[43,49,61],sampl:[10,16],sampling_id:16,sampling_id_lay:10,scale:[16,24],scaling_lay:10,scaling_project:[10,16],script:49,secur:48,selective_fc:16,selective_fc_lay:10,semant:61,sentiment:[20,62],seq_concat:16,seq_concat_lay:10,seq_reshap:16,seq_reshape_lay:10,seqtext_printer_evalu:9,sequenc:38,sequence_conv_pool:[11,17],sequencesoftmax:14,sequencesoftmaxactiv:6,sequenti:3,server:[24,45,48],servic:48,set:[12,46],setup:[32,48],sgd:45,share:23,shuffl:29,sigmoid:14,sigmoidactiv:6,simpl:[31,38],simple_attent:[11,17],simple_gru:[11,17],simple_img_conv_pool:[11,17],simple_lstm:[11,17],singl:29,slice:[10,16],slope_intercept:16,slope_intercept_lay:10,smooth_l1_cost:[10,16],softmax:14,softmaxactiv:6,softrelu:14,softreluactiv:6,sourc:[32,34],span:32,spars:47,specifi:[47,53],split:60,spp:16,spp_layer:10,squar:14,squareactiv:6,squarerootn:19,squarerootnpool:13,stack:62,standard:58,stanh:14,stanhactiv:6,start:[23,36,48,49,58],startup:49,store:24,structur:54,suffici:29,sum:19,sum_cost:[10,16],sum_evalu:9,sum_to_one_norm:16,sum_to_one_norm_lay:10,summar:23,summari:58,sumpool:13,system:48,tabl:28,table_project:[10,16],take:29,tanh:14,tanhactiv:6,task:24,tear:48,templat:48,tensor:16,tensor_lay:10,test:[40,45,47,60,61,62],text:63,text_conv_pool:[11,17],timer:42,tip:42,todo:[25,26],toi:54,toler:24,tool:42,train:[22,23,24,29,31,33,43,45,47,48,49,53,54,55,58,60,61,62,63],trainer:[22,24,48,60],tran:16,trans_full_matrix_project:[10,16],trans_lay:10,transfer:58,tune:[42,45],tutori:[53,55,57,61,62,63],ubuntu:32,uci_h:20,unit:[40,45],updat:[23,39,48],usag:[29,33,41],use:29,user:[24,53,59,60,62,63],util:9,value_printer_evalu:9,vector:45,verifi:48,version:39,vgg_16_network:[11,17],visual:56,volum:48,vpc:48,warp_ctc:16,warp_ctc_lay:10,what:42,why:[29,42],wmt14:20,word:[53,58],work:33,workflow:63,workspac:43,wrapper:40,write:[40,58],yaml:49,your:39,zoo:[56,57]}})
\ No newline at end of file
+Search.setIndex({docnames:["about/index_en","api/index_en","api/v1/data_provider/dataprovider_en","api/v1/data_provider/pydataprovider2_en","api/v1/index_en","api/v1/predict/swig_py_paddle_en","api/v2/config/activation","api/v2/config/attr","api/v2/config/evaluators","api/v2/config/layer","api/v2/config/networks","api/v2/config/optimizer","api/v2/config/pooling","api/v2/data","api/v2/model_configs","api/v2/run_logic","design/api","design/cluster_train/README","design/cluster_train/checkpointing","design/cluster_train/data_dispatch","design/multi_language_interface/00.why_plain_c","design/multi_language_interface/01.inference_implementation","design/reader/README","design/releasing_process","getstarted/basic_usage/index_en","getstarted/build_and_install/build_from_source_en","getstarted/build_and_install/docker_install_en","getstarted/build_and_install/index_en","getstarted/build_and_install/ubuntu_install_en","getstarted/index_en","howto/deep_model/rnn/index_en","howto/deep_model/rnn/rnn_config_en","howto/dev/contribute_to_paddle_en","howto/dev/new_layer_en","howto/index_en","howto/optimization/gpu_profiling_en","howto/usage/cluster/cluster_train_en","howto/usage/cmd_parameter/arguments_en","howto/usage/cmd_parameter/detail_introduction_en","howto/usage/cmd_parameter/index_en","howto/usage/cmd_parameter/use_case_en","howto/usage/k8s/k8s_aws_en","howto/usage/k8s/k8s_en","howto/usage/k8s/src/k8s_data/README","howto/usage/k8s/src/k8s_train/README","index_en","tutorials/embedding_model/index_en","tutorials/gan/index_en","tutorials/image_classification/index_en","tutorials/imagenet_model/resnet_model_en","tutorials/index_en","tutorials/quick_start/index_en","tutorials/rec/ml_dataset_en","tutorials/rec/ml_regression_en","tutorials/semantic_role_labeling/index_en","tutorials/sentiment_analysis/index_en","tutorials/text_generation/index_en"],envversion:50,filenames:["about/index_en.rst","api/index_en.rst","api/v1/data_provider/dataprovider_en.rst","api/v1/data_provider/pydataprovider2_en.rst","api/v1/index_en.rst","api/v1/predict/swig_py_paddle_en.rst","api/v2/config/activation.rst","api/v2/config/attr.rst","api/v2/config/evaluators.rst","api/v2/config/layer.rst","api/v2/config/networks.rst","api/v2/config/optimizer.rst","api/v2/config/pooling.rst","api/v2/data.rst","api/v2/model_configs.rst","api/v2/run_logic.rst","design/api.md","design/cluster_train/README.md","design/cluster_train/checkpointing.md","design/cluster_train/data_dispatch.md","design/multi_language_interface/00.why_plain_c.md","design/multi_language_interface/01.inference_implementation.md","design/reader/README.md","design/releasing_process.md","getstarted/basic_usage/index_en.rst","getstarted/build_and_install/build_from_source_en.md","getstarted/build_and_install/docker_install_en.rst","getstarted/build_and_install/index_en.rst","getstarted/build_and_install/ubuntu_install_en.rst","getstarted/index_en.rst","howto/deep_model/rnn/index_en.rst","howto/deep_model/rnn/rnn_config_en.rst","howto/dev/contribute_to_paddle_en.md","howto/dev/new_layer_en.rst","howto/index_en.rst","howto/optimization/gpu_profiling_en.rst","howto/usage/cluster/cluster_train_en.md","howto/usage/cmd_parameter/arguments_en.md","howto/usage/cmd_parameter/detail_introduction_en.md","howto/usage/cmd_parameter/index_en.rst","howto/usage/cmd_parameter/use_case_en.md","howto/usage/k8s/k8s_aws_en.md","howto/usage/k8s/k8s_en.md","howto/usage/k8s/src/k8s_data/README.md","howto/usage/k8s/src/k8s_train/README.md","index_en.rst","tutorials/embedding_model/index_en.md","tutorials/gan/index_en.md","tutorials/image_classification/index_en.md","tutorials/imagenet_model/resnet_model_en.md","tutorials/index_en.md","tutorials/quick_start/index_en.md","tutorials/rec/ml_dataset_en.md","tutorials/rec/ml_regression_en.rst","tutorials/semantic_role_labeling/index_en.md","tutorials/sentiment_analysis/index_en.md","tutorials/text_generation/index_en.md"],objects:{"paddle.trainer.PyDataProvider2":{provider:[3,0,1,""]}},objnames:{"0":["py","function","Python function"]},objtypes:{"0":"py:function"},terms:{"0000x":51,"00186201e":5,"00m":35,"02595v1":9,"03m":35,"0424m":35,"0473v3":10,"055ee37d":41,"05d":48,"0630u":35,"06u":35,"0810u":35,"08823112e":5,"0957m":35,"0ab":9,"0rc1":23,"0rc2":[23,26],"0th":56,"10007_10":55,"10014_7":55,"100gb":35,"100gi":41,"10m":35,"1150u":35,"11\u5b9e\u73b0\u4e86c":21,"11e6":42,"12194102e":5,"124n":35,"13m":42,"1490u":35,"15501715e":5,"1550u":35,"15mb":51,"1636k":56,"16mb":51,"16u":35,"173m":49,"173n":35,"1770u":35,"18ad":41,"18e457ce3d362ff5f3febf8e7f85ffec852f70f3b629add10aed84f930a68750":42,"197u":35,"1gb":35,"1st":[46,49,55,56],"202mb":56,"210u":35,"211839e770f7b538e2d8":10,"215n":35,"228u":35,"234m":49,"2520u":35,"252kb":51,"25639710e":5,"25k":51,"2680u":35,"27787406e":5,"279n":35,"27m":35,"285m":35,"2863m":35,"28m":35,"28x28":3,"2977m":35,"2cbf7385":41,"2nd":[9,55,56],"302n":35,"30u":35,"32777140e":5,"328n":35,"32u":35,"32x32":[13,48],"331n":35,"3320u":35,"36540484e":5,"365e":41,"36u":35,"3710m":35,"3768m":35,"387u":35,"38u":35,"3920u":35,"39u":35,"3rd":[53,55,56],"4035m":35,"4090u":35,"4096mb":38,"4279m":35,"43630644e":5,"43u":35,"448a5b355b84":42,"4560u":35,"4563m":35,"45u":35,"4650u":35,"4726m":35,"473m":42,"48565123e":5,"48684503e":5,"49316648e":5,"4gb":38,"50bd":41,"50gi":41,"51111044e":5,"514u":35,"525n":35,"526u":35,"53018653e":5,"536u":35,"5460u":35,"5470u":35,"54u":35,"55g":56,"5690m":35,"573u":35,"578n":35,"5798m":35,"586u":35,"58s":42,"5969m":35,"6080u":35,"6082v4":9,"6140u":35,"6305m":35,"639u":35,"655u":35,"6780u":35,"6810u":35,"682u":35,"6970u":35,"6ce9":41,"6node":36,"6th":56,"704u":35,"70634608e":5,"7090u":35,"72296313e":5,"72u":35,"73u":35,"75u":35,"760u":35,"767u":35,"783n":35,"784u":35,"78m":35,"7eamaa":13,"7kb":42,"8250u":35,"8300u":35,"830n":35,"849m":35,"85625684e":5,"861u":35,"864k":56,"8661m":35,"892m":35,"901n":35,"90u":35,"918u":35,"9247m":35,"924n":35,"9261m":35,"93137714e":5,"9330m":35,"94u":35,"9530m":35,"96644767e":5,"983m":35,"988u":35,"997u":35,"99982715e":5,"99m":49,"99u":35,"9f18":42,"\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u6a21\u578b\u7531\u5927\u91cf\u7684\u53c2\u6570\u7ec4\u6210":18,"\u4e00\u822c\u4e0d\u5141\u8bb8\u518d\u4ece":23,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u67d0\u4e00\u7c7b\u6587\u4ef6\u7684\u5b9e\u73b0\u65b9\u5f0f":21,"\u4e0d\u4f7f\u7528\u9759\u6001\u5e93":20,"\u4e0d\u4f7f\u7528c":20,"\u4e0d\u4f7f\u7528swig":20,"\u4e0d\u540c\u7248\u672c\u7684\u7f16\u8bd1\u5668\u4e4b\u95f4":20,"\u4e0d\u540c\u8bed\u8a00\u7684\u63a5\u53e3\u9002\u5e94\u4e0d\u540c\u8bed\u8a00\u7684\u7279\u6027":20,"\u4e0d\u5728":21,"\u4e0d\u5d4c\u5165\u5176\u4ed6\u8bed\u8a00\u89e3\u91ca\u5668":20,"\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":20,"\u4e0d\u663e\u793a\u7684\u5199\u6bcf\u4e2a\u7c7b\u5177\u4f53\u5305\u542b\u4ec0\u4e48":20,"\u4e0e\u529f\u80fd\u5206\u652f\u4e0d\u540c\u7684\u662f":23,"\u4e0e\u53ef\u80fd\u6709\u7684":23,"\u4e14\u589e\u52a0\u4e00\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00":20,"\u4e14\u8c03\u7528\u65f6\u4e0d\u80fd\u629b\u51fa\u5f02\u5e38\u6216\u51fa\u73b0\u8fd0\u884c\u65f6\u9519\u8bef":21,"\u4e14c99\u652f\u6301bool\u7c7b\u578b\u548c\u5b9a\u957f\u6574\u6570":20,"\u4e14c99\u76f8\u5bf9\u4e8ec11\u4f7f\u7528\u66f4\u52a0\u5e7f\u6cdb":20,"\u4e2a\u6027\u5316\u63a8\u8350":23,"\u4e2d":[20,21],"\u4e2d\u5199\u5165json\u5185\u5bb9":18,"\u4e2d\u5b8c\u5168\u4e00\u81f4":20,"\u4e2d\u5b9e\u73b0\u7684\u7ed3\u6784\u4f53":21,"\u4e3a\u4e86\u66b4\u9732\u7684\u63a5\u53e3\u5c3d\u91cf\u7b80\u5355":21,"\u4e4b\u5916\u7684\u6240\u6709\u5934\u6587\u4ef6":21,"\u4e5f\u4e0d\u4f7f\u7528\u5176\u4ed6\u52a8\u6001\u5e93":20,"\u4e5f\u4e0d\u5e94\u8be5\u62a5\u9519":21,"\u4e5f\u4e0d\u751f\u6210":21,"\u4e66\u5199":20,"\u4eba\u8138\u8bc6\u522b":19,"\u4ec5\u4ec5\u4f7f\u7528":20,"\u4ece":23,"\u4ece\u78c1\u76d8\u6587\u4ef6\u4e2d\u52a0\u8f7duuid\u6587\u4ef6\u540d\u7684\u68c0\u67e5\u70b9\u5feb\u7167\u6587\u4ef6":18,"\u4eceetcd\u4e2d\u8bfb\u53d6\u8282\u70b9":18,"\u4ed6\u4e3b\u8981\u5305\u542b\u4e86\u5b9e\u9645\u66b4\u9732\u7684\u7c7b\u578b\u7ed3\u6784":21,"\u4ed6\u662f\u5c06":21,"\u4ed6\u7684\u76ee\u6807\u662f\u4f7f\u7528c":20,"\u4ee3\u7801\u751f\u6210\u7684\u7b26\u53f7\u53ef\u80fd\u4e0d\u4e00\u81f4":20,"\u4ee3\u8868\u8fd9\u4e2ashard\u7684\u6700\u5927index":19,"\u4ee3\u8868shard\u7684index":19,"\u4ee5\u4e0a\u4ee3\u7801\u7684reader\u8f93\u51fa\u7684data":19,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u4e0b\u751f\u6210100\u4e2a\u6587\u4ef6":19,"\u4ee5\u4e0b":19,"\u4ee5\u53canumpi":19,"\u4efb\u610f\u65f6\u523b\u53ea\u53ef\u80fd\u540c\u65f6\u6709\u4e00\u53f0\u670d\u52a1\u5668\u6545\u969c":18,"\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7248\u672cpython\u5728\u4e00\u4e2a\u8fdb\u7a0b\u91cc\u7684bug":20,"\u4f1a\u76f4\u63a5\u62a5\u9519\u9000\u51fa":20,"\u4f1a\u88abpickle\u5e8f\u5217\u5316\u6210\u5b57\u7b26\u4e32":19,"\u4f20\u5165":19,"\u4f46":21,"\u4f46\u4e0d\u66b4\u9732":21,"\u4f46\u5e76\u6ca1\u6709\u7ecf\u8fc7\u56de\u5f52\u6d4b\u8bd5":23,"\u4f46\u6240\u6709fork\u7684\u7248\u672c\u5e93\u7684\u6240\u6709\u5206\u652f\u90fd\u76f8\u5f53\u4e8e\u7279\u6027\u5206\u652f":23,"\u4f46\u662f\u53c8\u8fc7\u4e8e\u7410\u788e":21,"\u4f46\u662f\u89e3\u91ca\u6027\u8bed\u8a00":20,"\u4f5c\u4e3a\u7c7b\u53e5\u67c4":20,"\u4f7f\u7528":[21,23],"\u4f7f\u7528\u4e0b\u9762\u547d\u4ee4":19,"\u4f7f\u7528\u52a8\u6001\u5e93":20,"\u4f7f\u7528\u540c\u6837\u7684\u8bad\u7ec3\u6570\u636eblock":18,"\u4f7f\u7528\u667a\u80fd\u6307\u9488\u7684\u539f\u56e0\u662f":21,"\u4f7f\u7528\u76f8\u5bf9\u8def\u5f84\u7684\u5f15\u7528\u65b9\u5f0f":21,"\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u5b8c\u6210\u5bf9\u65b0\u6570\u636e\u7684\u9884\u6d4b":18,"\u4f7f\u7528\u9759\u6001\u5e93\u548c\u52a8\u6001\u5e93\u96be\u5ea6\u5dee\u4e0d\u591a":20,"\u4f7f\u7528c":21,"\u4f7f\u7528c99\u505a\u63a5\u53e3":20,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c11\u7684\u539f\u56e0\u662f":20,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c89":20,"\u4f7f\u7528regress":23,"\u4f7f\u7528swig\u53ea\u652f\u6301cpython\u89e3\u91ca\u5668":20,"\u4f7f\u7528swig\u9700\u8981\u591a\u8bed\u8a00\u7ed1\u5b9a\u7684\u5f00\u53d1\u4eba\u5458\u719f\u7ec3\u638c\u63e1swig\u914d\u7f6e":20,"\u4f7f\u7528void":20,"\u4f8b\u5982":[20,21,23],"\u4f8b\u5982\u5bf9\u4e8ejava\u6216\u8005python":20,"\u4f8b\u5982\u5bf9\u4e8ejava\u6765\u8bf4":20,"\u4f8b\u5982\u5bf9\u4e8epython":20,"\u4f8b\u5982c":20,"\u4f8b\u5982java\u4e0epython\u7684\u9519\u8bef\u5904\u7406\u662f\u76f4\u63a5\u6254\u51fa\u6765except":20,"\u4f8b\u5982python\u53ef\u4ee5\u4f7f\u7528":20,"\u4f8b\u5982python\u7684":20,"\u4f9d\u6b21\u7c7b\u63a8":23,"\u4fbf\u662f\u5c06\u9759\u6001\u5e93\u52a0\u5165jvm\u4e2d":20,"\u4fee\u590d\u6240\u6709bug\u540e":23,"\u4fee\u590ddocker\u7f16\u8bd1\u955c\u50cf\u95ee\u9898":23,"\u4fee\u590dubuntu":23,"\u505a\u5982\u4e0b\u51e0\u4e2a\u64cd\u4f5c":23,"\u505a\u63a5\u53e3":20,"\u505c\u6b62\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":18,"\u5148\u5b9e\u73b0\u6a21\u578b\u63a8\u65ad\u7684api":21,"\u5176\u4e2d":[20,23],"\u5176\u4ed6\u51fd\u6570\u5747\u8fd4\u56de":21,"\u5176\u4ed6\u7528\u6237\u7684fork\u7248\u672c\u5e93\u5e76\u4e0d\u9700\u8981\u4e25\u683c\u9075\u5b88":23,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":21,"\u5177\u4f53\u539f\u56e0\u53c2\u8003":21,"\u5177\u4f53\u8bf7\u53c2\u8003":21,"\u5185\u90e8\u9a71\u52a8python\u89e3\u91ca\u5668\u8fdb\u884c\u6a21\u578b\u914d\u7f6e\u89e3\u6790\u548c\u6570\u636e\u8bfb\u53d6":20,"\u518d\u5728\u6bcf\u4e00\u4e2aapi\u4e2d\u81ea\u5df1\u68c0\u67e5\u7c7b\u578b":20,"\u518d\u57fa\u4e8e":23,"\u5199\u4ee3\u7801":20,"\u5199\u5165\u5feb\u7167\u6570\u636e":18,"\u51fd\u6570\u5373\u53ef\u5b8c\u6210\u8f6c\u6362":19,"\u51fd\u6570\u540d\u4e3a":21,"\u51fd\u6570\u547d\u540d":20,"\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":18,"\u5206\u652f":23,"\u5206\u652f\u4e00\u65e6\u5efa\u7acb":23,"\u5206\u652f\u4e2d":23,"\u5206\u652f\u4e3a\u5f00\u53d1":23,"\u5206\u652f\u4e3a\u6bcf\u4e00\u6b21release\u65f6\u5efa\u7acb\u7684\u4e34\u65f6\u5206\u652f":23,"\u5206\u652f\u4e3a\u7a33\u5b9a":23,"\u5206\u652f\u529f\u80fd\u7684\u5c01\u95ed":23,"\u5206\u652f\u5408\u5165":23,"\u5206\u652f\u5408\u5165master\u5206\u652f":23,"\u5206\u652f\u540c\u6b65\u4e3b\u7248\u672c\u5e93\u7684":23,"\u5206\u652f\u540d\u4e3a":23,"\u5206\u652f\u5b58\u5728\u7684\u65f6\u5019":23,"\u5206\u652f\u6d3e\u751f\u51fa\u65b0\u7684\u5206\u652f":23,"\u5206\u652f\u7684\u7248\u672c\u90fd\u662f\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5\u548c\u56de\u5f52\u6d4b\u8bd5\u7684\u7248\u672c":23,"\u5206\u652f\u7684\u7248\u672c\u90fd\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5":23,"\u5206\u7247":18,"\u5219\u4f7f\u7528\u542f\u52a8\u53c2\u6570\u5b9a\u4e49\u7684\u521d\u59cb\u5316\u65b9\u6cd5\u521d\u59cb\u5316\u53c2\u6570":18,"\u5219\u5ffd\u7565":18,"\u5219\u628a\u53e6\u4e00\u4e2a\u6162\u901f\u7684kill\u6389":18,"\u5219\u76f4\u63a5\u5f15\u5165\u53e6\u4e00\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5219\u9700\u8981\u56de\u6eda\u5230\u4e0a\u4e00\u4e2a\u68c0\u67e5\u70b9":18,"\u5220\u9664\u78c1\u76d8\u76ee\u5f55\u4e2d\u4e0d\u662f\u5f53\u524duuid\u7684\u5feb\u7167\u6587\u4ef6":18,"\u5230":18,"\u529f\u80fd\u7684\u6b63\u786e\u6027\u5305\u62ec\u9a8c\u8bc1paddle\u76ee\u524d\u7684":23,"\u52a8\u6001\u5e93":20,"\u5305\u542b\u4e86\u67d0\u79cd\u7c7b\u578b\u7684\u7c7b\u578b\u5b9a\u4e49\u548c\u66b4\u9732\u7684\u5168\u90e8\u51fd\u6570":21,"\u5305\u62ec":19,"\u5305\u62ec\u6743\u91cdw\u548c\u504f\u7f6eb":18,"\u534f\u540c\u5b8c\u6210releas":23,"\u5355\u4e2a\u503c":19,"\u5355\u70b9\u6545\u969c":18,"\u5373":21,"\u5373\u4f7f\u7528":21,"\u5373\u4f7f\u7528\u6237\u76f4\u63a5\u5f15\u7528\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5373\u4f7fc":21,"\u5373\u4f8b\u5982":21,"\u5373\u4fbfpaddl":21,"\u5373\u5b8c\u6210\u67d0\u4e00\u4e2a\u4efb\u52a1\u7684\u6700\u5c11\u51fd\u6570":21,"\u5373\u66b4\u9732":21,"\u5373\u8fd9\u4e2a\u52a8\u6001\u5e93\u662f\u4e0d\u4f9d\u8d56\u4e8e\u5176\u4ed6\u4efb\u4f55\u6587\u4ef6\u7684":20,"\u53c2\u6570":20,"\u53c2\u8003":20,"\u53d1\u5e03\u5230dockerhub":23,"\u53d1\u5e03\u5230github":23,"\u53ea\u5bf9\u7279\u6b8a\u5728\u7ebf\u7cfb\u7edf\u8003\u8651\u4e24\u53f0\u4ee5\u4e0a\u540c\u65f6\u6545\u969c\u7684\u5bb9\u707e":18,"\u53ea\u66b4\u9732\u6982\u5ff5\u7684\u63a5\u53e3":21,"\u53ea\u80fd\u8c03\u7528paddle\u7684\u52a8\u6001\u5e93":20,"\u53ea\u9700\u8981\u6062\u590d\u8fd9\u53f0\u8282\u70b9":18,"\u53ef\u4ee5\u51cf\u5c0f\u7cfb\u7edf\u590d\u6742\u6027":18,"\u53ef\u4ee5\u5728\u4efb\u4f55\u673a\u5668\u4e0a\u6267\u884c\u7684":20,"\u53ef\u4ee5\u628a\u672c\u5730\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5b58\u50a8\u96c6\u7fa4\u4e2d":19,"\u53ef\u4ee5\u6709\u6548\u7684\u907f\u514dparamet":18,"\u53ef\u4ee5\u7528\u4ee5\u4e0b\u6307\u4ee4":19,"\u53ef\u4ee5\u7ee7\u7eed\u5728\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f\u63d0\u4ea4\u4ee3\u7801":23,"\u53ef\u4ee5\u901a\u8fc7\u9636\u6bb5\u6027\u7684\u4fdd\u5b58\u6bcf\u4e2aparamet":18,"\u53ef\u4ee5\u901a\u8fc7volume\u6216persistentvolume\u6302\u8f7d\u5b58\u50a8\u7a7a\u95f4\u5230\u6bcf\u4e2a\u5bb9\u5668\u4e2d":19,"\u53ef\u80fd\u4f1a\u9020\u6210\u7f51\u7edc\u62e5\u585e":18,"\u540c\u65f6\u518d\u5c06":23,"\u540c\u65f6\u63d0\u8d77":23,"\u540d\u5b57\u4fee\u9970":20,"\u5411\u6307\u5b9a\u7684\u76ee\u5f55\u4e2d\u4e00\u4e2a\u65b0\u7684\u6587\u4ef6":18,"\u5411paddle\u7684\u4e3b\u7248\u672c\u5e93\u63d0\u4ea4":23,"\u5426\u5219\u5f97\u628apaddle\u9759\u6001\u5e93\u94fe\u63a5\u5230\u89e3\u91ca\u5668\u91cc":20,"\u542f\u52a8\u4e00\u4e2a\u65b0\u7684\u7ebf\u7a0b\u5f00\u59cb\u4fdd\u5b58\u68c0\u67e5\u70b9":18,"\u548c":[19,20,21,23],"\u548c\u79bb\u7ebf\u6570\u636e\u7684\u65b9\u5f0f":19,"\u54ea\u4e2atrainer\u5148\u5b8c\u6210block\u7684\u8bad\u7ec3":18,"\u56e0\u4e3a\u8fd9\u6837\u505a\u4e5f\u6ca1\u6cd5\u4fdd\u8bc1\u6d88\u9664\u968f\u673a\u6027":18,"\u56e0\u4e3aswig\u5728\u7b2c\u4e09\u65b9\u8bed\u8a00\u4e2d\u66b4\u9732\u7684\u51fd\u6570\u540d":20,"\u56fe\u50cf\u5206\u7c7b":23,"\u5728":[21,23],"\u5728\u4e00\u4e2a\u4e0d\u53ef\u4e2d\u65ad\u5e76\u7f3a\u5c11\u5907\u4efd\u7684\u8bad\u7ec3\u4efb\u52a1\u4e2d":18,"\u5728\u4e0a\u56fe\u4e2d\u663e\u793a\u4e86\u5728\u4e00\u4e2a\u5b9e\u9645\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u5e94\u7528":19,"\u5728\u51fa\u73b0\u5355\u70b9\u6545\u969c\u65f6":18,"\u5728\u5b9e\u73b0\u8fc7\u7a0b\u4e2d":21,"\u5728\u5f02\u6784\u96c6\u7fa4\u4e2d":18,"\u5728\u5f15\u5165\u5176\u4ed6\u7c7b\u578b\u7684\u5934\u6587\u4ef6\u65f6":21,"\u5728\u5feb\u7167\u5199\u5165\u5b8c\u6210\u540e":18,"\u5728\u60a8\u7684\u5b9e\u9645\u73af\u5883\u4e2d":18,"\u5728\u6570\u636e\u96c6\u53ef\u4ee5\u88ab\u8bad\u7ec3\u4e4b\u524d":19,"\u5728\u6837\u4f8b\u4e2d":21,"\u5728\u7528\u6237\u4f7f\u7528c":21,"\u5728\u7ebf\u6a21\u578b\u9884\u6d4b\u670d\u52a1":19,"\u5728\u8bc4\u5ba1\u8fc7\u7a0b\u4e2d":23,"\u5728\u8fd9\u4e2a":23,"\u5728\u8fd9\u4e2a\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":20,"\u5728\u8fd9\u4e2a\u9636\u6bb5\u7684\u4ee3\u7801\u6b63\u5728\u7ecf\u5386\u56de\u5f52\u6d4b\u8bd5":23,"\u5728\u8fd9\u4e9b\u5934\u6587\u4ef6\u4e2d":21,"\u5728\u8fd9\u4e9b\u6587\u4ef6\u4e2d":21,"\u5728\u96c6\u7fa4\u8fd0\u884cmapreduc":19,"\u5728c":20,"\u5728c\u7684\u5934\u6587\u4ef6":20,"\u5728cephfs\u5b58\u50a8\u7cfb\u7edf\u4e2d\u7684\u516c\u5f00\u76ee\u5f55":19,"\u5728kubernetes\u4e0a\u8fd0\u884c\u7684\u4e0d\u540c\u7684\u8ba1\u7b97\u6846\u67b6":19,"\u5728paddle\u4e4b\u4e0a\u8fd0\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u8f93\u51fa\u7684\u6a21\u578b\u4f1a\u63d0\u4f9b\u7ed9\u5728\u7ebf\u4eba\u8138\u8bc6\u522b\u7684\u5e94\u7528\u4f7f\u7528":19,"\u5728paramet":18,"\u5747\u4f1a\u88ab\u5b89\u88c5\u5230includ":21,"\u5747\u662f\u5728":21,"\u591a\u4e2a\u503c":19,"\u591a\u4e2aparamet":18,"\u5927\u591a\u6570\u8bed\u8a00\u90fd\u652f\u6301\u4f7f\u7528c\u8bed\u8a00api":20,"\u5982\u56fe\u4e2dtrainer":18,"\u5982\u679c\u4e0a\u9762\u4e24\u6b65\u51fa\u73b0\u9519\u8bef":18,"\u5982\u679c\u4f7f\u7528swig\u6211\u4eec\u9700\u8981\u5c06\u5728interface\u6587\u4ef6\u91cc":20,"\u5982\u679c\u5931\u8d25":23,"\u5982\u679c\u5b58\u5728\u67d0\u4e9btrainer\u6267\u884c\u901f\u5ea6\u8fc7\u6162\u4f1a\u5f71\u54cd\u6574\u4f53\u96c6\u7fa4\u7684\u901f\u5ea6":18,"\u5982\u679c\u5df2\u7ecf\u6b63\u5728\u6267\u884c\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":18,"\u5982\u679c\u662f\u5176\u5b83\u7c7b\u578b":19,"\u5982\u679c\u6709bugfix\u7684\u884c\u4e3a":23,"\u5982\u679c\u67d0\u4e00\u4e2a\u7c7b\u578b\u9700\u8981\u5f15\u7528\u53e6\u4e00\u4e2a\u7c7b\u578b":21,"\u5982\u679c\u67d0\u4e00\u4e2apaddl":21,"\u5982\u679c\u67d0\u4e00\u4e2apaddle\u6982\u5ff5\u5fc5\u987b\u8981\u66b4\u9732":21,"\u5982\u679c\u6ee1\u8db3\u6761\u4ef6":18,"\u5982\u679c\u7528\u6237\u8981\u628apaddle\u7684\u9759\u6001\u5e93":20,"\u5982\u679c\u8c03\u7528\u9759\u6001\u5e93\u53ea\u80fd\u5c06\u9759\u6001\u5e93\u4e0e\u89e3\u91ca\u5668\u94fe\u63a5":20,"\u5982\u679cparamet":18,"\u5b57\u7b26\u4e32":19,"\u5b58\u50a8":19,"\u5b66\u4e60\u6210\u672c\u9ad8":20,"\u5b83\u4eec\u7684\u6587\u4ef6\u540d\u662f":19,"\u5b89\u88c5\u540e\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":21,"\u5b8c\u6210\u6570\u636e\u7684\u9884\u5904\u7406":19,"\u5b9e\u73b0\u7b80\u5355":20,"\u5bf9\u4e8e\u4e0d\u540c\u8bed\u8a00":20,"\u5bf9\u4e8e\u540c\u4e00\u6bb5c":20,"\u5bf9\u4e8e\u591a\u8bed\u8a00\u63a5\u53e3":20,"\u5bf9\u4e8e\u5927\u591a\u6570\u8bed\u8a00":20,"\u5bf9\u4e8e\u6bcf\u79cd\u7c7b\u578b":21,"\u5bf9\u4e8e\u6bcf\u79cdc":21,"\u5bf9\u6bd4":20,"\u5bf9\u8f93\u5165\u53c2\u6570\u7684\u5b89\u5168\u6027\u8fdb\u884c\u4e86\u5fc5\u8981\u7684\u5224\u65ad":21,"\u5bf9\u8fd9\u4e2a\u7248\u672c\u7684\u63d0\u4ea4":23,"\u5bfc\u51fa\u8fd9\u4e9b\u63a5\u53e3":21,"\u5c06":23,"\u5c06\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u62c6\u5206\u6210\u591a\u4efd":18,"\u5c06\u5927\u91cf\u7684":20,"\u5c06\u65b0\u5206\u652f\u7684\u7248\u672c\u6253\u4e0atag":23,"\u5c06master\u5206\u652f\u7684\u5408\u5165commit\u6253\u4e0atag":23,"\u5c31\u9700\u8981\u5bf9\u8fd9\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00\u589e\u52a0\u4e00\u4e9b\u5b9a\u4e49":20,"\u5e73\u5747\u6545\u969c\u4fee\u590d\u65f6\u95f4":18,"\u5e73\u5747\u6545\u969c\u7387":18,"\u5e76\u4e14\u4f7f\u7528":21,"\u5e76\u4e14\u53ef\u4ee5\u88ab\u63d0\u4ea4\u7684job\u76f4\u63a5\u4f7f\u7528":19,"\u5e76\u4e14\u5728\u5e38\u89c1\u7684\u5e73\u53f0\u4e0a":20,"\u5e76\u4e14\u8ba9\u63a5\u53e3\u8131\u79bb\u5b9e\u73b0\u7ec6\u8282":20,"\u5e76\u5220\u9664":23,"\u5e76\u5220\u9664\u66f4\u65e9\u7684\u5feb\u7167":18,"\u5e76\u52a0\u8f7d\u5176\u4e2d\u7684\u53c2\u6570":18,"\u5e76\u5728\u96c6\u7fa4\u4e2d\u8fd0\u884c\u591a\u4e2a\u5206\u5e03\u5f0f\u6570\u636e\u5904\u7406\u4efb\u52a1":19,"\u5e76\u5c06c":21,"\u5e76\u628a\u5feb\u7167\u4fdd\u5b58\u5230\u8fd9\u4e2a\u76ee\u5f55\u4e0b":18,"\u5e76\u6ca1\u6709paddle\u7279\u522b\u9700\u8981\u7684\u7279\u6027":20,"\u5e76\u88ab\u5b58\u50a8\u5728\u8bf8\u5982hadoop":19,"\u5e76\u9002\u5e94github\u7684\u7279\u6027\u505a\u4e86\u4e00\u4e9b\u533a\u522b":23,"\u5efa\u8bae":23,"\u5f00\u53d1\u4e86\u6a21\u578b\u9884\u6d4b\u7684\u6837\u4f8b\u4ee3\u7801":21,"\u5f00\u53d1\u8005\u4fee\u6539\u81ea\u5df1\u7684\u4ee3\u7801":23,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4e2d":23,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4f7f\u7528":23,"\u5f00\u59cb\u63d0\u4f9b\u670d\u52a1":18,"\u5f15\u5165\u4e86\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5f53\u529f\u80fd\u5206\u652f\u5f00\u53d1\u5b8c\u6bd5\u540e":23,"\u5f53\u7528\u6237\u4f7f\u7528\u5b8c\u8fd9\u4e2a\u53c2\u6570\u540e":21,"\u5f88\u96be\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":20,"\u5f97\u4f7f\u7528":20,"\u5fc5\u8981":21,"\u60c5\u611f\u5206\u6790":23,"\u6211\u4eec\u4e5f\u53ef\u4ee5\u786e\u5b9a\u6bcf\u4e00\u4e2a\u53c2\u6570\u7684\u7c7b\u578b":21,"\u6211\u4eec\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":20,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u8f6c\u6362\u65b9\u5f0f":19,"\u6211\u4eec\u6700\u7ec8\u7684\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165python\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u8bed\u8a00\u7684\u89e3\u91ca\u5668":20,"\u6211\u4eec\u90fd\u63d0\u4f9bpython\u7684\u8f6c\u6362\u5e93":19,"\u6216\u8005":[20,21],"\u6216\u8005\u5c06\u8fd9\u53f0\u8282\u70b9\u8fc1\u79fb\u5230\u53e6\u4e00\u4e2a\u8282\u70b9\u5e76\u542f\u52a8\u5373\u53ef\u6062\u590d\u8bad\u7ec3\u4efb\u52a1":18,"\u6216\u8005\u7528tuple\u8868\u793a\u7684\u591a\u4e2a\u503c":19,"\u6216\u8005\u7531\u5b83\u4eec\u7ec4\u6210\u7684list":19,"\u6240\u4ee5\u5728\u5199\u5165\u5feb\u7167\u7684\u8fc7\u7a0b\u4e2d":18,"\u6240\u6709\u4e0e\u7c7b\u578b\u76f8\u5173\u7684\u51fd\u6570":21,"\u6240\u6709\u7684\u63a5\u53e3\u5747\u4e3ac\u63a5\u53e3":21,"\u6240\u6709\u7c7b\u578b\u540d\u4e3a":21,"\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":20,"\u6253\u5f00\u8fd9\u4e2a\u7f16\u8bd1\u9009\u9879":21,"\u628a":19,"\u628a\u4e4b\u524d\u793a\u4f8b\u4e2d\u8f6c\u6362\u5b8c\u6bd5\u7684random":19,"\u6307\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u4e4b\u540e\u5f97\u5230\u7684\u6240\u6709\u53c2\u6570":18,"\u63a5\u53e3":[20,21],"\u63a5\u53e3\u5c42\u505a\u8fc7\u591a\u5c01\u88c5":21,"\u63a5\u53e3\u662f":19,"\u63d0\u4f9b\u7ed9\u7528\u6237\u672c\u5730\u8f6c\u6362\u7684\u5e93":19,"\u63d0\u4f9b\u7ed9paddle\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e":19,"\u6570\u636e\u8bfb\u53d6\u5747\u4ea4\u7531\u5176\u4ed6\u8bed\u8a00\u5b8c\u6210":20,"\u6570\u636e\u9884\u5904\u7406\u4efb\u52a1":19,"\u6587\u4ef6":20,"\u6587\u4ef6\u5185\u5bb9\u4e3a":20,"\u6587\u4ef6\u540d\u4e3a\u6b64uuid":18,"\u6587\u4ef6\u9700\u8981\u9884\u5148\u88ab\u8f6c\u6362\u6210paddlepaddle\u96c6\u7fa4\u5185\u90e8\u7684\u5b58\u50a8\u683c\u5f0f":19,"\u65b0\u624b\u5165\u95e8\u7ae0\u8282":23,"\u65b9\u4fbf\u6d4b\u8bd5\u4eba\u5458\u6d4b\u8bd5paddle\u7684\u884c\u4e3a":23,"\u65e0\u6cd5\u505a\u5230\u5bf9\u4e8e\u5404\u79cd\u8bed\u8a00\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u7684\u9002\u914d":20,"\u65e0\u8bba\u5728\u672c\u5730\u8fd8\u662f\u5728\u4e91\u7aef":19,"\u65e0\u8bba\u662f\u5728\u672c\u5730\u6216\u662f\u4e91\u7aef\u8f6c\u6362":19,"\u65f6":18,"\u662f\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3\u7684\u4ee3\u7801\u751f\u6210\u5668":20,"\u662f\u4e00\u4e2a\u7c7b\u578b\u7684\u6807\u5fd7":21,"\u662f\u4e0d\u5e38\u89c1\u7684\u505a\u6cd5":20,"\u662f\u5404\u4e2a\u5b9e\u73b0\u4e2d\u5171\u4eab\u7684\u5934\u6587\u4ef6":21,"\u662f\u56e0\u4e3ac99\u652f\u6301":20,"\u662f\u6307":21,"\u662f\u7528\u6237\u4f7f\u7528c":21,"\u662fc":21,"\u6682\u65f6\u4e0d\u8003\u8651\u591a\u4e2aparamet":18,"\u66b4\u9732\u8fd9\u4e2a\u6982\u5ff5\u5fc5\u8981\u51fd\u6570":21,"\u6700\u540e\u5220\u9664":23,"\u6700\u5e38\u89c1\u7684\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662fexcept":20,"\u6709\u6807\u51c6\u7684":20,"\u6709\u7684\u65f6\u5019":20,"\u672c\u5217\u8868\u8bf4\u660epaddle\u53d1\u7248\u4e4b\u524d\u9700\u8981\u6d4b\u8bd5\u7684\u529f\u80fd\u70b9":23,"\u672c\u6587\u6863\u63cf\u8ff0paddl":21,"\u673a\u5668\u7ffb\u8bd1":23,"\u6765\u4fdd\u8bc1\u8bad\u7ec3\u8fc7\u7a0b\u53ef\u4ee5\u4ece\u4e2d\u95f4\u72b6\u6001\u91cd\u65b0\u542f\u52a8":18,"\u6765\u786e\u4fdd\u628a":20,"\u6765\u8868\u793apaddle\u5185\u90e8\u7c7b":20,"\u6765\u8fdb\u884c\u8ba8\u8bba":21,"\u6807\u51c6\u8868\u793apaddle\u7248\u672c\u53f7":23,"\u68c0\u67e5\u70b9\u4fdd\u5b58\u7a0b\u5e8f\u6d41\u7a0b":18,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\u901a\u8fc7\u5b9a\u671f\u5411\u78c1\u76d8\u4e0a\u4fdd\u5b58\u4e00\u4efd\u5b58\u50a8\u5728paramet":18,"\u6a21\u578b\u6570\u636e\u68c0\u67e5\u70b9\u7684\u5b9e\u73b0":18,"\u6a21\u578b\u914d\u7f6e\u89e3\u6790":20,"\u6b64\u65f6master\u5c06\u8d1f\u8d23\u542f\u52a8\u4e00\u4e2a\u65b0\u7684train":18,"\u6bcf\u4e00\u4e2a":23,"\u6bcf\u4e00\u4e2a\u6587\u4ef6\u662f\u6570\u636e\u96c6\u7684\u4e00\u4e2ashard":19,"\u6bcf\u4e2a\u503c\u7684\u7c7b\u578b\u53ef\u4ee5\u662f\u6574\u5f62":19,"\u6bcf\u4e2adata":19,"\u6bcf\u4e2aparamet":18,"\u6bcf\u4e2ashard\u5206\u522b\u5b58\u50a8\u5728\u5176\u4e2d\u4e00\u53f0paramet":18,"\u6bcf\u6b21\u8f93\u51fa\u4e00\u4e2adata":19,"\u6bcf\u969410\u5206\u949f":18,"\u6bd4\u5982":19,"\u6bd4\u5982\u6bcf\u969410\u5206\u949f\u6700\u65b0\u7684\u5feb\u7167":18,"\u6bd4\u5982\u6d41\u5f0f\u6570\u636e\u5904\u7406":19,"\u6bd4\u5982imagenet\u8fd9\u4e2a\u6570\u636e\u96c6\u53ef\u80fd\u88ab\u5206\u62101000\u4e2ashard":19,"\u6bd4\u5982mnist":19,"\u6ce8":18,"\u6d4b\u8bd5docker\u955c\u50cf":23,"\u6d6e\u70b9\u578b\u6570\u636e":19,"\u7136\u540e\u5728etcd\u7684":18,"\u7248\u672c\u5206\u652f":23,"\u7248\u672c\u53f7":23,"\u7248\u672c\u53f7rc":23,"\u7248\u672cfork\u51fa\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f":23,"\u73b0\u9636\u6bb5paddle\u6709\u4e00\u4e2a\u95ee\u9898\u662f":20,"\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u8bad\u7ec3\u6570\u636e\u96c6\u901a\u5e38\u4f53\u79ef\u5f88\u5927":19,"\u751f\u4ea7\u73af\u5883\u7684\u65e5\u5fd7\u6570\u636e\u4f1a\u901a\u8fc7\u5b9e\u65f6\u6d41\u7684\u65b9\u5f0f":19,"\u751f\u6210\u5404\u79cd\u8bed\u8a00\u7684\u7ed1\u5b9a\u4ee3\u7801":20,"\u751f\u6210\u6587\u6863":20,"\u751f\u6210\u7684":19,"\u751f\u6210\u7ed9\u5b9asstable\u6587\u4ef6\u5bf9\u5e94\u7684data":19,"\u751f\u6210api\u6587\u6863":20,"\u7528\u4e8e\u4e5f\u53ef\u4ee5\u4e0a\u4f20label":19,"\u7528\u6237\u53ef\u4ee5\u4e0a\u4f20\u81ea\u5df1\u7684\u6570\u636e\u96c6":19,"\u7528\u6237\u53ef\u4ee5\u5b89\u5168\u7684\u91ca\u653e\u67d0\u4e2ac":21,"\u7528\u6237\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u8fd9\u4e2a\u52a8\u6001\u5e93\u6765\u5f15\u5165paddl":21,"\u7528\u6237\u53ef\u4ee5\u7f16\u5199\u7a0b\u5e8f\u5b8c\u6210\u8f6c\u6362":19,"\u7528\u6237\u901a\u8fc7c":21,"\u7531\u4e8e\u5bf9parameters\u7684\u66f4\u65b0\u9700\u8981\u83b7\u53d6parameters\u5185\u5b58\u7684":18,"\u7531\u4e8e\u96c6\u7fa4\u4e2d\u540c\u65f6\u5b58\u5728\u4e24\u53f0\u673a\u5668\u6545\u969c\u7684\u6982\u7387\u6781\u4f4e":18,"\u7531\u4e8ec":20,"\u7684\u547d\u540d\u98ce\u683c\u5e76\u4e0d\u80fd\u9002\u5e94\u5176\u4ed6\u7b2c\u4e09\u65b9\u8bed\u8a00":20,"\u7684\u5934\u6587\u4ef6":20,"\u7684\u63a5\u53e3\u6837\u5f0f":20,"\u7684\u6570\u636e\u6d41\u56fe":19,"\u7684\u6e90\u7801\u91cc\u4f7f\u7528\u4e86":20,"\u7684\u89c4\u8303":20,"\u76ee\u524d\u53ea\u8003\u8651\u52a8\u6001\u6269\u5bb9trainer\u6570\u91cf":18,"\u76ee\u524d\u5d4c\u5165python\u89e3\u91ca\u5668":20,"\u76ee\u524dpaddle\u7684\u8fdb\u7a0b\u6a21\u578b\u662fc":20,"\u76ee\u5f55\u4e0b":21,"\u76f4\u63a5\u4f7f\u7528c\u8bed\u8a00\u7684":20,"\u76f4\u63a5\u5220\u9664\u8fd9\u4e2a\u53c2\u6570\u5373\u53ef":21,"\u76f4\u63a5\u5bfc\u51fa\u5230c\u7684\u63a5\u53e3\u6bd4\u8f83\u56f0\u96be":20,"\u793e\u533a\u53c2\u4e0e\u56f0\u96be":20,"\u793e\u533a\u8d21\u732e\u4ee3\u7801\u5b66\u4e60\u6210\u672c\u9ad8":20,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u7684\u53c2\u6570":18,"\u79bb\u7ebf\u6279\u5904\u7406":19,"\u7b2c\u4e00\u4e2atag\u4e3a":23,"\u7b2c\u4e09\u6b65\u5b8c\u6210\u540e":23,"\u7b2c\u4e8c\u4e2a\u4e3a":23,"\u7b49":21,"\u7b49\u5168\u90e8\u9759\u6001\u5e93\u4e2d\u7684\u76ee\u6807\u6587\u4ef6\u5168\u90e8\u6253\u5305\u540e\u4ea7\u751f\u7684\u6587\u4ef6":21,"\u7b49\u6587\u4ef6":21,"\u7c7b\u4f3c":21,"\u7c7b\u540d\u548cc":20,"\u7c7b\u578b":20,"\u7ea2\u697c\u68a6":46,"\u7ed3\u8bba":20,"\u7f16\u8bd1\u5668\u6ca1\u6709":20,"\u7f16\u8bd1\u578b\u8bed\u8a00":20,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684docker\u53d1\u884c\u955c\u50cf":23,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684ubuntu":23,"\u7f16\u8bd1c":21,"\u7f16\u8bd1master\u5206\u652f\u7684docker\u53d1\u884c\u955c\u50cf":23,"\u7f16\u8bd1ubuntu\u7684deb\u5305":23,"\u800c\u4e0d\u5fc5\u5728\u610fpaddl":21,"\u800c\u4e0d\u652f\u6301pypy\u89e3\u91ca\u5668":20,"\u800c\u4e0d\u66b4\u9732\u6982\u5ff5\u7684\u5b9e\u73b0":21,"\u800c\u51fa\u73b0\u9636\u6bb5\u6027\u7684\u8fd0\u884c\u505c\u6ede":18,"\u800c\u5728cpp\u91cc\u9762\u5b9e\u73b0\u8fd9\u4e2ac\u7684\u63a5\u53e3":20,"\u800c\u591a\u8bed\u8a00\u63a5\u53e3\u9700\u8981\u76f4\u63a5\u8bfb\u53d6\u751f\u6210\u7684\u4e8c\u8fdb\u5236":20,"\u800c\u5bf9\u4e8egolang":20,"\u800c\u5bf9\u4e8egolang\u9519\u8bef\u5904\u7406\u5e94\u8be5\u4f7f\u7528\u8fd4\u56de\u503c":20,"\u800c\u662f\u76f4\u63a5\u4fee\u6539paddl":21,"\u800cswig\u53ea\u80fd\u7b80\u5355\u7684\u66b4\u9732c":20,"\u81ea\u52a8\u6302\u8f7d\u5206\u5e03\u5f0f\u5b58\u50a8\u76ee\u5f55":18,"\u81f3\u4e8e\u4e3a\u4ec0\u4e48\u9700\u8981c":21,"\u826f\u597d\u7684\u6587\u6863":20,"\u83b7\u53d6\u6700\u65b0\u7684\u68c0\u67e5\u70b9\u7684\u6587\u4ef6uuid":18,"\u867d\u7136\u4e0d\u9f13\u52b1\u8fd9\u6837":21,"\u89e3\u91ca\u578b\u8bed\u8a00\u53ea\u80fd\u8c03\u7528\u52a8\u6001\u5e93":20,"\u89e3\u91ca\u6027\u8bed\u8a00\u5b9e\u9645\u8fd0\u884c\u7684\u4e8c\u8fdb\u5236\u662f\u89e3\u91ca\u5668\u672c\u8eab":20,"\u8ba1\u7b97\u8fd9\u4e2a\u6587\u4ef6\u7684md5":18,"\u8ba9paddle\u6838\u5fc3\u4e2d":21,"\u8bad\u7ec3\u4efb\u52a1\u7684\u8fd0\u884c\u53ef\u80fd\u4f1a\u5360\u6ee1trainer\u548cparamet":18,"\u8bad\u7ec3\u548c\u7eaf\u4f7f\u7528":23,"\u8bad\u7ec3\u6a21\u578b\u6b63\u786e\u6027":23,"\u8bb0\u5f55\u4e0b\u6240\u6709\u5931\u8d25\u7684\u4f8b\u5b50":23,"\u8bbe\u7f6e":21,"\u8bc6\u522b\u6570\u5b57":23,"\u8bcd\u5411\u91cf":23,"\u8bed\u610f\u89d2\u8272\u6807\u6ce8":23,"\u8bf4\u660e":18,"\u8bf7\u53c2\u8003":21,"\u8f6c\u6362\u751f\u6210\u7684\u6587\u4ef6\u540d\u4f1a\u662f\u4ee5\u4e0b\u683c\u5f0f":19,"\u8fbe\u5230\u5bb9\u707e\u7684\u76ee\u7684":18,"\u8fd4\u56de\u7b2c\u4e8c\u6b65":23,"\u8fd9\u4e00\u5c42\u8fdb\u884c\u5c01\u88c5":21,"\u8fd9\u4e00\u6982\u5ff5\u4e0d\u518d\u7410\u788e":21,"\u8fd9\u4e09\u4e2a\u5206\u652f":23,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u7684\u8fde\u63a5\u53c2\u6570\u4e0epaddle\u7684\u5176\u4ed6\u4e8c\u8fdb\u5236":21,"\u8fd9\u4e2a\u53c2\u6570\u4e5f\u4e0d\u4f1a\u4e00\u5e76\u5220\u9664":21,"\u8fd9\u4e2a\u5934\u6587\u4ef6\u4e0d\u5047\u8bbe\u5176\u4ed6\u6587\u4ef6\u7684\u5f15\u7528\u987a\u5e8f":21,"\u8fd9\u4e2a\u63a5\u53e3\u9700\u8981\u505a\u5230":20,"\u8fd9\u4e2a\u6587\u4ef6\u5177\u6709\u72ec\u7279\u7684\u8bed\u6cd5":20,"\u8fd9\u4e2a\u76ee\u5f55\u4e2d\u9664\u4e86":21,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u4e2d\u7684\u53e6\u4e00\u4e2a\u9879\u76ee\u662f":21,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u5305\u542b\u4e24\u4e2a\u9879\u76ee":21,"\u8fd9\u4e2a\u9759\u6001\u5e93\u5305\u542b\u4e86paddle\u7684\u5168\u90e8\u7b26\u53f7":21,"\u8fd9\u4e2ainstance\u53ef\u4ee5\u662f\u5355\u4e2a\u503c":19,"\u8fd9\u4e9b\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1\u901a\u5e38\u4f1a\u628a\u6570\u636e\u5207\u5272\u6210\u591a\u4e2a\u5206\u7247\u5206\u5e03\u5f0f\u7684\u5b58\u50a8\u5728\u591a\u4e2a\u8282\u70b9\u4e4b\u4e0a":19,"\u8fd9\u5bf9\u4e8e\u901a\u5e38\u7684java\u7684\u5f00\u53d1\u8005\u6765\u8bf4":20,"\u8fd9\u662f\u56e0\u4e3a":20,"\u8fd9\u6837":21,"\u8fd9\u6837\u4fdd\u8bc1":23,"\u8fd9\u6837\u5c31\u53ef\u4ee5\u5728\u4e91\u7aef\u6267\u884c\u591a\u79cd\u6570\u636e\u7c7b\u8ba1\u7b97\u4efb\u52a1":19,"\u8fd9\u90fd\u9700\u8981\u8fd9\u4e2a\u63a5\u53e3\u6309\u7167\u7ea6\u5b9a\u4fd7\u6210\u7684\u89c4\u5219\u6765\u6ce8\u91ca\u5b8c\u5907":20,"\u8fd9\u91cc\u9700\u8981\u7528\u6237\u989d\u5916\u6ce8\u610f":18,"\u8fdb\u800c\u8fdb\u884c\u4ee3\u7801\u8bc4\u5ba1":23,"\u9009\u62e9cephfs\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u670d\u52a1":19,"\u901a\u5e38":21,"\u901a\u5e38\u6307\u5c06\u4e00\u4e2a\u6574\u4f53\u62c6\u5206\u6210\u591a\u4efd\u7684\u5176\u4e2d\u7684\u4e00\u4efd":18,"\u901a\u8fc7\u6a21\u578b\u63a8\u65adapi\u7684\u5b9e\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u6837\u4f8b":21,"\u9075\u5faa\u4ee5\u4e0b\u6d41\u7a0b":23,"\u90a3\u4e48":21,"\u90fd\u662f\u4e94\u4f4d\u7684\u6570\u5b57":19,"\u90fd\u662fabi\u8c03\u7528\u6807\u51c6\u7684":20,"\u91ca\u653e\u5bf9paramters\u5185\u5b58\u7684\u9501\u5b9a":18,"\u91cc\u6240\u6709\u7684\u7b26\u53f7\u90fd\u5199\u5165\u81ea\u5df1\u7684\u7a0b\u5e8f\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cc":20,"\u91cd\u547d\u540d\u6210":20,"\u94fe\u63a5\u5230\u81ea\u5df1\u7684\u7a0b\u5e8f\u91cc":20,"\u9519\u8bef\u5904\u7406":20,"\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662f\u8fd4\u56de\u503c":20,"\u9519\u8bef\u5904\u7406\u7684\u65b9\u5f0f\u4e5f\u4e0d\u5c3d\u76f8\u540c":20,"\u9664\u6784\u9020\u67d0\u79cd\u7c7b\u578b\u7684\u51fd\u6570":21,"\u9700\u8981\u4fdd\u5b58\u4e00\u4e9b\u9884\u7f6e\u7684\u516c\u5f00\u6570\u636e\u96c6":19,"\u9700\u8981\u5728cmake\u7684\u65f6\u5019":21,"\u9700\u8981\u5c06bugfix\u7684\u5206\u652f\u540c\u65f6merge\u5230":23,"\u9700\u8981\u5f15\u7528":21,"\u9700\u8981\u6709\u7a33\u5b9a\u7684\u5bfc\u51fa\u7b26\u53f7":20,"\u9700\u8981\u6ce8\u610f\u7684\u662f":23,"\u9700\u8981\u88ab\u66b4\u9732\u5230\u5176\u4ed6\u8bed\u8a00":21,"\ufb01xed":56,"abstract":[33,38],"api\u4e2d\u4f7f\u7528":20,"api\u5bfc\u51fa\u7684\u52a8\u6001\u5e93":21,"api\u5bfc\u51fa\u7684\u9759\u6001\u5e93":21,"api\u63a5\u53d7\u7684\u7c7b\u578b\u5168\u662f":21,"api\u63a5\u53e3\u7684\u53c2\u6570\u8f6c\u53d1\u7ed9":21,"api\u65f6":21,"api\u65f6\u6240\u552f\u4e00\u9700\u8981\u5f15\u5165\u7684\u5934\u6587\u4ef6":21,"api\u662f\u591a\u8bed\u8a00api\u7684\u57fa\u7840\u90e8\u5206":21,"api\u66b4\u9732\u7684\u7c7b\u578b":21,"api\u751f\u6210\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u4f1a\u88ab\u5b89\u88c5\u5230":21,"api\u7684\u5b9e\u4f8b":21,"api\u7684\u5b9e\u73b0\u7ec6\u8282":21,"api\u7684\u63a5\u53e3":21,"api\u7684\u65f6\u5019\u63a8\u8350paddle\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":21,"api\u7684\u7f16\u8bd1\u9009\u9879\u9ed8\u8ba4\u5173\u95ed":21,"api\u76ee\u5f55\u7ed3\u6784\u5982\u4e0a\u56fe\u8868\u6240\u793a":21,"api\u83b7\u5f97\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u5b9e\u4f8b":21,"block\u6784\u6210\u4e00\u4e2amodel":18,"book\u4e2d\u6240\u6709\u7ae0\u8282\u529f\u80fd\u7684\u6b63\u786e\u6027":23,"boolean":[9,20],"break":51,"bugfix\u5206\u652f\u4e5f\u662f\u5728\u5f00\u53d1\u8005\u81ea\u5df1\u7684fork\u7248\u672c\u5e93\u7ef4\u62a4":23,"bugfix\u5206\u652f\u9700\u8981\u5206\u522b\u7ed9\u4e3b\u7248\u672c\u5e93\u7684":23,"c99\u662f\u76ee\u524dc\u6700\u5e7f\u6cdb\u7684\u4f7f\u7528\u6807\u51c6":20,"c\u6709\u6807\u51c6\u7684abi":20,"c\u8bed\u8a00\u662f\u6709\u5bfc\u51fa\u7b26\u53f7\u7684\u6807\u51c6\u7684":20,"case":[9,21,22,24,31,32,33,35,39,41,47,51],"char":53,"class":[5,6,7,8,9,10,11,12,13,15,16,20,37,48,55],"const":33,"core\u4e2d\u7684\u6a21\u578b\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u53c2\u6570":21,"core\u4e2d\u8fd9\u4e00\u7c7b\u578b\u63a5\u53e3\u7684\u667a\u80fd\u6307\u9488":21,"core\u662f\u5426\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u5b9e\u4f8b":21,"core\u6982\u5ff5":21,"data\u5230\u5206\u5e03\u5f0f\u5b58\u50a8\u8865\u5145\u8bad\u7ec3\u6570\u636e":19,"deb\u5305":23,"deb\u5305\u7f16\u8bd1\u95ee\u9898":23,"default":[3,7,8,9,10,12,13,15,16,26,36,38,40,41,42,51,53,55,56],"export":[25,48],"final":[10,24,25,33,53,55],"float":[3,7,8,9,11,13,24,33,35,40,46,49,53],"function":[3,5,9,10,11,13,16,17,22,24,31,33,35,36,38,47,48,51,54,55,56],"golang\u53ef\u4ee5\u4f7f\u7528":20,"golang\u7684":20,"h\u5e76\u4e0d\u56f0\u96be":20,"imagenet\u6570\u636e\u96c6\u7b49":19,"images\u6570\u636e\u96c6\u4e0a\u4f20\u5230\u4e91\u7aef\u7684":19,"import":[3,5,8,9,16,24,31,35,41,46,47,48,49,51,53,55,56],"instance\u4e0e\u751f\u6210\u6570\u636e\u96c6\u65f6":19,"instance\u5305\u6db5\u4e24\u4e2a\u503c":19,"instance\u662f\u4e00\u6a21\u4e00\u6837\u7684":19,"int":[3,7,8,9,10,13,20,21,22,33,40,51,53,54],"interface\u6587\u4ef6\u7684\u5199\u6cd5\u975e\u5e38":20,"job\u5b8c\u6210\u8f6c\u6362":19,"list\u4f5c\u4e3a\u68c0\u67e5\u5217\u8868":23,"long":[2,9,10,13,26,35,54,55],"model\u505a\u5206\u652f\u7ba1\u7406":23,"ndarray\u7c7b\u578b\u7684\u503c\u548c\u6574\u578b\u7684\u503c":19,"new":[3,9,13,17,22,32,34,41,42,47,51,54,55],"note\u7684\u4e66\u5199":23,"null":[9,33,38,53],"paddle\u4e00\u4e2a\u52a8\u6001\u5e93\u53ef\u4ee5\u5728\u4efb\u4f55linux\u7cfb\u7edf\u4e0a\u8fd0\u884c":20,"paddle\u4f7f\u7528git":23,"paddle\u5185\u5d4c\u7684python\u89e3\u91ca\u5668\u548c\u5916\u90e8\u4f7f\u7528\u7684python\u5982\u679c\u7248\u672c\u4e0d\u540c":20,"paddle\u5185\u90e8\u7684\u7c7b\u4e3ac":20,"paddle\u5f00\u53d1\u8fc7\u7a0b\u4f7f\u7528":23,"paddle\u6bcf\u6b21\u53d1\u65b0\u7684\u7248\u672c":23,"paddle\u6bcf\u6b21\u53d1\u7248\u672c\u9996\u5148\u8981\u4fdd\u8bc1paddl":23,"paddle\u7684\u4e3b\u7248\u672c\u5e93\u9075\u5faa":23,"paddle\u7684\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0\u5305\u62ec\u4e00\u4e0b\u51e0\u4e2a\u65b9\u9762":20,"paddle\u7684\u7c7b\u578b\u5168\u90e8\u9000\u5316\u6210":21,"paddle\u7684\u94fe\u63a5\u65b9\u5f0f\u6bd4\u8f83\u590d\u6742":20,"paddle\u7684c":21,"paddle\u8bad\u7ec3\u4efb\u52a1":19,"paddle\u8def\u5f84\u4e0b":21,"paddle\u9700\u8981\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3":20,"paddle\u9700\u8981\u66b4\u9732\u7684api\u5f88\u591a":21,"paddle\u9759\u6001\u5e93\u94fe\u63a5\u590d\u6742":20,"paddle_\u7c7b\u578b\u540d":21,"paddle_\u7c7b\u578b\u540d_\u51fd\u6570\u540d":21,"paddlepaddle\u63d0\u4f9b\u4e13\u7528\u7684":19,"patch\u53f7":23,"patch\u53f7\u52a0\u4e00":23,"public":[13,33,36,41,42,55],"reader\u7684\u4f7f\u7528\u65b9\u5f0f\u90fd\u662f\u4e00\u81f4\u7684":19,"reader\u8f93\u51fa\u7684data":19,"release\u9875\u9762":23,"return":[3,8,9,10,12,13,15,16,19,24,31,33,41,47,49,51,52,53,56],"s3\u4e4b\u7c7b\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u4e4b\u4e0a":19,"server\u4e4b\u4e0a":18,"server\u4e4b\u95f4\u7684\u7f51\u7edc\u5e26\u5bbd":18,"server\u4f1a\u6682\u505c\u53c2\u6570\u66f4\u65b0\u5e76\u7b49\u5f85":18,"server\u4f1a\u83b7\u53d6parameters\u5185\u5b58\u7684":18,"server\u5185\u5b58\u4e2d\u7684\u6a21\u578b\u6570\u636e\u7684\u5b8c\u6574\u955c\u50cf":18,"server\u540c\u6b65\u7684\u4fdd\u5b58\u4e00\u4e2a\u7279\u5b9a\u65f6\u95f4\u70b9\u7684\u5168\u5c40\u68c0\u67e5\u70b9":18,"server\u5728\u96c6\u7fa4\u4e2d\u542f\u52a8\u540e":18,"server\u6545\u969c\u540e\u88abkubernetes\u91cd\u65b0\u542f\u52a8":18,"server\u6b64\u65f6\u8fd8\u9700\u8981\u901a\u8fc7\u7f51\u7edc\u8bbf\u95ee\u5206\u5e03\u5f0f\u5b58\u50a8\u4ee5\u4fdd\u5b58\u5feb\u7167":18,"server\u751f\u6210\u4e00\u4e2auuid":18,"server\u7684\u5355\u70b9\u6216\u591a\u70b9\u540c\u65f6\u6545\u969c":18,"server\u7684\u6570\u636e\u5feb\u7167":18,"server\u7684\u68c0\u67e5\u70b9\u5404\u81ea\u72ec\u7acb\u4fdd\u5b58":18,"server\u7b2c\u4e00\u6b21\u542f\u52a8\u6216\u4efb\u610f\u65f6\u95f4paramet":18,"short":[9,10,24,53,54,55],"static":[21,41],"super":33,"swig\u652f\u6301\u7684\u8bed\u8a00\u6216\u8005\u89e3\u91ca\u5668\u6709\u5c40\u9650":20,"swig\u66b4\u9732\u7684\u63a5\u53e3\u4fdd\u7559\u4e86c":20,"swig\u751f\u6210\u7684\u4ee3\u7801\u4e0d\u80fd\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":20,"swig\u76f4\u63a5\u8bfb\u53d6c":20,"swig\u9700\u8981\u5199\u4e00\u4e2ainterface\u6587\u4ef6":20,"switch":[21,41,55],"tag\u4e3a":23,"throw":41,"true":[3,7,8,9,10,12,13,15,16,22,24,31,33,38,40,41,49,53,54,55,56],"try":[11,17,22,35,47,53],"type\u5b57\u6bb5\u5747\u4e0d\u5c3d\u76f8\u540c":21,"ubuntu\u5b89\u88c5\u5305\u7684\u529f\u80fd\u6b63\u786e\u6027":23,"void":[20,21,33],"while":[2,3,7,13,22,26,31,38,47,51,55,56],AGE:[41,42],AND:53,ARE:53,AWS:[19,34,43,44],Age:52,And:[3,8,9,11,13,22,28,32,40,41,42,46,49,53,55,56],But:[3,9,10],EOS:9,For:[2,3,8,9,11,13,16,22,24,25,26,31,33,35,36,37,38,40,46,48,49,51,55,56],Going:55,Has:3,IDs:[13,51],Ids:51,Into:41,Its:[3,31,41,53],Not:[16,17,36],ONE:3,One:[8,10,15,31,33,38,47,51,55,56],QoS:42,THE:3,TLS:[16,41],That:[9,13,22,26,38,40],The:[2,3,5,6,7,8,9,10,11,13,15,16,17,21,22,24,25,26,27,28,31,32,33,35,36,38,40,41,42,46,47,48,49,51,52,53,54,55,56],Their:[3,9,17],Then:[5,9,25,26,31,32,33,35,41,42,46,48,53,54,55],There:[8,9,13,15,16,17,24,26,28,35,41,47,48,49,50,51,53,56],These:[36,40,48,54],USE:53,USING:53,Use:[3,16,22,26,33,35,38,39,41,53],Used:10,Useful:3,Using:[17,42,55],VPS:41,WITH:32,Will:[13,15],With:[3,9,10,24,47,54],Yes:26,___fc_layer_0__:41,__init__:33,__list_to_map__:53,__main__:49,__meta__:53,__name__:49,__rnn_step__:31,_error:47,_link:10,_proj:9,_res2_1_branch1_bn:49,_source_language_embed:[31,46],_target_language_embed:[31,46],aaaaa:19,aaaaaaaaaaaaa:41,abc:9,abl:[9,16,47,55],about:[5,9,10,24,35,37,38,41,45,54,55,56],abov:[3,5,9,16,17,24,26,35,41,42,47,49,51,54],abs:[10,47],absolut:[2,36],academ:52,acceler:[18,40],accept:[3,5,13,16,22,51,54],acceptor:54,access:[2,9,10,16,31,56],accessmod:41,accident:52,accomplish:26,accord:[2,3,8,9,31,32,36,37,38,40],accordingli:[5,33],accordingto:54,accrod:10,accumul:17,accuraci:[8,33,51,52,55],achiev:[35,48],ack:38,acl:55,aclimdb:55,across:9,act:[9,10,24,31,51],act_typ:51,action:[41,52],activ:[0,5,9,10,14,24,25,33,38,51,55],activi:10,actual:[3,9,24],adadelta:51,adagrad:51,adam:[16,51,55,56],adamax:51,adamoptim:[46,51,55,56],adapt:[8,11,24,55,56],add:[3,9,10,13,24,25,32,33,35,40,51,53],add_input:33,add_test:33,add_to:9,add_unittest_without_exec:33,addbia:33,added:[3,8,33],adding:49,addit:[9,10,26,51],address:[17,26,35,38],addrow:33,addtion:36,addtolay:9,adject:55,adjust:24,admin:52,adopt:54,advanc:[31,35,38],advantag:55,adventur:52,adverb:55,adversari:22,advic:35,affect:9,afi:3,aforement:36,after:[9,13,25,28,31,33,36,38,40,41,42,47,48,49,51,53,54,55,56],again:[16,17,35],against:41,age:[13,53],agg_level:9,aggreg:41,aggregatelevel:9,aid:35,aim:[55,56],aircraft:56,airplan:48,aistat:9,alex:[9,55],alexnet_pass1:40,alexnet_pass2:40,algorithm:[9,11,24,31,46,48,55,56],alia:[6,7],align:[9,10,13,56],all:[0,3,7,8,9,11,16,17,21,24,26,31,32,33,35,36,37,38,40,41,42,46,47,49,51,52,53,54,55,56],alloc:[7,33,40],allow:[16,26,32,33,35,38,41,51],allow_only_one_model_on_one_gpu:[37,38,40],almost:[10,24,36,46],along:55,alreadi:[17,26,35,36,38,41,42,55],alreali:[37,56],also:[2,3,9,10,13,16,22,25,26,31,33,35,36,42,47,48,49,51,54,55],although:24,alwai:[5,9,10,15,22,24,38,41,56],amaz:48,amazon:[41,42,51,55],amazonaw:41,amazonec2fullaccess:41,amazonelasticfilesystemfullaccess:41,amazonroute53domainsfullaccess:41,amazonroute53fullaccess:41,amazons3fullaccess:41,amazonvpcfullaccess:41,ambigu:[22,54],amd64:41,amend:32,american:48,among:[41,55],amount:[35,55],analysi:[24,35,50,54],analyz:[51,55],andd:41,ani:[2,3,9,10,13,16,17,22,31,32,35,41,51,53,56],anim:52,annot:54,annual:54,anoth:[3,9,16,26,38,41,54,55],ans:41,answer:[24,41,54],anyth:[13,22,32,41,54],api:[13,15,16,23,25,33,35,41,45,47,51,53,55],api_trainer_config_helpers_data_sourc:53,api_trainer_config_helpers_lay:31,api_trainer_config_helpers_layers_context_project:53,api_trainer_config_helpers_layers_cos_sim:53,api_trainer_config_helpers_layers_data_lay:53,api_trainer_config_helpers_layers_embedding_lay:53,api_trainer_config_helpers_layers_fc_lay:53,api_trainer_config_helpers_layers_pooling_lay:53,apiserv:41,apivers:[41,42],apo:56,appar:56,appear:54,append:[3,15,22,31,33,36,53],append_gradient_machin:15,appleclang:25,appleyard:35,appli:[0,9,10,31,33,48,51],applic:[26,35,41,42,55],appreci:[32,55],approach:9,apt:[25,28,48],arbitrari:9,architectur:[46,54,55,56],architecur:55,archiv:[13,20,21],arg:[3,8,9,10,13,24,37,47,48,49,51,53,54,55],arg_nam:9,argu:54,argument:[3,5,9,13,31,33,38,39,46,47,48,49,53,54,55,56],argv:49,arn:41,around:[3,9,41],arrai:[5,9,13,15,22,24,49],art:[24,54],articl:[36,42],artifact:41,artifici:47,artist:52,arxiv:[9,10,47,55],ask:17,aspect:55,assign:[9,38,41],associ:[54,55,56],assum:[9,31,40,46],assur:2,astyp:[22,47],asyc:17,async:[17,37],async_count:38,async_lagged_grad_discard_ratio:38,async_lagged_ratio_default:[37,38],async_lagged_ratio_min:[37,38],asynchron:[17,38],atla:25,atlas_root:25,attenion:10,attent:[9,10,26,56],attitud:55,attr:[7,9,10],attribut:[3,9,10,14,33,46,54],auc:37,auc_evalu:8,aucvalidationlay:38,authent:41,author:[41,49],authorized_kei:36,autmot:32,auto:[20,33,35,50,53],autom:[41,56],automak:25,automat:[9,16,25,31,33,36,37,38,41,53,54,56],automaticli:9,automobil:48,avail:[17,25,41],availabel:25,averag:[8,9,12,38,49,51,53,54,55,56],average_test_period:[37,38,54],average_window:55,averagepool:9,avg:[35,51],avgcost:[51,53,55,56],avgpool:[9,51],avoid:[17,35],avx:[25,26,28],await:42,awar:[16,17,26,41],aws_account_id:41,awsaccountid:41,awskeymanagementservicepowerus:41,b2t:46,b363:42,b8561f5c79193550d64fa47418a9e67ebdd71546186e840f88de5026b8097465:42,ba5f:41,back:[3,17,26],backward:[6,9,10,31,33,38,40],backward_first:31,backwardactiv:33,bag:[51,55],baidu:[0,9,24,28,32,42,46],baik:46,balanc:[38,41,47],balasubramanyan:55,bank:54,bardward:10,bare:42,barrier:38,barrierstatset:35,base:[9,10,12,13,16,24,28,31,32,33,35,36,38,41,46,47,51,53,55,56],baseactiv:9,baseev:15,basematrix:33,basenam:8,basepoolingtyp:[9,10],basestr:[7,8,9,10,12,15,53],bash:[26,41,42],bashrc:25,basic:[3,9,15,26,32,33,51,52,55],batch:[3,9,10,11,13,15,16,17,19,33,36,38,41,42,47,48,49,51,53,54,55,56],batch_0:49,batch_id:15,batch_norm:10,batch_norm_typ:9,batch_read:[19,22],batch_siz:[3,13,24,36,46,47,48,51,53,55,56],batchsiz:[9,33],bbbbb:19,bcd:9,bcebo:13,beam:[9,31,38,54,56],beam_gen:[9,31],beam_search:[15,31],beam_siz:[9,31,37,38,40],beamsiz:56,becaus:[5,9,13,16,17,22,31,32,33,40,41,48,51,54],becom:[32,35],been:[3,25,32,48,51,54,55,56],befor:[5,9,10,17,22,26,32,36,41,48,53,55,56],begin:[5,8,9,33],beginiter:[15,16],beginn:31,beginpass:[15,16],begintrain:16,behavior:35,being:[17,22,47],belong:[9,56],below:[3,9,13,17,22,31,33,35,36,41,47,48,51,53],benefit:10,bengio:9,bertolami:55,besid:[2,9,13,56],best:[9,25,26,38,51,53,55,56],best_model_path:54,besteffort:42,beta1:11,beta2:11,beta:49,better:[9,10,24,36,41,47,53],between:[9,11,17,21,24,32,41,47,51,52,55,56],bgr:49,bi_gru:10,bi_lstm:10,bia:[9,10,11,31,33,49],bias:[9,33],bias_attr:[9,10,24,31],bias_param_attr:10,biases_:33,biasparameter_:33,biassiz:33,bidi:42,bidirect:[10,31,54,56],bidirectional_lstm_net:55,big:35,bigger:17,biggest:55,bilinear:9,bilinear_interpol:9,bilinearfwdbwd:35,bin:[25,26,36,41,42,53],binari:[3,8,9,13,35,41,46,51,55],bird:48,bison:25,bit:51,bitext:56,bla:25,blank:[9,41],block:[9,18,24,33,35,38,49,55],block_i:9,block_x:9,blog:55,bn_attr:10,bn_bias_attr:10,bn_param_attr:10,bollen:55,book:13,bool:[3,7,8,9,10,12,13,33,38,40,51,53,55],boot:[9,31],boot_bia:9,boot_bias_active_typ:9,boot_lay:[9,31],boot_with_const_id:9,bootstrap:25,bos_id:[9,31],both:[0,6,7,9,10,16,17,26,31,33,35,41,47,49,51],bottleneck:[35,49],bottom:55,bow:[19,51,55],box:35,branch:[9,16,23,32],breadth:[38,56],brendan:55,brew:25,briefli:35,broadcast:17,brows:26,browser:[26,41],bryan:55,bucket_nam:41,buf_siz:13,buffer:[3,13,22,38],buffered_read:22,bug:41,bui:55,build:[0,13,24,26,29,38,41,43,44,46,48,49,51,53,55,56],build_dict:13,built:[0,25,26,47,54],bunch:[35,51],bunk:55,button:[32,41],c11:20,c99:21,c99e:41,cach:[51,53,54],cache_pass_in_mem:[3,51,53,54],cachetyp:[3,51,53,54],calc_batch_s:[3,54],calcul:[3,8,9,10,11,17,31,33,35,38,40,47,53],call:[3,9,10,16,24,26,31,33,35,38,41,48,49,51,55,56],callabl:[3,9,13],callback:33,caller:41,caltech:48,can:[2,3,5,7,8,9,10,13,16,17,22,24,25,26,28,31,32,33,35,36,37,38,40,41,42,46,47,48,49,51,53,54,55,56],can_over_batch_s:[3,54],candid:9,cannot:33,caoi:56,capabl:[25,55],capac:41,capi:20,capi_prvi:21,caption:[24,56],captur:[24,36],card:36,care:[10,22,37,38,52],carefulli:[36,38,49],cat:[26,48,49,55],categor:54,categori:[9,13,17,51,55],categorig:13,categoryfil:42,caus:17,caution:[41,42],ccb2_pc30:56,cde:9,cdn:13,ceil:9,ceil_mod:9,cell:[9,10,55],center:3,ceph:[19,42],certain:[2,37,54],certif:[16,41],cffi:20,cfg:42,cgo:20,chain:[13,33],challeng:17,chanc:[16,33,51],chang:[9,13,22,24,26,31,32,33,35,38,41,51,55],channel:[9,35,36,49],channl:[36,49],char_bas:53,charact:[51,53],character:24,characterist:[40,48],check:[3,13,24,25,26,32,38,40,41,52],check_align:13,check_eq:33,check_fail_continu:3,check_l:33,check_sparse_distribution_batch:[37,38],check_sparse_distribution_in_pserv:[37,38],check_sparse_distribution_ratio:[37,38],check_sparse_distribution_unbalance_degre:[37,38],checkgrad:38,checkgrad_ep:38,checkout:32,children:52,china:26,chines:50,chmod:[25,41],choic:[26,52],choos:[38,51,53],chosen:[2,52,56],chunk:[47,54],chunk_evalu:8,chunk_schem:8,chunktyp:8,cifar:[47,48],cifar_vgg_model:48,claim:41,claimnam:41,clang:[20,25,26,32],class1:55,class2:55,class_dim:55,classfic:[49,55],classfiic:48,classic:[9,24],classif:[3,5,9,40,49,50,51,55,56],classifc:55,classifi:[47,48,49,51,55],classification_cost:[48,51],classification_error_evalu:[8,47,51,55,56],classification_error_printer_evalu:8,claster:41,clean:[5,53],cleric:52,cli:41,click:[32,35,41],client:32,clip:[7,38,51,55],clock:9,clone:[25,26],close:[3,22],closer:24,cloud:17,cls:51,cludform:41,cluster:[16,17,37,38,42,51,56],cluster_train:36,cm469:41,cmake3:25,cmake:[21,25,33,35],cmakelist:33,cmatrix:[20,21],cmd:42,cna:9,cname:41,cnn:[9,42,49,51],code:[0,3,5,13,16,22,24,25,26,27,31,33,34,35,36,41,42,47,51,52],coeff:9,coeffici:9,collabor:17,collect:[9,13,15,24,52],collectbia:33,colleg:52,color:[48,49],colour:13,column:[8,9,22,33,46,56],column_sum_evalu:8,colunm:56,com:[9,10,13,25,26,28,32,41,42,49],combin:[9,10,13,15,47,53,55],come:55,comedi:52,comma:[38,46],command:[2,5,24,25,26,28,32,33,34,35,36,41,42,43,44,46,47,48,49,53,54,55],commandlin:[35,55],commenc:51,comment:[10,32,51,55],commnun:36,common:[31,33,37],common_util:[36,53],commonli:[31,35,40],commun:[0,17,33,36,41],compani:55,compar:[33,47,51],compat:3,compet:55,competit:47,compil:[25,26,32,33],complet:[0,5,9,10,13,15,17,33,41,42,51],complex:[2,3,10,22,31,35,51],complic:9,compon:33,compos:[13,16,47,54],composenotalign:13,comput:[9,10,16,17,24,25,26,31,33,35,40,41,51,53,54,55],computation:31,conat:9,concat:56,concat_lay:31,concaten:10,concept:[3,16,26,31],concern:16,concurr:17,concurrentremoteparameterupdat:38,condit:[9,31,36,42,56],conduct:35,conf:[5,9,36,46,47,49,56],conf_paddle_gradient_num:41,conf_paddle_n:41,conf_paddle_port:41,conf_paddle_ports_num:41,conf_paddle_ports_num_spars:41,confid:55,config:[3,7,9,10,24,33,36,37,38,41,42,46,47,48,49,51,55,56],config_:38,config_arg:[37,38,40,49,51,54,55],config_bas:[8,9,10,15],config_fil:54,config_gener:[36,53],config_lay:33,config_pars:[5,33],configur:[1,2,3,5,9,24,32,33,35,38,46,48,49,55,56],conflict:32,confront:56,congest:38,conll05st:54,conll:[13,54],connect:[2,10,24,33,41,42,47,48,49,51,53,55],connectionist:[9,55],connor:55,consequ:[9,10],consid:[8,9,11,25,26,35,40,48],consider:[3,10],consist:[9,13,22,48,49,51,54,56],consol:[35,41],constant:33,construct:[3,5,16,31,53],construct_featur:53,constructor:33,consum:[17,55],contact:17,contain:[3,8,9,10,12,13,15,16,27,28,31,32,36,41,48,49,51,52,55,56],containerport:41,contemporan:55,content:[42,54,55],context:[9,10,13,31,46,51,53,54,55,56],context_attr:10,context_len:[9,10,51,53],context_proj_nam:10,context_proj_param_attr:10,context_project:[10,53],context_start:[9,10,51],contibut:32,contin:41,continu:[3,17,28,38],contrast:[9,56],contribut:[0,27,34,55],contributor:0,control:[7,26,38,41,42,56],conv:10,conv_act:10,conv_attr:10,conv_batchnorm_drop_r:10,conv_bias_attr:10,conv_filter_s:10,conv_num_filt:10,conv_op:9,conv_pad:10,conv_param_attr:10,conv_strid:10,conv_with_batchnorm:10,conveni:[16,36],converg:[36,47,55],convert:[3,5,13,19,22,31,46,48,49,51,53],convlay:9,convolut:[9,10,47,49,53],convoper:9,convtran:9,convtranslay:9,cool:[3,32],coordin:17,copi:[15,16,26,41,47,53],copy_shared_paramet:47,copytonumpymat:47,core:[3,7,21,38,56],coreo:41,corespond:54,corpora:56,corpu:[13,54],correct:[3,8,9,33,41],correctli:[8,13,33,47],correl:[24,48,55],correspoind:16,correspond:[3,5,16,24,31,33,48,52,54,55,56],corss_entropi:16,cos:9,cos_sim:53,cosin:[9,53],cost:[5,11,15,16,24,38,47,51,53,55,56],cost_id:9,could:[3,5,9,13,15,16,22,26,35,36,41,51,53],count:[17,22,35,38,40,42,46,53,54,55,56],counter:17,coupl:24,coverag:25,coveral:25,coveralls_uploadpackag:25,cpickl:[49,53],cpp:[20,21,32,33,35,51,53,56],cpu:[2,3,7,9,23,25,28,35,38,42,47,54,55,56],cpuinfo:26,cpusparsematrix:21,craftsman:52,crash:[17,35,36,38],crazi:36,creat:[5,7,9,13,15,16,17,24,25,26,33,36,38,46,47,48,56],create_bias_paramet:33,create_input_paramet:33,createargu:47,createfromconfigproto:[5,47],createstack:41,creation:41,creationd:41,creator:[13,19],credit:47,crf:54,crime:52,critic:55,crop:49,crop_siz:49,cross:[9,51,54],cross_entropi:[9,16,47],cross_entropy_with_selfnorm:9,csc:33,cslm:56,csr:33,csv:52,ctc:8,ctc_error_evalu:8,ctest:26,ctrl:[36,53],ctx:54,ctx_0:54,ctx_0_slot:54,ctx_n1:54,ctx_n1_slot:54,ctx_n2:54,ctx_n2_slot:54,ctx_p1:54,ctx_p1_slot:54,ctx_p2:54,ctx_p2_slot:54,cub:48,cuda:[25,26,28,35,36,38],cuda_dir:[37,38],cudaconfigurecal:35,cudadevicegetattribut:35,cudaeventcr:35,cudaeventcreatewithflag:35,cudafre:35,cudagetdevic:35,cudagetdevicecount:35,cudagetdeviceproperti:35,cudagetlasterror:35,cudahostalloc:35,cudalaunch:35,cudamalloc:35,cudamemcpi:35,cudaprofilerstart:35,cudaprofilerstop:35,cudaruntimegetvers:35,cudasetdevic:35,cudasetupargu:35,cudastreamcr:35,cudastreamcreatewithflag:35,cudastreamsynchron:35,cudeviceget:35,cudevicegetattribut:35,cudevicegetcount:35,cudevicegetnam:35,cudevicetotalmem:35,cudnn:[9,12,25,28,38],cudnn_batch_norm:9,cudnn_conv:9,cudnn_conv_workspace_limit_in_mb:[37,38],cudnn_convt:9,cudnn_dir:[37,38],cudrivergetvers:35,cuinit:35,cumul:9,curl:[25,41],current:[3,9,17,24,26,31,32,33,36,38,41,51,55,56],current_word:31,currentcost:[51,53,55,56],currentev:[51,53,55,56],curv:[16,48,54],custom:[2,3,16,33,41,52,55],custom_batch_read:22,cutoff:13,cycl:17,cyclic:9,cython:20,d3e0:41,dai:56,daili:55,dalla:3,dan:54,danger:3,darwin:41,dat:[13,36,53],data:[1,2,3,5,8,10,11,15,16,19,25,26,33,35,36,37,38,40,43,49,52],data_batch_gen:47,data_dir:[46,48,55,56],data_feed:13,data_fil:24,data_initialz:51,data_lay:[3,24,31,47,48,51,53,54],data_nam:13,data_read:[13,22],data_reader_creator_random_imag:22,data_sourc:47,data_typ:[9,13],databas:[13,55],datadim:9,datalay:9,dataprovid:[2,24,31,36,53,54],dataprovider_bow:51,dataprovider_emb:51,dataproviderconvert:5,datasci:9,dataset:[1,3,19,22,24,38,46,48,49,51,54,55],date:54,db_lstm:54,dcgan:47,dcmake_install_prefix:25,dead:17,deal:[32,47],deb:[27,28],debian:[26,27],debug:3,decai:[11,48],decid:[16,22],declar:[9,53],decod:[9,10,31,54,56],decoder_boot:31,decoder_group_nam:31,decoder_input:31,decoder_mem:31,decoder_prev:10,decoder_s:31,decoder_st:[10,31],deconv:9,deconvolut:9,decor:[3,13,33],decreas:24,decrypt:41,deep:[0,9,24,26,29,35,47,48,49,51,54],deeper:[24,26,49],deer:48,def:[3,9,13,16,19,22,24,31,33,47,49,51,53,54],defalut:[9,38,40],default_devic:40,default_valu:40,defferenct:3,defin:[2,3,9,10,13,16,17,22,24,31,33,36,38,46,47,48,53,54],define_py_data_sources2:[3,24,48,49,51,53],defini:56,definit:[3,13,17,24,26,46,51,55],degre:9,del:53,delai:38,delar:51,delet:17,deletestack:41,delimit:[8,52,53],demand:17,demo:[9,13,31,36,42,43,46,47,48,49,50,51,52,53,54,55,56],demograph:52,demolish:42,demonstr:[24,31,47,53],denot:[40,51,52,54],dens:[3,9,13,33,41,51,53],dense_vector:[3,5,9,13,24,53],dense_vector_sequ:13,depend:[17,24,26,28,36,40,48,52],deploi:[36,40],deploy:[36,41],deriv:[6,16],descent:[9,17],describ:[16,24,33,41,42,47,51,54],describestack:41,describestackev:41,describestackresourc:41,descript:[5,25,31,39,41,48,53],deseri:15,design:[3,9,13,20,55],desir:[17,41,42,46],destructor:33,detail:[3,5,7,9,10,11,31,32,33,35,36,39,40,41,42,46,47,49,51,53,55,56],detect:8,determin:[3,9,13,33,47],dev:[25,26,48,53,56],devel:25,develop:[0,23,25,32,37,38,56],deverlop:38,deviat:7,devic:[7,38,56],deviceid:40,devid:[9,38],dez:55,diagnos:36,diagram:49,dict:[3,13,15,51,53,55,56],dict_dim:55,dict_fil:[8,31,51,54],dict_siz:13,dictionai:51,dictionari:[3,8,9,13,15,16,31,40,49,51,53,54,55,56],dictsiz:56,did:3,differ:[3,8,9,17,24,26,31,32,33,36,38,41,42,46,48,49,51,55,56],difficult:24,dig:[26,35,41],digit:[3,9],dim:[13,33,46,49,51,55],dimens:[6,9,12,13,33,40,46,51,53,55],dimension:[3,24,31,33,47,51],dimenst:46,dimes:9,din:53,dir:[36,49,51,53,54,55,56],dirctori:26,direct:[9,10,26,49,54],directli:[2,3,10,24,26,36,42,55],directori:[2,19,25,26,32,35,36,38,42,48,49,51,53,54,55,56],diretcoti:49,dis_conf:47,dis_train:47,dis_training_machin:47,disabl:3,discard:[13,17,38],discount:9,discov:[17,54],discoveri:41,discrep:35,discrimin:47,discriminator_train:47,discuss:16,disk:42,dispatch:[36,38],disput:56,dist_train:16,distanc:8,distibut:46,distinguish:[36,47,56],distribut:[9,25,34,42,43,44,47,51,54],distribute_test:[37,38],distributedli:33,disucss:16,divid:[11,37,48,56],diy_beam_search_prob_so:[37,38],dmkl_root:25,dns:41,do_forward_backward:22,doc:[5,10,13,25,26,36],docker:[23,27,41,43,44],docker_build:16,docker_push:16,dockerfil:26,dockerhub:26,doctor:52,document:[3,5,10,25,32,40,48,51,53,54,55],documentari:[3,52],doe:[3,5,10,17,22,24,28,31,33,35,51,53,54],doesn:[7,9,13,16,22,32,35,42,56],dog:[48,49],doing:35,domain:41,don:[10,16,22,24,41,55],done:[9,10,17,31,35,41,47,55],dopenblas_root:25,dot:[38,49,56],dot_period:[38,40,47,48,53,55,56],dotmuloper:9,dotmulproject:9,doubl:[3,25,38],down:[35,51],download:[13,17,26,28,47,48,51,54,55],download_cifar:48,downsampl:48,doxygen:[25,32],dpkg:28,drama:52,drop:3,drop_rat:7,dropout:[7,9,33,51],dropout_r:10,drwxr:42,dtoh:35,dtype:[5,24,49],dubai:56,due:[52,53],duplic:52,durat:35,dure:[2,3,9,17,24,32,33,37,38,41,51,53,54,56],durn:3,dwith_c_api:21,dwith_doc:25,dwith_profil:35,dwith_python:21,dwith_swig_pi:21,dwith_tim:35,dynam:[2,3,21,22,25,35,38],dynamic_cast:33,each:[2,3,5,8,9,12,13,15,17,22,24,26,31,32,33,36,38,40,41,46,48,49,51,52,53,54,55,56],each_feature_vector:6,each_meta:53,each_pixel_str:3,each_sequ:9,each_time_step_output:6,each_timestep:9,each_word:3,eaqual:9,eas:[13,22,49],easi:[0,22,26,33,36,51],easier:[16,22,33],easili:[16,22,24],echo:[26,53,55],edit:[8,26,41],editor:[26,32],edu:[13,41,42,48],educ:52,eeoi3ezpr86c:41,effect:[3,38,41],effici:[0,2,3,31,33],efg:9,efs:41,efs_dns_nam:41,efsvol:41,eight:54,either:[9,13,15,16,35,51,53],elb:41,elbapis:41,elec:51,electron:[42,51],elem_dim:9,element:[3,5,8,9,10,13,15,22,51,55,56],elif:[16,53],elimin:54,els:[9,16,26,33,49,51,53],emac:[26,32],emb:[42,51],embed:[16,31,50,53,55],embedd:54,embedding_lay:[31,51,53],embedding_nam:[9,31],embedding_s:[9,31],emphas:35,empir:9,emplace_back:33,emploi:[31,52],empti:[8,13,17,24],emul:56,enabl:[3,7,35,36,38,41],enable_grad_shar:[37,38],enable_parallel_vector:38,enc_proj:[10,31],enc_seq:10,enc_vec:31,encod:[10,31,56],encoded_proj:[10,31],encoded_sequ:[10,31],encoded_vector:31,encoder_last:9,encoder_proj:31,encoder_s:31,encrypt:41,encrypt_decrypt:41,end:[3,8,9,22,24,31,38,46,54,55,56],end_pass:16,enditer:[15,16],endpass:[15,16],endpoint:41,endtrain:16,engin:[0,35,52],english:[3,9,56],enjoi:26,enough:24,ensembl:10,ensur:[3,17,33],enter:[26,52],entir:[9,10,55],entri:[13,33,41,52],entropi:[9,51,54],enumer:[6,51,53],enumerate_data_types_of_data_lay:13,env:[32,41],environ:[16,25,26,28,35,36,37,38,41,42,47,48,53],eol:32,eos_id:[9,31],epel:25,epoch:52,epsilon:11,equal:[9,10,17,38],equat:[9,10,11,26],equilibrium:47,equip:[25,31],equival:[9,16],error:[7,8,9,11,16,24,28,33,36,38,41,48,49,51,52,53,55,56],error_clipping_threshold:7,especi:[3,10,54],essenc:16,essenti:[9,16,25,54,56],estat:24,estim:[9,16],eta:42,etc:[13,22,26,36,37,40,41,55,56],etcd:17,eth0:[36,41],ethternet:36,eval:[8,51,53,55,56],eval_bleu:56,evalu:[2,9,14,15,35,36,51,55,56],evaluate_pass:55,evalut:[24,56],even:[16,22,35,38,55],evenli:41,event:42,event_handl:[15,16],everi:[2,3,8,9,10,13,16,17,31,32,33,38,51,54,55,56],everyth:[24,32],exactli:[3,8,9,10,26,41,54],exampl:[2,3,9,10,11,13,15,22,24,25,26,31,33,35,36,37,38,40,41,42,48,49,50,51,55,56],exceed:9,except:[3,13,40,46,53,55],excluded_chunk_typ:8,exconv:9,exconvt:9,exdb:13,exec:[26,38],execut:[17,33,35,41,52,54,55],exist:[16,17,22,33,38,41,52,55],exit:[38,42],expand:[33,54,55,56],expand_a:9,expand_level:9,expandconvlay:9,expandlevel:9,expect:[9,35,55],expens:56,experi:40,expir:17,explain:[3,8,17,36,47,55],explan:[9,51,56],explanatori:[24,26],explicit:33,explicitli:[3,16],exploit:48,explor:9,exponenti:6,expos:[26,41],express:[16,41,55],extend:[0,53],extens:[52,53,56],extent:21,extern:[3,20,21],extra:[7,9,10,24],extra_input:9,extra_lay:15,extraattr:[7,40],extraattribut:[9,10],extraattributenon:9,extract:[9,41,48,54,55],extract_fea_c:49,extract_fea_pi:49,extract_para:46,extralayerattribut:[7,9],extrapaddl:10,extrem:[9,35],extremli:2,f120da72:42,f7e3:41,fa0wx:42,fabric:36,facotr:9,fact:49,factor:[7,9,11],factori:20,fail:[3,17,38,40,42,48],failur:17,fake:47,fake_imag:22,fals:[3,7,8,9,10,11,13,22,24,31,33,38,40,42,46,51,53,54,55,56],false_label:22,false_read:22,famili:56,familiar:[3,24],fanscin:3,fantasi:52,fantast:51,far:0,farmer:52,fascinatingli:2,fast:[9,32,35],faster:[9,10,17,31,35,55],favorit:32,favourit:26,fbd1f2bb71f4:42,fc1:[33,40],fc2:40,fc3:40,fc4:40,fc8a365:41,fc8a:41,fc_act:10,fc_attr:10,fc_bias_attr:10,fc_layer:[9,24,33,40,51,53],fc_mat:15,fc_name:10,fc_param_attr:10,fclayer:33,fdata:54,fea:49,fea_output:49,feat:55,featur:[3,6,9,13,32,38,48,51,55,56],feature_map:53,feed:[10,13,15,16,24,55],feedback:0,feeder:13,feedforward:48,femal:52,fernan:55,festiv:3,fetch:[13,17,31,33],few:[3,17,22,26],fewer:9,fg0:9,field:[9,15,35,41],figur:[16,31,33,35,46,47,48,49,54,55,56],file1:56,file2:56,file:[2,3,5,8,9,13,15,16,17,19,21,22,24,25,26,31,32,33,36,38,46,48,49,54,55,56],file_list:3,file_nam:[3,24,49,51,54],filenam:[3,19,53],filer:9,filesystem:[26,41],fill:[9,17,41,51],film:52,filter:[9,49],filter_s:[9,10],filter_size_i:9,finali:36,find:[9,11,17,26,35,48,55,56],fine:[7,53],fingerprint:41,finish:[3,17,26,36,41,42,48],finit:33,first:[3,9,13,16,17,24,26,28,31,32,33,35,38,40,41,46,47,48,49,51,53,54,55,56],first_seq:31,firstn:13,firstseen:42,fit:[2,13,32],five:[35,51],fix:[3,7,20,56],flag:[13,38,47,48,54],flexiabl:22,flexibl:[0,2,9,10,16,31],flight:56,float32:[5,13,22,24,47,49],floor:9,flow:[23,32],fly:[24,51],fnt03:41,focu:[3,35],folder:[19,25,41,48,55,56],follow:[2,3,8,9,10,11,13,16,17,22,25,26,28,31,32,33,35,36,40,41,42,43,44,46,47,48,49,51,52,53,54,55,56],fool:47,forbid:16,force_load:20,forecast:55,forget:[11,16,55],form:[2,3,10,11,35,54],format:[2,3,8,24,32,33,38,41,46,48,52,53,55],former:[16,56],formula:[9,10],formular:9,forward:[6,10,31,32,33,40,47,54,55],forwardactiv:33,forwardtest:5,found:[3,5,9,25,31,47,48,51,55],four:[3,28,46,49,51,53,54,55],frame:8,framework:[16,33,49,51,55],free:[13,56],french:56,frequenc:[13,35,46,51,55],frequent:[22,36,56],frog:48,from:[0,3,5,9,10,13,15,17,19,22,24,26,31,32,33,35,36,38,40,41,42,46,47,48,49,51,52,53,54,55,56],from_timestep:9,fromfil:[22,24,49],fulfil:35,full:[9,17,26,31,33],full_matrix_project:[10,31],fulli:[24,32,33,35,47,48,49,51,53,55],fullmatrixproject:9,fully_matrix_project:10,fullyconnect:46,fullyconnectedlay:33,func:13,fundament:24,further:9,fusion:53,gain:9,game:47,gamma:49,gan:16,gan_train:47,gap:38,gate:[9,10,55],gate_act:[9,10],gate_recurr:9,gather:[9,33,53],gauss:7,gaussian:47,gcc:[20,25,26],gdebi:28,gen:[9,56],gen_conf:[47,56],gen_data:56,gen_result:56,gen_train:47,gen_training_machin:47,gen_trans_fil:31,gender:[13,52,53],gener:[2,3,5,8,9,10,13,15,16,17,19,22,24,25,26,35,36,38,40,41,46,49,50,51,53,55],generated_word_embed:9,generatedinput:[9,31],generator_conf:47,generator_machin:47,generator_train:47,genert:3,genr:[52,53],gereat:8,get:[3,9,10,13,15,24,25,28,31,33,35,36,41,45,48,49,51,53,54,55],get_batch_s:54,get_best_pass:55,get_config_arg:[40,51,53,55],get_data:[42,51,54],get_dict:13,get_embed:13,get_imdb:55,get_input_lay:33,get_mnist_data:47,get_model:49,get_movie_title_dict:13,get_nois:47,get_output_attr:10,get_shap:15,get_training_loss:47,get_word_dict:13,getbatchs:33,getenv:16,getinput:33,getinputgrad:33,getinputvalu:33,getoutputgrad:33,getoutputvalu:33,getparameterptr:33,getsiz:33,getslotvalu:47,gettempl:41,gettranspos:33,getw:33,getweight:33,getwgrad:33,gfortran:25,gildea:54,gist:10,git:[23,25,26,32],github:[9,10,25,26,28,49],give:[3,17,24,26,33,35,41,51],given:[3,13,15,22,33,38,47,51,54,55,56],global:[3,7,16,17,35,38,41,53,55],global_learning_r:7,globalstat:35,globalstatinfo:35,globe:3,goal:[35,54],godoc:20,goe:[9,10,17,24],going:[51,55],good:[9,22,35,55,56],goodfellow13:9,googl:16,googleapi:41,gpg2:41,gpg:41,gpu:[2,3,7,9,12,23,25,28,34,36,47,48,49,53,54,55,56],gpu_id:[38,40,47],gpugpu_id:37,grab:[17,55],grad:[38,52],grad_share_block_num:[37,38],gradient:[7,8,9,11,15,17,38,51,55],gradient_clipping_threshold:[7,51,55],gradient_machin:[15,21],gradient_printer_evalu:8,gradientmachin:[5,15,21,47,53,56],gradual:[24,35],grai:48,gram:[46,55],grant:41,graph:[9,15,17,46],graphviz:49,grave:55,grayscal:3,greater:9,grep:[26,55],groudtruth:31,ground:[8,9,51,56],group:[10,55],group_id:53,group_input:31,grouplen:[13,52],gru:[9,31,51,56],gru_attr:10,gru_bias_attr:10,gru_decod:31,gru_decoder_with_attent:31,gru_encoder_decod:[46,56],gru_memori:10,gru_siz:51,gru_step:[10,31],gru_step_lay:31,grumemori:[10,31],gserver:[9,33],gsizex:35,guarante:33,guess:[24,55],gui:35,guid:[27,31,32,33,35,41,42,46,48,55,56],guidenc:24,gur_group:10,gzip:42,hack:[27,36],hadoop:16,half:41,hand:[52,53,55],handl:[16,22,36,53,55],handler:15,handwrit:[3,55],hard:[41,51],hardwar:[26,35],has:[3,5,9,10,11,13,16,17,26,31,33,35,41,42,46,48,51,52,53,54,55,56],has_kei:15,have:[2,3,5,8,9,10,13,16,17,22,24,25,26,31,32,33,35,36,38,40,41,46,48,51,52,53,55,56],hdf:[2,19],head:[32,46,55],header:[21,24,33,46,49,53],health:52,heavi:36,height:[9,13,20,22,33,48],held:17,hello:16,help:[3,5,32,36],helper:[9,10,33],here:[3,5,7,9,10,13,16,17,22,24,25,31,36,37,40,41,42,46,48,49,50,51,52,53,54,55,56],heurist:[9,38,56],hidden:[9,10,31,41,51,53,55],hidden_s:[10,53],hierarch:[9,31],high:[7,33,47],higher:2,highest:[13,56],highli:[2,3,13,31,40,53,55],him:16,hint:24,histor:55,hl_get_sync_flag:33,hold:[16,17,41],home:[19,36,41,42],homemak:52,hook:[3,53,54],hope:0,horizont:[9,49],horror:52,hors:48,horst:55,host:[25,26,36,41,42],hostnam:[36,41],hostpath:42,hostport:41,hot:53,hour:56,hous:[3,13,24,46],how:[2,3,7,9,16,17,24,31,36,38,41,42,45,48,49,51,53],howev:[3,10,22,24,31,32,37,38,41,55,56],hpp:20,html:[13,26,48],htod:35,http:[9,10,13,25,26,28,32,41,42,47,48,49,56],huber:9,huge:[9,32],huina:55,human:56,hyper:[9,33],hyperplan:13,i0601:53,i0706:56,i0719:56,i1117:35,iamfullaccess:41,iamusersshkei:41,ib0:36,ics:13,icwsm:55,id_input:[8,31],idea:[9,22],ident:[24,26,41,52],identifi:[31,33],identityoffsetproject:9,identityproject:9,ids:[8,9,33,51,53],idx:33,ieee:55,ignor:[3,8,9,38,46],ijcnlp:55,illustr:[3,17,31,33,35,51],ilsvrc:49,imag:[3,12,13,16,22,24,27,40,41,43,44,47,49,50,56],image_a:22,image_b:22,image_classif:48,image_fil:22,image_lay:22,image_list_provid:49,image_nam:16,image_path:22,image_provid:48,image_reader_cr:22,image_s:49,imagenet:[19,50],imagepullpolici:41,imageri:9,images_reader_cr:22,imdb:52,imdber:55,img:[3,9,48],img_conv:10,img_featur:3,img_pool:10,img_siz:48,imgsiz:35,imgsizei:35,imgsizex:35,immedi:41,immutable_paramet:16,implement:[3,9,10,11,13,20,21,31,51,54],importerror:53,improv:[0,35,41,55,56],inbound:41,includ:[2,3,9,10,13,16,20,21,25,26,31,33,35,38,41,42,46,51,52,54,56],inconsist:52,incorrect:9,increas:[17,38,56],increment:38,incupd:33,inde:[13,22,26],independ:[9,51],index:[3,8,9,12,13,15,17,31,36,41,53],indexslot:[9,54],indic:[3,8,9,24,36,41,54],individu:[17,24,41],industri:17,infer:[1,16,17,20,25],infiniband:36,info:[8,9,13,33,36],infom:32,inform:[5,13,33,35,38,41,52,53,54,55,56],infrastructur:[41,47],ingor:38,ininst:16,init:[7,33,40,41,47,51,53,54],init_hook:[51,53,54],init_model_path:[37,38,40,46,51,54],initi:[3,5,7,9,13,31,33,38,46,47,51,54],initial_max:7,initial_mean:[7,9],initial_min:7,initial_std:[7,9],initpaddl:[5,47],inlcud:10,inlin:41,inner:33,inner_param_attr:10,input1:[9,10],input2:9,input:[3,5,6,8,9,10,12,13,15,22,24,31,33,40,46,47,48,49,51,53,54,55,56],input_data:33,input_data_target:33,input_featur:6,input_fil:[24,54],input_hassub_sequence_data:33,input_id:9,input_imag:[10,48],input_index:33,input_label:33,input_lay:33,input_nam:16,input_sequence_data:33,input_sequence_label:33,input_sparse_float_value_data:33,input_sparse_non_value_data:33,input_t:33,input_typ:[24,31,51,53],inputdef:33,inputlayers_:33,inputtyp:[3,13],insid:[8,9,17,22,26,41],inspir:46,instal:[26,29,32,36,42,48,49,53,54,55],instanc:[9,17,19,31,33,35,38,54],instance_ip:41,instanti:17,instead:[9,10,12,22,26,32,36,51,56],instruct:[26,28,35,51],int32:38,integ:[3,8,9,13,20,31,33,51,55],integer_valu:[3,13,51],integer_value_sequ:[3,13,31,51,54],integr:[25,54],intend:0,inter:[9,36],interact:[26,41],intercept:9,interest:[35,55],interfac:[1,5,7,9,10,36,41,48,53,55],interg:51,intergr:9,intermedi:54,intern:[9,10,13,15,41],internet:[17,55],interpret:[3,8,25,35],interv:55,intrins:25,introduc:[3,17,42,53,55],introduct:[4,47],invalid:22,invari:48,invok:[3,9,15,35,41,53],involv:47,iob:8,ioe:8,ips:41,ipt:[9,31],ipython:16,is_discriminator_train:47,is_gener:[9,46,47,56],is_generator_train:47,is_kei:53,is_loc:15,is_predict:[51,53,55],is_seq:[9,31,53],is_sequ:53,is_stat:7,is_test:[49,54,55],is_train:3,isn:35,isol:26,isspars:33,issu:[25,26,35],item:[9,13,15,22],iter:[9,10,11,13,15,16,17,22,48,54,55],its:[3,8,9,10,16,17,33,35,38,41,46,47,48,51,55,56],itself:[10,17],java:20,jeremi:35,jie:[54,55],jmlr:9,job:[5,13,37,38,40,49,51,53,54,55,56],job_dispatch_packag:36,job_id:13,job_mod:46,job_nam:41,job_namespac:41,job_path:41,job_workspac:36,jobpath:41,jobport0:41,jobport1:41,jobport2:41,jobport3:41,johan:55,join:17,joint:[46,56],jointli:[10,56],journal:[54,55],journei:26,jpeg:48,jpg:49,json:[36,41,42,53],jth:10,judg:56,jupyt:26,just:[3,6,8,9,10,13,24,32,36,40,41,46,48,53,54,55],jx4xr:41,jypyt:16,k8s_data:41,k8s_job:16,k8s_token:16,k8s_train:41,k8s_user:16,kafka:19,kaim:9,kaimingh:49,kebilinearinterpbw:35,kebilinearinterpfw:35,keep:[3,9,17],kei:[3,13,15,17,35,36,53,55],kernel:[9,35,51],key1:38,key2:38,key_pair_nam:41,keyid:41,keymetadata:41,keypair:41,keyserv:41,keystat:41,keyusag:41,keyword:3,kill:[17,41],kind:[2,3,16,17,24,41,42,47,51,53],kingsburi:54,kms:41,know:[3,10,16,24,33,35,41,53],knowledg:55,known:[47,55,56],kriz:[13,48],ksimonyan:10,kube_cluster_tl:16,kube_ctrl_start_job:16,kube_list_containers_in_job_and_return_current_containers_rank:16,kubeconfig:41,kubectl:42,kuberent:[17,41],kubernet:[16,17,34,36,43,44],kubernetes_service_host:16,kwarg:[3,8,9,10,11,13,51,53,54],l1_rate:7,l2_rate:7,l2regular:[48,51,55],label:[3,5,8,9,11,13,15,22,24,31,42,47,48,49,50,51,53,55],label_dict:54,label_dim:[9,51],label_fil:[22,54],label_lay:22,label_list:54,label_path:22,label_slot:54,labeledbow:55,labl:55,lag:38,lake:3,lambdacost:9,lambdarank:9,languag:[9,13,40,46,54,55,56],larg:[12,13,54,55,56],larger:[3,7,8,9,36],last:[8,9,10,24,31,36,38,51,55,56],last_time_step_output:9,lastseen:42,late:55,latenc:[36,41],later:[25,32,41,51],latest:[9,17,26,32,42,55],latter:56,launch:[38,41,55],launcher:16,lawyer:52,layer1:[9,10],layer2:9,layer3:9,layer:[5,7,8,10,12,13,14,15,22,24,31,34,37,38,46,47,48,49,51,53,54,55],layer_0:33,layer_attr:[9,31,40],layer_num:[40,49],layer_s:9,layer_typ:9,layerbas:33,layerconfig:33,layergradutil:33,layermap:33,layeroutout:9,layeroutput:[9,53],lbl:[8,48],ld_library_path:[25,28,36],lead:35,learn:[0,7,8,9,10,11,13,16,22,24,26,29,31,33,35,48,49,51,54,55,56],learnabl:9,learning_method:[24,46,48,51,53,55,56],learning_r:[7,24,46,48,51,53,55,56],leas:17,least:[8,9,17,25,52],leav:[3,41],lecun:13,left:[9,24,49],leman:56,len:[3,9,31,33,51,53,54],length:[9,10,13,31,38,42,55,56],less:[9,16,36,56],less_than:16,let02:42,let:[5,9,16,24,41,53],level:[7,9,36,38,47,53,55,56],lib64:[25,36,38],lib:21,libari:21,libcudnn:25,libjpeg:48,libpaddl:[20,21],libpaddle_capi:21,libpaddle_gserv:21,libpaddle_math:21,libpython:25,librari:[9,21,25,26,36,38,53],licens:54,life:17,like:[3,8,9,13,17,22,24,25,31,35,36,37,40,41,46,49,51,53,55,56],limit:[9,13,35,38],line:[2,3,5,8,13,24,32,34,35,36,40,41,46,48,49,53,54,55,56],linear:9,linearactiv:[9,24],linguist:54,link:[9,10,25,41,51,55],linux:[25,26,28,41,56],lipeng:46,lipton:55,list:[2,3,8,9,13,15,16,24,26,31,33,36,38,40,41,48,49,51,53,54,55,56],listen:38,literatur:55,littl:[2,3,38,51,55],lium:56,live:[17,26],liwicki:55,load:[2,3,5,9,16,17,24,38,41,49,53,54,55,56],load_featur:49,load_feature_c:49,load_feature_pi:49,load_missing_parameter_strategi:[37,38,40,46,54],load_uniform_data:47,loadparamet:5,loadsave_parameters_in_pserv:[37,38],local:[7,17,25,26,32,36,37,38,42,48,55],localhost:26,locat:[31,33,51,54],lock:17,log:[3,32,33,36,38,41,42,48,53,54,55,56],log_barrier_abstract:38,log_barrier_lowest_nod:[37,38],log_barrier_show_log:[37,38],log_clip:[37,38],log_error_clip:[37,38],log_period:[38,40,42,47,48,51,53,54,55,56],log_period_serv:[37,38],logarithm:6,logger:3,logic:[3,36],login:26,longer:56,look:[3,24,36,37,41,42,47,51],lookup:51,loop:22,loss:[9,33,47,51,55,56],lot:37,low:9,lower:36,lowest:38,lpaddle_capi_shar:21,lpaddle_capi_whol:21,lst:53,lstm:[9,31,42,51],lstm_attr:10,lstm_bias_attr:10,lstm_cell_attr:10,lstm_group:10,lstm_size:51,lstm_step:10,lstmemori:[10,31],lstmemory_group:9,ltr:9,lucki:24,mac:[21,25,26],machan:10,machin:[9,10,13,15,24,32,33,37,38,40,41,42,51,53,55,56],made:[3,17,24,31,52],mai:[3,9,22,26,32,35,41,52],main:[3,5,32,41,48,54,55],mainli:38,maintain:[9,41],major:[26,32,47,49,55,56],make:[3,16,17,22,25,26,32,33,35,36,41,48,51,53,55],male:52,malloc:33,manag:[17,32,36],manageri:52,mandarin:9,mani:[0,9,10,24,26,38,51,52,53,55],mannal:36,manual:32,manufactur:56,mao:55,map:[3,9,13,15,16,38,48,49,53],map_read:13,mapreduc:16,marcu:55,mark:[3,31,54],mark_slot:54,market:[24,52,55],martha:54,mask:[7,9],master:[16,23,32,38,55],mat:[20,21],mat_param_attr:10,match:35,math:[10,20,33,35],matirx:9,matplotlib:48,matric:[5,31,33],matrix:[8,9,10,13,15,20,21,31,33,37,40,49,54],matrixptr:33,matrixtyp:21,matter:3,max:[3,7,9,13,35,38,40,48,51,53],max_id:[9,15,51],max_job_id:13,max_length:[9,31],max_movie_id:13,max_sort_s:9,max_user_id:13,maxframe_printer_evalu:8,maxid:[8,51],maxid_lay:51,maxid_printer_evalu:8,maxim:[9,56],maximum:[8,13,31,35,38,51,54,55],maxinum:12,maxpool:9,mayb:[9,10,48],md5:[13,18],mean:[3,7,8,9,10,11,12,13,15,22,24,31,35,36,38,40,41,46,47,48,49,51,53,54,55,56],mean_img_s:48,mean_meta:49,mean_meta_224:49,mean_valu:49,measur:[24,35],mechan:[9,10,31,41,55],media:55,meet:54,mem:9,member:16,memcpi:35,memor:55,memori:[2,3,10,31,33,35,38,40,42,51,54,55,56],memory_nam:9,memory_threshold_on_load_data:38,mere:10,merg:[32,38,46,56],mergedict:[46,56],messag:[24,38,42,53,55,56],meta:[36,48,49,51],meta_config:[36,53],meta_fil:53,meta_gener:[36,53],meta_path:48,meta_to_head:53,metadata:[41,42],metaplotlib:16,method:[3,9,11,15,26,33,35,38,40,51,53,55,56],might:[9,26,33,41],mileag:35,million:[13,40,52],min:[7,35,40,41,53],min_pool_s:3,mind:36,mini:[3,9,13,15,17],mini_batch:22,minibatch:9,minibatch_data:13,minim:[3,11,24,38],minimum:9,minimun:38,minst:3,minut:[17,41,56],mirror:26,miss:[38,46,54],mit:41,mix:[10,31,54],mixed_attr:10,mixed_bias_attr:10,mixed_lay:[9,31,54],mixedlayertyp:9,mkdir:[25,26,41],mkl:25,mkl_path:25,mkl_root:25,ml_data:[36,53],mnist:[3,5,19,22],mnist_provid:3,mnist_random_image_batch_read:22,mnist_train:[3,22],mnist_train_batch_read:22,mod:54,modal:54,mode:[9,38,47,48,49,53,55,56],model:[1,2,5,9,10,13,17,18,32,33,34,38,41,53,54,55],model_config:[5,47],model_list:[38,40,54,55],model_output:55,model_path:40,model_zoo:[46,49],modifi:[5,31,32,33,36,41],modul:[2,3,5,10,13,15,24,25,48,49,51,53,54],modulo:9,momentum:[7,24,51],momentumoptim:[24,48],mon:42,monitor:[51,55],mono:9,month:[51,56],mood:55,more:[2,3,5,8,9,10,13,16,17,22,24,26,31,33,35,36,40,42,48,51,54,55,56],morin:9,mose:[55,56],moses_bleu:56,mosesdecod:55,most:[3,5,9,13,16,22,24,31,33,35,37,53,54,55,56],mostli:[48,52],mount:[26,41,42],mountpath:[41,42],move:[9,17,35,41,53,55],movement:[35,55],movi:[3,13,55],movie_categori:13,movie_featur:53,movie_head:53,movie_id:53,movie_info:13,movie_meta:53,movie_nam:53,movie_review:13,movieid:52,movieinfo:13,movielen:50,moving_average_fract:9,mpi:36,mse_cost:[24,53],much:[9,17,22,35],mul:33,mulit:36,multi:[9,33,37,38,49,56],multi_binary_label_cross_entropi:9,multi_crop:49,multinomi:9,multipl:[8,9,10,13,16,26,31,33,38,40,41,47,51,53,55],multipli:[8,9,33,48],multithread:3,music:52,must:[3,6,8,9,10,22,25,26,31,32,33,36,38,40,41,56],my_cluster_nam:41,my_cool_stuff_branch:32,my_external_dns_nam:41,mypaddl:42,mysteri:52,name:[3,7,8,9,10,12,13,15,16,17,19,21,24,26,31,33,35,36,38,40,42,43,44,46,47,48,49,51,53,55,56],name_prefix:19,namespac:[20,26,33,42],nano:32,nativ:9,natur:[40,54,55],nchw:9,ndarrai:[15,19],ndarri:15,ndcg:9,ndcg_num:9,nearest:51,necessari:[3,9,25,33,36,51,55],necessarili:33,need:[3,9,10,13,16,24,25,26,28,31,32,33,36,37,38,40,41,42,47,48,49,51,53,54,55,56],neg:[3,8,9,51,54,55],neg_distribut:9,negat:54,neighbor:51,nest:[3,13],net:[9,10],net_conf:55,net_diagram:49,network:[2,3,5,7,8,9,11,13,14,15,16,17,22,24,26,33,35,36,38,46,55,56],network_config:40,networkadministr:41,neural:[3,5,9,10,11,13,15,16,17,24,35,38,46,47,49,55,56],neuralnetwork:9,neuron:[5,33,51,55],never:[13,22,41,42],newest:32,newtork:55,next:[9,13,17,31,33,35,38,41,42,54,55,56],nfs4:41,nfs:41,nfsver:41,nginx:26,nic:[36,37,38],nine:[13,54],nlp:[3,9],nltk:13,nmt:56,nnz:33,no_cach:3,no_sequ:[3,53],noah:55,noavx:[26,28],node:[9,33,36,38,41,42,55,56],node_0:41,node_1:41,node_2:41,nodefil:36,noir:52,nois:[9,17,47],noise_dim:47,non:[9,17,33,38,41],none:[2,3,5,7,8,9,10,11,12,13,15,16,24,31,49,51],nonlinear:33,norm:47,norm_by_tim:9,normal:[3,5,9,10,13,28,31,33,36,38,42,46,47,49],normzal:49,north:48,notat:9,note:[3,5,7,9,10,12,15,16,22,25,35,38,40,41,46,48,53,55],notebook:26,noth:[6,15,38],notic:[31,33],novel:55,now:[0,3,9,17,24,26,32,38,41,47,53,54],np_arrai:13,nproc:25,ntst1213:56,ntst14:56,nullptr:33,num:[9,36,38,51,54,55,56],num_channel:[9,10,48],num_chunk_typ:8,num_class:[9,10,48],num_filt:[9,10],num_gradient_serv:[37,38],num_group:9,num_neg_sampl:9,num_parameter_serv:16,num_pass:[15,24,37,38,40,42,51,53,54,55,56],num_repeat:9,num_result:8,num_results_per_sampl:9,num_shard:19,number:[3,8,9,13,17,19,22,24,33,36,38,41,46,48,49,51,54,55,56],numchunktyp:8,numdevices_:40,numlogicaldevices_:40,numofallsampl:8,numofwrongpredict:8,numpi:[13,15,19,22,24,25,47,49],numsampl:35,numtagtyp:8,nvcc:26,nvidia:[25,26,35,38],obj:[3,24,48,49,51,53],object:[3,5,7,9,10,11,13,15,16,20,35,47,48,49,51,54],observ:[11,24,33,35,56],obtain:[51,54,55],occup:[52,53],occur:[13,15,32],oct:42,odd:9,off:[21,26],offer:[5,54],offici:[26,41,48],offlin:[17,19],offset:[9,53],often:[36,51,56],ograd:33,old:[26,32,38],omit:51,on_init:3,on_travisexclud:25,onc:[3,9,17,26,32,33,41,51],one:[3,6,8,9,10,11,12,13,16,17,22,24,26,32,33,36,38,40,41,42,46,47,48,49,51,53,54,55,56],one_host_dens:53,one_hot_dens:53,onli:[2,3,5,8,9,10,12,13,15,16,24,25,31,32,33,35,37,38,40,41,42,46,49,51,52,55,56],onlin:[11,17,19,22],onto:41,open:[0,3,9,16,22,24,26,41,49,51,53,54],openbla:25,openblas_path:25,openblas_root:25,oper:[9,10,11,26,31,33,35,38,41,46,48,53],opinion:55,opt:[16,25],optim:[3,7,14,15,24,33,35,55],option:[3,8,9,16,24,32,33,36,40],order:[3,9,10,13,22,33,38,41,42,47,49,51,55,56],ordinari:55,oregon:41,org:[9,10,13,25,26,47],organ:[9,48,55,56],origin:[0,2,3,9,13,32,47,54,56],other:[3,8,9,10,13,17,25,26,28,31,32,40,41,42,46,47,48,49,51,52,53,54,55,56],otherchunktyp:8,otherwis:[2,9,13,16,17,22,31,36,40,53,56],our:[16,26,31,33,41,42,46,48,51,54,55,56],out:[9,15,16,24,31,35,38,41,42,48,55],out_dir:41,out_left:9,out_mem:31,out_right:9,out_size_i:9,out_size_x:9,outlin:39,outperform:54,outptut_lay:15,output:[5,6,7,8,9,10,12,13,15,16,19,22,24,31,33,35,38,40,42,46,47,48,49,51,53,54,55,56],output_:[9,33],output_dir:49,output_fil:54,output_id:9,output_lay:[15,49],output_max_index:12,output_mem:[9,31],output_path:19,outputh:9,outputw:9,outsid:[3,9,10],outter_kwarg:3,outv:33,over:[2,9,10,16,32,33,35,51,54,55],overcom:55,overhead:35,overlap:33,overrid:[17,33],owe:0,own:[26,32,36,41],pacakg:28,pack:26,packag:[3,13,26,27,41],pad:[31,51],pad_c:9,pad_h:9,pad_w:9,paddepaddl:2,padding_attr:9,padding_i:9,padding_x:9,paddl:[3,5,6,7,8,9,10,11,12,13,15,16,17,19,20,21,24,25,26,27,28,32,33,34,35,36,38,40,41,47,48,51,53,54,55,56],paddle_error:[20,21],paddle_matrix:[20,21],paddle_matrix_cr:21,paddle_matrix_get_shap:20,paddle_matrix_shap:20,paddle_n:36,paddle_output:42,paddle_port:36,paddle_ports_num:36,paddle_ports_num_for_spars:36,paddle_pserver2:36,paddle_root:46,paddle_source_root:46,paddle_train:[21,23,36],paddledev:[41,42],paddlepaddl:[0,2,3,5,9,10,13,17,22,24,25,28,31,32,33,34,35,36,43,44,49,51,53,54,55],paddlepadl:3,paddlpaddl:0,paddpepaddl:3,page:[32,41,53],pai:26,pair:[8,54],palceholder_just_ignore_the_embed:46,palmer:54,paper:[9,46,47,49,54,55,56],paraconvert:46,paragraph:55,parallel:[35,38,40,41,42,56],parallel_nn:[7,37,38],param:[7,9,53],param_attr:[9,10,24,31],paramattr:[7,9,24,31],paramet:[2,3,5,8,9,10,11,12,13,14,18,22,24,33,34,40,47,48,51,53,54,55,56],parameter_attribut:9,parameter_block_s:[37,38],parameter_block_size_for_spars:[37,38],parameter_learning_r:7,parameter_nam:[15,16],parameter_serv:16,parameterattribut:[7,9,10],parametermap:33,parameters_:33,parameterset:16,parametris:11,paramt:[41,46],paramutil:53,paraphras:56,paraphrase_data:46,paraphrase_model:46,paraspars:33,parent:33,pars:[5,13,40,41,47,53,54],parse_config:[5,47],parser:53,part:[3,24,31,32,33,35,47,51,53,54,55,56],parti:[35,53],partial:[9,47],participl:46,particular:35,partit:[17,19,41],pass:[3,9,13,15,17,22,24,32,33,35,36,38,41,42,47,48,51,53,54,55,56],pass_id:15,pass_idx:22,pass_test:47,passtyp:33,password:36,past:[16,26,41],path:[2,3,8,13,15,17,22,24,25,31,36,38,40,41,42,46,48,49,51,54,55,56],pattern:[13,17,20,24,41,53,55],paul:54,paus:17,pave:56,pdf:[9,10],pem:[16,41],pend:17,penn:54,per:[9,13,17,22,38,48,51],perfom:[38,40],perform:[2,9,10,24,31,32,33,34,36,37,47,48,51,55,56],period:[2,17,38,51,53,54,55,56],perl:[55,56],permiss:41,peroid:9,persist:41,persistentvolum:41,persistentvolumeclaim:41,person:16,perspect:35,perturb:33,pfs:19,pgp:41,phase:24,photo:48,pick:[3,41],pickl:53,pictur:51,piec:[9,10,24],pillow:48,pip:[25,32,36,48,53],pipe:52,pipelin:54,pixel:[3,9,13],pixels_float:3,pixels_str:3,place:[2,3,17,33,35,36,49,56],placehold:[24,46],plai:[54,55],plain:[2,8,9,15,21],plan:[17,33],platform:[0,24,41],pleas:[3,5,7,9,10,11,16,17,22,25,26,27,31,32,33,41,46,48,51,53,54],plot:[16,48],plotcurv:48,png:[48,49],pnpair_evalu:8,pnpairvalidationlay:38,pnpairvalidationpredict_fil:37,pod:[41,42],pod_nam:41,point:[17,24,35],polar:[13,55],polici:41,polit:55,poll:55,poo:48,pool3:33,pool:[3,10,14,48,51,53],pool_attr:10,pool_bias_attr:10,pool_pad:10,pool_siz:[3,9,10],pool_size_i:9,pool_strid:10,pool_typ:[9,10],pooling_lay:51,pooling_typ:[9,51],poolingtyp:12,popular:[24,49],port:[26,36,37,38,41,42],port_num:37,ports_num:38,ports_num_for_spars:[37,38,40],pos:[53,55],pose:17,posit:[3,8,9,13,51,54,55,56],positive_label:8,possibl:[16,32,35,47],post1:25,potenti:35,power:[51,56],practic:[9,24,31,33],pre:[3,9,10,13,16,26,41,42,46,48,54,55,56],pre_dictandmodel:46,precis:[8,25],precision_recall_evalu:8,pred:[51,54],predefin:55,predetermin:[9,38,56],predic:[13,54],predicate_dict:54,predicate_dict_fil:54,predicate_slot:54,predict:[3,4,8,9,11,15,24,31,36,38,46,51,56],predict_fil:38,predict_output_dir:[37,38,51],predict_sampl:5,predicted_label_id:51,predictor:53,predin:48,prefer:55,prefetch:33,prefix:[17,19,41],pregrad:33,preinstal:25,premodel:46,prepar:[5,43,51],preprcess:55,preprocess:[13,31,36,42,55],prerequisit:25,present:[16,49,54,56],pretti:24,prev_batch_st:[37,38],prevent:[2,11,16,17],previou:[9,10,17,33,38,41,54,56],previous:[9,42,49],price:[13,24],primarili:55,principl:16,print:[7,15,16,24,31,38,46,51,53,54,55,56],printallstatu:35,printer:8,printstatu:35,prite:8,privat:21,privileg:41,prob:[8,15,47],probabilist:[9,46],probability_of_label_0:51,probability_of_label_1:51,probabl:[8,9,15,31,32,49,51,54],problem:[5,9,11,16,51,54,55],proc:26,proc_from_raw_data:51,proce:[13,17,22,41],procedur:[46,54,56],proceed:[9,54],process:[2,3,5,7,9,10,16,19,24,26,31,36,38,40,41,42,46,48,49,51,53,54,55,56],process_pr:51,processdata:[48,49],processor:35,prod:26,produc:[10,13,17,22,26,49,51],product:[0,26,33,41,51,55],productgraph:42,profil:25,proflier:35,program:[2,13,16,19,22,26,35,36,38],programm:52,progress:[17,38],proivid:3,proj:9,project:[9,10,21,25,31,33,53],promis:[9,10],prompt:32,prone:16,prop:54,propag:[11,38,40],properli:51,properti:[3,38],propos:56,proposit:54,protect:33,proto:12,protobuf:25,protocol:38,prove:51,proven:56,provid:[0,9,13,16,24,26,31,35,36,41,46,47,48,49,52,55],providermemory_threshold_on_load_data:37,provis:41,provod:3,prune:9,ps_desir:17,pserver:[36,37,38,41],pserver_id:18,pserver_num_thread:[37,38],pserverstart_pserv:37,pseudo:16,psize:33,ptr:21,pull:[23,26,46,56],punctuat:55,purchas:51,purpos:[0,17,35],push_back:33,put:[17,26,33,36,42,51],pvc:41,pwd:26,py_paddl:[5,13,47],pydataprovid:[2,3,51],pydataprovider2:[4,5,24,31,51,53,55],pyramid:9,pyramid_height:9,python:[2,3,4,15,16,20,23,24,25,32,36,46,47,48,54,55,56],pythonpath:48,pzo:55,qualifi:25,qualiti:51,queri:[9,41,56],question:[9,16,41,54],quick:[38,42,50,56],quick_start:[41,42,43,51],quick_start_data:42,quickli:24,quickstart:42,quit:35,quot:52,rac:9,rais:13,ramnath:55,ran:35,rand:[35,38,40,47,54],random:[3,7,9,13,19,22,24,38,47,48,54],random_imag:19,randomli:55,randomnumberse:37,rang:[3,9,13,19,22,38,40,48,52,54],rank:[9,16,41,49,51],rare:3,rate:[7,8,11,13,33,36,48,51,53,55,56],rather:[5,41,55],ratio:38,raw:[9,24,51,55],raw_meta:53,rdma:[25,38],rdma_tcp:[37,38],reach:[17,35,54],read:[2,3,13,15,16,17,19,22,24,31,36,41,49,51,53],read_from_realistic_imag:16,read_from_rng:16,read_lock:18,read_mnist_imag:16,read_ranking_model_data:16,reader:[1,15,19,56],reader_cr:19,reader_creator_bool:22,reader_creator_random_imag:[13,22],reader_creator_random_image_and_label:[13,22],readi:[17,24,41,42,48],readm:[21,52,53,55],readonesamplefromfil:3,readwritemani:41,real:[3,22,24,47],realist:16,reason:[10,16,17,26,42],rebas:32,recal:8,receiv:17,recent:56,reciev:38,recogn:48,recognit:[3,9,49,55],recommand:3,recommend:[2,10,16,26,31,33,36,38,53],recommonmark:25,recompil:35,record:[41,53,54],recordio:16,recov:[17,24,47],rectangular:9,recurr:[54,55],recurrent_group:[10,31],recurrentgradientmachin:21,recurrentgroup:8,recurrentlay:38,recv:41,reduc:[11,36,38,40],refer:[2,5,7,9,10,11,17,31,33,36,42,46,48,51,53,56],referenc:9,regard:54,regardless:56,regex:53,region:[35,54],regist:[33,35],register_gpu_profil:35,register_lay:33,register_timer_info:35,registri:[26,42],regress:50,regular:[7,33,41,48,51,55],rel:[2,10,36],relat:[3,17,26,28,42,53,55],relationship:[13,24,47],releas:[23,25,26,28,41,52,54],relev:[54,56],reli:25,reliabl:17,relu:[9,33],remain:51,rememb:9,remot:[7,32,33,36,38,40,41],remoteparameterupdat:38,remov:[13,36,38,55],renam:56,reorgan:9,replac:[26,55],repo:32,report:[35,36],repositori:32,repres:[3,5,9,13,31,33,41,48,51,52],represent:[51,55],reproduc:56,request:[17,23,41,42,46,56],requir:[2,8,9,16,17,33,36,41,42,47,48,51,53],requrest:32,res5_3_branch2c_bn:49,res5_3_branch2c_conv:49,res:54,research:[9,13,48,52,55],resembl:55,reserv:3,reserveoutput:33,reset:[9,17],reshap:22,reshape_s:9,residu:49,resnet:50,resnet_101:49,resnet_152:49,resnet_50:49,resolv:[32,42],resourc:[26,41],respect:[3,24,31,33,38,48,49,54,56],respons:[9,41,42],rest:[3,9,24],restart:[17,41,42],restartpolici:[41,42],restrict:38,resu:22,result:[5,6,8,9,15,31,35,38,41,48,49,51,53,54,55],result_fil:[8,31],ret_val:53,retir:52,retran:41,retriev:[33,42],return_seq:10,reuqest:23,reus:[22,33],reveal:16,revers:[9,10,31,54,55],review:[13,32,42,51,55],reviews_electronics_5:42,revis:51,rewrit:56,rgb:9,rgen:55,rho:11,rich:24,right:[3,9,49],rmsprop:51,rmspropoptim:53,rnn:[9,10,34,37,51,55],rnn_bias_attr:31,rnn_layer_attr:31,rnn_out:31,rnn_step:9,rnn_use_batch:[37,38],rnnlm:13,robot:48,role:[13,16,31,41,50,55],roman:55,romanc:52,root:[11,12,36,41,42],root_dir:36,rot:9,roughli:[3,47],routin:53,routledg:55,row:[5,8,9,13,33,49],row_id:9,rsize:41,rtype:[9,53],rule:[33,41],run:[16,17,26,32,33,34,35,38,41,43,44,46,48,49,51,53,55,56],runinitfunct:35,runtim:[2,3,25,26,36],s_fusion:53,s_id:53,s_param:47,s_recurrent_group:31,sacrif:2,sai:[24,38,40],sake:33,sale:52,same:[3,5,8,9,10,16,31,36,40,41,46,51,53,54,55,56],samping_id:9,sampl:[3,5,8,13,36,38,40,46,47,49,51,53,54,55,56],sample_dim:47,sample_id:8,sample_num:8,santiago:55,satisfi:[36,41,51],save:[3,9,13,17,19,24,38,40,41,42,48,49,51,53,54,55,56],save_dir:[24,38,40,42,47,48,51,53,54,55,56],save_only_on:[37,38],saving_period:[37,38],saving_period_by_batch:[37,38,40,51],saw:3,scalabl:0,scalar:[3,9],scale:[0,6,49,52,53],scalingproject:9,scatter:9,scenario:[24,37],scene:37,schdule:41,schedul:[41,47],scheduler_factor:7,schema:46,scheme:[8,54],schmidhub:55,schwenk:56,sci:52,scienc:55,scientist:[0,52],score:[8,9,53,55,56],screen:53,scrip:51,script:[5,13,36,41,48,49,51,54,55,56],seaplane_s_000978:48,search:[9,17,25,31,38,54,56],seat:56,second:[3,9,13,16,22,24,26,32,36,46,49,51,52,53,55],secret:41,section:[3,31,33,36,41,51],sed:55,see:[3,5,9,10,16,17,24,26,32,35,41,46,47,49,51,53,55,56],seed:[35,38],segment:8,segmentor:46,sel_fc:9,select:[9,32,41,52,56],selectiv:9,selector:42,self:[24,33,52,55],selfnorm:9,semant:[13,16,23,31,50,55],semat:16,sen_len:54,send:[17,38,41],sent:[16,42],sent_id:31,sentenc:[3,9,13,31,51,54,55,56],sentiment:[3,24,50,51,54],sentiment_data:55,sentiment_net:55,sentimental_provid:3,separ:[3,8,38,46,51,52,53,54,56],seq:[9,53],seq_pool:9,seq_text_print:8,seq_to_seq_data:[46,56],seq_typ:[3,13,53],seqtext_printer_evalu:[8,31],seqtoseq:[9,31,46,56],seqtoseq_net:[9,31,46,56],sequel:3,sequenc:[3,6,8,9,10,12,13,33,46,51,53,54,55,56],sequence_conv_pool:51,sequence_layer_group:9,sequence_nest_layer_group:9,sequencestartposit:9,sequencetextprint:8,sequencetyp:3,sequenti:[9,31,51,54],seri:[10,55],serial:[3,15],serv:[26,35,41,47],server:[16,26,33,36,37],serverless:17,servic:[26,52],session:[26,35],set:[2,3,5,7,8,9,10,13,15,16,17,24,25,26,28,31,33,34,35,36,37,38,40,41,42,46,48,49,51,52,53,54,55,56],set_active_typ:33,set_default_parameter_nam:7,set_drop_r:33,set_input:9,set_siz:33,set_typ:33,setp:41,settup:33,setup:[3,26,33,51],sever:[3,9,36,40,41,50,51,53,54,55,56],sgd:[11,15,16,17,36,47,55,56],sgdasync_count:37,shallow:54,shape:[9,15,49],shard:[17,18,19,41],share:[9,21,25,26,35,38,42,54],shared_bia:10,shared_bias:9,shared_ptr:[20,21],shell:[41,49],shift:49,ship:48,shold:55,shop:55,shorten:9,shorter:49,should:[3,5,8,9,13,15,16,22,24,28,31,32,36,41,48,51,53,54,55,56],should_be_fals:16,should_be_tru:16,should_shuffl:[3,54],shouldn:32,show:[5,11,13,17,24,32,38,41,42,46,49,51,53,54,55,56],show_check_sparse_distribution_log:[37,38],show_layer_stat:[37,38],show_parameter_stats_period:[37,38,40,42,51,54,55,56],shown:[3,8,9,16,31,33,35,41,47,48,49,51,53,55,56],shrink:33,shuf:53,shuffl:[3,13,53,55],sid:41,side:[9,15,49],sig:41,sigint:36,sigmoid:[9,10,33],sign:41,signal:36,signatur:41,signific:35,similar:[9,22,41,51,53],similarli:[9,54],simpl:[2,3,6,8,9,10,13,15,25,26,32,35,38,51,53,54,55],simple_attent:31,simple_gru:51,simple_lstm:[9,51],simple_rnn:[9,31],simplest:41,simpli:[2,9,16,25,26,31,32,35,46,49,53,55,56],simplifi:[16,33,42],simultan:41,sinc:[9,17,22,24,26,35,41,47,51,52,56],sincer:[32,55],singl:[3,8,10,13,17,26,33,36,42,49,51,54,56],site:41,six:[46,54,56],size:[3,8,9,10,11,13,17,19,22,24,31,33,36,38,47,48,49,51,52,53,54,55,56],size_a:9,size_b:9,size_t:33,sizeof:46,skill:56,skip:[22,24,36,41,49],slide:[9,11,13,17],slightli:48,slope:9,slot:[53,54],slot_dim:53,slot_nam:53,slottyp:53,slow:[3,35],small:[3,13,33,36,38,48,56],small_messag:[37,38],small_vgg:48,smaller:[9,17],smith:55,smooth:9,snap:42,snapshot:[18,41],snippet:[31,33,35,41,51],social:55,sock_recv_buf_s:[37,38],sock_send_buf_s:[37,38],socket:38,softmax:[9,10,16,31,33,46,51,54,55],softmax_param_attr:10,softmax_selfnorm_alpha:9,softmaxactiv:[31,51],softwar:[26,35],solv:[16,54],solver:56,some:[3,7,9,13,15,16,24,25,32,33,35,37,38,40,41,47,51,52,53,54,55,56],some_c_api_funct:21,some_inst:21,some_python_class:20,somecppclass:20,somedata:15,somegotyp:20,someth:[3,9],sometim:[11,22,35,55],soon:17,sophist:[24,33,36],sort:[9,13,38,41,53,55,56],sourc:[0,9,13,21,22,24,26,31,32,41,42,46,51,53,56],source_dict_dim:31,source_language_word:31,space:[8,26,31,35],space_seperated_tokens_from_dictionary_according_to_seq:8,space_seperated_tokens_from_dictionary_according_to_sub_seq:8,spars:[3,7,9,11,13,33,36,38,41,51],sparse_binary_vector:[3,13,51],sparse_binary_vector_sequ:13,sparse_float_vector:3,sparse_non_value_slot:13,sparse_upd:7,sparse_value_slot:13,sparse_vector:13,sparse_vector_sequ:13,sparseparam:33,sparseprefetchrowcpumatrix:33,spatial:[9,48],speak:[31,56],spec:[41,42],specfii:38,speci:48,special:[9,25,46,51,56],specif:[2,40,48,51,53],specifi:[2,3,8,9,13,16,24,25,31,33,38,41,47,48,49,51,52,53,55,56],speech:9,speed:[10,26],spefici:49,sphinx:[20,25,26],sphinx_rtd_them:25,split:[3,9,36,40,41,46,49,51,54],split_count:41,sql:2,squar:[9,11,12,24],squarerootnpool:9,squash:56,srand:38,src:56,src_backward:31,src_dict:31,src_embed:31,src_forward:31,src_id:31,src_root:5,src_word_id:31,srl:[13,54],ssd:9,ssh:[26,36,41,42],sshd:26,ssl:25,sstabl:[16,19],stabl:[23,41],stack:[24,41,51,54],stacked_lstm_net:55,stacked_num:55,stackexchang:9,stage:36,stake:56,stale:[17,32],stamp:35,standard:[7,46,48,54,55,56],stanford:[13,42],star:52,start:[9,15,17,24,26,31,32,35,36,38,45,46,50,53,56],start_pass:[37,38],start_pserv:38,startup:[17,41],stat:[25,35,38,54,55,56],state:[9,10,17,24,31,38,42,47,54,56],state_act:[9,10],statement:[33,41],staticinput:[9,31],statist:[9,38,51,54,55,56],statset:35,statu:[32,35,41,42],status:42,std:[20,21,33,38],stderr:36,stdout:36,step:[5,9,10,12,17,26,31,33,35,36,41,42,51,53,54,55,56],still:49,stmt1482205552000:41,stmt1482205746000:41,stochast:[11,17],stock:55,stop:[9,26,36,38,42,53],storag:[41,42,48],store:[8,9,13,33,36,38,41,42,46,48,49,51,53,54,55,56],str:[15,40],straight:32,strategi:[3,12,17,38,54],street:[9,54],strength:47,strict:22,stride:9,stride_i:9,stride_x:9,string:[2,3,8,9,33,38,41,55],strip:[51,53,54],struct:21,structur:[13,36,41,46,48,51,53,54,55,56],sts:41,stub:9,student:52,stuff:32,stun:3,style:[3,9,25,32],sub:[8,9,13,16,31,33,48,51,56],sub_sequ:3,subgradi:11,submit:[32,37,38,41],subnet0:41,subnet:[16,41],subobjectpath:42,subsequenceinput:9,subset:[33,56],substanti:49,substitut:56,succe:55,succeed:42,success:[41,42,49,54],successfulcr:42,successfuli:55,successfulli:[49,53,55],successor:[38,56],sucessfulli:56,sudo:[25,28,41,48],suffic:[22,24],suffici:38,suffix:56,suggest:[9,35],suitabl:[32,38,48],sum:[9,11,18,31,33],sum_:9,sum_evalu:8,summar:[51,55],sumpool:9,support:[7,8,9,12,13,17,22,25,26,28,31,33,35,38,41,54],suppos:[24,33,51],sure:[32,33,41,48,55],survei:55,swap_channel:49,swig:[5,20,21,25],swig_paddl:[5,13,47],symbol:[9,21],sync:[17,32,38,47],syncflag:33,synchron:[11,17,36,38,41],syntact:54,syntax:[22,53],synthect:24,synthes:47,synthet:24,sys:49,system:[17,25,26,36,42,51,54,55,56],t2b:46,t_i:9,tab:[26,51],tabl:[3,9,49,51,56],tableproject:9,tag:[8,13,26,31],tagtyp:8,take:[3,5,8,9,10,16,31,33,35,41,42,47,54,56],taken:[3,54],tanh:[9,10,33],tanhactiv:31,taobao:55,tar:[25,41],tarbal:41,target:[9,13,15,31,46,51,56],target_dict_dim:31,target_dictionary_dim:9,target_language_embed:9,target_language_word:31,targetinlink:9,task:[3,8,9,24,31,40,46,49,54,55,56],tconf:55,tcp:[38,41],teach:51,tear:35,technic:17,technician:52,techniqu:[31,33],tee:[42,48,53,54,55,56],tell:[17,26,35,53],tellig:55,templat:[42,54],tempor:[9,51,54],term:[9,10,17,54,55],termin:42,terminolog:24,tese:2,tesh:54,test100:13,test10:13,test:[2,3,9,13,15,16,21,22,23,25,26,28,32,35,36,37,46,48,49,51,52,56],test_all_data_in_one_period:[42,48,53,54,55],test_data:56,test_fcgrad:33,test_gpuprofil:35,test_layergrad:33,test_list:[3,24,48,51],test_part_000:55,test_pass:[37,38,40,56],test_period:[37,38,40],test_ratio:53,test_wait:[37,38],testa:16,testb:16,testbilinearfwdbwd:35,testconfig:33,tester:[53,56],testfcgrad:33,testfclay:33,testlayergrad:33,testmodel_list:37,testq:16,testresult:15,testsave_dir:37,testutil:33,text:[2,3,8,10,13,16,26,31,41,46,50,51,53,55],text_conv:51,text_conv_pool:53,text_fil:[13,55],tflop:35,tgz:[13,25],than:[3,5,7,8,9,10,17,25,26,31,33,36,41,49,54,55,56],thank:[0,46,56],thei:[3,16,17,24,26,31,33,35,36,37,41,49,55],them:[2,3,10,16,17,22,24,26,31,35,37,38,41,48,49,51,53,55,56],theori:35,therefor:25,therein:9,therun:49,thi:[2,3,7,8,9,10,11,13,15,16,17,22,24,25,26,28,31,32,33,35,36,38,40,41,42,46,47,48,49,51,52,53,54,55,56],thing:[3,24,31,32,35,53,54],think:16,third:[9,17,35,49,55],those:[17,49,54],thought:35,thread:[33,35,38,40,53,54,55,56],thread_local_rand_use_global_se:[37,38],threadid:40,threadloc:35,three:[3,8,9,17,22,24,31,38,47,49,55,56],threshold:[7,8,17,38,55],thriller:52,through:[5,9,17,31,33,35,36,46,47,48,55,56],throughout:51,throughput:35,thu:[3,9,24,33,41,56],tier:42,tight:25,time:[3,9,10,12,13,16,17,22,24,31,35,38,40,42,51,52,54,55,56],timelin:[9,35],timeo:41,timeout:17,timer:25,timestamp:[9,18,52],timestep:[3,9],titil:53,titl:[13,32,52,53],tmall:55,todo:[8,10,13,17],toend:9,togeth:[3,9,10,13,15,31],token:[8,9,16,31,46,55,56],too:[13,26,28],tool:[26,31,32,41,55],toolchain:25,toolkit:[25,28],top:[8,49,54],top_k:8,topolog:[13,16,17],topolopi:15,toronto:[13,48],total:[8,15,17,22,35,36,42,46,56],total_pass:22,touch:55,tourism:55,tourist:56,toward:24,tra:56,track:17,tractabl:9,tradesman:52,tradit:9,trail:13,train100:13,train10:13,train:[1,2,3,5,7,9,11,13,19,31,33,34,35,37,43,44,49],train_conf:[46,56],train_config_dir:41,train_data:56,train_id:41,train_list:[3,24,48,49,51],train_part_000:55,trainabl:9,traindot_period:37,trainer:[3,5,16,18,19,24,33,36,38,40,47,51,54,55,56],trainer_config:[2,3,24,36,41,42,51,53,55],trainer_config_help:[3,24,33,48,51,53],trainer_count:[37,38,40,41,42,53,54,55,56],trainer_id:[38,41],trainerintern:[51,53,56],training_machin:47,trainingtest_period:37,trainonedatabatch:47,tran:[33,38],trane:3,transact:[17,55],transfer:[2,3],transform:[9,31,33,47,48,51,54],transform_param_attr:10,translat:[9,10,24,46,53,55,56],transpar:36,transport:38,transpos:[9,33,47],transposedfullmatrixproject:9,travel:3,travi:[25,32],treat:[9,31],tree:[9,32,38,56],trg:56,trg_dict:31,trg_dict_path:31,trg_embed:31,trg_id:31,trg_ids_next:31,triain:2,tricki:20,trivial:3,trn:51,truck:48,true_imag:22,true_label:22,true_read:22,truth:[8,9,51,56],tst:51,tune:[7,34,51,53,56],tuninglog_barrier_abstract:37,tupl:[3,9,13,15,22],ture:9,turn:[9,22,47],tutori:[26,31,32,33,35,36,41,42,43,44,49,51],tweet:55,twelv:56,twitter:55,two:[2,3,9,10,16,22,24,26,31,35,36,40,41,46,47,48,49,51,53,54,55,56],txt:[3,33,36,41,51,53,55],type:[3,8,9,10,12,13,15,16,17,20,21,22,24,26,31,33,38,40,41,42,48,49,51,53,54],type_nam:53,typedef:[20,21],typic:[5,8,26,35,55],typo_freq:13,ubuntu:[23,28],ubyt:22,uci:13,ufldl:9,uid:42,uint64:20,uint64_t:20,unbalanc:38,unbound:31,unconstrain:55,under:[24,25,26,41,52,55],underli:24,understand:[26,35,46,48,55],understudi:56,undeterminist:35,unemploi:52,unexist:54,uniform:[7,9,13,19,22,38,47],uniqu:[16,17,32,38,41],unique_ptr:33,unit:[9,10,24,25,26,31,32,54],unittest:21,unittestcheckgrad_ep:37,univ:56,unix:36,unk:[46,56],unk_idx:[51,54],unknown:9,unlabel:55,unlik:[54,55,56],unseg:9,unsup:55,unsupbow:55,until:[17,36,41,54],unus:53,unzip:53,updat:[7,9,17,25,33,36,38,40,55],update_equ:15,updatecallback:33,updatestack:41,upload:17,upon:[0,54],upstream:32,uri:41,url:[13,28,55],urls_neg:55,urls_po:55,urls_unsup:55,usag:[2,3,8,9,10,13,15,24,35,46,47,53],use:[0,2,3,5,7,8,9,10,12,13,15,16,17,24,25,26,27,28,31,32,33,35,36,38,40,41,42,46,47,48,49,51,52,53,54,55,56],use_global_stat:9,use_gpu:[37,38,40,42,47,48,49,51,53,54,55,56],use_jpeg:48,use_old_updat:[37,38],use_seq:[24,53],use_seq_or_not:53,used:[2,3,5,8,9,10,11,12,13,15,16,17,22,24,27,28,31,33,35,36,37,38,40,41,46,48,49,51,53,54,55,56],useful:[2,3,9,10,31,33,40,51,54,55],usegpu:[33,47],useless:36,user:[2,3,7,9,10,13,15,16,22,24,26,32,36,37,38,41,49,51,54],user_featur:53,user_head:53,user_id:53,user_info:13,user_meta:53,user_nam:53,userid:52,userinfo:13,usernam:32,uses:[3,17,31,32,33,38,41,48,49,51,53,56],using:[2,3,5,7,9,10,13,16,17,22,24,26,31,32,33,35,38,40,41,42,46,47,48,49,51,54,55],usr:[25,36,38,41],usrdict:46,usrmodel:46,usual:[9,13,15,24,25,35,38,40,41,55],utf:46,util:[5,25,31,33,35,48,53,55],uuid:18,v28:9,valid:[22,41,49,55],valu:[3,5,7,8,9,11,12,13,15,17,24,31,33,38,40,41,47,48,49,54,55],value1:38,value2:38,value_printer_evalu:8,value_rang:13,vanilla:31,vanish:55,vari:[35,41],variabl:[3,9,13,16,24,25,28,33,36,41,42,55],varianc:[9,49],variant:26,vast:32,vector:[3,9,10,13,16,31,33,46,51,53,55,56],vectorenable_parallel_vector:37,verb:[13,54],veri:[3,9,12,31,35,48,51,55],verifi:[32,33],versa:25,version:[9,10,23,25,26,28,33,35,36,37,38,41,42,46,48,52,54,55,56],versu:16,vertic:[9,49],vgg:[10,48],vgg_16_cifar:48,via:[17,22,25,35,36,41,51],vice:25,view:9,vim:32,virtual:26,virtualenv:53,visibl:26,vision:48,visipedia:48,visual:[9,26,35],viterbi:54,voc_dim:51,vocab:55,volum:[26,42],volumemount:[41,42],volumn:41,voluntarili:52,vutbr:13,wai:[3,9,10,16,24,26,31,33,36,40,53,54,56],wait:[11,17,38],walk:[5,47],wall:54,want:[3,9,10,16,22,24,25,26,33,38,40,46,49,51,53,54,55],war:52,warn:9,warp:[9,35],watch:17,wbia:[41,49],web:26,websit:[48,51,54,55],wei:[54,55],weight:[8,9,10,11,31,33,38,40,48,49],weight_act:10,weightlist:33,weights_:33,weights_t:33,welcom:[53,55],well:[26,33,38,41,48,51],west:41,western:52,wether:9,what:[7,9,10,11,24,36,51,53],wheel:25,when:[2,3,7,8,9,13,15,17,28,31,32,33,35,38,40,41,42,46,47,48,54,55,56],whenev:53,where:[3,9,10,11,16,17,24,31,33,35,36,38,40,46,49,54,56],whether:[8,9,10,22,33,38,47,48,53,55,56],which:[0,2,3,5,8,9,10,11,13,16,17,19,22,24,28,31,33,35,36,38,40,41,47,48,49,51,52,53,54,55,56],whichev:47,whl:25,who:[46,49,52],whole:[3,8,13,20,21,41,42,51,52,53,56],whole_cont:53,whose:[3,9,13,17,31,53,54],why:[10,21],wide:54,width:[8,9,13,20,22,33,48,56],wiki:9,wikipedia:[9,13],wilder:3,window:[9,12,13,26,55],wise:9,with_avx:26,with_avxcompil:25,with_coveragecompil:25,with_doccompil:25,with_doubl:33,with_doublecompil:25,with_dsocompil:25,with_gpu:26,with_gpucompil:25,with_profil:35,with_profilercompil:25,with_pythoncompil:25,with_rdmacompil:25,with_style_checkcompil:25,with_swig_pycompil:25,with_test:26,with_testingcompil:25,with_tim:35,with_timercompil:25,within:[9,24],without:[8,9,17,22,36,55],wmt14:56,wmt14_data:56,wmt14_model:56,wmt:56,wmt_shrinked_data:13,woboq:26,won:[35,49],wonder:3,word:[3,8,9,13,31,40,50,53,54,55,56],word_dict:[51,54],word_dim:51,word_id:3,word_idx:13,word_slot:54,word_vector:51,word_vector_dim:[9,31,46],words_freq_sort:13,work:[3,5,13,16,17,22,25,31,32,33,35,36,38,41,42,51,53],worker:41,workercount:41,workflow:[26,32,41],workspac:[26,38,53],worri:24,wors:47,would:[15,17,22,26,36,41,47,51,54],wrap:54,wrapper:[10,35],writ:53,write:[3,13,16,17,22,26,31,32,34,36,41,48,53,54,56],write_lock:18,writelin:24,writer:[16,52],written:[53,55],wrong:[3,22],wsize:41,wsj:54,www:[9,13,48,56],x64:25,xarg:33,xgbe0:38,xgbe1:38,xiaojun:55,xrang:[22,24,33],xxbow:55,xxx:[16,49,56],xxxx:18,xxxxxxxxx:41,xxxxxxxxxx:41,xxxxxxxxxxxxx:41,xxxxxxxxxxxxxxxxxxx:41,xzf:25,y_i:9,y_predict:24,yaml:[41,53],yann:13,year:52,yeild:[15,48],yield:[3,13,16,19,22,24,31,51,53,54,55],you:[2,3,5,7,9,10,24,25,26,28,31,32,33,35,36,38,40,41,46,47,48,49,51,53,54,55,56],your:[3,9,16,25,26,33,35,36,40,41,51,55],your_access_key_id:41,your_secrete_access_kei:41,your_source_root:21,yum:25,yuyang18:[10,13],zachari:55,zeng:55,zero:[3,7,9,11,13,17,33,38,41,51],zhidao:46,zhou:[54,55],zip:[13,52],zone:41,zxvf:41},titles:["ABOUT","API","Introduction","PyDataProvider2","API","Python Prediction","Activation","Parameter Attribute","Evaluators","Layers","Networks","Optimizer","Pooling","Data Reader Interface and DataSets","Model Configuration","Training and Inference","PaddlePaddle Design Doc","Design Doc: Distributed Training","\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\uff08Checkpointing\uff09","\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Python Data Reader Design Doc","Paddle\u53d1\u884c\u89c4\u8303","Simple Linear Regression","Installing from Sources","PaddlePaddle in Docker Containers","Install and Build","Debian Package installation guide","GET STARTED","RNN Models","RNN Configuration","Contribute Code","Write New Layers","HOW TO","Tune GPU Performance","Run Distributed Training","Argument Outline","Detail Description","Set Command-line Parameters","Use Case","Distributed PaddlePaddle Training on AWS with Kubernetes","Paddle On Kubernetes","&lt;no title&gt;","&lt;no title&gt;","PaddlePaddle Documentation","Chinese Word Embedding Model Tutorial","Generative Adversarial Networks (GAN)","Image Classification Tutorial","Model Zoo - ImageNet","TUTORIALS","Quick Start","MovieLens Dataset","Regression MovieLens Ratting","Semantic Role labeling Tutorial","Sentiment Analysis Tutorial","Text generation Tutorial"],titleterms:{"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":19,"\u4e0d\u4f7f\u7528":20,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":20,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":20,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":20,"\u4ec5\u4ec5\u4f7f\u7528void":20,"\u4ece\u5feb\u7167\u6062\u590d":18,"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":20,"\u4f7f\u7528\u8f6c\u6362\u5e93":19,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":21,"\u5206\u652f\u89c4\u8303":23,"\u52a0\u901f\u6267\u884c":18,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":20,"\u52a8\u6001\u6269\u5bb9":18,"\u539f\u56e0":20,"\u539f\u56e0\u5217\u8868":20,"\u57fa\u672c\u8981\u6c42":20,"\u5b9e\u73b0":20,"\u5b9e\u73b0\u65b9\u5f0f":21,"\u5bfc\u51fac":20,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":18,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":20,"\u63a8\u6d4b\u6267\u884c":18,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":19,"\u6587\u4ef6\u9884\u5904\u7406":19,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":21,"\u672f\u8bed":18,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":18,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":21,"\u6d41\u7a0b\u4ecb\u7ecd":19,"\u76ee\u5f55\u7ed3\u6784":21,"\u793a\u4f8b\u7a0b\u5e8f":19,"\u7b26\u53f7":20,"\u7c7b":20,"\u7f16\u8bd1\u9009\u9879":21,"\u7f29\u5bb9":18,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":20,"\u80cc\u666f":20,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8":19,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":19,"\u8f6c\u6362\u5e93":19,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":20,"\u8fdb\u884c\u8bad\u7ec3":19,"book\u4e2d\u6240\u6709\u7ae0\u8282":23,"case":40,"class":33,"function":46,"new":33,"paddle\u52a8\u6001\u5e93\u4e2d":20,"paddle\u53d1\u884c\u89c4\u8303":23,"paddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":23,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":20,"return":22,AWS:41,Abs:6,DNS:41,EFS:41,For:42,KMS:41,Use:[40,42],Using:[26,32],With:26,about:0,access:41,account:41,activ:6,adadelta:11,adagrad:11,adam:11,adamax:11,add:41,address:41,addto:9,adversari:47,aggreg:9,algorithm:[17,51],analysi:55,api:[1,4,21,26],appendix:51,applic:4,approach:35,architectur:[31,51],argument:[22,37,40,51],asset:41,associ:41,async:38,attent:31,attribut:7,auc:8,avg:12,aws:41,background:24,basepool:12,batch:22,batch_norm:9,batch_siz:22,beam_search:9,between:16,bidirect:55,bidirectional_gru:10,bidirectional_lstm:10,bilinear_interp:9,bleu:56,block_expand:9,book:26,brelu:6,bucket:41,build:[25,27,42],built:35,cach:3,capi:21,capi_priv:21,cento:25,check:[9,33,36],checkpoint:[17,18],chines:46,choos:41,chunk:8,cifar:13,classif:[8,48],classification_error:8,classification_error_print:8,clone:32,cloudform:41,cluster:[36,40,41],code:32,column_sum:8,command:[39,40,51,56],commit:[32,42],common:38,commun:38,compos:22,concat:9,concept:41,config:[4,40,53,54],configur:[14,31,34,36,41,51,53],conll05:13,connect:9,contain:[26,42],content:[21,35,41],context_project:9,contribut:32,conv:9,conv_oper:9,conv_project:9,conv_shift:9,convolut:[48,51],core:41,cos_sim:9,cost:9,cpu:[26,40],creat:[22,32,41,42],creator:22,credenti:41,credit:0,crf:9,crf_decod:9,cross_channel_norm:9,cross_entropy_cost:9,cross_entropy_with_selfnorm_cost:9,ctc:9,ctc_error:8,cudnnavg:12,cudnnmax:12,custom:22,dat:52,data:[9,13,17,22,24,31,41,42,46,47,48,51,53,54,55,56],datafeed:13,dataprovid:[3,4,38],dataset:[13,17,52,53,56],datatyp:13,date:32,debian:28,decayedadagrad:11,decor:22,defin:[41,51,55,56],delet:41,delv:48,demo:41,depend:25,deriv:33,descript:[38,47,52,54],design:[16,17,22],destroi:41,detail:[38,48],develop:[26,34],devic:40,dictionari:[22,46],differ:40,directori:41,dispatch:17,distribut:[16,17,36,38,41],doc:[16,17,22],docker:[26,42],document:[26,45],dotmul_oper:9,dotmul_project:9,down:41,download:[25,41,42,46,49,53,56],dropout_lay:10,dylib:21,dynam:17,ec2:41,elast:41,embed:[9,46,51],entri:22,eos:9,equat:33,evalu:[8,24,53],evalutaion:56,event:[15,16],exampl:[16,21,46,47],exercis:48,exp:6,expand:9,extern:41,extract:[46,49,53,56],fault:17,featur:[49,52,53,54],field:53,file:[41,42,51,52,53],find:41,first_seq:9,fork:32,format:[17,51],from:[16,25,27],full_matrix_project:9,fulli:9,gan:47,gate:31,gener:[31,47,56],get:[29,42],get_output:9,github:32,gpu:[26,35,38,40],gradient:33,gradient_print:8,group:[9,41],gru:[10,38],gru_group:10,gru_step:9,gru_unit:10,grumemori:9,guid:28,hand:35,handler:[16,20],hook:32,how:[22,34,35],hsigmoid:9,huber_cost:9,iam:41,ident:6,identity_project:9,imag:[9,10,26,42,48],imagenet:49,imdb:[13,55],img_cmrnorm:9,img_conv:9,img_conv_bn_pool:10,img_conv_group:10,img_pool:9,imikolov:13,implement:[22,33,47],infer:[15,51],info:49,ingredi:16,init_hook:3,initi:[40,41],input_typ:3,inspect:41,instal:[25,27,28,41,51],instanc:41,integr:41,interfac:[13,17,22,49],interpol:9,introduct:[2,46,49,55,56],isn:22,job:[17,36,41,42],join:9,keep:32,kei:41,kill:36,kube:41,kubectl:41,kubernet:[41,42],label:54,lambda_cost:9,last_seq:9,lastest:32,launch:36,layer:[9,16,33,40],libpaddle_capi_shar:21,libpaddle_capi_whol:21,line:[39,51],linear:[6,24],linear_comb:9,list:[18,22],local:[40,41],log:[6,51],logist:51,lstm:[10,38,54,55],lstm_step:9,lstmemori:9,lstmemory_group:10,lstmemory_unit:10,map:22,master:17,math:9,matrix:38,max:12,maxframe_print:8,maxid:9,maxid_print:8,maxout:9,memori:9,meta:53,mini:22,minibatch:13,misc:10,mix:[9,40],mnist:[13,47],model:[3,4,14,16,24,26,30,31,36,40,46,47,48,49,50,51,56],modifi:42,momentum:11,movi:[52,53],movielen:[13,52,53],mse_cost:9,multi_binary_label_cross_entropy_cost:9,multipl:22,name:41,nce:9,need:[22,35],network:[10,31,40,47,48,49,51,53,54],neural:[31,48,51,53,54],neuralnetwork:24,nlp:[10,38],non:3,norm:9,nvprof:35,nvvp:35,object:[17,53],observ:[46,49],onli:[22,26],optim:[11,17,34,51],option:[25,46],outlin:37,output:[36,41],overview:51,packag:28,pad:9,paddl:[22,23,42],paddlepaddl:[16,26,27,41,45,46,56],pair:41,parallel_nn:40,paramet:[7,15,16,17,38,39,41,46,49],paraphras:46,pass:40,perform:[35,38],pnpair:8,point:41,pool:[9,12],power:9,pre:32,precision_recal:8,predict:[5,48,49,53,54,55],prefetch:22,prepar:[24,31,36,41,46,47,48,53,55,56],preprocess:[46,48,51,53,56],prerequisit:36,pretrain:[46,56],print:8,privat:41,problem:24,process:17,profil:35,provid:[3,22,51,53,54],pull:32,push:32,pydataprovider2:3,python:[5,22,26,33,49,51,53],queue:17,quick:51,randomnumb:38,rank:8,rank_cost:9,rat:53,rate:52,reader:[13,16,22],recoveri:17,recurr:[9,10,31,51],recurrent_group:9,refer:[3,35,54,55],region:41,regress:[24,51,53],relu:6,render:41,repeat:9,request:32,requir:[25,32],reshap:9,resnet:49,result:[36,42,56],revis:[32,46],rmsprop:11,rnn:[30,31,38],role:54,rotat:9,route53:41,run:[36,42,54],sampl:9,sampling_id:9,scale:[9,17],scaling_project:9,script:42,secur:41,selective_fc:9,semant:54,sentiment:[13,55],seq_concat:9,seq_reshap:9,seqtext_print:8,sequenc:31,sequence_conv_pool:10,sequencesoftmax:6,sequenti:3,server:[17,38,41],servic:41,set:39,setup:[25,41],sgd:38,share:16,shuffl:22,sigmoid:6,simpl:[24,31],simple_attent:10,simple_gru2:10,simple_gru:10,simple_img_conv_pool:10,simple_lstm:10,singl:22,slice:9,slope_intercept:9,small_vgg:10,smooth_l1_cost:9,softmax:6,softrelu:6,sourc:[25,27],span:25,spars:40,specifi:[40,46],split:53,spp:9,squar:6,squarerootn:12,stack:55,standard:51,stanh:6,start:[16,29,41,42,51],startup:42,store:17,structur:47,suffici:22,sum:[8,12],sum_cost:9,sum_to_one_norm:9,summar:16,summari:51,system:41,tabl:21,table_project:9,take:22,tanh:6,task:17,tear:41,templat:41,tensor:9,test:[33,38,40,53,54,55],text:56,text_conv_pool:10,timer:35,tip:35,todo:[18,19],toi:47,toler:17,tool:35,train:[15,16,17,22,24,26,36,38,40,41,42,46,47,48,51,53,54,55,56],trainer:[15,17,41,53],tran:9,trans_full_matrix_project:9,transfer:51,tune:[35,38],tutori:[46,48,50,54,55,56],ubuntu:25,uci_h:13,unit:[33,38],updat:[16,32,41],usag:[22,26,34],use:22,user:[17,46,52,53,55,56],util:8,value_print:8,vector:38,verifi:41,version:32,vgg_16_network:10,visual:49,volum:41,vpc:41,warp_ctc:9,what:35,why:[22,35],wmt14:13,word:[46,51],work:26,workflow:56,workspac:36,wrapper:33,write:[33,51],yaml:42,your:32,zoo:[49,50]}})
\ No newline at end of file
diff --git a/develop/doc/tutorials/embedding_model/index_en.html b/develop/doc/tutorials/embedding_model/index_en.html
index 0f063ffcd7067366a99f6b79390466d2348dbc34..556a1327ecfe1e4e9832befb601476a73dc6af92 100644
--- a/develop/doc/tutorials/embedding_model/index_en.html
+++ b/develop/doc/tutorials/embedding_model/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/gan/index_en.html b/develop/doc/tutorials/gan/index_en.html
index d26e1d6ca64237efb80cb295a6a1ed4d647e8642..715b3539a4eab5663507d03738582847cb3ce073 100644
--- a/develop/doc/tutorials/gan/index_en.html
+++ b/develop/doc/tutorials/gan/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/image_classification/index_en.html b/develop/doc/tutorials/image_classification/index_en.html
index 854d1b9f4deca83a1d5d9c64388cb40a0074719d..d38f28a8df2f2484fd04c4ff5538b304cdd3eda0 100644
--- a/develop/doc/tutorials/image_classification/index_en.html
+++ b/develop/doc/tutorials/image_classification/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/imagenet_model/resnet_model_en.html b/develop/doc/tutorials/imagenet_model/resnet_model_en.html
index 85a8e7cb3ec069923f876a5d0c7d3bf8063628c8..128776ca8e8ac7a4dee83f751168a0a6f25c9ddf 100644
--- a/develop/doc/tutorials/imagenet_model/resnet_model_en.html
+++ b/develop/doc/tutorials/imagenet_model/resnet_model_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/index_en.html b/develop/doc/tutorials/index_en.html
index e44145f43c5a9885e122b1d26aba6d9dda379960..fb095fd82972331f1bc9e9db1ed52101adecd821 100644
--- a/develop/doc/tutorials/index_en.html
+++ b/develop/doc/tutorials/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/quick_start/index_en.html b/develop/doc/tutorials/quick_start/index_en.html
index 7136215114fbe84276b81dc5c079074570bd2997..34a5f25c3018eed7015cbd3e95d6b055898964a4 100644
--- a/develop/doc/tutorials/quick_start/index_en.html
+++ b/develop/doc/tutorials/quick_start/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/rec/ml_dataset_en.html b/develop/doc/tutorials/rec/ml_dataset_en.html
index f0617fed5105b741b307a7151262d0da23bad59a..72637504e6ecaa1efe29978f30b0ee3e14a859d7 100644
--- a/develop/doc/tutorials/rec/ml_dataset_en.html
+++ b/develop/doc/tutorials/rec/ml_dataset_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/rec/ml_regression_en.html b/develop/doc/tutorials/rec/ml_regression_en.html
index 42f55c2406bb0cca2307ad456df4f613099ba016..7b9751c945e1bacf80a7bf8939166db21306bf23 100644
--- a/develop/doc/tutorials/rec/ml_regression_en.html
+++ b/develop/doc/tutorials/rec/ml_regression_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
@@ -611,14 +612,14 @@ get one user feature. Then we calculate the cosine similarity of these two
 features.</p>
 <p>In these networks, we use several APIs in <a class="reference internal" href="../../api/v1/index_en.html#api-trainer-config"><span class="std std-ref">Model Config API</span></a> . There are</p>
 <ul class="simple">
-<li>Data Layer, <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-data-layer"><span class="std std-ref">data_layer</span></a></li>
-<li>Fully Connected Layer, <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-fc-layer"><span class="std std-ref">fc_layer</span></a></li>
-<li>Embedding Layer, <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-embedding-layer"><span class="std std-ref">embedding_layer</span></a></li>
-<li>Context Projection Layer, <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-context-projection"><span class="std std-ref">context_projection</span></a></li>
-<li>Pooling Layer, <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-pooling-layer"><span class="std std-ref">pooling_layer</span></a></li>
-<li>Cosine Similarity Layer, <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-cos-sim"><span class="std std-ref">cos_sim</span></a></li>
+<li>Data Layer, <span class="xref std std-ref">api_trainer_config_helpers_layers_data_layer</span></li>
+<li>Fully Connected Layer, <span class="xref std std-ref">api_trainer_config_helpers_layers_fc_layer</span></li>
+<li>Embedding Layer, <span class="xref std std-ref">api_trainer_config_helpers_layers_embedding_layer</span></li>
+<li>Context Projection Layer, <span class="xref std std-ref">api_trainer_config_helpers_layers_context_projection</span></li>
+<li>Pooling Layer, <span class="xref std std-ref">api_trainer_config_helpers_layers_pooling_layer</span></li>
+<li>Cosine Similarity Layer, <span class="xref std std-ref">api_trainer_config_helpers_layers_cos_sim</span></li>
 <li>Text Convolution Pooling Layer, <a class="reference internal" href="../../api/v2/config/networks.html#api-trainer-config-helpers-network-text-conv-pool"><span class="std std-ref">text_conv_pool</span></a></li>
-<li>Declare Python Data Sources <a class="reference internal" href="../../api/v1/trainer_config_helpers/data_sources.html#api-trainer-config-helpers-data-sources"><span class="std std-ref">DataSources</span></a>.</li>
+<li>Declare Python Data Sources <span class="xref std std-ref">api_trainer_config_helpers_data_sources</span>.</li>
 </ul>
 </div>
 <div class="section" id="data-provider">
diff --git a/develop/doc/tutorials/semantic_role_labeling/index_en.html b/develop/doc/tutorials/semantic_role_labeling/index_en.html
index 8a1a9d5b66c67dbab52ddf8d058e0519ccc65496..da2715ecd55a1e70c5f7a09b9bce182725ba2a7c 100644
--- a/develop/doc/tutorials/semantic_role_labeling/index_en.html
+++ b/develop/doc/tutorials/semantic_role_labeling/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/sentiment_analysis/index_en.html b/develop/doc/tutorials/sentiment_analysis/index_en.html
index 86f11ded6f39daa34175d63987fb95ca75069618..67a7007a7ce25497e41639d0031d99201871eba2 100644
--- a/develop/doc/tutorials/sentiment_analysis/index_en.html
+++ b/develop/doc/tutorials/sentiment_analysis/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc/tutorials/text_generation/index_en.html b/develop/doc/tutorials/text_generation/index_en.html
index 6f7d9f54ced24a520311505e733697cc97fe3bcb..c33e58feef6d2ddf591b91922b1044aefac2a5b2 100644
--- a/develop/doc/tutorials/text_generation/index_en.html
+++ b/develop/doc/tutorials/text_generation/index_en.html
@@ -135,6 +135,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/activations.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/activations.rst.txt
deleted file mode 100644
index 269e6491e7ebe3899c3fb24fca756a393043473b..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/activations.rst.txt
+++ /dev/null
@@ -1,108 +0,0 @@
-===========
-Activations
-===========
-
-BaseActivation
-==============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: BaseActivation
-    :noindex:
-    
-AbsActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: AbsActivation
-    :noindex:
-    
-ExpActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: ExpActivation
-    :noindex:
-    
-IdentityActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: IdentityActivation
-    :noindex:
-    
-LinearActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: LinearActivation
-    :noindex:
-
-LogActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: LogActivation
-    :noindex:
-    
-SquareActivation
-================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SquareActivation
-    :noindex:
-    
-SigmoidActivation
-=================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SigmoidActivation
-    :noindex:
-    
-SoftmaxActivation
-=================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SoftmaxActivation
-    :noindex:
-    
-SequenceSoftmaxActivation
-=========================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SequenceSoftmaxActivation
-    :noindex:
-    
-ReluActivation
-==============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: ReluActivation
-    :noindex:
-    
-BReluActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: BReluActivation
-    :noindex:
-    
-SoftReluActivation
-==================
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: SoftReluActivation
-    :noindex:
-    
-TanhActivation
-==============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: TanhActivation
-    :noindex:
-    
-STanhActivation
-===============
-
-..  automodule:: paddle.trainer_config_helpers.activations
-    :members: STanhActivation
-    :noindex:
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/attrs.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/attrs.rst.txt
deleted file mode 100644
index ac63127bf7d9db6351365ab7b58f43db12347a8e..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/attrs.rst.txt
+++ /dev/null
@@ -1,5 +0,0 @@
-Parameter Attributes
-=======================
-
-..  automodule:: paddle.trainer_config_helpers.attrs
-    :members:
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/data_sources.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/data_sources.rst.txt
deleted file mode 100644
index b9dd4dda01ae59d1260356aff50ddf298d02c87f..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/data_sources.rst.txt
+++ /dev/null
@@ -1,7 +0,0 @@
-..  _api_trainer_config_helpers_data_sources:
-
-DataSources
-===========
-
-..  automodule:: paddle.trainer_config_helpers.data_sources
-    :members:
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/evaluators.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/evaluators.rst.txt
deleted file mode 100644
index 11dc735164284d6ed1d661fab1e7690d263b3a7c..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/evaluators.rst.txt
+++ /dev/null
@@ -1,108 +0,0 @@
-..  _api_trainer_config_helpers_evaluators:
-
-==========
-Evaluators
-==========
-
-Base
-====
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: evaluator_base
-    :noindex:
-
-Classification 
-==============
-
-classification_error_evaluator
-------------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: classification_error_evaluator
-    :noindex:
-
-auc_evaluator
--------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: auc_evaluator
-    :noindex:
-
-ctc_error_evaluator
--------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: ctc_error_evaluator
-    :noindex:
-
-chunk_evaluator
----------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: chunk_evaluator
-    :noindex:
-
-precision_recall_evaluator
---------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  precision_recall_evaluator
-    :noindex:
-
-Rank
-====
-
-pnpair_evaluator
-----------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  pnpair_evaluator
-    :noindex:
-
-Utils
-=====
-
-sum_evaluator
--------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: sum_evaluator
-    :noindex:
-
-column_sum_evaluator
---------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members: column_sum_evaluator
-    :noindex:
-
-Print
-=====
-
-classification_error_printer_evaluator
---------------------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  classification_error_printer_evaluator
-    :noindex:
-
-gradient_printer_evaluator
---------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  gradient_printer_evaluator
-    :noindex:
-
-maxid_printer_evaluator
------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  maxid_printer_evaluator
-    :noindex:
-
-maxframe_printer_evaluator
----------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  maxframe_printer_evaluator
-    :noindex:
-
-seqtext_printer_evaluator
--------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  seqtext_printer_evaluator
-    :noindex:
-
-value_printer_evaluator
------------------------
-..  automodule:: paddle.trainer_config_helpers.evaluators
-    :members:  value_printer_evaluator
-    :noindex:
-
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/layers.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/layers.rst.txt
deleted file mode 100644
index 75c1b35246486528524fd5baa04704249d5a9275..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/layers.rst.txt
+++ /dev/null
@@ -1,514 +0,0 @@
-..  _api_trainer_config_helpers_layers:
-
-======
-Layers
-======
-
-Base
-======
-
-LayerType
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: LayerType
-    :noindex:
-
-LayerOutput
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: LayerOutput
-    :noindex:
-
-Data layer
-===========
-
-..  _api_trainer_config_helpers_layers_data_layer:
-
-data_layer
-----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: data_layer
-    :noindex:
-
-Fully Connected Layers
-======================
-
-..  _api_trainer_config_helpers_layers_fc_layer:
-
-fc_layer
---------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: fc_layer
-    :noindex:
-
-selective_fc_layer
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: selective_fc_layer
-    :noindex:
-
-Conv Layers
-===========
-
-conv_operator
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: conv_operator
-    :noindex:
-
-conv_projection
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: conv_projection
-    :noindex:
-
-conv_shift_layer
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: conv_shift_layer
-    :noindex:
-
-img_conv_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: img_conv_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_context_projection:
-
-context_projection 
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: context_projection
-    :noindex:
-
-Image Pooling Layer
-===================
-
-img_pool_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: img_pool_layer
-    :noindex:   
-
-spp_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: spp_layer
-    :noindex:
-
-maxout_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: maxout_layer
-    :noindex:
-
-Norm Layer
-==========
-
-img_cmrnorm_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: img_cmrnorm_layer
-    :noindex:
-
-batch_norm_layer
----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: batch_norm_layer
-    :noindex:
-
-sum_to_one_norm_layer
----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: sum_to_one_norm_layer
-    :noindex:
-    
-Recurrent Layers
-================
-
-recurrent_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: recurrent_layer
-    :noindex:
-
-lstmemory
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: lstmemory
-    :noindex:
-
-grumemory
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: grumemory
-    :noindex:
-
-Recurrent Layer Group
-=====================
-
-memory
-------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: memory
-    :noindex:
-
-recurrent_group
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: recurrent_group
-    :noindex:
-    
-lstm_step_layer
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: lstm_step_layer
-    :noindex:
-
-gru_step_layer
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: gru_step_layer
-    :noindex:
-
-beam_search
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: beam_search
-    :noindex:
-    
-get_output_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: get_output_layer
-    :noindex:
-    
-Mixed Layer
-===========
-
-..  _api_trainer_config_helpers_layers_mixed_layer:
-
-mixed_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: mixed_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_embedding_layer:
-
-embedding_layer
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: embedding_layer
-    :noindex:
-
-scaling_projection
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: scaling_projection
-    :noindex:
-
-dotmul_projection
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: dotmul_projection
-    :noindex:
-
-dotmul_operator
----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: dotmul_operator
-    :noindex:
-
-full_matrix_projection
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: full_matrix_projection
-    :noindex:
-
-identity_projection
--------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: identity_projection
-    :noindex:
-
-
-table_projection
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: table_projection
-    :noindex:
-
-trans_full_matrix_projection
-----------------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: trans_full_matrix_projection
-    :noindex:
-    
-Aggregate Layers
-================
-
-..  _api_trainer_config_helpers_layers_pooling_layer:
-
-pooling_layer
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: pooling_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_last_seq:
-
-last_seq
---------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: last_seq
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_first_seq:
-
-first_seq
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: first_seq
-    :noindex:
-
-concat_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: concat_layer
-    :noindex:
-
-seq_concat_layer
-----------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: seq_concat_layer
-    :noindex:
-
-Reshaping Layers
-================
-
-block_expand_layer
-------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: block_expand_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_expand_layer:
-
-expand_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: expand_layer
-    :noindex:
-
-repeat_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: repeat_layer
-    :noindex:
-
-rotate_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: rotate_layer
-    :noindex:
-
-seq_reshape_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: seq_reshape_layer
-    :noindex:
-
-Math Layers
-===========
-
-addto_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: addto_layer
-    :noindex:
-
-linear_comb_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: linear_comb_layer
-    :noindex:
-
-interpolation_layer
--------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: interpolation_layer
-    :noindex:
-
-bilinear_interp_layer
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: bilinear_interp_layer
-    :noindex:
-
-power_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: power_layer
-    :noindex:
-
-scaling_layer
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: scaling_layer
-    :noindex:
-
-slope_intercept_layer
-----------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: slope_intercept_layer
-    :noindex:
-
-tensor_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: tensor_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_cos_sim:
-
-cos_sim
--------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: cos_sim
-    :noindex:
-
-trans_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: trans_layer
-    :noindex:
-
-Sampling Layers
-===============
-
-maxid_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: maxid_layer
-    :noindex:
-
-sampling_id_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: sampling_id_layer
-    :noindex:
-
-Slicing and Joining Layers
-==========================
-
-pad_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: pad_layer
-    :noindex:
-
-..  _api_trainer_config_helpers_layers_cost_layers:
-
-Cost Layers
-===========
-
-cross_entropy
--------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: cross_entropy
-    :noindex:
-
-cross_entropy_with_selfnorm
----------------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: cross_entropy_with_selfnorm
-    :noindex:
-
-multi_binary_label_cross_entropy
---------------------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: multi_binary_label_cross_entropy
-    :noindex:
-
-mse_cost
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: mse_cost
-    :noindex:
-
-huber_cost
-----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: huber_cost
-    :noindex:
-
-lambda_cost
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: lambda_cost
-    :noindex:
-
-rank_cost
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: rank_cost
-    :noindex:
-
-sum_cost
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: sum_cost
-    :noindex:
-
-crf_layer
------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: crf_layer
-    :noindex:
-
-crf_decoding_layer
--------------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: crf_decoding_layer
-    :noindex:
-
-ctc_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: ctc_layer
-    :noindex:
-
-warp_ctc_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: warp_ctc_layer
-    :noindex:
-
-nce_layer
------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: nce_layer
-    :noindex:
-
-hsigmoid
----------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: hsigmoid
-    :noindex:
-
-smooth_l1_cost
---------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: smooth_l1_cost
-    :noindex:
-
-Check Layer 
-============
-
-eos_layer
-------------
-..  automodule:: paddle.trainer_config_helpers.layers
-    :members: eos_layer
-    :noindex:
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/networks.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/networks.rst.txt
deleted file mode 100644
index edb53acbf0c31532aa34bda044066fed72eaa426..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/networks.rst.txt
+++ /dev/null
@@ -1,123 +0,0 @@
-========
-Networks
-========
-
-The networks module contains pieces of neural network that combine multiple layers.
-
-NLP
-===
-
-sequence_conv_pool
-------------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: sequence_conv_pool
-    :noindex:
-
-..  _api_trainer_config_helpers_network_text_conv_pool:
-
-text_conv_pool
---------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: text_conv_pool
-    :noindex:
-
-Images
-======
-
-img_conv_bn_pool
-----------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: img_conv_bn_pool
-    :noindex:
-
-img_conv_group
---------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: img_conv_group
-    :noindex:
-
-..  _api_trainer_config_helpers_network_simple_img_conv_pool:
-
-simple_img_conv_pool
---------------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_img_conv_pool
-    :noindex:
-
-vgg_16_network
----------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: vgg_16_network
-    :noindex:
-
-Recurrent
-=========
-
-LSTM
-----
-
-lstmemory_unit
-``````````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: lstmemory_unit
-    :noindex:
-
-lstmemory_group
-```````````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: lstmemory_group
-    :noindex:
-
-simple_lstm
-```````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_lstm
-    :noindex:
-
-bidirectional_lstm
-``````````````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: bidirectional_lstm
-    :noindex:
-
-GRU
----
-
-gru_unit
-````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: gru_unit
-    :noindex:
-
-gru_group
-`````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: gru_group
-    :noindex:
-
-simple_gru
-``````````
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_gru
-    :noindex:
-
-simple_attention
-----------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: simple_attention
-    :noindex:
-
-Miscs
-=====
-
-dropout_layer
---------------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: dropout_layer
-    :noindex:
-
-outputs
--------
-..  automodule:: paddle.trainer_config_helpers.networks
-    :members: outputs
-    :noindex:
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/optimizers.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/optimizers.rst.txt
deleted file mode 100644
index d2f4958c92b8e3b7426945f1af07112ab4071136..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/optimizers.rst.txt
+++ /dev/null
@@ -1,61 +0,0 @@
-..  _api_trainer_config_helpers_optimizers:
-
-==========
-Optimizers
-==========
-
-BaseSGDOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: BaseSGDOptimizer
-    :noindex:
-
-MomentumOptimizer
-=================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: MomentumOptimizer
-    :noindex:
-
-AdamOptimizer
-=============
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdamOptimizer
-    :noindex:
-
-AdamaxOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdamaxOptimizer
-    :noindex:
-
-AdaGradOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdaGradOptimizer
-    :noindex:
-
-DecayedAdaGradOptimizer
-=======================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: DecayedAdaGradOptimizer
-    :noindex:
-
-AdaDeltaOptimizer
-=================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: AdaDeltaOptimizer
-    :noindex:
-
-RMSPropOptimizer
-================
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: RMSPropOptimizer
-    :noindex:
-
-..  _api_trainer_config_helpers_optimizers_settings:
-
-settings
-========
-..  automodule:: paddle.trainer_config_helpers.optimizers
-    :members: settings
-    :noindex:
diff --git a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/poolings.rst.txt b/develop/doc_cn/_sources/api/v1/trainer_config_helpers/poolings.rst.txt
deleted file mode 100644
index 66566809d26f59263597b5286c5b27e0bbc9415a..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/v1/trainer_config_helpers/poolings.rst.txt
+++ /dev/null
@@ -1,33 +0,0 @@
-========
-Poolings
-========
-
-BasePoolingType
-===============
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: BasePoolingType
-    :noindex:
-
-AvgPooling
-==========
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: AvgPooling
-    :noindex:
-
-MaxPooling
-==========
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: MaxPooling
-    :noindex:
-
-SumPooling
-==========
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: SumPooling
-    :noindex:
-
-SquareRootNPooling
-==================
-..  automodule:: paddle.trainer_config_helpers.poolings
-    :members: SquareRootNPooling
-    :noindex:
diff --git a/develop/doc_cn/_sources/api/v2/config/evaluators.rst.txt b/develop/doc_cn/_sources/api/v2/config/evaluators.rst.txt
new file mode 100644
index 0000000000000000000000000000000000000000..39db51fa4abc370855ca3f2778b47464f33b6fce
--- /dev/null
+++ b/develop/doc_cn/_sources/api/v2/config/evaluators.rst.txt
@@ -0,0 +1,101 @@
+..  _api_v2:
+
+==========
+Evaluators
+==========
+
+Classification 
+==============
+
+classification_error
+--------------------
+..  automodule:: paddle.v2.evaluator
+    :members: classification_error
+    :noindex:
+
+auc
+---
+..  automodule:: paddle.v2.evaluator
+    :members: auc
+    :noindex:
+
+ctc_error
+---------
+..  automodule:: paddle.v2.evaluator
+    :members: ctc_error
+    :noindex:
+
+chunk
+-----
+..  automodule:: paddle.v2.evaluator
+    :members: chunk
+    :noindex:
+
+precision_recall
+----------------
+..  automodule:: paddle.v2.evaluator
+    :members:  precision_recall
+    :noindex:
+
+Rank
+====
+
+pnpair
+------
+..  automodule:: paddle.v2.evaluator
+    :members:  pnpair
+    :noindex:
+
+Utils
+=====
+
+sum
+---
+..  automodule:: paddle.v2.evaluator
+    :members: sum
+    :noindex:
+
+column_sum
+----------
+..  automodule:: paddle.v2.evaluator
+    :members: column_sum
+    :noindex:
+
+Print
+=====
+
+classification_error_printer
+----------------------------
+..  automodule:: paddle.v2.evaluator
+    :members:  classification_error_printer
+    :noindex:
+
+gradient_printer
+----------------
+..  automodule:: paddle.v2.evaluator
+    :members:  gradient_printer
+    :noindex:
+
+maxid_printer
+-------------
+..  automodule:: paddle.v2.evaluator
+    :members:  maxid_printer
+    :noindex:
+
+maxframe_printer
+----------------
+..  automodule:: paddle.v2.evaluator
+    :members:  maxframe_printer
+    :noindex:
+
+seqtext_printer
+---------------
+..  automodule:: paddle.v2.evaluator
+    :members:  seqtext_printer
+    :noindex:
+
+value_printer
+-------------
+..  automodule:: paddle.v2.evaluator
+    :members:  value_printer
+    :noindex:
diff --git a/develop/doc_cn/_sources/api/v2/config/networks.rst.txt b/develop/doc_cn/_sources/api/v2/config/networks.rst.txt
index 6f209bc95bec7279051118bb857a96515e0371a9..b2a617fff134035c04eeabbbaf6d9cbe2a525f1c 100644
--- a/develop/doc_cn/_sources/api/v2/config/networks.rst.txt
+++ b/develop/doc_cn/_sources/api/v2/config/networks.rst.txt
@@ -44,6 +44,12 @@ simple_img_conv_pool
     :members: simple_img_conv_pool
     :noindex:
 
+small_vgg
+---------
+..  automodule:: paddle.v2.networks
+    :members: small_vgg
+    :noindex:
+
 vgg_16_network
 ---------------
 ..  automodule:: paddle.v2.networks
@@ -101,6 +107,18 @@ simple_gru
     :members: simple_gru
     :noindex:
 
+simple_gru2
+```````````
+..  automodule:: paddle.v2.networks
+    :members: simple_gru2
+    :noindex:
+
+bidirectional_gru
+``````````````````
+..  automodule:: paddle.v2.networks
+    :members: bidirectional_gru
+    :noindex:
+
 simple_attention
 ----------------
 ..  automodule:: paddle.v2.networks
diff --git a/develop/doc_cn/_sources/api/v2/model_configs.rst.txt b/develop/doc_cn/_sources/api/v2/model_configs.rst.txt
index a5fae7e29e56d9e236d489353a5c8967d1954641..992b559cbd87244612521d4c96f84f997d6c4196 100644
--- a/develop/doc_cn/_sources/api/v2/model_configs.rst.txt
+++ b/develop/doc_cn/_sources/api/v2/model_configs.rst.txt
@@ -6,6 +6,7 @@ Model Configuration
 
     config/activation.rst
     config/layer.rst
+    config/evaluators.rst
     config/optimizer.rst
     config/pooling.rst
     config/networks.rst
diff --git a/develop/doc_cn/about/index_cn.html b/develop/doc_cn/about/index_cn.html
index 68443a239f3b7a45a80bdfbdff1eff32c6b8f24d..ef2297b6c9dc99bb3aa33cd6baaef23bcc717f45 100644
--- a/develop/doc_cn/about/index_cn.html
+++ b/develop/doc_cn/about/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/index_cn.html b/develop/doc_cn/api/index_cn.html
index 5f65e1170968cf4bfed070ecddd1601703fbfb86..cc189acabe329ac4bdaae75e5656ea085577d21d 100644
--- a/develop/doc_cn/api/index_cn.html
+++ b/develop/doc_cn/api/index_cn.html
@@ -144,6 +144,7 @@
 <li class="toctree-l2"><a class="reference internal" href="v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v1/data_provider/dataprovider_cn.html b/develop/doc_cn/api/v1/data_provider/dataprovider_cn.html
index bdba7a46d0eb7fa49d3a1c318faf481d03fe9f0b..19d3b0e51803d6f7d2200a51a2a8ea32f4afd5fc 100644
--- a/develop/doc_cn/api/v1/data_provider/dataprovider_cn.html
+++ b/develop/doc_cn/api/v1/data_provider/dataprovider_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v1/data_provider/pydataprovider2_cn.html b/develop/doc_cn/api/v1/data_provider/pydataprovider2_cn.html
index e797913bdfa8de9ee178d279b381e56930b0c969..3ca2c8e81d8042c080829a2b7e130c03c590e08a 100644
--- a/develop/doc_cn/api/v1/data_provider/pydataprovider2_cn.html
+++ b/develop/doc_cn/api/v1/data_provider/pydataprovider2_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v1/index_cn.html b/develop/doc_cn/api/v1/index_cn.html
index f0753c7b8624456b7fae327e11cb914c7cb289ab..c35f503dc6d0cd150f485571a3e95b9eeaaaa4d2 100644
--- a/develop/doc_cn/api/v1/index_cn.html
+++ b/develop/doc_cn/api/v1/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../v2/config/networks.html">Networks</a></li>
@@ -196,16 +197,6 @@
 <div class="section" id="model-config-api">
 <span id="api-trainer-config"></span><h2>Model Config API<a class="headerlink" href="#model-config-api" title="永久链接至标题">¶</a></h2>
 <div class="toctree-wrapper compound">
-<ul>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/optimizers.html">Optimizers</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/data_sources.html">DataSources</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/layers.html">Layers</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/activations.html">Activations</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/poolings.html">Poolings</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/networks.html">Networks</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/evaluators.html">Evaluators</a></li>
-<li class="toctree-l1"><a class="reference internal" href="trainer_config_helpers/attrs.html">Parameter Attributes</a></li>
-</ul>
 </div>
 </div>
 <div class="section" id="applications-api">
diff --git a/develop/doc_cn/api/v1/predict/swig_py_paddle_cn.html b/develop/doc_cn/api/v1/predict/swig_py_paddle_cn.html
index 6f8eb18264a8ff70713f4293a7bef29de3f5e027..7a4c49a03989df1a2730b53c1ba15eeb250ccc51 100644
--- a/develop/doc_cn/api/v1/predict/swig_py_paddle_cn.html
+++ b/develop/doc_cn/api/v1/predict/swig_py_paddle_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/activations.html b/develop/doc_cn/api/v1/trainer_config_helpers/activations.html
deleted file mode 100644
index 82d624172112a71ae23b58be5b66c16d0df4a314..0000000000000000000000000000000000000000
--- a/develop/doc_cn/api/v1/trainer_config_helpers/activations.html
+++ /dev/null
@@ -1,381 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Activations &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="../../../genindex.html"/>
-        <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Activations</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="activations">
-<h1>Activations<a class="headerlink" href="#activations" title="永久链接至标题">¶</a></h1>
-<div class="section" id="baseactivation">
-<h2>BaseActivation<a class="headerlink" href="#baseactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">BaseActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Base</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="absactivation">
-<h2>AbsActivation<a class="headerlink" href="#absactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">AbsActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Abs</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="expactivation">
-<h2>ExpActivation<a class="headerlink" href="#expactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">ExpActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Exp</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="identityactivation">
-<h2>IdentityActivation<a class="headerlink" href="#identityactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">IdentityActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Linear</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="linearactivation">
-<h2>LinearActivation<a class="headerlink" href="#linearactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">LinearActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Linear</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="logactivation">
-<h2>LogActivation<a class="headerlink" href="#logactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">LogActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Log</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="squareactivation">
-<h2>SquareActivation<a class="headerlink" href="#squareactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SquareActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Square</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="sigmoidactivation">
-<h2>SigmoidActivation<a class="headerlink" href="#sigmoidactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SigmoidActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Sigmoid</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="softmaxactivation">
-<h2>SoftmaxActivation<a class="headerlink" href="#softmaxactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SoftmaxActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Softmax</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="sequencesoftmaxactivation">
-<h2>SequenceSoftmaxActivation<a class="headerlink" href="#sequencesoftmaxactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SequenceSoftmaxActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">SequenceSoftmax</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="reluactivation">
-<h2>ReluActivation<a class="headerlink" href="#reluactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">ReluActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Relu</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="breluactivation">
-<h2>BReluActivation<a class="headerlink" href="#breluactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">BReluActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">BRelu</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="softreluactivation">
-<h2>SoftReluActivation<a class="headerlink" href="#softreluactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">SoftReluActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">SoftRelu</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="tanhactivation">
-<h2>TanhActivation<a class="headerlink" href="#tanhactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">TanhActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Tanh</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="stanhactivation">
-<h2>STanhActivation<a class="headerlink" href="#stanhactivation" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.activations.</code><code class="descname">STanhActivation</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">STanh</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="../../../_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/attrs.html b/develop/doc_cn/api/v1/trainer_config_helpers/attrs.html
deleted file mode 100644
index ae86b433d3c25b326eb6ff9d3ef0cec0491df3b2..0000000000000000000000000000000000000000
--- a/develop/doc_cn/api/v1/trainer_config_helpers/attrs.html
+++ /dev/null
@@ -1,336 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Parameter Attributes &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="../../../genindex.html"/>
-        <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Parameter Attributes</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="module-paddle.trainer_config_helpers.attrs">
-<span id="parameter-attributes"></span><h1>Parameter Attributes<a class="headerlink" href="#module-paddle.trainer_config_helpers.attrs" title="永久链接至标题">¶</a></h1>
-<dl class="class">
-<dt id="paddle.trainer_config_helpers.attrs.ParameterAttribute">
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ParameterAttribute</code><span class="sig-paren">(</span><em>name=None</em>, <em>is_static=False</em>, <em>initial_std=None</em>, <em>initial_mean=None</em>, <em>initial_max=None</em>, <em>initial_min=None</em>, <em>l1_rate=None</em>, <em>l2_rate=None</em>, <em>learning_rate=None</em>, <em>momentum=None</em>, <em>gradient_clipping_threshold=None</em>, <em>sparse_update=False</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="永久链接至目标">¶</a></dt>
-<dd><p>Parameter Attributes object. To fine-tuning network training process, user
-can set attribute to control training details, such as l1,l2 rate / learning
-rate / how to init param.</p>
-<p>NOTE: IT IS A HIGH LEVEL USER INTERFACE.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>is_static</strong> (<em>bool</em>) &#8211; True if this parameter will be fixed while training.</li>
-<li><strong>initial_std</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Gauss Random initialization standard deviation.
-None if not using Gauss Random initialize parameter.</li>
-<li><strong>initial_mean</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Gauss Random initialization mean.
-None if not using Gauss Random initialize parameter.</li>
-<li><strong>initial_max</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Uniform initialization max value.</li>
-<li><strong>initial_min</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; Uniform initialization min value.</li>
-<li><strong>l1_rate</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; the l1 regularization factor</li>
-<li><strong>l2_rate</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; the l2 regularization factor</li>
-<li><strong>learning_rate</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; The parameter learning rate. None means 1.
-The learning rate when optimize is LEARNING_RATE =
-GLOBAL_LEARNING_RATE * PARAMETER_LEARNING_RATE
-* SCHEDULER_FACTOR.</li>
-<li><strong>momentum</strong> (<em>float</em><em> or </em><em>None</em>) &#8211; The parameter momentum. None means use global value.</li>
-<li><strong>gradient_clipping_threshold</strong> (<em>float</em>) &#8211; gradient clipping threshold. If gradient
-value larger than some value, will be
-clipped.</li>
-<li><strong>sparse_update</strong> (<em>bool</em>) &#8211; Enable sparse update for this parameter. It will
-enable both local and remote sparse update.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-<dl class="method">
-<dt id="paddle.trainer_config_helpers.attrs.ParameterAttribute.set_default_parameter_name">
-<code class="descname">set_default_parameter_name</code><span class="sig-paren">(</span><em>name</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ParameterAttribute.set_default_parameter_name" title="永久链接至目标">¶</a></dt>
-<dd><p>Set default parameter name. If parameter not set, then will use default
-parameter name.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><strong>name</strong> (<em>basestring</em>) &#8211; default parameter name.</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</dd></dl>
-
-<dl class="class">
-<dt id="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute">
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ExtraLayerAttribute</code><span class="sig-paren">(</span><em>error_clipping_threshold=None</em>, <em>drop_rate=None</em>, <em>device=None</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="永久链接至目标">¶</a></dt>
-<dd><p>Some high level layer attributes config. You can set all attributes here,
-but some layer doesn&#8217;t support all attributes. If you set an attribute to a
-layer that not support this attribute, paddle will print an error and core.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>error_clipping_threshold</strong> (<em>float</em>) &#8211; Error clipping threshold.</li>
-<li><strong>drop_rate</strong> (<em>float</em>) &#8211; Dropout rate. Dropout will create a mask on layer output.
-The dropout rate is the zero rate of this mask. The
-details of what dropout is please refer to <a class="reference external" href="https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf">here</a>.</li>
-<li><strong>device</strong> (<em>int</em>) &#8211; <p>device ID of layer. device=-1, use CPU. device&gt;=0, use GPU.
-The details allocation in parallel_nn please refer to <a class="reference external" href="http://www.paddlepaddle.org/doc/ui/cmd_argument/use_case.html#case-2-specify-layers-in-different-devices">here</a>.</p>
-</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-<dl class="attribute">
-<dt id="paddle.trainer_config_helpers.attrs.ParamAttr">
-<code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ParamAttr</code><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ParamAttr" title="永久链接至目标">¶</a></dt>
-<dd><p><a class="reference internal" href="#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><code class="xref py py-class docutils literal"><span class="pre">ParameterAttribute</span></code></a> 的别名</p>
-</dd></dl>
-
-<dl class="attribute">
-<dt id="paddle.trainer_config_helpers.attrs.ExtraAttr">
-<code class="descclassname">paddle.trainer_config_helpers.attrs.</code><code class="descname">ExtraAttr</code><a class="headerlink" href="#paddle.trainer_config_helpers.attrs.ExtraAttr" title="永久链接至目标">¶</a></dt>
-<dd><p><a class="reference internal" href="#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><code class="xref py py-class docutils literal"><span class="pre">ExtraLayerAttribute</span></code></a> 的别名</p>
-</dd></dl>
-
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="../../../_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/data_sources.html b/develop/doc_cn/api/v1/trainer_config_helpers/data_sources.html
deleted file mode 100644
index 8ef79e2a18f0ebab2db68d999e5a6f0065e57b3f..0000000000000000000000000000000000000000
--- a/develop/doc_cn/api/v1/trainer_config_helpers/data_sources.html
+++ /dev/null
@@ -1,292 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>DataSources &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="../../../genindex.html"/>
-        <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>DataSources</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="module-paddle.trainer_config_helpers.data_sources">
-<span id="datasources"></span><span id="api-trainer-config-helpers-data-sources"></span><h1>DataSources<a class="headerlink" href="#module-paddle.trainer_config_helpers.data_sources" title="永久链接至标题">¶</a></h1>
-<p>Data Sources are helpers to define paddle training data or testing data.</p>
-<dl class="function">
-<dt id="paddle.trainer_config_helpers.data_sources.define_py_data_sources2">
-<code class="descclassname">paddle.trainer_config_helpers.data_sources.</code><code class="descname">define_py_data_sources2</code><span class="sig-paren">(</span><em>train_list</em>, <em>test_list</em>, <em>module</em>, <em>obj</em>, <em>args=None</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer_config_helpers.data_sources.define_py_data_sources2" title="永久链接至目标">¶</a></dt>
-<dd><p>Define python Train/Test data sources in one method. If train/test use
-the same Data Provider configuration, module/obj/args contain one argument,
-otherwise contain a list or tuple of arguments. For example:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">define_py_data_sources2</span><span class="p">(</span><span class="n">train_list</span><span class="o">=</span><span class="s2">&quot;train.list&quot;</span><span class="p">,</span>
-                        <span class="n">test_list</span><span class="o">=</span><span class="s2">&quot;test.list&quot;</span><span class="p">,</span>
-                        <span class="n">module</span><span class="o">=</span><span class="s2">&quot;data_provider&quot;</span>
-                        <span class="c1"># if train/test use different configurations,</span>
-                        <span class="c1"># obj=[&quot;process_train&quot;, &quot;process_test&quot;]</span>
-                        <span class="n">obj</span><span class="o">=</span><span class="s2">&quot;process&quot;</span><span class="p">,</span>
-                        <span class="n">args</span><span class="o">=</span><span class="p">{</span><span class="s2">&quot;dictionary&quot;</span><span class="p">:</span> <span class="n">dict_name</span><span class="p">})</span>
-</pre></div>
-</div>
-<p>The related data provider can refer to <span class="xref std std-ref">api_pydataprovider2_sequential_model</span> .</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>train_list</strong> (<em>basestring</em>) &#8211; Train list name.</li>
-<li><strong>test_list</strong> (<em>basestring</em>) &#8211; Test list name.</li>
-<li><strong>module</strong> (<em>basestring</em><em> or </em><em>tuple</em><em> or </em><em>list</em>) &#8211; python module name. If train and test is different, then
-pass a tuple or list to this argument.</li>
-<li><strong>obj</strong> (<em>basestring</em><em> or </em><em>tuple</em><em> or </em><em>list</em>) &#8211; python object name. May be a function name if using
-PyDataProviderWrapper. If train and test is different, then pass
-a tuple or list to this argument.</li>
-<li><strong>args</strong> (<em>string</em><em> or </em><em>picklable object</em><em> or </em><em>list</em><em> or </em><em>tuple.</em>) &#8211; The best practice is using dict() to pass arguments into
-DataProvider, and use <code class="code docutils literal"><span class="pre">&#64;init_hook_wrapper</span></code> to receive
-arguments. If train and test is different, then pass a tuple
-or list to this argument.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">None</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">None</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="../../../_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/layers.html b/develop/doc_cn/api/v1/trainer_config_helpers/layers.html
deleted file mode 100644
index 4061d2cc353c8eb6e557150da6e68e4c02f2ac2b..0000000000000000000000000000000000000000
--- a/develop/doc_cn/api/v1/trainer_config_helpers/layers.html
+++ /dev/null
@@ -1,3569 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Layers &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="../../../genindex.html"/>
-        <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Layers</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="layers">
-<span id="api-trainer-config-helpers-layers"></span><h1>Layers<a class="headerlink" href="#layers" title="永久链接至标题">¶</a></h1>
-<div class="section" id="base">
-<h2>Base<a class="headerlink" href="#base" title="永久链接至标题">¶</a></h2>
-<div class="section" id="layertype">
-<h3>LayerType<a class="headerlink" href="#layertype" title="永久链接至标题">¶</a></h3>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">LayerType</code></dt>
-<dd><p>Layer type enumerations.</p>
-<dl class="staticmethod">
-<dt>
-<em class="property">static </em><code class="descname">is_layer_type</code><span class="sig-paren">(</span><em>type_name</em><span class="sig-paren">)</span></dt>
-<dd><p>If type_name is a layer type.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><strong>type_name</strong> (<em>basestring</em>) &#8211; layer type name. Because layer type enumerations are
-strings.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body">True if is a layer_type</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body">bool</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</dd></dl>
-
-</div>
-<div class="section" id="layeroutput">
-<h3>LayerOutput<a class="headerlink" href="#layeroutput" title="永久链接至标题">¶</a></h3>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">LayerOutput</code><span class="sig-paren">(</span><em>name</em>, <em>layer_type</em>, <em>parents=None</em>, <em>activation=None</em>, <em>num_filters=None</em>, <em>img_norm_type=None</em>, <em>size=None</em>, <em>outputs=None</em>, <em>reverse=None</em><span class="sig-paren">)</span></dt>
-<dd><p>LayerOutput is output for layer function. It is used internally by several
-reasons.</p>
-<ul>
-<li><p class="first">Check layer connection make sense.</p>
-<blockquote>
-<div><ul class="simple">
-<li>FC(Softmax) =&gt; Cost(MSE Error) is not good for example.</li>
-</ul>
-</div></blockquote>
-</li>
-<li><p class="first">Tracking layer connection.</p>
-</li>
-<li><p class="first">Pass to layer methods as input.</p>
-</li>
-</ul>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer output name.</li>
-<li><strong>layer_type</strong> (<em>basestring</em>) &#8211; Current Layer Type. One of LayerType enumeration.</li>
-<li><strong>activation</strong> (<em>BaseActivation.</em>) &#8211; Layer Activation.</li>
-<li><strong>parents</strong> (<em>list|tuple|collections.Sequence</em>) &#8211; Layer&#8217;s parents.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-<dl class="method">
-<dt>
-<code class="descname">set_input</code><span class="sig-paren">(</span><em>input</em><span class="sig-paren">)</span></dt>
-<dd><p>Set the input for a memory layer. Can only be used for memory layer</p>
-</dd></dl>
-
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="data-layer">
-<h2>Data layer<a class="headerlink" href="#data-layer" title="永久链接至标题">¶</a></h2>
-<div class="section" id="api-trainer-config-helpers-layers-data-layer">
-<span id="id1"></span><h3>data_layer<a class="headerlink" href="#api-trainer-config-helpers-layers-data-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">data_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define DataLayer For NeuralNetwork.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">data</span> <span class="o">=</span> <span class="n">data_layer</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s2">&quot;input&quot;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">1000</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Name of this data layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; Size of this data layer.</li>
-<li><strong>height</strong> (<em>int|None</em>) &#8211; Height of this data layer, used for image</li>
-<li><strong>width</strong> (<em>int|None</em>) &#8211; Width of this data layer, used for image</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="fully-connected-layers">
-<h2>Fully Connected Layers<a class="headerlink" href="#fully-connected-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="fc-layer">
-<span id="api-trainer-config-helpers-layers-fc-layer"></span><h3>fc_layer<a class="headerlink" href="#fc-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">fc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Helper for declare fully connected layer.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">fc</span> <span class="o">=</span> <span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-              <span class="n">size</span><span class="o">=</span><span class="mi">1024</span><span class="p">,</span>
-              <span class="n">act</span><span class="o">=</span><span class="n">LinearActivation</span><span class="p">(),</span>
-              <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>which is equal to:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">1024</span><span class="p">)</span> <span class="k">as</span> <span class="n">fc</span><span class="p">:</span>
-    <span class="n">fc</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; The input layer. Could be a list/tuple of input layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The layer dimension.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Default is tanh.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; The Parameter Attribute|list.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|Any</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="selective-fc-layer">
-<h3>selective_fc_layer<a class="headerlink" href="#selective-fc-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">selective_fc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Selectived fully connected layer. Different from fc_layer, the output
-of this layer maybe sparse. It requires an additional input to indicate
-several selected columns for output. If the selected columns is not
-specified, selective_fc_layer acts exactly like fc_layer.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">sel_fc</span> <span class="o">=</span> <span class="n">selective_fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">128</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; The input layer.</li>
-<li><strong>select</strong> (<em>LayerOutput</em>) &#8211; The select layer. The output of select layer should be a
-sparse binary matrix, and treat as the mask of selective fc.
-If is None, acts exactly like fc_layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The layer dimension.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Default is tanh.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; The Parameter Attribute.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|Any</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="conv-layers">
-<h2>Conv Layers<a class="headerlink" href="#conv-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="conv-operator">
-<h3>conv_operator<a class="headerlink" href="#conv-operator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">conv_operator</code><span class="sig-paren">(</span><em>img</em>, <em>filter</em>, <em>filter_size</em>, <em>num_filters</em>, <em>num_channels=None</em>, <em>stride=1</em>, <em>padding=0</em>, <em>filter_size_y=None</em>, <em>stride_y=None</em>, <em>padding_y=None</em>, <em>trans=False</em><span class="sig-paren">)</span></dt>
-<dd><p>Different from img_conv_layer, conv_op is an Operator, which can be used
-in mixed_layer. And conv_op takes two inputs to perform convolution.
-The first input is the image and the second is filter kernel. It only
-support GPU mode.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">op</span> <span class="o">=</span> <span class="n">conv_operator</span><span class="p">(</span><span class="n">img</span><span class="o">=</span><span class="n">input1</span><span class="p">,</span>
-                   <span class="nb">filter</span><span class="o">=</span><span class="n">input2</span><span class="p">,</span>
-                   <span class="n">filter_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
-                   <span class="n">num_filters</span><span class="o">=</span><span class="mi">64</span><span class="p">,</span>
-                   <span class="n">num_channels</span><span class="o">=</span><span class="mi">64</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>img</strong> (<em>LayerOutput</em>) &#8211; input image</li>
-<li><strong>filter</strong> (<em>LayerOutput</em>) &#8211; input filter</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; The x dimension of a filter kernel.</li>
-<li><strong>filter_size_y</strong> (<em>int</em>) &#8211; The y dimension of a filter kernel. Since
-PaddlePaddle now supports rectangular filters,
-the filter&#8217;s shape can be (filter_size, filter_size_y).</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; channel of output data.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; channel of input data.</li>
-<li><strong>stride</strong> (<em>int</em>) &#8211; The x dimension of the stride.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The y dimension of the stride.</li>
-<li><strong>padding</strong> (<em>int</em>) &#8211; The x dimension of padding.</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The y dimension of padding.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A ConvOperator Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">ConvOperator</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="conv-projection">
-<h3>conv_projection<a class="headerlink" href="#conv-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">conv_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Different from img_conv_layer and conv_op, conv_projection is an Projection,
-which can be used in mixed_layer and conat_layer. It use cudnn to implement
-conv and only support GPU mode.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">conv_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input1</span><span class="p">,</span>
-                       <span class="n">filter_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
-                       <span class="n">num_filters</span><span class="o">=</span><span class="mi">64</span><span class="p">,</span>
-                       <span class="n">num_channels</span><span class="o">=</span><span class="mi">64</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; The x dimension of a filter kernel.</li>
-<li><strong>filter_size_y</strong> (<em>int</em>) &#8211; The y dimension of a filter kernel. Since
-PaddlePaddle now supports rectangular filters,
-the filter&#8217;s shape can be (filter_size, filter_size_y).</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; channel of output data.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; channel of input data.</li>
-<li><strong>stride</strong> (<em>int</em>) &#8211; The x dimension of the stride.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The y dimension of the stride.</li>
-<li><strong>padding</strong> (<em>int</em>) &#8211; The x dimension of padding.</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The y dimension of padding.</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; The group number.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Convolution param attribute. None means default attribute</li>
-<li><strong>trans</strong> (<em>boolean</em>) &#8211; whether it is convTrans or conv</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A DotMulProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">DotMulProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="conv-shift-layer">
-<h3>conv_shift_layer<a class="headerlink" href="#conv-shift-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">conv_shift_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>This layer performs cyclic convolution for two input. For example:</dt>
-<dd><ul class="first last simple">
-<li>a[in]: contains M elements.</li>
-<li>b[in]: contains N elements (N should be odd).</li>
-<li>c[out]: contains M elements.</li>
-</ul>
-</dd>
-</dl>
-<div class="math">
-\[c[i] = \sum_{j=-(N-1)/2}^{(N-1)/2}a_{i+j} * b_{j}\]</div>
-<dl class="docutils">
-<dt>In this formular:</dt>
-<dd><ul class="first last simple">
-<li>a&#8217;s index is computed modulo M. When it is negative, then get item from
-the right side (which is the end of array) to the left.</li>
-<li>b&#8217;s index is computed modulo N. When it is negative, then get item from
-the right size (which is the end of array) to the left.</li>
-</ul>
-</dd>
-</dl>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">conv_shift</span> <span class="o">=</span> <span class="n">conv_shift_layer</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; Input layer a.</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input layer b.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="img-conv-layer">
-<h3>img_conv_layer<a class="headerlink" href="#img-conv-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">img_conv_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Convolution layer for image. Paddle can support both square and non-square
-input currently.</p>
-<p>The details of convolution layer, please refer UFLDL&#8217;s <a class="reference external" href="http://ufldl.stanford.edu/tutorial/supervised/FeatureExtractionUsingConvolution/">convolution</a> .</p>
-<p>Convolution Transpose (deconv) layer for image. Paddle can support both square
-and non-square input currently.</p>
-<p>The details of convolution transpose layer,
-please refer to the following explanation and references therein
-&lt;<a class="reference external" href="http://datascience.stackexchange.com/questions/6107/">http://datascience.stackexchange.com/questions/6107/</a>
-what-are-deconvolutional-layers/&gt;`_ .
-The num_channel means input image&#8217;s channel number. It may be 1 or 3 when
-input is raw pixels of image(mono or RGB), or it may be the previous layer&#8217;s
-num_filters * num_group.</p>
-<p>There are several group of filter in PaddlePaddle implementation.
-Each group will process some channel of the inputs. For example, if an input
-num_channel = 256, group = 4, num_filter=32, the PaddlePaddle will create
-32*4 = 128 filters to process inputs. The channels will be split into 4
-pieces. First 256/4 = 64 channels will process by first 32 filters. The
-rest channels will be processed by rest group of filters.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">conv</span> <span class="o">=</span> <span class="n">img_conv_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">data</span><span class="p">,</span> <span class="n">filter_size</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">filter_size_y</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                      <span class="n">num_channels</span><span class="o">=</span><span class="mi">8</span><span class="p">,</span>
-                      <span class="n">num_filters</span><span class="o">=</span><span class="mi">16</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                      <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">,</span>
-                      <span class="n">act</span><span class="o">=</span><span class="n">ReluActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Layer Input.</li>
-<li><strong>filter_size</strong> (<em>int|tuple|list</em>) &#8211; The x dimension of a filter kernel. Or input a tuple for
-two image dimension.</li>
-<li><strong>filter_size_y</strong> (<em>int|None</em>) &#8211; The y dimension of a filter kernel. Since PaddlePaddle
-currently supports rectangular filters, the filter&#8217;s
-shape will be (filter_size, filter_size_y).</li>
-<li><strong>num_filters</strong> &#8211; Each filter group&#8217;s number of filter</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type. Default is tanh</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; Group size of filters.</li>
-<li><strong>stride</strong> (<em>int|tuple|list</em>) &#8211; The x dimension of the stride. Or input a tuple for two image
-dimension.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The y dimension of the stride.</li>
-<li><strong>padding</strong> (<em>int|tuple|list</em>) &#8211; The x dimension of the padding. Or input a tuple for two
-image dimension</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The y dimension of the padding.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Convolution bias attribute. None means default bias.
-False means no bias.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; number of input channels. If None will be set
-automatically from previous output.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Convolution param attribute. None means default attribute</li>
-<li><strong>shared_biases</strong> (<em>bool</em>) &#8211; Is biases will be shared between filters or not.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Layer Extra Attribute.</li>
-<li><strong>trans</strong> (<em>bool</em>) &#8211; true if it is a convTransLayer, false if it is a convLayer</li>
-<li><strong>layer_type</strong> (<em>String</em>) &#8211; specify the layer_type, default is None. If trans=True,
-layer_type has to be &#8220;exconvt&#8221; or &#8220;cudnn_convt&#8221;,
-otherwise layer_type has to be either &#8220;exconv&#8221; or
-&#8220;cudnn_conv&#8221;</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="context-projection">
-<span id="api-trainer-config-helpers-layers-context-projection"></span><h3>context_projection<a class="headerlink" href="#context-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">context_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Context Projection.</p>
-<p>It just simply reorganizes input sequence, combines &#8220;context_len&#8221; sequence
-to one context from context_start. &#8220;context_start&#8221; will be set to
--(context_len - 1) / 2 by default. If context position out of sequence
-length, padding will be filled as zero if padding_attr = False, otherwise
-it is trainable.</p>
-<p>For example, origin sequence is [A B C D E F G], context len is 3, then
-after context projection and not set padding_attr, sequence will
-be [ 0AB ABC BCD CDE DEF EFG FG0 ].</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Sequence.</li>
-<li><strong>context_len</strong> (<em>int</em>) &#8211; context length.</li>
-<li><strong>context_start</strong> (<em>int</em>) &#8211; context start position. Default is
--(context_len - 1)/2</li>
-<li><strong>padding_attr</strong> (<em>bool|ParameterAttribute</em>) &#8211; Padding Parameter Attribute. If false, it means padding
-always be zero. Otherwise Padding is learnable, and
-parameter attribute is set by this parameter.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">Projection</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">Projection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="image-pooling-layer">
-<h2>Image Pooling Layer<a class="headerlink" href="#image-pooling-layer" title="永久链接至标题">¶</a></h2>
-<div class="section" id="img-pool-layer">
-<h3>img_pool_layer<a class="headerlink" href="#img-pool-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">img_pool_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Image pooling Layer.</p>
-<p>The details of pooling layer, please refer ufldl&#8217;s <a class="reference external" href="http://ufldl.stanford.edu/tutorial/supervised/Pooling/">pooling</a> .</p>
-<ul class="simple">
-<li>ceil_mode=True:</li>
-</ul>
-<div class="math">
-\[w = 1 + int(ceil(input\_width + 2 * padding - pool\_size) / float(stride))
-h = 1 + int(ceil(input\_height + 2 * padding\_y - pool\_size\_y) / float(stride\_y))\]</div>
-<ul class="simple">
-<li>ceil_mode=False:</li>
-</ul>
-<div class="math">
-\[w = 1 + int(floor(input\_width + 2 * padding - pool\_size) / float(stride))
-h = 1 + int(floor(input\_height + 2 * padding\_y - pool\_size\_y) / float(stride\_y))\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">maxpool</span> <span class="o">=</span> <span class="n">img_pool_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">conv</span><span class="p">,</span>
-                         <span class="n">pool_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
-                         <span class="n">pool_size_y</span><span class="o">=</span><span class="mi">5</span><span class="p">,</span>
-                         <span class="n">num_channels</span><span class="o">=</span><span class="mi">8</span><span class="p">,</span>
-                         <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                         <span class="n">stride_y</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span>
-                         <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                         <span class="n">padding_y</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span>
-                         <span class="n">pool_type</span><span class="o">=</span><span class="n">MaxPooling</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>padding</strong> (<em>int</em>) &#8211; pooling padding width.</li>
-<li><strong>padding_y</strong> (<em>int|None</em>) &#8211; pooling padding height. It&#8217;s equal to padding by default.</li>
-<li><strong>name</strong> (<em>basestring.</em>) &#8211; name of pooling layer</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input</li>
-<li><strong>pool_size</strong> (<em>int</em>) &#8211; pooling window width</li>
-<li><strong>pool_size_y</strong> (<em>int|None</em>) &#8211; pooling window height. It&#8217;s eaqual to pool_size by default.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; number of input channel.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType</em>) &#8211; pooling type. MaxPooling or AvgPooling. Default is
-MaxPooling.</li>
-<li><strong>stride</strong> (<em>int</em>) &#8211; stride width of pooling.</li>
-<li><strong>stride_y</strong> (<em>int|None</em>) &#8211; stride height of pooling. It is equal to stride by default.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-<li><strong>ceil_mode</strong> (<em>bool</em>) &#8211; Wether to use ceil mode to calculate output height and with.
-Defalut is True. If set false, Otherwise use floor.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="spp-layer">
-<h3>spp_layer<a class="headerlink" href="#spp-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">spp_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition.
-The details please refer to
-<a class="reference external" href="https://arxiv.org/abs/1406.4729">Kaiming He&#8217;s paper</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">spp</span> <span class="o">=</span> <span class="n">spp_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">data</span><span class="p">,</span>
-                <span class="n">pyramid_height</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span>
-                <span class="n">num_channels</span><span class="o">=</span><span class="mi">16</span><span class="p">,</span>
-                <span class="n">pool_type</span><span class="o">=</span><span class="n">MaxPooling</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; number of input channel.</li>
-<li><strong>pool_type</strong> &#8211; Pooling type. MaxPooling or AveragePooling. Default is MaxPooling.</li>
-<li><strong>pyramid_height</strong> (<em>int</em>) &#8211; pyramid height.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="maxout-layer">
-<h3>maxout_layer<a class="headerlink" href="#maxout-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">maxout_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>A layer to do max out on conv layer output.</dt>
-<dd><ul class="first last simple">
-<li>Input: output of a conv layer.</li>
-<li>Output: feature map size same as input. Channel is (input channel) / groups.</li>
-</ul>
-</dd>
-</dl>
-<p>So groups should be larger than 1, and the num of channels should be able
-to devided by groups.</p>
-<dl class="docutils">
-<dt>Please refer to Paper:</dt>
-<dd><ul class="first last simple">
-<li>Maxout Networks: <a class="reference external" href="http://www.jmlr.org/proceedings/papers/v28/goodfellow13.pdf">http://www.jmlr.org/proceedings/papers/v28/goodfellow13.pdf</a></li>
-<li>Multi-digit Number Recognition from Street View         Imagery using Deep Convolutional Neural Networks:         <a class="reference external" href="https://arxiv.org/pdf/1312.6082v4.pdf">https://arxiv.org/pdf/1312.6082v4.pdf</a></li>
-</ul>
-</dd>
-</dl>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">maxout</span> <span class="o">=</span> <span class="n">maxout_layer</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span>
-                      <span class="n">num_channels</span><span class="o">=</span><span class="mi">128</span><span class="p">,</span>
-                      <span class="n">groups</span><span class="o">=</span><span class="mi">4</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>num_channels</strong> (<em>int|None</em>) &#8211; The channel number of input layer. If None will be set
-automatically from previous output.</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; The group number of input layer.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layer, which can not specify.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="norm-layer">
-<h2>Norm Layer<a class="headerlink" href="#norm-layer" title="永久链接至标题">¶</a></h2>
-<div class="section" id="img-cmrnorm-layer">
-<h3>img_cmrnorm_layer<a class="headerlink" href="#img-cmrnorm-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">img_cmrnorm_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Response normalization across feature maps.
-The details please refer to
-<a class="reference external" href="http://www.cs.toronto.edu/~fritz/absps/imagenet.pdf">Alex&#8217;s paper</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">norm</span> <span class="o">=</span> <span class="n">img_cmrnorm_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">net</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">5</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; Normalize in number of <span class="math">\(size\)</span> feature maps.</li>
-<li><strong>scale</strong> (<em>float</em>) &#8211; The hyper-parameter.</li>
-<li><strong>power</strong> (<em>float</em>) &#8211; The hyper-parameter.</li>
-<li><strong>num_channels</strong> &#8211; input layer&#8217;s filers number or channels. If
-num_channels is None, it will be set automatically.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="batch-norm-layer">
-<h3>batch_norm_layer<a class="headerlink" href="#batch-norm-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">batch_norm_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Batch Normalization Layer. The notation of this layer as follow.</p>
-<p><span class="math">\(x\)</span> is the input features over a mini-batch.</p>
-<div class="math">
-\[\begin{split}\mu_{\beta} &amp;\gets \frac{1}{m} \sum_{i=1}^{m} x_i \qquad &amp;//\
-\ mini-batch\ mean \\
-\sigma_{\beta}^{2} &amp;\gets \frac{1}{m} \sum_{i=1}^{m}(x_i - \
-\mu_{\beta})^2 \qquad &amp;//\ mini-batch\ variance \\
-\hat{x_i} &amp;\gets \frac{x_i - \mu_\beta} {\sqrt{\
-\sigma_{\beta}^{2} + \epsilon}} \qquad &amp;//\ normalize \\
-y_i &amp;\gets \gamma \hat{x_i} + \beta \qquad &amp;//\ scale\ and\ shift\end{split}\]</div>
-<p>The details of batch normalization please refer to this
-<a class="reference external" href="http://arxiv.org/abs/1502.03167">paper</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">norm</span> <span class="o">=</span> <span class="n">batch_norm_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">net</span><span class="p">,</span> <span class="n">act</span><span class="o">=</span><span class="n">ReluActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; batch normalization input. Better be linear activation.
-Because there is an activation inside batch_normalization.</li>
-<li><strong>batch_norm_type</strong> (<em>None|string</em><em>, </em><em>None</em><em> or </em><em>&quot;batch_norm&quot;</em><em> or </em><em>&quot;cudnn_batch_norm&quot;</em>) &#8211; We have batch_norm and cudnn_batch_norm. batch_norm
-supports both CPU and GPU. cudnn_batch_norm requires
-cuDNN version greater or equal to v4 (&gt;=v4). But
-cudnn_batch_norm is faster and needs less memory
-than batch_norm. By default (None), we will
-automaticly select cudnn_batch_norm for GPU and
-batch_norm for CPU. Otherwise, select batch norm
-type based on the specified type. If you use cudnn_batch_norm,
-we suggested you use latest version, such as v5.1.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Better be relu. Because batch
-normalization will normalize input near zero.</li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; num of image channels or previous layer&#8217;s number of
-filters. None will automatically get from layer&#8217;s
-input.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; <span class="math">\(\beta\)</span>, better be zero when initialize. So the
-initial_std=0, initial_mean=1 is best practice.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; <span class="math">\(\gamma\)</span>, better be one when initialize. So the
-initial_std=0, initial_mean=1 is best practice.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-<li><strong>use_global_stats</strong> (<em>bool|None.</em>) &#8211; whether use moving mean/variance statistics
-during testing peroid. If None or True,
-it will use moving mean/variance statistics during
-testing. If False, it will use the mean
-and variance of current batch of test data for
-testing.</li>
-<li><strong>moving_average_fraction</strong> (<em>float.</em>) &#8211; Factor used in the moving average
-computation, referred to as facotr,
-<span class="math">\(runningMean = newMean*(1-factor)
-+ runningMean*factor\)</span></li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="sum-to-one-norm-layer">
-<h3>sum_to_one_norm_layer<a class="headerlink" href="#sum-to-one-norm-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">sum_to_one_norm_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for sum-to-one normalization,
-which is used in NEURAL TURING MACHINE.</p>
-<div class="math">
-\[out[i] = \frac {in[i]} {\sum_{k=1}^N in[k]}\]</div>
-<p>where <span class="math">\(in\)</span> is a (batchSize x dataDim) input vector,
-and <span class="math">\(out\)</span> is a (batchSize x dataDim) output vector.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">sum_to_one_norm</span> <span class="o">=</span> <span class="n">sum_to_one_norm_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="recurrent-layers">
-<h2>Recurrent Layers<a class="headerlink" href="#recurrent-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="recurrent-layer">
-<h3>recurrent_layer<a class="headerlink" href="#recurrent-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">recurrent_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Simple recurrent unit layer. It is just a fully connect layer through both
-time and neural network.</p>
-<p>For each sequence [start, end] it performs the following computation:</p>
-<div class="math">
-\[\begin{split}out_{i} = act(in_{i})     \      \      \text{for} \ i = start \\
-out_{i} = act(in_{i} + out_{i-1} * W) \ \ \text{for} \ start &lt; i &lt;= end\end{split}\]</div>
-<p>If reversed is true, the order is reversed:</p>
-<div class="math">
-\[\begin{split}out_{i} = act(in_{i})           \    \   \text{for} \ i = end  \\
-out_{i} = act(in_{i} + out_{i+1} * W) \ \ \text{for} \ start &lt;= i &lt; end\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; activation.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; bias attribute.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; parameter attribute.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the layer</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lstmemory">
-<h3>lstmemory<a class="headerlink" href="#lstmemory" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">lstmemory</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Long Short-term Memory Cell.</p>
-<p>The memory cell was implemented as follow equations.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>NOTE: In PaddlePaddle&#8217;s implementation, the multiplications
-<span class="math">\(W_{xi}x_{t}\)</span> , <span class="math">\(W_{xf}x_{t}\)</span>,
-<span class="math">\(W_{xc}x_t\)</span>, <span class="math">\(W_{xo}x_{t}\)</span> are not done in the lstmemory layer,
-so an additional mixed_layer with full_matrix_projection or a fc_layer must
-be included in the configuration file to complete the input-to-hidden
-mappings before lstmemory is called.</p>
-<p>NOTE: This is a low level user interface. You can use network.simple_lstm
-to config a simple plain lstm layer.</p>
-<p>Please refer to <strong>Generating Sequences With Recurrent Neural Networks</strong> for
-more details about LSTM.</p>
-<p><a class="reference external" href="http://arxiv.org/abs/1308.0850">Link</a> goes as below.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The lstmemory layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; is sequence process reversed or not.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; activation type, TanhActivation by default. <span class="math">\(h_t\)</span></li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; gate activation type, SigmoidActivation by default.</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; state activation type, TanhActivation by default.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias attribute. None means default bias. False means no
-bias.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Parameter Attribute.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer attribute</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="grumemory">
-<h3>grumemory<a class="headerlink" href="#grumemory" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">grumemory</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Gate Recurrent Unit Layer.</p>
-<p>The memory cell was implemented as follow equations.</p>
-<p>1. update gate <span class="math">\(z\)</span>: defines how much of the previous memory to
-keep around or the unit updates its activations. The update gate
-is computed by:</p>
-<div class="math">
-\[z_t = \sigma(W_{z}x_{t} + U_{z}h_{t-1} + b_z)\]</div>
-<p>2. reset gate <span class="math">\(r\)</span>: determines how to combine the new input with the
-previous memory. The reset gate is computed similarly to the update gate:</p>
-<div class="math">
-\[r_t = \sigma(W_{r}x_{t} + U_{r}h_{t-1} + b_r)\]</div>
-<p>3. The candidate activation <span class="math">\(\tilde{h_t}\)</span> is computed similarly to
-that of the traditional recurrent unit:</p>
-<div class="math">
-\[{\tilde{h_t}} = tanh(W x_{t} + U (r_{t} \odot h_{t-1}) + b)\]</div>
-<p>4. The hidden activation <span class="math">\(h_t\)</span> of the GRU at time t is a linear
-interpolation between the previous activation <span class="math">\(h_{t-1}\)</span> and the
-candidate activation <span class="math">\(\tilde{h_t}\)</span>:</p>
-<div class="math">
-\[h_t = (1 - z_t) h_{t-1} + z_t {\tilde{h_t}}\]</div>
-<p>NOTE: In PaddlePaddle&#8217;s implementation, the multiplication operations
-<span class="math">\(W_{r}x_{t}\)</span>, <span class="math">\(W_{z}x_{t}\)</span> and <span class="math">\(W x_t\)</span> are not computed in
-gate_recurrent layer. Consequently, an additional mixed_layer with
-full_matrix_projection or a fc_layer must be included before grumemory
-is called.</p>
-<p>More details can be found by referring to <a class="reference external" href="https://arxiv.org/abs/1412.3555">Empirical Evaluation of Gated
-Recurrent Neural Networks on Sequence Modeling.</a></p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">grumemory</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The gru layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; input layer.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; Whether sequence process is reversed or not.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; activation type, TanhActivation by default. This activation
-affects the <span class="math">\({\tilde{h_t}}\)</span>.</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; gate activation type, SigmoidActivation by default.
-This activation affects the <span class="math">\(z_t\)</span> and <span class="math">\(r_t\)</span>. It is the
-<span class="math">\(\sigma\)</span> in the above formula.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias attribute. None means default bias. False means no
-bias.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Parameter Attribute.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer attribute</li>
-<li><strong>size</strong> (<em>None</em>) &#8211; Stub parameter of size, but actually not used. If set this size
-will get a warning.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="recurrent-layer-group">
-<h2>Recurrent Layer Group<a class="headerlink" href="#recurrent-layer-group" title="永久链接至标题">¶</a></h2>
-<div class="section" id="memory">
-<h3>memory<a class="headerlink" href="#memory" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">memory</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>The memory layers is a layer cross each time step. Reference this output
-as previous time step layer <code class="code docutils literal"><span class="pre">name</span></code> &#8216;s output.</p>
-<p>The default memory is zero in first time step, previous time step&#8217;s
-output in the rest time steps.</p>
-<p>If boot_bias, the first time step value is this bias and
-with activation.</p>
-<p>If boot_with_const_id, then the first time stop is a IndexSlot, the
-Arguments.ids()[0] is this <code class="code docutils literal"><span class="pre">cost_id</span></code>.</p>
-<p>If boot_layer is not null, the memory is just the boot_layer&#8217;s output.
-Set <code class="code docutils literal"><span class="pre">is_seq</span></code> is true boot layer is sequence.</p>
-<p>The same name layer in recurrent group will set memory on each time
-step.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">mem</span> <span class="o">=</span> <span class="n">memory</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;state&#39;</span><span class="p">)</span>
-<span class="n">state</span> <span class="o">=</span> <span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">mem</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;state&#39;</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>If you do not want to specify the name, you can equivalently use set_input()
-to specify the layer needs to be remembered as the following:</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; the name of the layer which this memory remembers.
-If name is None, user should call set_input() to specify the
-name of the layer which this memory remembers.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; size of memory.</li>
-<li><strong>memory_name</strong> (<em>basestring</em>) &#8211; the name of the memory.
-It is ignored when name is provided.</li>
-<li><strong>is_seq</strong> (<em>bool</em>) &#8211; is sequence for boot_layer</li>
-<li><strong>boot_layer</strong> (<em>LayerOutput|None</em>) &#8211; boot layer of memory.</li>
-<li><strong>boot_bias</strong> (<em>ParameterAttribute|None</em>) &#8211; boot layer&#8217;s bias</li>
-<li><strong>boot_bias_active_type</strong> (<em>BaseActivation</em>) &#8211; boot layer&#8217;s active type.</li>
-<li><strong>boot_with_const_id</strong> (<em>int</em>) &#8211; boot layer&#8217;s id.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object which is a memory.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="recurrent-group">
-<h3>recurrent_group<a class="headerlink" href="#recurrent-group" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">recurrent_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Recurrent layer group is an extremely flexible recurrent unit in
-PaddlePaddle. As long as the user defines the calculation done within a
-time step, PaddlePaddle will iterate such a recurrent calculation over
-sequence input. This is extremely usefull for attention based model, or
-Neural Turning Machine like models.</p>
-<p>The basic usage (time steps) is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">step</span><span class="p">(</span><span class="nb">input</span><span class="p">):</span>
-    <span class="n">output</span> <span class="o">=</span> <span class="n">fc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                      <span class="n">size</span><span class="o">=</span><span class="mi">1024</span><span class="p">,</span>
-                      <span class="n">act</span><span class="o">=</span><span class="n">LinearActivation</span><span class="p">(),</span>
-                      <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-    <span class="k">return</span> <span class="n">output</span>
-
-<span class="n">group</span> <span class="o">=</span> <span class="n">recurrent_group</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                        <span class="n">step</span><span class="o">=</span><span class="n">step</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>You can see following configs for further usages:</p>
-<ul class="simple">
-<li>time steps: lstmemory_group, paddle/gserver/tests/sequence_layer_group.conf,                   demo/seqToseq/seqToseq_net.py</li>
-<li>sequence steps: paddle/gserver/tests/sequence_nest_layer_group.conf</li>
-</ul>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>step</strong> (<em>callable</em>) &#8211; <p>recurrent one time step function.The input of this function is
-input of the group. The return of this function will be
-recurrent group&#8217;s return value.</p>
-<p>The recurrent group scatter a sequence into time steps. And
-for each time step, will invoke step function, and return
-a time step result. Then gather each time step of output into
-layer group&#8217;s output.</p>
-</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; recurrent_group&#8217;s name.</li>
-<li><strong>input</strong> (<em>LayerOutput|StaticInput|SubsequenceInput|list|tuple</em>) &#8211; <p>Input links array.</p>
-<p>LayerOutput will be scattered into time steps.
-SubsequenceInput will be scattered into sequence steps.
-StaticInput will be imported to each time step, and doesn&#8217;t change
-through time. It&#8217;s a mechanism to access layer outside step function.</p>
-</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; If reverse is set true, the recurrent unit will process the
-input sequence in a reverse order.</li>
-<li><strong>targetInlink</strong> (<em>LayerOutput|SubsequenceInput</em>) &#8211; <p>the input layer which share info with layer group&#8217;s output</p>
-<p>Param input specifies multiple input layers. For
-SubsequenceInput inputs, config should assign one input
-layer that share info(the number of sentences and the number
-of words in each sentence) with all layer group&#8217;s outputs.
-targetInlink should be one of the layer group&#8217;s input.</p>
-</li>
-<li><strong>is_generating</strong> &#8211; If is generating, none of input type should be LayerOutput;
-else, for training or testing, one of the input type must
-be LayerOutput.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-<p>: type is_generating: bool</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">返回:</th><td class="field-body">LayerOutput object.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回类型:</th><td class="field-body">LayerOutput</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lstm-step-layer">
-<h3>lstm_step_layer<a class="headerlink" href="#lstm-step-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">lstm_step_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>LSTM Step Layer. It used in recurrent_group. The lstm equations are shown
-as follow.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>The input of lstm step is <span class="math">\(Wx_t + Wh_{t-1}\)</span>, and user should use
-<code class="code docutils literal"><span class="pre">mixed_layer</span></code> and <code class="code docutils literal"><span class="pre">full_matrix_projection</span></code> to calculate these
-input vector.</p>
-<p>The state of lstm step is <span class="math">\(c_{t-1}\)</span>. And lstm step layer will do</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t = \sigma(input + W_{ci}c_{t-1} + b_i)\\...\end{aligned}\end{align} \]</div>
-<p>This layer contains two outputs. Default output is <span class="math">\(h_t\)</span>. The other
-output is <span class="math">\(o_t\)</span>, which name is &#8216;state&#8217; and can use
-<code class="code docutils literal"><span class="pre">get_output_layer</span></code> to extract this output.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer&#8217;s name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; Layer&#8217;s size. NOTE: lstm layer&#8217;s size, should be equal as
-<code class="code docutils literal"><span class="pre">input.size/4</span></code>, and should be equal as
-<code class="code docutils literal"><span class="pre">state.size</span></code>.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer. <span class="math">\(Wx_t + Wh_{t-1}\)</span></li>
-<li><strong>state</strong> (<em>LayerOutput</em>) &#8211; State Layer. <span class="math">\(c_{t-1}\)</span></li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type. Default is tanh</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; Gate Activation Type. Default is sigmoid, and should
-be sigmoid only.</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; State Activation Type. Default is sigmoid, and should
-be sigmoid only.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Bias Attribute.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="gru-step-layer">
-<h3>gru_step_layer<a class="headerlink" href="#gru-step-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">gru_step_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; </li>
-<li><strong>output_mem</strong> &#8211; </li>
-<li><strong>size</strong> &#8211; </li>
-<li><strong>act</strong> &#8211; </li>
-<li><strong>name</strong> &#8211; </li>
-<li><strong>gate_act</strong> &#8211; </li>
-<li><strong>bias_attr</strong> &#8211; </li>
-<li><strong>param_attr</strong> &#8211; the parameter_attribute for transforming the output_mem
-from previous step.</li>
-<li><strong>layer_attr</strong> &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="beam-search">
-<h3>beam_search<a class="headerlink" href="#beam-search" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">beam_search</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Beam search is a heuristic search algorithm used in sequence generation.
-It explores a graph by expanding the most promising nodes in a limited set
-to maintain tractability.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">rnn_step</span><span class="p">(</span><span class="nb">input</span><span class="p">):</span>
-    <span class="n">last_time_step_output</span> <span class="o">=</span> <span class="n">memory</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;rnn&#39;</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">)</span>
-    <span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">,</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;rnn&#39;</span><span class="p">)</span> <span class="k">as</span> <span class="n">simple_rnn</span><span class="p">:</span>
-        <span class="n">simple_rnn</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
-        <span class="n">simple_rnn</span> <span class="o">+=</span> <span class="n">last_time_step_output</span>
-    <span class="k">return</span> <span class="n">simple_rnn</span>
-
-<span class="n">generated_word_embedding</span> <span class="o">=</span> <span class="n">GeneratedInput</span><span class="p">(</span>
-                       <span class="n">size</span><span class="o">=</span><span class="n">target_dictionary_dim</span><span class="p">,</span>
-                       <span class="n">embedding_name</span><span class="o">=</span><span class="s2">&quot;target_language_embedding&quot;</span><span class="p">,</span>
-                       <span class="n">embedding_size</span><span class="o">=</span><span class="n">word_vector_dim</span><span class="p">)</span>
-
-<span class="n">beam_gen</span> <span class="o">=</span> <span class="n">beam_search</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s2">&quot;decoder&quot;</span><span class="p">,</span>
-                       <span class="n">step</span><span class="o">=</span><span class="n">rnn_step</span><span class="p">,</span>
-                       <span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">StaticInput</span><span class="p">(</span><span class="n">encoder_last</span><span class="p">),</span>
-                              <span class="n">generated_word_embedding</span><span class="p">],</span>
-                       <span class="n">bos_id</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span>
-                       <span class="n">eos_id</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                       <span class="n">beam_size</span><span class="o">=</span><span class="mi">5</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>Please see the following demo for more details:</p>
-<ul class="simple">
-<li>machine translation : demo/seqToseq/translation/gen.conf                             demo/seqToseq/seqToseq_net.py</li>
-</ul>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>base string</em>) &#8211; Name of the recurrent unit that generates sequences.</li>
-<li><strong>step</strong> (<em>callable</em>) &#8211; <p>A callable function that defines the calculation in a time
-step, and it is applied to sequences with arbitrary length by
-sharing a same set of weights.</p>
-<p>You can refer to the first parameter of recurrent_group, or
-demo/seqToseq/seqToseq_net.py for more details.</p>
-</li>
-<li><strong>input</strong> (<em>list</em>) &#8211; Input data for the recurrent unit, which should include the
-previously generated words as a GeneratedInput object.</li>
-<li><strong>bos_id</strong> (<em>int</em>) &#8211; Index of the start symbol in the dictionary. The start symbol
-is a special token for NLP task, which indicates the
-beginning of a sequence. In the generation task, the start
-symbol is essential, since it is used to initialize the RNN
-internal state.</li>
-<li><strong>eos_id</strong> (<em>int</em>) &#8211; Index of the end symbol in the dictionary. The end symbol is
-a special token for NLP task, which indicates the end of a
-sequence. The generation process will stop once the end
-symbol is generated, or a pre-defined max iteration number
-is exceeded.</li>
-<li><strong>max_length</strong> (<em>int</em>) &#8211; Max generated sequence length.</li>
-<li><strong>beam_size</strong> (<em>int</em>) &#8211; Beam search for sequence generation is an iterative search
-algorithm. To maintain tractability, every iteration only
-only stores a predetermined number, called the beam_size,
-of the most promising next words. The greater the beam
-size, the fewer candidate words are pruned.</li>
-<li><strong>num_results_per_sample</strong> (<em>int</em>) &#8211; Number of the generated results per input
-sequence. This number must always be less than
-beam size.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">The generated word index.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="get-output-layer">
-<h3>get_output_layer<a class="headerlink" href="#get-output-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">get_output_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Get layer&#8217;s output by name. In PaddlePaddle, a layer might return multiple
-values, but returns one layer&#8217;s output. If the user wants to use another
-output besides the default one, please use get_output_layer first to get
-the output from input.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer&#8217;s name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; get output layer&#8217;s input. And this layer should contains
-multiple outputs.</li>
-<li><strong>arg_name</strong> (<em>basestring</em>) &#8211; Output name from input.</li>
-<li><strong>layer_attr</strong> &#8211; Layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="mixed-layer">
-<h2>Mixed Layer<a class="headerlink" href="#mixed-layer" title="永久链接至标题">¶</a></h2>
-<div class="section" id="api-trainer-config-helpers-layers-mixed-layer">
-<span id="id2"></span><h3>mixed_layer<a class="headerlink" href="#api-trainer-config-helpers-layers-mixed-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">mixed_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Mixed Layer. A mixed layer will add all inputs together, then activate.
-Each inputs is a projection or operator.</p>
-<p>There are two styles of usages.</p>
-<ol class="arabic simple">
-<li>When not set inputs parameter, use mixed_layer like this:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">)</span> <span class="k">as</span> <span class="n">m</span><span class="p">:</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">)</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">identity_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<ol class="arabic simple" start="2">
-<li>You can also set all inputs when invoke mixed_layer as follows:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">m</span> <span class="o">=</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                <span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">),</span>
-                       <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer2</span><span class="p">)])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; mixed layer name. Can be referenced by other layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; layer size.</li>
-<li><strong>input</strong> &#8211; inputs layer. It is an optional parameter. If set,
-then this function will just return layer&#8217;s name.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; The extra layer config. Default is None.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">MixedLayerType object can add inputs or layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">MixedLayerType</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="embedding-layer">
-<span id="api-trainer-config-helpers-layers-embedding-layer"></span><h3>embedding_layer<a class="headerlink" href="#embedding-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">embedding_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define a embedding Layer.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Name of this embedding layer.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer for this embedding. NOTE: must be Index Data.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The embedding dimension.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None</em>) &#8211; The embedding parameter attribute. See ParameterAttribute
-for details.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra layer Config. Default is None.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="scaling-projection">
-<h3>scaling_projection<a class="headerlink" href="#scaling-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">scaling_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>scaling_projection multiplies the input with a scalar parameter and add to
-the output.</p>
-<div class="math">
-\[out += w * in\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">scaling_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A ScalingProjection object</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">ScalingProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="dotmul-projection">
-<h3>dotmul_projection<a class="headerlink" href="#dotmul-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">dotmul_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>DotMulProjection with a layer as input.
-It performs element-wise multiplication with weight.</p>
-<div class="math">
-\[out.row[i] += in.row[i] .* weight\]</div>
-<p>where <span class="math">\(.*\)</span> means element-wise multiplication.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">dotmul_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A DotMulProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">DotMulProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="dotmul-operator">
-<h3>dotmul_operator<a class="headerlink" href="#dotmul-operator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">dotmul_operator</code><span class="sig-paren">(</span><em>a=None</em>, <em>b=None</em>, <em>scale=1</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>DotMulOperator takes two inputs and performs element-wise multiplication:</p>
-<div class="math">
-\[out.row[i] += scale * (a.row[i] .* b.row[i])\]</div>
-<p>where <span class="math">\(.*\)</span> means element-wise multiplication, and
-scale is a config scalar, its default value is one.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">op</span> <span class="o">=</span> <span class="n">dotmul_operator</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">scale</span><span class="o">=</span><span class="mf">0.5</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; Input layer1</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; Input layer2</li>
-<li><strong>scale</strong> (<em>float</em>) &#8211; config scalar, default value is one.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A DotMulOperator Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">DotMulOperator</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="full-matrix-projection">
-<h3>full_matrix_projection<a class="headerlink" href="#full-matrix-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">full_matrix_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Full Matrix Projection. It performs full matrix multiplication.</p>
-<div class="math">
-\[out.row[i] += in.row[i] * weight\]</div>
-<p>There are two styles of usage.</p>
-<ol class="arabic simple">
-<li>When used in mixed_layer like this, you can only set the input:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span> <span class="k">as</span> <span class="n">m</span><span class="p">:</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<ol class="arabic simple" start="2">
-<li>When used as an independant object like this, you must set the size:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                              <span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                              <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;_proj&#39;</span><span class="p">))</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The parameter size. Means the width of parameter.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A FullMatrixProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">FullMatrixProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="identity-projection">
-<h3>identity_projection<a class="headerlink" href="#identity-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">identity_projection</code><span class="sig-paren">(</span><em>input</em>, <em>offset=None</em><span class="sig-paren">)</span></dt>
-<dd><ol class="arabic simple">
-<li>IdentityProjection if offset=None. It performs:</li>
-</ol>
-<div class="math">
-\[out.row[i] += in.row[i]\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">identity_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>2. IdentityOffsetProjection if offset!=None. It likes IdentityProjection,
-but layer size may be smaller than input size.
-It select dimesions [offset, offset+layer_size) from input:</p>
-<div class="math">
-\[out.row[i] += in.row[i + \textrm{offset}]\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">identity_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                           <span class="n">offset</span><span class="o">=</span><span class="mi">10</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>Note that both of two projections should not have any parameter.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer.</li>
-<li><strong>offset</strong> (<em>int</em>) &#8211; Offset, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A IdentityProjection or IdentityOffsetProjection object</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">IdentityProjection or IdentityOffsetProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="table-projection">
-<h3>table_projection<a class="headerlink" href="#table-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">table_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Table Projection. It selects rows from parameter where row_id
-is in input_ids.</p>
-<div class="math">
-\[out.row[i] += table.row[ids[i]]\]</div>
-<p>where <span class="math">\(out\)</span> is output, <span class="math">\(table\)</span> is parameter, <span class="math">\(ids\)</span> is input_ids,
-and <span class="math">\(i\)</span> is row_id.</p>
-<p>There are two styles of usage.</p>
-<ol class="arabic simple">
-<li>When used in mixed_layer like this, you can only set the input:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">with</span> <span class="n">mixed_layer</span><span class="p">(</span><span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span> <span class="k">as</span> <span class="n">m</span><span class="p">:</span>
-    <span class="n">m</span> <span class="o">+=</span> <span class="n">table_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<ol class="arabic simple" start="2">
-<li>When used as an independant object like this, you must set the size:</li>
-</ol>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">table_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                        <span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                        <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span><span class="n">name</span><span class="o">=</span><span class="s1">&#39;_proj&#39;</span><span class="p">))</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer, which must contains id fields.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The parameter size. Means the width of parameter.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A TableProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">TableProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="trans-full-matrix-projection">
-<h3>trans_full_matrix_projection<a class="headerlink" href="#trans-full-matrix-projection" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">trans_full_matrix_projection</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Different from full_matrix_projection, this projection performs matrix
-multiplication, using transpose of weight.</p>
-<div class="math">
-\[out.row[i] += in.row[i] * w^\mathrm{T}\]</div>
-<p><span class="math">\(w^\mathrm{T}\)</span> means transpose of weight.
-The simply usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">proj</span> <span class="o">=</span> <span class="n">trans_full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                                    <span class="n">size</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                                    <span class="n">param_attr</span><span class="o">=</span><span class="n">ParamAttr</span><span class="p">(</span>
-                                         <span class="n">name</span><span class="o">=</span><span class="s1">&#39;_proj&#39;</span><span class="p">,</span>
-                                         <span class="n">initial_mean</span><span class="o">=</span><span class="mf">0.0</span><span class="p">,</span>
-                                         <span class="n">initial_std</span><span class="o">=</span><span class="mf">0.01</span><span class="p">))</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The parameter size. Means the width of parameter.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">A TransposedFullMatrixProjection Object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">TransposedFullMatrixProjection</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="aggregate-layers">
-<h2>Aggregate Layers<a class="headerlink" href="#aggregate-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="pooling-layer">
-<span id="api-trainer-config-helpers-layers-pooling-layer"></span><h3>pooling_layer<a class="headerlink" href="#pooling-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">pooling_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Pooling layer for sequence inputs, not used for Image.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">seq_pool</span> <span class="o">=</span> <span class="n">pooling_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                         <span class="n">pooling_type</span><span class="o">=</span><span class="n">AvgPooling</span><span class="p">(),</span>
-                         <span class="n">agg_level</span><span class="o">=</span><span class="n">AggregateLevel</span><span class="o">.</span><span class="n">EACH_SEQUENCE</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>agg_level</strong> (<em>AggregateLevel</em>) &#8211; AggregateLevel.EACH_TIMESTEP or
-AggregateLevel.EACH_SEQUENCE</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>pooling_type</strong> (<em>BasePoolingType|None</em>) &#8211; Type of pooling, MaxPooling(default), AvgPooling,
-SumPooling, SquareRootNPooling.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias parameter attribute. False if no bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; The Extra Attributes for layer, such as dropout.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="last-seq">
-<span id="api-trainer-config-helpers-layers-last-seq"></span><h3>last_seq<a class="headerlink" href="#last-seq" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">last_seq</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Get Last Timestamp Activation of a sequence.</p>
-<p>If stride &gt; 0, this layer slides a window whose size is determined by stride,
-and return the last value of the window as the output. Thus, a long sequence
-will be shorten. Note that for sequence with sub-sequence, the default value
-of stride is -1.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">seq</span> <span class="o">=</span> <span class="n">last_seq</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>agg_level</strong> &#8211; Aggregated level</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>stride</strong> (<em>Int</em>) &#8211; window size.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="first-seq">
-<span id="api-trainer-config-helpers-layers-first-seq"></span><h3>first_seq<a class="headerlink" href="#first-seq" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">first_seq</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Get First Timestamp Activation of a sequence.</p>
-<p>If stride &gt; 0, this layer slides a window whose size is determined by stride,
-and return the first value of the window as the output. Thus, a long sequence
-will be shorten. Note that for sequence with sub-sequence, the default value
-of stride is -1.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">seq</span> <span class="o">=</span> <span class="n">first_seq</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>agg_level</strong> &#8211; aggregation level</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>stride</strong> (<em>Int</em>) &#8211; window size.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="concat-layer">
-<h3>concat_layer<a class="headerlink" href="#concat-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">concat_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Concat all input vector into one huge vector.
-Inputs can be list of LayerOutput or list of projection.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">concat</span> <span class="o">=</span> <span class="n">concat_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>list|tuple|collections.Sequence</em>) &#8211; input layers or projections</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="seq-concat-layer">
-<h3>seq_concat_layer<a class="headerlink" href="#seq-concat-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">seq_concat_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Concat sequence a with sequence b.</p>
-<dl class="docutils">
-<dt>Inputs:</dt>
-<dd><ul class="first last simple">
-<li>a = [a1, a2, ..., an]</li>
-<li>b = [b1, b2, ..., bn]</li>
-<li>Note that the length of a and b should be the same.</li>
-</ul>
-</dd>
-</dl>
-<p>Output: [a1, b1, a2, b2, ..., an, bn]</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">concat</span> <span class="o">=</span> <span class="n">seq_concat_layer</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; input sequence layer</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input sequence layer</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="reshaping-layers">
-<h2>Reshaping Layers<a class="headerlink" href="#reshaping-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="block-expand-layer">
-<h3>block_expand_layer<a class="headerlink" href="#block-expand-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">block_expand_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>Expand feature map to minibatch matrix.</dt>
-<dd><ul class="first last simple">
-<li>matrix width is: block_y * block_x * num_channels</li>
-<li>matirx height is: outputH * outputW</li>
-</ul>
-</dd>
-</dl>
-<div class="math">
-\[ \begin{align}\begin{aligned}outputH = 1 + (2 * padding_y + imgSizeH - block_y + stride_y - 1) / stride_y\\outputW = 1 + (2 * padding_x + imgSizeW - block_x + stride_x - 1) / stride_x\end{aligned}\end{align} \]</div>
-<p>The expand method is the same with ExpandConvLayer, but saved the transposed
-value. After expanding, output.sequenceStartPositions will store timeline.
-The number of time steps are outputH * outputW and the dimension of each
-time step is block_y * block_x * num_channels. This layer can be used after
-convolution neural network, and before recurrent neural network.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">block_expand</span> <span class="o">=</span> <span class="n">block_expand_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                                  <span class="n">num_channels</span><span class="o">=</span><span class="mi">128</span><span class="p">,</span>
-                                  <span class="n">stride_x</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                                  <span class="n">stride_y</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                                  <span class="n">block_x</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
-                                  <span class="n">block_x</span><span class="o">=</span><span class="mi">3</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>num_channels</strong> (<em>int|None</em>) &#8211; The channel number of input layer.</li>
-<li><strong>block_x</strong> (<em>int</em>) &#8211; The width of sub block.</li>
-<li><strong>block_y</strong> (<em>int</em>) &#8211; The width of sub block.</li>
-<li><strong>stride_x</strong> (<em>int</em>) &#8211; The stride size in horizontal direction.</li>
-<li><strong>stride_y</strong> (<em>int</em>) &#8211; The stride size in vertical direction.</li>
-<li><strong>padding_x</strong> (<em>int</em>) &#8211; The padding size in horizontal direction.</li>
-<li><strong>padding_y</strong> (<em>int</em>) &#8211; The padding size in vertical direction.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layer, which can not specify.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="expand-layer">
-<span id="api-trainer-config-helpers-layers-expand-layer"></span><h3>expand_layer<a class="headerlink" href="#expand-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">expand_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for &#8220;Expand Dense data or (sequence data where the length of each
-sequence is one) to sequence data.&#8221;</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">expand</span> <span class="o">=</span> <span class="n">expand_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span>
-                      <span class="n">expand_as</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span>
-                      <span class="n">expand_level</span><span class="o">=</span><span class="n">ExpandLevel</span><span class="o">.</span><span class="n">FROM_TIMESTEP</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer</li>
-<li><strong>expand_as</strong> (<em>LayerOutput</em>) &#8211; Expand as this layer&#8217;s sequence info.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias attribute. None means default bias. False means no
-bias.</li>
-<li><strong>expand_level</strong> (<em>ExpandLevel</em>) &#8211; whether input layer is timestep(default) or sequence.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="repeat-layer">
-<h3>repeat_layer<a class="headerlink" href="#repeat-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">repeat_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for repeating the input for num_repeats times. This is equivalent
-to apply concat_layer() with num_repeats same input.</p>
-<div class="math">
-\[y  = [x, x, \cdots, x]\]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">expand</span> <span class="o">=</span> <span class="n">repeat_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span> <span class="n">num_repeats</span><span class="o">=</span><span class="mi">4</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer</li>
-<li><strong>num_repeats</strong> (<em>int</em>) &#8211; Repeat the input so many times</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="rotate-layer">
-<h3>rotate_layer<a class="headerlink" href="#rotate-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">rotate_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for rotating 90 degrees (clock-wise) for each feature channel,
-usually used when the input sample is some image or feature map.</p>
-<div class="math">
-\[y(j,i,:) = x(M-i-1,j,:)\]</div>
-<p>where <span class="math">\(x\)</span> is (M x N x C) input, and <span class="math">\(y\)</span> is (N x M x C) output.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">rot</span> <span class="o">=</span> <span class="n">rotate_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span>
-                   <span class="n">height</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span>
-                   <span class="n">width</span><span class="o">=</span><span class="mi">100</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>height</strong> (<em>int</em>) &#8211; The height of the sample matrix</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="seq-reshape-layer">
-<h3>seq_reshape_layer<a class="headerlink" href="#seq-reshape-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">seq_reshape_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for reshaping the sequence. Assume the input sequence has T instances,
-the dimension of each instance is M, and the input reshape_size is N, then the
-output sequence has T*M/N instances, the dimension of each instance is N.</p>
-<p>Note that T*M/N must be an integer.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">reshape</span> <span class="o">=</span> <span class="n">seq_reshape_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span> <span class="n">reshape_size</span><span class="o">=</span><span class="mi">4</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>reshape_size</strong> (<em>int</em>) &#8211; the size of reshaped sequence.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation type.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="math-layers">
-<h2>Math Layers<a class="headerlink" href="#math-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="addto-layer">
-<h3>addto_layer<a class="headerlink" href="#addto-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">addto_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>AddtoLayer.</p>
-<div class="math">
-\[y = f(\sum_{i} x_i + b)\]</div>
-<p>where <span class="math">\(y\)</span> is output, <span class="math">\(x\)</span> is input, <span class="math">\(b\)</span> is bias,
-and <span class="math">\(f\)</span> is activation function.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">addto</span> <span class="o">=</span> <span class="n">addto_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">],</span>
-                    <span class="n">act</span><span class="o">=</span><span class="n">ReluActivation</span><span class="p">(),</span>
-                    <span class="n">bias_attr</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>This layer just simply add all input layers together, then activate the sum
-inputs. Each input of this layer should be the same size, which is also the
-output size of this layer.</p>
-<p>There is no weight matrix for each input, because it just a simple add
-operation. If you want a complicated operation before add, please use
-mixed_layer.</p>
-<p>It is a very good way to set dropout outside the layers. Since not all
-PaddlePaddle layer support dropout, you can add an add_to layer, set
-dropout here.
-Please refer to dropout_layer for details.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; Input layers. It could be a LayerOutput or list/tuple of
-LayerOutput.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type, default is tanh.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|bool</em>) &#8211; Bias attribute. If False, means no bias. None is default
-bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="linear-comb-layer">
-<h3>linear_comb_layer<a class="headerlink" href="#linear-comb-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">linear_comb_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><dl class="docutils">
-<dt>A layer for weighted sum of vectors takes two inputs.</dt>
-<dd><ul class="first last simple">
-<li><dl class="first docutils">
-<dt>Input: size of weights is M</dt>
-<dd>size of vectors is M*N</dd>
-</dl>
-</li>
-<li>Output: a vector of size=N</li>
-</ul>
-</dd>
-</dl>
-<div class="math">
-\[z(i) = \sum_{j=0}^{M-1} x(j) y(i+Nj)\]</div>
-<p>where <span class="math">\(0 \le i \le N-1\)</span></p>
-<p>Or in the matrix notation:</p>
-<div class="math">
-\[z = x^\mathrm{T} Y\]</div>
-<dl class="docutils">
-<dt>In this formular:</dt>
-<dd><ul class="first last simple">
-<li><span class="math">\(x\)</span>: weights</li>
-<li><span class="math">\(y\)</span>: vectors.</li>
-<li><span class="math">\(z\)</span>: the output.</li>
-</ul>
-</dd>
-</dl>
-<p>Note that the above computation is for one sample. Multiple samples are
-processed in one batch.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">linear_comb</span> <span class="o">=</span> <span class="n">linear_comb_layer</span><span class="p">(</span><span class="n">weights</span><span class="o">=</span><span class="n">weight</span><span class="p">,</span> <span class="n">vectors</span><span class="o">=</span><span class="n">vectors</span><span class="p">,</span>
-                                <span class="n">size</span><span class="o">=</span><span class="n">elem_dim</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>weights</strong> (<em>LayerOutput</em>) &#8211; The weight layer.</li>
-<li><strong>vectors</strong> (<em>LayerOutput</em>) &#8211; The vector layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; the dimension of this layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="interpolation-layer">
-<h3>interpolation_layer<a class="headerlink" href="#interpolation-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">interpolation_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer is for linear interpolation with two inputs,
-which is used in NEURAL TURING MACHINE.</p>
-<div class="math">
-\[y.row[i] = w[i] * x_1.row[i] + (1 - w[i]) * x_2.row[i]\]</div>
-<p>where <span class="math">\(x_1\)</span> and <span class="math">\(x_2\)</span> are two (batchSize x dataDim) inputs,
-<span class="math">\(w\)</span> is (batchSize x 1) weight vector, and <span class="math">\(y\)</span> is
-(batchSize x dataDim) output.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">interpolation</span> <span class="o">=</span> <span class="n">interpolation_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">],</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer3</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>list|tuple</em>) &#8211; Input layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="bilinear-interp-layer">
-<h3>bilinear_interp_layer<a class="headerlink" href="#bilinear-interp-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">bilinear_interp_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer is to implement bilinear interpolation on conv layer output.</p>
-<p>Please refer to Wikipedia: <a class="reference external" href="https://en.wikipedia.org/wiki/Bilinear_interpolation">https://en.wikipedia.org/wiki/Bilinear_interpolation</a></p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">bilinear</span> <span class="o">=</span> <span class="n">bilinear_interp_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">out_size_x</span><span class="o">=</span><span class="mi">64</span><span class="p">,</span> <span class="n">out_size_y</span><span class="o">=</span><span class="mi">64</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; A input layer.</li>
-<li><strong>out_size_x</strong> (<em>int|None</em>) &#8211; bilinear interpolation output width.</li>
-<li><strong>out_size_y</strong> (<em>int|None</em>) &#8211; bilinear interpolation output height.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The layer&#8217;s name, which cna not be specified.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="power-layer">
-<h3>power_layer<a class="headerlink" href="#power-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">power_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer applies a power function to a vector element-wise,
-which is used in NEURAL TURING MACHINE.</p>
-<div class="math">
-\[y = x^w\]</div>
-<p>where <span class="math">\(x\)</span> is a input vector, <span class="math">\(w\)</span> is scalar weight,
-and <span class="math">\(y\)</span> is a output vector.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">power</span> <span class="o">=</span> <span class="n">power_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="scaling-layer">
-<h3>scaling_layer<a class="headerlink" href="#scaling-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">scaling_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for multiplying input vector by weight scalar.</p>
-<div class="math">
-\[y  = w x\]</div>
-<p>where <span class="math">\(x\)</span> is size=dataDim input, <span class="math">\(w\)</span> is size=1 weight,
-and <span class="math">\(y\)</span> is size=dataDim output.</p>
-<p>Note that the above computation is for one sample. Multiple samples are
-processed in one batch.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">scale</span> <span class="o">=</span> <span class="n">scaling_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer2</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="slope-intercept-layer">
-<h3>slope_intercept_layer<a class="headerlink" href="#slope-intercept-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">slope_intercept_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer for applying a slope and an intercept to the input
-element-wise. There is no activation and weight.</p>
-<div class="math">
-\[y = slope * x + intercept\]</div>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">scale</span> <span class="o">=</span> <span class="n">slope_intercept_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">slope</span><span class="o">=-</span><span class="mf">1.0</span><span class="p">,</span> <span class="n">intercept</span><span class="o">=</span><span class="mf">1.0</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>slope</strong> (<em>float.</em>) &#8211; the scale factor.</li>
-<li><strong>intercept</strong> (<em>float.</em>) &#8211; the offset.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="tensor-layer">
-<h3>tensor_layer<a class="headerlink" href="#tensor-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">tensor_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This layer performs tensor operation for two input.
-For example, each sample:</p>
-<div class="math">
-\[y_{i} = a * W_{i} * {b^\mathrm{T}}, i=0,1,...,K-1\]</div>
-<dl class="docutils">
-<dt>In this formular:</dt>
-<dd><ul class="first last simple">
-<li><span class="math">\(a\)</span>: the first input contains M elements.</li>
-<li><span class="math">\(b\)</span>: the second input contains N elements.</li>
-<li><span class="math">\(y_{i}\)</span>: the i-th element of y.</li>
-<li><span class="math">\(W_{i}\)</span>: the i-th learned weight, shape if [M, N]</li>
-<li><span class="math">\(b^\mathrm{T}\)</span>: the transpose of <span class="math">\(b_{2}\)</span>.</li>
-</ul>
-</dd>
-</dl>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">tensor</span> <span class="o">=</span> <span class="n">tensor_layer</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">1000</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; Input layer a.</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input layer b.</li>
-<li><strong>size</strong> (<em>int.</em>) &#8211; the layer dimension.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type. Default is tanh.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; The Parameter Attribute.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|Any</em>) &#8211; The Bias Attribute. If no bias, then pass False or
-something not type of ParameterAttribute. None will get a
-default Bias.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="cos-sim">
-<span id="api-trainer-config-helpers-layers-cos-sim"></span><h3>cos_sim<a class="headerlink" href="#cos-sim" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">cos_sim</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Cosine Similarity Layer. The cosine similarity equation is here.</p>
-<div class="math">
-\[similarity = cos(\theta) = {\mathbf{a} \cdot \mathbf{b}
-\over \|\mathbf{a}\| \|\mathbf{b}\|}\]</div>
-<p>The size of a is M, size of b is M*N,
-Similarity will be calculated N times by step M. The output size is
-N. The scale will be multiplied to similarity.</p>
-<p>Note that the above computation is for one sample. Multiple samples are
-processed in one batch.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cos</span> <span class="o">=</span> <span class="n">cos_sim</span><span class="p">(</span><span class="n">a</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">size</span><span class="o">=</span><span class="mi">3</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>a</strong> (<em>LayerOutput</em>) &#8211; input layer a</li>
-<li><strong>b</strong> (<em>LayerOutput</em>) &#8211; input layer b</li>
-<li><strong>scale</strong> (<em>float</em>) &#8211; scale for cosine value. default is 5.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; layer size. NOTE size_a * size should equal size_b.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="trans-layer">
-<h3>trans_layer<a class="headerlink" href="#trans-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">trans_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for transposing a minibatch matrix.</p>
-<div class="math">
-\[y = x^\mathrm{T}\]</div>
-<p>where <span class="math">\(x\)</span> is (M x N) input, and <span class="math">\(y\)</span> is (N x M) output.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">trans</span> <span class="o">=</span> <span class="n">trans_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="sampling-layers">
-<h2>Sampling Layers<a class="headerlink" href="#sampling-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="maxid-layer">
-<h3>maxid_layer<a class="headerlink" href="#maxid-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">maxid_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for finding the id which has the maximal value for each sample.
-The result is stored in output.ids.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">maxid</span> <span class="o">=</span> <span class="n">maxid_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="sampling-id-layer">
-<h3>sampling_id_layer<a class="headerlink" href="#sampling-id-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">sampling_id_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for sampling id from multinomial distribution from the input layer.
-Sampling one id for one sample.</p>
-<p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">samping_id</span> <span class="o">=</span> <span class="n">sampling_id_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; The Layer Name.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="slicing-and-joining-layers">
-<h2>Slicing and Joining Layers<a class="headerlink" href="#slicing-and-joining-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="pad-layer">
-<h3>pad_layer<a class="headerlink" href="#pad-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">pad_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This operation pads zeros to the input data according to pad_c,pad_h
-and pad_w. pad_c, pad_h, pad_w specifies the which dimension and size
-of padding. And the input data shape is NCHW.</p>
-<p>For example, pad_c=[2,3] means padding 2 zeros before the
-input data and 3 zeros after the input data in channel dimension.
-pad_h means padding zeros in height dimension. pad_w means padding zeros
-in width dimension.</p>
-<p>For example,</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">input</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">)</span>  <span class="o">=</span> <span class="p">[</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">1</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">],</span> <span class="p">[</span><span class="mi">3</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">5</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">6</span><span class="p">,</span><span class="mi">7</span><span class="p">]]</span> <span class="p">],</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">4</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">7</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">3</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">9</span><span class="p">],</span> <span class="p">[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">]]</span> <span class="p">]</span>
-                  <span class="p">]</span>
-
-<span class="n">pad_c</span><span class="o">=</span><span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">1</span><span class="p">],</span> <span class="n">pad_h</span><span class="o">=</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="n">pad_w</span><span class="o">=</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]</span>
-
-<span class="n">output</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">)</span> <span class="o">=</span> <span class="p">[</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">1</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">],</span> <span class="p">[</span><span class="mi">3</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">5</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">6</span><span class="p">,</span><span class="mi">7</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]]</span> <span class="p">],</span>
-                    <span class="p">[</span> <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">4</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">7</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">3</span><span class="p">,</span><span class="mi">8</span><span class="p">,</span><span class="mi">9</span><span class="p">],</span> <span class="p">[</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">5</span><span class="p">]],</span>
-                      <span class="p">[[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">]]</span> <span class="p">]</span>
-                  <span class="p">]</span>
-</pre></div>
-</div>
-<p>The simply usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">pad</span> <span class="o">=</span> <span class="n">pad_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">ipt</span><span class="p">,</span>
-                <span class="n">pad_c</span><span class="o">=</span><span class="p">[</span><span class="mi">4</span><span class="p">,</span><span class="mi">4</span><span class="p">],</span>
-                <span class="n">pad_h</span><span class="o">=</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span><span class="mi">0</span><span class="p">],</span>
-                <span class="n">pad_w</span><span class="o">=</span><span class="p">[</span><span class="mi">2</span><span class="p">,</span><span class="mi">2</span><span class="p">])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input.</li>
-<li><strong>pad_c</strong> (<em>list|None</em>) &#8211; padding size in channel dimension.</li>
-<li><strong>pad_h</strong> (<em>list|None</em>) &#8211; padding size in height dimension.</li>
-<li><strong>pad_w</strong> (<em>list|None</em>) &#8211; padding size in width dimension.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="cost-layers">
-<span id="api-trainer-config-helpers-layers-cost-layers"></span><h2>Cost Layers<a class="headerlink" href="#cost-layers" title="永久链接至标题">¶</a></h2>
-<div class="section" id="cross-entropy">
-<h3>cross_entropy<a class="headerlink" href="#cross-entropy" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">cross_entropy</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for multi class entropy.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">cross_entropy</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                     <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float.</em>) &#8211; The cost is multiplied with coeff.
-The coefficient affects the gradient in the backward.</li>
-<li><strong>weight</strong> (<em>LayerOutout</em>) &#8211; The cost of each sample is multiplied with each weight.
-The weight should be a layer with size=1. Note that gradient
-will not be calculated for weight.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="cross-entropy-with-selfnorm">
-<h3>cross_entropy_with_selfnorm<a class="headerlink" href="#cross-entropy-with-selfnorm" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">cross_entropy_with_selfnorm</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for multi class entropy with selfnorm.
-Input should be a vector of positive numbers, without normalization.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">cross_entropy_with_selfnorm</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                                   <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float.</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>softmax_selfnorm_alpha</strong> (<em>float.</em>) &#8211; The scale factor affects the cost.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="multi-binary-label-cross-entropy">
-<h3>multi_binary_label_cross_entropy<a class="headerlink" href="#multi-binary-label-cross-entropy" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">multi_binary_label_cross_entropy</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for multi binary label cross entropy.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">multi_binary_label_cross_entropy</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                                        <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="mse-cost">
-<h3>mse_cost<a class="headerlink" href="#mse-cost" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">mse_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><blockquote>
-<div><p>mean squared error cost:</p>
-<div class="math">
-\[\]</div>
-</div></blockquote>
-<p>rac{1}{N}sum_{i=1}^N(t_i-y_i)^2</p>
-<blockquote>
-<div><table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">param name:</th><td class="field-body">layer name.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type name:</th><td class="field-body">basestring</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">param input:</th><td class="field-body">Network prediction.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type input:</th><td class="field-body">LayerOutput</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">param label:</th><td class="field-body">Data label.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type label:</th><td class="field-body">LayerOutput</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">param weight:</th><td class="field-body">The weight affects the cost, namely the scale of cost.
-It is an optional argument.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">type weight:</th><td class="field-body">LayerOutput</td>
-</tr>
-<tr class="field-odd field"><th class="field-name" colspan="2">param layer_attr:</th></tr>
-<tr class="field-odd field"><td>&#160;</td><td class="field-body">layer&#8217;s extra attribute.</td>
-</tr>
-<tr class="field-even field"><th class="field-name" colspan="2">type layer_attr:</th></tr>
-<tr class="field-even field"><td>&#160;</td><td class="field-body">ExtraLayerAttribute</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">return:</th><td class="field-body">LayerOutput object.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">rtype:</th><td class="field-body">LayerOutput</td>
-</tr>
-</tbody>
-</table>
-</div></blockquote>
-</dd></dl>
-
-</div>
-<div class="section" id="huber-cost">
-<h3>huber_cost<a class="headerlink" href="#huber-cost" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">huber_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer for huber loss.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">huber_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                  <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float.</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lambda-cost">
-<h3>lambda_cost<a class="headerlink" href="#lambda-cost" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">lambda_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>lambdaCost for lambdaRank LTR approach.</p>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">lambda_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                   <span class="n">score</span><span class="o">=</span><span class="n">score</span><span class="p">,</span>
-                   <span class="n">NDCG_num</span><span class="o">=</span><span class="mi">8</span><span class="p">,</span>
-                   <span class="n">max_sort_size</span><span class="o">=-</span><span class="mi">1</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Samples of the same query should be loaded as sequence.</li>
-<li><strong>score</strong> &#8211; The 2nd input. Score of each sample.</li>
-<li><strong>NDCG_num</strong> (<em>int</em>) &#8211; The size of NDCG (Normalized Discounted Cumulative Gain),
-e.g., 5 for NDCG&#64;5. It must be less than for equal to the
-minimum size of lists.</li>
-<li><strong>max_sort_size</strong> (<em>int</em>) &#8211; The size of partial sorting in calculating gradient.
-If max_sort_size = -1, then for each list, the
-algorithm will sort the entire list to get gradient.
-In other cases, max_sort_size must be greater than or
-equal to NDCG_num. And if max_sort_size is greater
-than the size of a list, the algorithm will sort the
-entire list of get gradient.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="rank-cost">
-<h3>rank_cost<a class="headerlink" href="#rank-cost" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">rank_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A cost Layer for learning to rank using gradient descent. Details can refer
-to <a class="reference external" href="http://research.microsoft.com/en-us/um/people/cburges/papers/ICML_ranking.pdf">papers</a>.
-This layer contains at least three inputs. The weight is an optional
-argument, which affects the cost.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}C_{i,j} &amp; = -\tilde{P_{ij}} * o_{i,j} + log(1 + e^{o_{i,j}})\\o_{i,j} &amp; =  o_i - o_j\\\tilde{P_{i,j}} &amp; = \{0, 0.5, 1\} \ or \ \{0, 1\}\end{aligned}\end{align} \]</div>
-<dl class="docutils">
-<dt>In this formula:</dt>
-<dd><ul class="first last simple">
-<li><span class="math">\(C_{i,j}\)</span> is the cross entropy cost.</li>
-<li><span class="math">\(\tilde{P_{i,j}}\)</span> is the label. 1 means positive order
-and 0 means reverse order.</li>
-<li><span class="math">\(o_i\)</span> and <span class="math">\(o_j\)</span>: the left output and right output.
-Their dimension is one.</li>
-</ul>
-</dd>
-</dl>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">rank_cost</span><span class="p">(</span><span class="n">left</span><span class="o">=</span><span class="n">out_left</span><span class="p">,</span>
-                 <span class="n">right</span><span class="o">=</span><span class="n">out_right</span><span class="p">,</span>
-                 <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>left</strong> (<em>LayerOutput</em>) &#8211; The first input, the size of this layer is 1.</li>
-<li><strong>right</strong> (<em>LayerOutput</em>) &#8211; The right input, the size of this layer is 1.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label is 1 or 0, means positive order and reverse order.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; The weight affects the cost, namely the scale of cost.
-It is an optional argument.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>coeff</strong> (<em>float</em>) &#8211; The coefficient affects the gradient in the backward.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="sum-cost">
-<h3>sum_cost<a class="headerlink" href="#sum-cost" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">sum_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A loss layer which calculate the sum of the input as loss</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">sum_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput.</em>) &#8211; The first input layer.</li>
-<li><strong>name</strong> (<em>None|basestring.</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput.</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="crf-layer">
-<h3>crf_layer<a class="headerlink" href="#crf-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">crf_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for calculating the cost of sequential conditional random
-field model.</p>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">crf</span> <span class="o">=</span> <span class="n">crf_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">,</span>
-                <span class="n">size</span><span class="o">=</span><span class="n">label_dim</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The first input layer is the feature.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; The second input layer is label.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; The category number.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; The third layer is &#8220;weight&#8221; of each sample, which is an
-optional argument.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter attribute. None means default attribute</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="crf-decoding-layer">
-<h3>crf_decoding_layer<a class="headerlink" href="#crf-decoding-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">crf_decoding_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for calculating the decoding sequence of sequential conditional
-random field model. The decoding sequence is stored in output.ids.
-If a second input is provided, it is treated as the ground-truth label, and
-this layer will also calculate error. output.value[i] is 1 for incorrect
-decoding or 0 for correct decoding.</p>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">crf_decoding</span> <span class="o">=</span> <span class="n">crf_decoding_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                                  <span class="n">size</span><span class="o">=</span><span class="n">label_dim</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The first input layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; size of this layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em><em> or </em><em>None</em>) &#8211; None or ground-truth label.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter attribute. None means default attribute</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="ctc-layer">
-<h3>ctc_layer<a class="headerlink" href="#ctc-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">ctc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Connectionist Temporal Classification (CTC) is designed for temporal
-classication task. That is, for sequence labeling problems where the
-alignment between the inputs and the target labels is unknown.</p>
-<p>More details can be found by referring to <a class="reference external" href="http://machinelearning.wustl.edu/mlpapers/paper_files/icml2006_GravesFGS06.pdf">Connectionist Temporal
-Classification: Labelling Unsegmented Sequence Data with Recurrent
-Neural Networks</a></p>
-<div class="admonition note">
-<p class="first admonition-title">注解</p>
-<p class="last">Considering the &#8216;blank&#8217; label needed by CTC, you need to use
-(num_classes + 1) as the input size. num_classes is the category number.
-And the &#8216;blank&#8217; is the last category index. So the size of &#8216;input&#8217; layer, such as
-fc_layer with softmax activation, should be num_classes + 1. The size of ctc_layer
-should also be num_classes + 1.</p>
-</div>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">ctc</span> <span class="o">=</span> <span class="n">ctc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">,</span>
-                <span class="n">size</span><span class="o">=</span><span class="mi">9055</span><span class="p">,</span>
-                <span class="n">norm_by_times</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; The data layer of label with variable length.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; category numbers + 1.</li>
-<li><strong>name</strong> (<em>basestring|None</em>) &#8211; The name of this layer</li>
-<li><strong>norm_by_times</strong> (<em>bool</em>) &#8211; Whether to normalization by times. False by default.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="warp-ctc-layer">
-<h3>warp_ctc_layer<a class="headerlink" href="#warp-ctc-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">warp_ctc_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer intergrating the open-source <cite>warp-ctc
-&lt;https://github.com/baidu-research/warp-ctc&gt;</cite> library, which is used in
-<cite>Deep Speech 2: End-toEnd Speech Recognition in English and Mandarin
-&lt;https://arxiv.org/pdf/1512.02595v1.pdf&gt;</cite>, to compute Connectionist Temporal
-Classification (CTC) loss.</p>
-<p>More details of CTC can be found by referring to <a class="reference external" href="http://machinelearning.wustl.edu/mlpapers/paper_files/icml2006_GravesFGS06.pdf">Connectionist Temporal
-Classification: Labelling Unsegmented Sequence Data with Recurrent
-Neural Networks</a></p>
-<div class="admonition note">
-<p class="first admonition-title">注解</p>
-<ul class="last simple">
-<li>Let num_classes represent the category number. Considering the &#8216;blank&#8217;
-label needed by CTC, you need to use (num_classes + 1) as the input
-size. Thus, the size of both warp_ctc_layer and &#8216;input&#8217; layer should
-be set to num_classes + 1.</li>
-<li>You can set &#8216;blank&#8217; to any value ranged in [0, num_classes], which
-should be consistent as that used in your labels.</li>
-<li>As a native &#8216;softmax&#8217; activation is interated to the warp-ctc library,
-&#8216;linear&#8217; activation is expected instead in the &#8216;input&#8217; layer.</li>
-</ul>
-</div>
-<p>The simple usage:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">ctc</span> <span class="o">=</span> <span class="n">warp_ctc_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span>
-                     <span class="n">label</span><span class="o">=</span><span class="n">label</span><span class="p">,</span>
-                     <span class="n">size</span><span class="o">=</span><span class="mi">1001</span><span class="p">,</span>
-                     <span class="n">blank</span><span class="o">=</span><span class="mi">1000</span><span class="p">,</span>
-                     <span class="n">norm_by_times</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; The data layer of label with variable length.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; category numbers + 1.</li>
-<li><strong>name</strong> (<em>basestring|None</em>) &#8211; The name of this layer, which can not specify.</li>
-<li><strong>blank</strong> (<em>int</em>) &#8211; the &#8216;blank&#8217; label used in ctc</li>
-<li><strong>norm_by_times</strong> (<em>bool</em>) &#8211; Whether to normalization by times. False by default.</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute|None</em>) &#8211; Extra Layer config.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="nce-layer">
-<h3>nce_layer<a class="headerlink" href="#nce-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">nce_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Noise-contrastive estimation.
-Implements the method in the following paper:
-A fast and simple algorithm for training neural probabilistic language models.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">nce_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer1</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="n">layer2</span><span class="p">,</span> <span class="n">weight</span><span class="o">=</span><span class="n">layer3</span><span class="p">,</span>
-                 <span class="n">num_classes</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">neg_distribution</span><span class="o">=</span><span class="p">[</span><span class="mf">0.1</span><span class="p">,</span><span class="mf">0.3</span><span class="p">,</span><span class="mf">0.6</span><span class="p">])</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>input</strong> (<em>LayerOutput|list|tuple|collections.Sequence</em>) &#8211; input layers. It could be a LayerOutput of list/tuple of LayerOutput.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; label layer</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; weight layer, can be None(default)</li>
-<li><strong>num_classes</strong> (<em>int</em>) &#8211; number of classes.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation, default is Sigmoid.</li>
-<li><strong>num_neg_samples</strong> (<em>int</em>) &#8211; number of negative samples. Default is 10.</li>
-<li><strong>neg_distribution</strong> (<em>list|tuple|collections.Sequence|None</em>) &#8211; The distribution for generating the random negative labels.
-A uniform distribution will be used if not provided.
-If not None, its length must be equal to num_classes.</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|None|False</em>) &#8211; Bias parameter attribute. True if no bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="hsigmoid">
-<h3>hsigmoid<a class="headerlink" href="#hsigmoid" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">hsigmoid</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Organize the classes into a binary tree. At each node, a sigmoid function
-is used to calculate the probability of belonging to the right branch.
-This idea is from &#8220;F. Morin, Y. Bengio (AISTATS 05):
-Hierarchical Probabilistic Neural Network Language Model.&#8221;</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">hsigmoid</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">,</span> <span class="n">layer2</span><span class="p">],</span>
-                <span class="n">label</span><span class="o">=</span><span class="n">data_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput|list|tuple</em>) &#8211; Input layers. It could be a LayerOutput or list/tuple of
-LayerOutput.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label layer.</li>
-<li><strong>num_classes</strong> (<em>int|None</em>) &#8211; number of classes.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; layer name</li>
-<li><strong>bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Bias attribute. None means default bias.
-False means no bias.</li>
-<li><strong>param_attr</strong> (<em>ParameterAttribute|None</em>) &#8211; Parameter Attribute. None means default parameter.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="smooth-l1-cost">
-<h3>smooth_l1_cost<a class="headerlink" href="#smooth-l1-cost" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">smooth_l1_cost</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>This is a L1 loss but more smooth. It requires that the
-size of input and label are equal. The formula is as follows,</p>
-<div class="math">
-\[L = \sum_{i} smooth_{L1}(input_i - label_i)\]</div>
-<p>in which</p>
-<div class="math">
-\[\begin{split}smooth_{L1}(x) = \begin{cases} 0.5x^2&amp; \text{if}  \ |x| &lt; 1 \\ |x|-0.5&amp; \text{otherwise} \end{cases}\end{split}\]</div>
-<p>More details can be found by referring to <a class="reference external" href="https://arxiv.org/pdf/1504.08083v2.pdf">Fast R-CNN</a></p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">cost</span> <span class="o">=</span> <span class="n">smooth_l1_cost</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">input_layer</span><span class="p">,</span>
-                      <span class="n">label</span><span class="o">=</span><span class="n">label_layer</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layer.</li>
-<li><strong>label</strong> &#8211; The input label.</li>
-<li><strong>name</strong> (<em>None|basestring</em>) &#8211; The name of this layers. It is not necessary.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; Extra Layer Attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="check-layer">
-<h2>Check Layer<a class="headerlink" href="#check-layer" title="永久链接至标题">¶</a></h2>
-<div class="section" id="eos-layer">
-<h3>eos_layer<a class="headerlink" href="#eos-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.layers.</code><code class="descname">eos_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A layer for checking EOS for each sample:
-- output_id = (input_id == conf.eos_id)</p>
-<p>The result is stored in output_.ids.
-It is used by recurrent layer group.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">eos</span> <span class="o">=</span> <span class="n">eos_layer</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">layer</span><span class="p">,</span> <span class="n">eos_id</span><span class="o">=</span><span class="nb">id</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; Layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer name.</li>
-<li><strong>eos_id</strong> (<em>int</em>) &#8211; end id of sequence</li>
-<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute.</em>) &#8211; extra layer attributes.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="../../../_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/networks.html b/develop/doc_cn/api/v1/trainer_config_helpers/networks.html
deleted file mode 100644
index db9ee2f1e4639cba9aa792517b6be40bac5e3dae..0000000000000000000000000000000000000000
--- a/develop/doc_cn/api/v1/trainer_config_helpers/networks.html
+++ /dev/null
@@ -1,957 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Networks &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="../../../genindex.html"/>
-        <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Networks</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="networks">
-<h1>Networks<a class="headerlink" href="#networks" title="永久链接至标题">¶</a></h1>
-<p>The networks module contains pieces of neural network that combine multiple layers.</p>
-<div class="section" id="nlp">
-<h2>NLP<a class="headerlink" href="#nlp" title="永久链接至标题">¶</a></h2>
-<div class="section" id="sequence-conv-pool">
-<h3>sequence_conv_pool<a class="headerlink" href="#sequence-conv-pool" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">sequence_conv_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Text convolution pooling layers helper.</p>
-<p>Text input =&gt; Context Projection =&gt; FC Layer =&gt; Pooling =&gt; Output.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of output layer(pooling layer name)</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; name of input layer</li>
-<li><strong>context_len</strong> (<em>int</em>) &#8211; context projection length. See
-context_projection&#8217;s document.</li>
-<li><strong>hidden_size</strong> (<em>int</em>) &#8211; FC Layer size.</li>
-<li><strong>context_start</strong> (<em>int</em><em> or </em><em>None</em>) &#8211; context projection length. See
-context_projection&#8217;s context_start.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType.</em>) &#8211; pooling layer type. See pooling_layer&#8217;s document.</li>
-<li><strong>context_proj_layer_name</strong> (<em>basestring</em>) &#8211; context projection layer name.
-None if user don&#8217;t care.</li>
-<li><strong>context_proj_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; context projection parameter attribute.
-None if user don&#8217;t care.</li>
-<li><strong>fc_layer_name</strong> (<em>basestring</em>) &#8211; fc layer name. None if user don&#8217;t care.</li>
-<li><strong>fc_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc layer parameter attribute. None if user don&#8217;t care.</li>
-<li><strong>fc_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc bias parameter attribute. False if no bias,
-None if user don&#8217;t care.</li>
-<li><strong>fc_act</strong> (<em>BaseActivation</em>) &#8211; fc layer activation type. None means tanh</li>
-<li><strong>pool_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; pooling layer bias attr. None if don&#8217;t care.
-False if no bias.</li>
-<li><strong>fc_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; fc layer extra attribute.</li>
-<li><strong>context_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; context projection layer extra attribute.</li>
-<li><strong>pool_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; pooling layer extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">output layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="text-conv-pool">
-<span id="api-trainer-config-helpers-network-text-conv-pool"></span><h3>text_conv_pool<a class="headerlink" href="#text-conv-pool" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">text_conv_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Text convolution pooling layers helper.</p>
-<p>Text input =&gt; Context Projection =&gt; FC Layer =&gt; Pooling =&gt; Output.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of output layer(pooling layer name)</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; name of input layer</li>
-<li><strong>context_len</strong> (<em>int</em>) &#8211; context projection length. See
-context_projection&#8217;s document.</li>
-<li><strong>hidden_size</strong> (<em>int</em>) &#8211; FC Layer size.</li>
-<li><strong>context_start</strong> (<em>int</em><em> or </em><em>None</em>) &#8211; context projection length. See
-context_projection&#8217;s context_start.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType.</em>) &#8211; pooling layer type. See pooling_layer&#8217;s document.</li>
-<li><strong>context_proj_layer_name</strong> (<em>basestring</em>) &#8211; context projection layer name.
-None if user don&#8217;t care.</li>
-<li><strong>context_proj_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; context projection parameter attribute.
-None if user don&#8217;t care.</li>
-<li><strong>fc_layer_name</strong> (<em>basestring</em>) &#8211; fc layer name. None if user don&#8217;t care.</li>
-<li><strong>fc_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc layer parameter attribute. None if user don&#8217;t care.</li>
-<li><strong>fc_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em>) &#8211; fc bias parameter attribute. False if no bias,
-None if user don&#8217;t care.</li>
-<li><strong>fc_act</strong> (<em>BaseActivation</em>) &#8211; fc layer activation type. None means tanh</li>
-<li><strong>pool_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None.</em>) &#8211; pooling layer bias attr. None if don&#8217;t care.
-False if no bias.</li>
-<li><strong>fc_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; fc layer extra attribute.</li>
-<li><strong>context_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; context projection layer extra attribute.</li>
-<li><strong>pool_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; pooling layer extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">output layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="images">
-<h2>Images<a class="headerlink" href="#images" title="永久链接至标题">¶</a></h2>
-<div class="section" id="img-conv-bn-pool">
-<h3>img_conv_bn_pool<a class="headerlink" href="#img-conv-bn-pool" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">img_conv_bn_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Convolution, batch normalization, pooling group.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; group name</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; layer&#8217;s input</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document</li>
-<li><strong>pool_size</strong> (<em>int</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; see batch_norm_layer&#8217;s document.</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document</li>
-<li><strong>conv_stride</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_padding</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>num_channel</strong> (<em>int</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>shared_bias</strong> (<em>bool</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>conv_layer_attr</strong> (<em>ExtraLayerOutput</em>) &#8211; see img_conv_layer&#8217;s document.</li>
-<li><strong>bn_param_attr</strong> (<em>ParameterAttribute.</em>) &#8211; see batch_norm_layer&#8217;s document.</li>
-<li><strong>bn_bias_attr</strong> &#8211; see batch_norm_layer&#8217;s document.</li>
-<li><strong>bn_layer_attr</strong> &#8211; ParameterAttribute.</li>
-<li><strong>pool_stride</strong> (<em>int</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>pool_padding</strong> (<em>int</em>) &#8211; see img_pool_layer&#8217;s document.</li>
-<li><strong>pool_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; see img_pool_layer&#8217;s document.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">Layer groups output</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="img-conv-group">
-<h3>img_conv_group<a class="headerlink" href="#img-conv-group" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">img_conv_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Image Convolution Group, Used for vgg net.</p>
-<p>TODO(yuyang18): Complete docs</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>conv_batchnorm_drop_rate</strong> &#8211; </li>
-<li><strong>input</strong> &#8211; </li>
-<li><strong>conv_num_filter</strong> &#8211; </li>
-<li><strong>pool_size</strong> &#8211; </li>
-<li><strong>num_channels</strong> &#8211; </li>
-<li><strong>conv_padding</strong> &#8211; </li>
-<li><strong>conv_filter_size</strong> &#8211; </li>
-<li><strong>conv_act</strong> &#8211; </li>
-<li><strong>conv_with_batchnorm</strong> &#8211; </li>
-<li><strong>pool_stride</strong> &#8211; </li>
-<li><strong>pool_type</strong> &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first last"></p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="simple-img-conv-pool">
-<span id="api-trainer-config-helpers-network-simple-img-conv-pool"></span><h3>simple_img_conv_pool<a class="headerlink" href="#simple-img-conv-pool" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_img_conv_pool</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Simple image convolution and pooling group.</p>
-<p>Input =&gt; conv =&gt; pooling</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; group name</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>filter_size</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>num_filters</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>pool_size</strong> (<em>int</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>pool_type</strong> (<em>BasePoolingType</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>groups</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>conv_stride</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>conv_padding</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer for details</li>
-<li><strong>num_channel</strong> (<em>int</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; see img_conv_layer for details</li>
-<li><strong>shared_bias</strong> (<em>bool</em>) &#8211; see img_conv_layer for details</li>
-<li><strong>conv_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; see img_conv_layer for details</li>
-<li><strong>pool_stride</strong> (<em>int</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>pool_padding</strong> (<em>int</em>) &#8211; see img_pool_layer for details</li>
-<li><strong>pool_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; see img_pool_layer for details</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">Layer&#8217;s output</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="vgg-16-network">
-<h3>vgg_16_network<a class="headerlink" href="#vgg-16-network" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">vgg_16_network</code><span class="sig-paren">(</span><em>input_image</em>, <em>num_channels</em>, <em>num_classes=1000</em><span class="sig-paren">)</span></dt>
-<dd><p>Same model from <a class="reference external" href="https://gist.github.com/ksimonyan/211839e770f7b538e2d8">https://gist.github.com/ksimonyan/211839e770f7b538e2d8</a></p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>num_classes</strong> &#8211; </li>
-<li><strong>input_image</strong> (<em>LayerOutput</em>) &#8211; </li>
-<li><strong>num_channels</strong> (<em>int</em>) &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first last"></p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="recurrent">
-<h2>Recurrent<a class="headerlink" href="#recurrent" title="永久链接至标题">¶</a></h2>
-<div class="section" id="lstm">
-<h3>LSTM<a class="headerlink" href="#lstm" title="永久链接至标题">¶</a></h3>
-<div class="section" id="lstmemory-unit">
-<h4>lstmemory_unit<a class="headerlink" href="#lstmemory-unit" title="永久链接至标题">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">lstmemory_unit</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define calculations that a LSTM unit performs in a single time step.
-This function itself is not a recurrent layer, so that it can not be
-directly applied to sequence input. This function is always used in
-recurrent_group (see layers.py for more details) to implement attention
-mechanism.</p>
-<p>Please refer to  <strong>Generating Sequences With Recurrent Neural Networks</strong>
-for more details about LSTM. The link goes as follows:
-.. _Link: <a class="reference external" href="https://arxiv.org/abs/1308.0850">https://arxiv.org/abs/1308.0850</a></p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">lstm_step</span> <span class="o">=</span> <span class="n">lstmemory_unit</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span>
-                           <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                           <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">(),</span>
-                           <span class="n">gate_act</span><span class="o">=</span><span class="n">SigmoidActivation</span><span class="p">(),</span>
-                           <span class="n">state_act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; lstmemory unit name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstmemory unit size.</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; lstm final activiation type</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; lstm gate activiation type</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; lstm state activiation type.</li>
-<li><strong>mixed_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of mixed layer.
-False means no bias, None means default bias.</li>
-<li><strong>lstm_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of lstm layer.
-False means no bias, None means default bias.</li>
-<li><strong>mixed_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; mixed layer&#8217;s extra attribute.</li>
-<li><strong>lstm_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; lstm layer&#8217;s extra attribute.</li>
-<li><strong>get_output_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; get output layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">lstmemory unit name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="lstmemory-group">
-<h4>lstmemory_group<a class="headerlink" href="#lstmemory-group" title="永久链接至标题">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">lstmemory_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>lstm_group is a recurrent layer group version of Long Short Term Memory. It
-does exactly the same calculation as the lstmemory layer (see lstmemory in
-layers.py for the maths) does. A promising benefit is that LSTM memory
-cell states, or hidden states in every time step are accessible to the
-user. This is especially useful in attention model. If you do not need to
-access the internal states of the lstm, but merely use its outputs,
-it is recommended to use the lstmemory, which is relatively faster than
-lstmemory_group.</p>
-<p>NOTE: In PaddlePaddle&#8217;s implementation, the following input-to-hidden
-multiplications:
-<span class="math">\(W_{xi}x_{t}\)</span> , <span class="math">\(W_{xf}x_{t}\)</span>,
-<span class="math">\(W_{xc}x_t\)</span>, <span class="math">\(W_{xo}x_{t}\)</span> are not done in lstmemory_unit to
-speed up the calculations. Consequently, an additional mixed_layer with
-full_matrix_projection must be included before lstmemory_unit is called.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">lstm_step</span> <span class="o">=</span> <span class="n">lstmemory_group</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span>
-                            <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                            <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">(),</span>
-                            <span class="n">gate_act</span><span class="o">=</span><span class="n">SigmoidActivation</span><span class="p">(),</span>
-                            <span class="n">state_act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; lstmemory group name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstmemory group size.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; is lstm reversed</li>
-<li><strong>param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; Parameter config, None if use default.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; lstm final activiation type</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; lstm gate activiation type</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; lstm state activiation type.</li>
-<li><strong>mixed_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of mixed layer.
-False means no bias, None means default bias.</li>
-<li><strong>lstm_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute of lstm layer.
-False means no bias, None means default bias.</li>
-<li><strong>mixed_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; mixed layer&#8217;s extra attribute.</li>
-<li><strong>lstm_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; lstm layer&#8217;s extra attribute.</li>
-<li><strong>get_output_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; get output layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">the lstmemory group.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="simple-lstm">
-<h4>simple_lstm<a class="headerlink" href="#simple-lstm" title="永久链接至标题">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_lstm</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Simple LSTM Cell.</p>
-<p>It just combine a mixed layer with fully_matrix_projection and a lstmemory
-layer. The simple lstm cell was implemented as follow equations.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}i_t &amp; = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t &amp; = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t &amp; = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t &amp; = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t &amp; = o_t tanh(c_t)\end{aligned}\end{align} \]</div>
-<p>Please refer <strong>Generating Sequences With Recurrent Neural Networks</strong> if you
-want to know what lstm is. <a class="reference external" href="http://arxiv.org/abs/1308.0850">Link</a> is here.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; lstm layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstm layer size.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
-<li><strong>mat_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; mixed layer&#8217;s matrix projection parameter attribute.</li>
-<li><strong>bias_param_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias parameter attribute. False means no bias, None
-means default bias.</li>
-<li><strong>inner_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; lstm cell parameter attribute.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; lstm final activiation type</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; lstm gate activiation type</li>
-<li><strong>state_act</strong> (<em>BaseActivation</em>) &#8211; lstm state activiation type.</li>
-<li><strong>mixed_layer_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; mixed layer&#8217;s extra attribute.</li>
-<li><strong>lstm_cell_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; lstm layer&#8217;s extra attribute.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">lstm layer name.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="bidirectional-lstm">
-<h4>bidirectional_lstm<a class="headerlink" href="#bidirectional-lstm" title="永久链接至标题">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">bidirectional_lstm</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>A bidirectional_lstm is a recurrent unit that iterates over the input
-sequence both in forward and bardward orders, and then concatenate two
-outputs form a final output. However, concatenation of two outputs
-is not the only way to form the final output, you can also, for example,
-just add them together.</p>
-<p>Please refer to  <strong>Neural Machine Translation by Jointly Learning to Align
-and Translate</strong> for more details about the bidirectional lstm.
-The link goes as follows:
-.. _Link: <a class="reference external" href="https://arxiv.org/pdf/1409.0473v3.pdf">https://arxiv.org/pdf/1409.0473v3.pdf</a></p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">bi_lstm</span> <span class="o">=</span> <span class="n">bidirectional_lstm</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">input1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; bidirectional lstm layer name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; lstm layer size.</li>
-<li><strong>return_seq</strong> (<em>bool</em>) &#8211; If set False, outputs of the last time step are
-concatenated and returned.
-If set True, the entire output sequences that are
-processed in forward and backward directions are
-concatenated and returned.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">LayerOutput object accroding to the return_seq.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="gru">
-<h3>GRU<a class="headerlink" href="#gru" title="永久链接至标题">¶</a></h3>
-<div class="section" id="gru-unit">
-<h4>gru_unit<a class="headerlink" href="#gru-unit" title="永久链接至标题">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">gru_unit</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Define calculations that a gated recurrent unit performs in a single time
-step. This function itself is not a recurrent layer, so that it can not be
-directly applied to sequence input. This function is almost always used in
-the recurrent_group (see layers.py for more details) to implement attention
-mechanism.</p>
-<p>Please see grumemory in layers.py for the details about the maths.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; type of the activation</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; type of the gate activation</li>
-<li><strong>gru_layer_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">the gru output layer.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="gru-group">
-<h4>gru_group<a class="headerlink" href="#gru-group" title="永久链接至标题">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">gru_group</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>gru_group is a recurrent layer group version of Gated Recurrent Unit. It
-does exactly the same calculation as the grumemory layer does. A promising
-benefit is that gru hidden states are accessible to the user. This is
-especially useful in attention model. If you do not need to access
-any internal state, but merely use the outputs of a GRU, it is recommended
-to use the grumemory, which is relatively faster.</p>
-<p>Please see grumemory in layers.py for more detail about the maths.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">gur_group</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span>
-                <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">,</span>
-                <span class="n">act</span><span class="o">=</span><span class="n">TanhActivation</span><span class="p">(),</span>
-                <span class="n">gate_act</span><span class="o">=</span><span class="n">SigmoidActivation</span><span class="p">())</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; type of the activiation</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; type of the gate activiation</li>
-<li><strong>gru_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias. False means no bias, None means default bias.</li>
-<li><strong>gru_layer_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">the gru group.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="simple-gru">
-<h4>simple_gru<a class="headerlink" href="#simple-gru" title="永久链接至标题">¶</a></h4>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_gru</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>You maybe see gru_step_layer, grumemory in layers.py, gru_unit, gru_group,
-simple_gru in network.py. The reason why there are so many interfaces is
-that we have two ways to implement recurrent neural network. One way is to
-use one complete layer to implement rnn (including simple rnn, gru and lstm)
-with multiple time steps, such as recurrent_layer, lstmemory, grumemory. But,
-the multiplication operation <span class="math">\(W x_t\)</span> is not computed in these layers.
-See details in their interfaces in layers.py.
-The other implementation is to use an recurrent group which can ensemble a
-series of layers to compute rnn step by step. This way is flexible for
-attenion mechanism or other complex connections.</p>
-<ul class="simple">
-<li>gru_step_layer: only compute rnn by one step. It needs an memory as input
-and can be used in recurrent group.</li>
-<li>gru_unit: a wrapper of gru_step_layer with memory.</li>
-<li>gru_group: a GRU cell implemented by a combination of multiple layers in
-recurrent group.
-But <span class="math">\(W x_t\)</span> is not done in group.</li>
-<li>gru_memory: a GRU cell implemented by one layer, which does same calculation
-with gru_group and is faster than gru_group.</li>
-<li>simple_gru: a complete GRU implementation inlcuding <span class="math">\(W x_t\)</span> and
-gru_group. <span class="math">\(W\)</span> contains <span class="math">\(W_r\)</span>, <span class="math">\(W_z\)</span> and <span class="math">\(W\)</span>, see
-formula in grumemory.</li>
-</ul>
-<p>The computational speed is that, grumemory is relatively better than
-gru_group, and gru_group is relatively better than simple_gru.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">simple_gru</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; input layer name.</li>
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
-<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
-<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
-<li><strong>act</strong> (<em>BaseActivation</em>) &#8211; type of the activiation</li>
-<li><strong>gate_act</strong> (<em>BaseActivation</em>) &#8211; type of the gate activiation</li>
-<li><strong>gru_bias_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; bias. False means no bias, None means default bias.</li>
-<li><strong>gru_layer_attr</strong> (<em>ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">the gru group.</p>
-</td>
-</tr>
-<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">LayerOutput</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="simple-attention">
-<h3>simple_attention<a class="headerlink" href="#simple-attention" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">simple_attention</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Calculate and then return a context vector by attention machanism.
-Size of the context vector equals to size of the encoded_sequence.</p>
-<div class="math">
-\[ \begin{align}\begin{aligned}a(s_{i-1},h_{j}) &amp; = v_{a}f(W_{a}s_{t-1} + U_{a}h_{j})\\e_{i,j} &amp; = a(s_{i-1}, h_{j})\\a_{i,j} &amp; = \frac{exp(e_{i,j})}{\sum_{k=1}^{T_x}{exp(e_{i,k})}}\\c_{i} &amp; = \sum_{j=1}^{T_{x}}a_{i,j}h_{j}\end{aligned}\end{align} \]</div>
-<p>where <span class="math">\(h_{j}\)</span> is the jth element of encoded_sequence,
-<span class="math">\(U_{a}h_{j}\)</span> is the jth element of encoded_proj
-<span class="math">\(s_{i-1}\)</span> is decoder_state
-<span class="math">\(f\)</span> is weight_act, and is set to tanh by default.</p>
-<p>Please refer to <strong>Neural Machine Translation by Jointly Learning to
-Align and Translate</strong> for more details. The link is as follows:
-<a class="reference external" href="https://arxiv.org/abs/1409.0473">https://arxiv.org/abs/1409.0473</a>.</p>
-<p>The example usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">context</span> <span class="o">=</span> <span class="n">simple_attention</span><span class="p">(</span><span class="n">encoded_sequence</span><span class="o">=</span><span class="n">enc_seq</span><span class="p">,</span>
-                           <span class="n">encoded_proj</span><span class="o">=</span><span class="n">enc_proj</span><span class="p">,</span>
-                           <span class="n">decoder_state</span><span class="o">=</span><span class="n">decoder_prev</span><span class="p">,)</span>
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the attention model.</li>
-<li><strong>softmax_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; parameter attribute of sequence softmax
-that is used to produce attention weight</li>
-<li><strong>weight_act</strong> (<em>Activation</em>) &#8211; activation of the attention model</li>
-<li><strong>encoded_sequence</strong> (<em>LayerOutput</em>) &#8211; output of the encoder</li>
-<li><strong>encoded_proj</strong> (<em>LayerOutput</em>) &#8211; attention weight is computed by a feed forward neural
-network which has two inputs : decoder&#8217;s hidden state
-of previous time step and encoder&#8217;s output.
-encoded_proj is output of the feed-forward network for
-encoder&#8217;s output. Here we pre-compute it outside
-simple_attention for speed consideration.</li>
-<li><strong>decoder_state</strong> (<em>LayerOutput</em>) &#8211; hidden state of decoder in previous time step</li>
-<li><strong>transform_param_attr</strong> (<a class="reference internal" href="attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a>) &#8211; parameter attribute of the feed-forward
-network that takes decoder_state as inputs to
-compute attention weight.</li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first last">a context vector</p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-<div class="section" id="miscs">
-<h2>Miscs<a class="headerlink" href="#miscs" title="永久链接至标题">¶</a></h2>
-<div class="section" id="dropout-layer">
-<h3>dropout_layer<a class="headerlink" href="#dropout-layer" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">dropout_layer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>&#64;TODO(yuyang18): Add comments.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>name</strong> &#8211; </li>
-<li><strong>input</strong> &#8211; </li>
-<li><strong>dropout_rate</strong> &#8211; </li>
-</ul>
-</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first last"></p>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="outputs">
-<h3>outputs<a class="headerlink" href="#outputs" title="永久链接至标题">¶</a></h3>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">outputs</code><span class="sig-paren">(</span><em>layers</em>, <em>*args</em><span class="sig-paren">)</span></dt>
-<dd><p>Declare the outputs of network. If user have not defined the inputs of
-network, this method will calculate the input order by dfs travel.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><strong>layers</strong> (<em>list|tuple|LayerOutput</em>) &#8211; Output layers.</td>
-</tr>
-<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"></td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="../../../_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/optimizers.html b/develop/doc_cn/api/v1/trainer_config_helpers/optimizers.html
deleted file mode 100644
index b2780e0bd32c1a69ec929e0a36119ee6c5b5ca75..0000000000000000000000000000000000000000
--- a/develop/doc_cn/api/v1/trainer_config_helpers/optimizers.html
+++ /dev/null
@@ -1,482 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Optimizers &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="../../../genindex.html"/>
-        <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Optimizers</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="optimizers">
-<span id="api-trainer-config-helpers-optimizers"></span><h1>Optimizers<a class="headerlink" href="#optimizers" title="永久链接至标题">¶</a></h1>
-<div class="section" id="basesgdoptimizer">
-<h2>BaseSGDOptimizer<a class="headerlink" href="#basesgdoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">BaseSGDOptimizer</code></dt>
-<dd><p>SGD Optimizer.</p>
-<p>SGD is an optimization method, trying to find a neural network that
-minimize the &#8220;cost/error&#8221; of it by iteration. In paddle&#8217;s implementation
-SGD Optimizer is synchronized, which means all gradients will be wait to
-calculate and reduced into one gradient, then do optimize operation.</p>
-<p>The neural network consider the learning problem of minimizing an objective
-function, that has the form of a sum</p>
-<div class="math">
-\[Q(w) = \sum_{i}^{n} Q_i(w)\]</div>
-<p>The value of function Q sometimes is the cost of neural network (Mean
-Square Error between prediction and label for example). The function Q is
-parametrised by w, the weight/bias of neural network. And weights is what to
-be learned. The i is the i-th observation in (trainning) data.</p>
-<p>So, the SGD method will optimize the weight by</p>
-<div class="math">
-\[w = w - \eta \nabla Q(w) = w - \eta \sum_{i}^{n} \nabla Q_i(w)\]</div>
-<p>where <span class="math">\(\eta\)</span> is learning rate. And <span class="math">\(n\)</span> is batch size.</p>
-</dd></dl>
-
-</div>
-<div class="section" id="momentumoptimizer">
-<h2>MomentumOptimizer<a class="headerlink" href="#momentumoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">MomentumOptimizer</code><span class="sig-paren">(</span><em>momentum=None</em>, <em>sparse=False</em><span class="sig-paren">)</span></dt>
-<dd><p>MomentumOptimizer.</p>
-<p>When sparse=True, the update scheme:</p>
-<div class="math">
-\[\begin{split}\alpha_t &amp;= \alpha_{t-1} / k \\
-\beta_t &amp;= \beta_{t-1} / (1 + \lambda \gamma_t) \\
-u_t &amp;= u_{t-1} - \alpha_t \gamma_t g_t \\
-v_t &amp;= v_{t-1} + \tau_{t-1} \alpha_t \gamma_t g_t \\
-\tau_t &amp;= \tau_{t-1} + \beta_t / \alpha_t\end{split}\]</div>
-<p>where <span class="math">\(k\)</span> is momentum, <span class="math">\(\lambda\)</span> is decay rate,
-<span class="math">\(\gamma_t\)</span> is learning rate at the t&#8217;th step.</p>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><strong>sparse</strong> (<em>bool</em>) &#8211; with sparse support or not.</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adamoptimizer">
-<h2>AdamOptimizer<a class="headerlink" href="#adamoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdamOptimizer</code><span class="sig-paren">(</span><em>beta1=0.9</em>, <em>beta2=0.999</em>, <em>epsilon=1e-08</em><span class="sig-paren">)</span></dt>
-<dd><p>Adam optimizer.
-The details of please refer <a class="reference external" href="https://arxiv.org/abs/1412.6980">Adam: A Method for Stochastic Optimization</a></p>
-<div class="math">
-\[\begin{split}m(w, t) &amp; = \beta_1 m(w, t-1) + (1 - \beta_1) \nabla Q_i(w) \\
-v(w, t) &amp; = \beta_2 v(w, t-1) + (1 - \beta_2)(\nabla Q_i(w)) ^2 \\
-w &amp; = w - \frac{\eta}{\sqrt{v(w,t) + \epsilon}}\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>beta1</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_1\)</span> in equation.</li>
-<li><strong>beta2</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_2\)</span> in equation.</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; the <span class="math">\(\epsilon\)</span> in equation. It is used to prevent
-divided by zero.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adamaxoptimizer">
-<h2>AdamaxOptimizer<a class="headerlink" href="#adamaxoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdamaxOptimizer</code><span class="sig-paren">(</span><em>beta1</em>, <em>beta2</em><span class="sig-paren">)</span></dt>
-<dd><p>Adamax optimizer.</p>
-<p>The details of please refer this <a class="reference external" href="https://arxiv.org/abs/1412.6980">Adam: A Method for Stochastic Optimization</a></p>
-<div class="math">
-\[\begin{split}m_t &amp; = \beta_1 * m_{t-1} + (1-\beta_1)* \nabla Q_i(w) \\
-u_t &amp; = max(\beta_2*u_{t-1}, abs(\nabla Q_i(w))) \\
-w_t &amp; = w_{t-1} - (\eta/(1-\beta_1^t))*m_t/u_t\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>beta1</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_1\)</span> in the equation.</li>
-<li><strong>beta2</strong> (<em>float</em>) &#8211; the <span class="math">\(\beta_2\)</span> in the equation.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adagradoptimizer">
-<h2>AdaGradOptimizer<a class="headerlink" href="#adagradoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdaGradOptimizer</code></dt>
-<dd><p>Adagrad(for ADAptive GRAdient algorithm) optimizer.</p>
-<p>For details please refer this <a class="reference external" href="http://www.magicbroom.info/Papers/DuchiHaSi10.pdf">Adaptive Subgradient Methods for
-Online Learning and Stochastic Optimization</a>.</p>
-<div class="math">
-\[\begin{split}G &amp;= \sum_{\tau=1}^{t} g_{\tau} g_{\tau}^T \\
-w &amp; = w - \eta diag(G)^{-\frac{1}{2}} \circ g\end{split}\]</div>
-</dd></dl>
-
-</div>
-<div class="section" id="decayedadagradoptimizer">
-<h2>DecayedAdaGradOptimizer<a class="headerlink" href="#decayedadagradoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">DecayedAdaGradOptimizer</code><span class="sig-paren">(</span><em>rho=0.95</em>, <em>epsilon=1e-06</em><span class="sig-paren">)</span></dt>
-<dd><p>AdaGrad method with decayed sum gradients. The equations of this method
-show as follow.</p>
-<div class="math">
-\[\begin{split}E(g_t^2) &amp;= \rho * E(g_{t-1}^2) + (1-\rho) * g^2 \\
-learning\_rate &amp;= 1/sqrt( ( E(g_t^2) + \epsilon )\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>rho</strong> (<em>float</em>) &#8211; The <span class="math">\(\rho\)</span> parameter in that equation</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; The <span class="math">\(\epsilon\)</span> parameter in that equation.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="adadeltaoptimizer">
-<h2>AdaDeltaOptimizer<a class="headerlink" href="#adadeltaoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">AdaDeltaOptimizer</code><span class="sig-paren">(</span><em>rho=0.95</em>, <em>epsilon=1e-06</em><span class="sig-paren">)</span></dt>
-<dd><p>AdaDelta method. The details of adadelta please refer to this
-<a class="reference external" href="http://www.matthewzeiler.com/pubs/googleTR2012/googleTR2012.pdf">ADADELTA: AN ADAPTIVE LEARNING RATE METHOD</a>.</p>
-<div class="math">
-\[\begin{split}E(g_t^2) &amp;= \rho * E(g_{t-1}^2) + (1-\rho) * g^2 \\
-learning\_rate &amp;= sqrt( ( E(dx_{t-1}^2) + \epsilon ) / ( \
-                  E(g_t^2) + \epsilon ) ) \\
-E(dx_t^2) &amp;= \rho * E(dx_{t-1}^2) + (1-\rho) * (-g*learning\_rate)^2\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>rho</strong> (<em>float</em>) &#8211; <span class="math">\(\rho\)</span> in equation</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; <span class="math">\(\rho\)</span> in equation</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="rmspropoptimizer">
-<h2>RMSPropOptimizer<a class="headerlink" href="#rmspropoptimizer" title="永久链接至标题">¶</a></h2>
-<dl class="class">
-<dt>
-<em class="property">class </em><code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">RMSPropOptimizer</code><span class="sig-paren">(</span><em>rho=0.95</em>, <em>epsilon=1e-06</em><span class="sig-paren">)</span></dt>
-<dd><p>RMSProp(for Root Mean Square Propagation) optimizer. For details please
-refer this <a class="reference external" href="http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf">slide</a>.</p>
-<p>The equations of this method as follows:</p>
-<div class="math">
-\[\begin{split}v(w, t) &amp; = \rho v(w, t-1) + (1 - \rho)(\nabla Q_{i}(w))^2 \\
-w &amp; = w - \frac{\eta} {\sqrt{v(w,t) + \epsilon}} \nabla Q_{i}(w)\end{split}\]</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>rho</strong> (<em>float</em>) &#8211; the <span class="math">\(\rho\)</span> in the equation. The forgetting factor.</li>
-<li><strong>epsilon</strong> (<em>float</em>) &#8211; the <span class="math">\(\epsilon\)</span> in the equation.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-<div class="section" id="settings">
-<span id="api-trainer-config-helpers-optimizers-settings"></span><h2>settings<a class="headerlink" href="#settings" title="永久链接至标题">¶</a></h2>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.optimizers.</code><code class="descname">settings</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
-<dd><p>Set the optimization method, learning rate, batch size, and other training
-settings. The currently supported algorithms are SGD and Async-SGD.</p>
-<div class="admonition warning">
-<p class="first admonition-title">警告</p>
-<p class="last">Note that the &#8216;batch_size&#8217; in PaddlePaddle is not equal to global
-training batch size. It represents the single training process&#8217;s batch
-size. If you use N processes to train one model, for example use three
-GPU machines, the global batch size is N*&#8217;batch_size&#8217;.</p>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>batch_size</strong> (<em>int</em>) &#8211; batch size for one training process.</li>
-<li><strong>learning_rate</strong> (<em>float</em>) &#8211; learning rate for SGD</li>
-<li><strong>learning_method</strong> (<em>BaseSGDOptimizer</em>) &#8211; The extension optimization algorithms of gradient
-descent, such as momentum, adagrad, rmsprop, etc.
-Note that it should be instance with base type
-BaseSGDOptimizer.</li>
-<li><strong>regularization</strong> (<em>BaseRegularization</em>) &#8211; The regularization method.</li>
-<li><strong>is_async</strong> (<em>bool</em>) &#8211; Is Async-SGD or not. Default value is False.</li>
-<li><strong>model_average</strong> (<em>ModelAverage</em>) &#8211; Model Average Settings.</li>
-<li><strong>gradient_clipping_threshold</strong> (<em>float</em>) &#8211; gradient clipping threshold. If gradient
-value larger than some value, will be
-clipped.</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="../../../_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/poolings.html b/develop/doc_cn/api/v1/trainer_config_helpers/poolings.html
deleted file mode 100644
index bb4fa3a0f3c76a71dd074596273b2871bbedfc43..0000000000000000000000000000000000000000
--- a/develop/doc_cn/api/v1/trainer_config_helpers/poolings.html
+++ /dev/null
@@ -1,291 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Poolings &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="../../../genindex.html"/>
-        <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-  
-  <script src="../../../_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="../../../_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../../howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
- 
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Poolings</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-  <div class="section" id="poolings">
-<h1>Poolings<a class="headerlink" href="#poolings" title="永久链接至标题">¶</a></h1>
-<div class="section" id="basepoolingtype">
-<h2>BasePoolingType<a class="headerlink" href="#basepoolingtype" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">BasePoolingType</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">BasePool</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="avgpooling">
-<h2>AvgPooling<a class="headerlink" href="#avgpooling" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">AvgPooling</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Avg</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="maxpooling">
-<h2>MaxPooling<a class="headerlink" href="#maxpooling" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">MaxPooling</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Max</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="sumpooling">
-<h2>SumPooling<a class="headerlink" href="#sumpooling" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">SumPooling</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">Sum</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-<div class="section" id="squarerootnpooling">
-<h2>SquareRootNPooling<a class="headerlink" href="#squarerootnpooling" title="永久链接至标题">¶</a></h2>
-<dl class="attribute">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.poolings.</code><code class="descname">SquareRootNPooling</code></dt>
-<dd><p><code class="xref py py-class docutils literal"><span class="pre">SquareRootN</span></code> 的别名</p>
-</dd></dl>
-
-</div>
-</div>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'../../../',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="../../../_static/jquery.js"></script>
-      <script type="text/javascript" src="../../../_static/underscore.js"></script>
-      <script type="text/javascript" src="../../../_static/doctools.js"></script>
-      <script type="text/javascript" src="../../../_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="../../../_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/api/v2/config/activation.html b/develop/doc_cn/api/v2/config/activation.html
index 934ea840269ade43b1c24e47883a1ccbd94581b6..f6ea43b8d4ce14909bee13427ce06cd5d7561c90 100644
--- a/develop/doc_cn/api/v2/config/activation.html
+++ b/develop/doc_cn/api/v2/config/activation.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">模型配置</a><ul class="current">
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v2/config/attr.html b/develop/doc_cn/api/v2/config/attr.html
index 5bf5f54945f77dd150bbb104fdf6af87b7138c43..40abd3cca43ed52b80491c1e0267bd80fda78cec 100644
--- a/develop/doc_cn/api/v2/config/attr.html
+++ b/develop/doc_cn/api/v2/config/attr.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">模型配置</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v1/trainer_config_helpers/evaluators.html b/develop/doc_cn/api/v2/config/evaluators.html
similarity index 69%
rename from develop/doc_cn/api/v1/trainer_config_helpers/evaluators.html
rename to develop/doc_cn/api/v2/config/evaluators.html
index 3be98a90dc2b901ec656d711932358f84e44a7be..de16ba3f95ec3588d27376cff19932df0d1036d7 100644
--- a/develop/doc_cn/api/v1/trainer_config_helpers/evaluators.html
+++ b/develop/doc_cn/api/v2/config/evaluators.html
@@ -32,7 +32,10 @@
         <link rel="index" title="索引"
               href="../../../genindex.html"/>
         <link rel="search" title="搜索" href="../../../search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/> 
+    <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/>
+        <link rel="up" title="Model Configuration" href="../model_configs.html"/>
+        <link rel="next" title="Optimizer" href="optimizer.html"/>
+        <link rel="prev" title="Layers" href="layer.html"/> 
 
   <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
   <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
@@ -80,10 +83,10 @@
       </div>
       <div class="doc-module">
         
-        <ul>
+        <ul class="current">
 <li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a></li>
 <li class="toctree-l1"><a class="reference internal" href="../../../howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a></li>
+<li class="toctree-l1 current"><a class="reference internal" href="../../index_cn.html">API</a></li>
 <li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
 </ul>
 
@@ -105,7 +108,7 @@
     <nav class="doc-menu-vertical" role="navigation">
         
           
-          <ul>
+          <ul class="current">
 <li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_cn.html">新手入门</a><ul>
 <li class="toctree-l2"><a class="reference internal" href="../../../getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
@@ -138,18 +141,19 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
 </ul>
 </li>
-<li class="toctree-l1"><a class="reference internal" href="../../index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="../../v2/config/attr.html">Parameter Attribute</a></li>
+<li class="toctree-l1 current"><a class="reference internal" href="../../index_cn.html">API</a><ul class="current">
+<li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">模型配置</a><ul class="current">
+<li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
+<li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3 current"><a class="current reference internal" href="#">Evaluators</a></li>
+<li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
+<li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
+<li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
+<li class="toctree-l3"><a class="reference internal" href="attr.html">Parameter Attribute</a></li>
 </ul>
 </li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="../../v2/run_logic.html">训练与应用</a></li>
+<li class="toctree-l2"><a class="reference internal" href="../data.html">数据访问</a></li>
+<li class="toctree-l2"><a class="reference internal" href="../run_logic.html">训练与应用</a></li>
 </ul>
 </li>
 <li class="toctree-l1"><a class="reference internal" href="../../../faq/index_cn.html">FAQ</a></li>
@@ -173,6 +177,10 @@
 <div role="navigation" aria-label="breadcrumbs navigation">
   <ul class="wy-breadcrumbs">
       
+        <li><a href="../../index_cn.html">API</a> > </li>
+      
+        <li><a href="../model_configs.html">Model Configuration</a> > </li>
+      
     <li>Evaluators</li>
   </ul>
 </div>
@@ -183,54 +191,14 @@
            <div itemprop="articleBody">
             
   <div class="section" id="evaluators">
-<span id="api-trainer-config-helpers-evaluators"></span><h1>Evaluators<a class="headerlink" href="#evaluators" title="永久链接至标题">¶</a></h1>
-<div class="section" id="base">
-<h2>Base<a class="headerlink" href="#base" title="永久链接至标题">¶</a></h2>
-<dl class="function">
-<dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">evaluator_base</code><span class="sig-paren">(</span><em>input</em>, <em>type</em>, <em>label=None</em>, <em>weight=None</em>, <em>name=None</em>, <em>chunk_scheme=None</em>, <em>num_chunk_types=None</em>, <em>classification_threshold=None</em>, <em>positive_label=None</em>, <em>dict_file=None</em>, <em>result_file=None</em>, <em>num_results=None</em>, <em>delimited=None</em>, <em>top_k=None</em>, <em>excluded_chunk_types=None</em><span class="sig-paren">)</span></dt>
-<dd><p>Evaluator will evaluate the network status while training/testing.</p>
-<p>User can use evaluator by classify/regression job. For example.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">classify</span><span class="p">(</span><span class="n">prediction</span><span class="p">,</span> <span class="n">output</span><span class="p">,</span> <span class="n">evaluator</span><span class="o">=</span><span class="n">classification_error_evaluator</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>And user could define evaluator separately as follow.</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">classification_error_evaluator</span><span class="p">(</span><span class="s2">&quot;ErrorRate&quot;</span><span class="p">,</span> <span class="n">prediction</span><span class="p">,</span> <span class="n">label</span><span class="p">)</span>
-</pre></div>
-</div>
-<p>The evaluator often contains a name parameter. It will also be printed when
-evaluating network. The printed information may look like the following.</p>
-<div class="highlight-text"><div class="highlight"><pre><span></span>Batch=200 samples=20000 AvgCost=0.679655 CurrentCost=0.662179 Eval:
-classification_error_evaluator=0.4486
-CurrentEval: ErrorRate=0.3964
-</pre></div>
-</div>
-<table class="docutils field-list" frame="void" rules="none">
-<col class="field-name" />
-<col class="field-body" />
-<tbody valign="top">
-<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>list|LayerOutput</em>) &#8211; Input layers, a object of LayerOutput or a list of
-LayerOutput.</li>
-<li><strong>label</strong> (<em>LayerOutput|None</em>) &#8211; An input layer containing the ground truth label.</li>
-<li><strong>weight</strong> (<em>LayerOutput.</em>) &#8211; An input layer which is a weight for each sample.
-Each evaluator may calculate differently to use this weight.</li>
-<li><strong>top_k</strong> (<em>int</em>) &#8211; number k in top-k error rate</li>
-</ul>
-</td>
-</tr>
-</tbody>
-</table>
-</dd></dl>
-
-</div>
+<span id="api-v2"></span><h1>Evaluators<a class="headerlink" href="#evaluators" title="永久链接至标题">¶</a></h1>
 <div class="section" id="classification">
 <h2>Classification<a class="headerlink" href="#classification" title="永久链接至标题">¶</a></h2>
-<div class="section" id="classification-error-evaluator">
-<h3>classification_error_evaluator<a class="headerlink" href="#classification-error-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="classification-error">
+<h3>classification_error<a class="headerlink" href="#classification-error" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">classification_error_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">classification_error</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Classification Error Evaluator. It will print error rate for classification.</p>
 <p>The classification error is:</p>
 <div class="math">
@@ -245,9 +213,9 @@ Each evaluator may calculate differently to use this weight.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
 <li><strong>name</strong> (<em>basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
 <li><strong>label</strong> (<em>basestring</em>) &#8211; Label layer name.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. And will just multiply to NumOfWrongPredicts
 and NumOfAllSamples. So, the elements of weight are all one,
 then means not set weight. The larger weight it is, the more
@@ -265,11 +233,11 @@ important this sample is.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="auc-evaluator">
-<h3>auc_evaluator<a class="headerlink" href="#auc-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="auc">
+<h3>auc<a class="headerlink" href="#auc" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">auc_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">auc</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Auc Evaluator which adapts to binary classification.</p>
 <p>The simple usage:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">auc_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">label</span><span class="p">)</span>
@@ -281,9 +249,9 @@ important this sample is.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
 <li><strong>label</strong> (<em>None|basestring</em>) &#8211; Label layer name.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1].</li>
 </ul>
 </td>
@@ -293,11 +261,11 @@ important this sample is.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="ctc-error-evaluator">
-<h3>ctc_error_evaluator<a class="headerlink" href="#ctc-error-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="ctc-error">
+<h3>ctc_error<a class="headerlink" href="#ctc-error" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">ctc_error_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">ctc_error</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This evaluator is to calculate sequence-to-sequence edit distance.</p>
 <p>The simple usage is :</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">ctc_error_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="nb">input</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="n">lbl</span><span class="p">)</span>
@@ -309,9 +277,9 @@ important this sample is.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer. Should be the same as the input for ctc_layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; input label, which is a data_layer. Should be the same as the
-label for ctc_layer</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer. Should be the same as the input for ctc.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; input label, which is a data. Should be the same as the
+label for ctc</li>
 </ul>
 </td>
 </tr>
@@ -320,11 +288,11 @@ label for ctc_layer</li>
 </dd></dl>
 
 </div>
-<div class="section" id="chunk-evaluator">
-<h3>chunk_evaluator<a class="headerlink" href="#chunk-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="chunk">
+<h3>chunk<a class="headerlink" href="#chunk" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">chunk_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">chunk</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Chunk evaluator is used to evaluate segment labelling accuracy for a
 sequence. It calculates the chunk detection F1 score.</p>
 <p>A chunk is correctly detected if its beginning, end and type are correct.
@@ -355,8 +323,8 @@ The tag type for each of the scheme is shown as follows:</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; The input layers.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; An input layer containing the ground truth label.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; The input layers.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; An input layer containing the ground truth label.</li>
 <li><strong>chunk_scheme</strong> (<em>basestring</em>) &#8211; The labelling schemes support 4 types. It is one of
 &#8220;IOB&#8221;, &#8220;IOE&#8221;, &#8220;IOBES&#8221;, &#8220;plain&#8221;. It is required.</li>
 <li><strong>num_chunk_types</strong> &#8211; number of chunk types other than &#8220;other&#8221;</li>
@@ -370,11 +338,11 @@ The tag type for each of the scheme is shown as follows:</p>
 </dd></dl>
 
 </div>
-<div class="section" id="precision-recall-evaluator">
-<h3>precision_recall_evaluator<a class="headerlink" href="#precision-recall-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="precision-recall">
+<h3>precision_recall<a class="headerlink" href="#precision-recall" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">precision_recall_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">precision_recall</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>An Evaluator to calculate precision and recall, F1-score.
 It is adapt to the task with multiple labels.</p>
 <ul class="simple">
@@ -393,10 +361,10 @@ F1-score of this label.</li>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label layer name.</li>
-<li><strong>positive_label</strong> (<em>LayerOutput.</em>) &#8211; The input label layer.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Label layer name.</li>
+<li><strong>positive_label</strong> (<em>paddle.v2.config_base.Layer.</em>) &#8211; The input label layer.</li>
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. (TODO, explaination)</li>
 </ul>
 </td>
@@ -409,11 +377,11 @@ F1-score of this label.</li>
 </div>
 <div class="section" id="rank">
 <h2>Rank<a class="headerlink" href="#rank" title="永久链接至标题">¶</a></h2>
-<div class="section" id="pnpair-evaluator">
-<h3>pnpair_evaluator<a class="headerlink" href="#pnpair-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="pnpair">
+<h3>pnpair<a class="headerlink" href="#pnpair" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">pnpair_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">pnpair</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Positive-negative pair rate Evaluator which adapts to rank task like
 learning to rank. This evaluator must contain at least three layers.</p>
 <p>The simple usage:</p>
@@ -426,10 +394,10 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name. The output prediction of network.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Label layer name.</li>
-<li><strong>info</strong> (<em>LayerOutput</em>) &#8211; Label layer name. (TODO, explaination)</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name. The output prediction of network.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Label layer name.</li>
+<li><strong>info</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Label layer name. (TODO, explaination)</li>
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. (TODO, explaination)</li>
 </ul>
 </td>
@@ -442,11 +410,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </div>
 <div class="section" id="utils">
 <h2>Utils<a class="headerlink" href="#utils" title="永久链接至标题">¶</a></h2>
-<div class="section" id="sum-evaluator">
-<h3>sum_evaluator<a class="headerlink" href="#sum-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="sum">
+<h3>sum<a class="headerlink" href="#sum" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">sum_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">sum</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>An Evaluator to sum the result of input.</p>
 <p>The simple usage:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">sum_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
@@ -458,8 +426,8 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name.</li>
-<li><strong>weight</strong> (<em>LayerOutput</em>) &#8211; Weight Layer name. It should be a matrix with size
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name.</li>
+<li><strong>weight</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Weight Layer name. It should be a matrix with size
 [sample_num, 1]. (TODO, explaination)</li>
 </ul>
 </td>
@@ -469,11 +437,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="column-sum-evaluator">
-<h3>column_sum_evaluator<a class="headerlink" href="#column-sum-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="column-sum">
+<h3>column_sum<a class="headerlink" href="#column-sum" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">column_sum_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">column_sum</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to sum the last column of input.</p>
 <p>The simple usage is:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">column_sum_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">label</span><span class="p">)</span>
@@ -485,7 +453,7 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input Layer name.</li>
 </ul>
 </td>
 </tr>
@@ -497,11 +465,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </div>
 <div class="section" id="print">
 <h2>Print<a class="headerlink" href="#print" title="永久链接至标题">¶</a></h2>
-<div class="section" id="classification-error-printer-evaluator">
-<h3>classification_error_printer_evaluator<a class="headerlink" href="#classification-error-printer-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="classification-error-printer">
+<h3>classification_error_printer<a class="headerlink" href="#classification-error-printer" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">classification_error_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">classification_error_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the classification error of each sample.</p>
 <p>The simple usage is:</p>
 <div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">classification_error_printer_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span>
@@ -512,8 +480,8 @@ learning to rank. This evaluator must contain at least three layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput</em>) &#8211; Input layer.</li>
-<li><strong>label</strong> (<em>LayerOutput</em>) &#8211; Input label layer.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input layer.</li>
+<li><strong>label</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Input label layer.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -523,11 +491,11 @@ learning to rank. This evaluator must contain at least three layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="gradient-printer-evaluator">
-<h3>gradient_printer_evaluator<a class="headerlink" href="#gradient-printer-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="gradient-printer">
+<h3>gradient_printer<a class="headerlink" href="#gradient-printer" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">gradient_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">gradient_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the gradient of input layers. It contains
 one or more input layers.</p>
 <p>The simple usage is:</p>
@@ -539,7 +507,7 @@ one or more input layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; One or more input layers.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; One or more input layers.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -549,11 +517,11 @@ one or more input layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="maxid-printer-evaluator">
-<h3>maxid_printer_evaluator<a class="headerlink" href="#maxid-printer-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="maxid-printer">
+<h3>maxid_printer<a class="headerlink" href="#maxid-printer" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">maxid_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">maxid_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print maximum top k values and their indexes
 of each row of input layers. It contains one or more input layers.
 k is specified by num_results.</p>
@@ -566,7 +534,7 @@ k is specified by num_results.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; Input Layer name.</li>
 <li><strong>num_results</strong> (<em>int.</em>) &#8211; This number is used to specify the top k numbers.
 It is 1 by default.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
@@ -578,11 +546,11 @@ It is 1 by default.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="maxframe-printer-evaluator">
-<h3>maxframe_printer_evaluator<a class="headerlink" href="#maxframe-printer-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="maxframe-printer">
+<h3>maxframe_printer<a class="headerlink" href="#maxframe-printer" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">maxframe_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">maxframe_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the top k frames of each input layers.
 The input layers should contain sequences info or sequences type.
 k is specified by num_results.
@@ -600,7 +568,7 @@ It contains one or more input layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; Input Layer name.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -610,11 +578,11 @@ It contains one or more input layers.</p>
 </dd></dl>
 
 </div>
-<div class="section" id="seqtext-printer-evaluator">
-<h3>seqtext_printer_evaluator<a class="headerlink" href="#seqtext-printer-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="seqtext-printer">
+<h3>seqtext_printer<a class="headerlink" href="#seqtext-printer" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">seqtext_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">seqtext_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>Sequence text printer will print text according to index matrix and a
 dictionary. There can be multiple input to this layer:</p>
 <p>1. If there is no id_input, the input must be a matrix containing
@@ -646,7 +614,7 @@ the sequence of indices;</p>
 <p>Typically SequenceTextPrinter layer takes output of maxid or RecurrentGroup
 with maxid (when generating) as an input.</p>
 <p>The simple usage is:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">seqtext_printer_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">maxid_layer</span><span class="p">,</span>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="nb">eval</span> <span class="o">=</span> <span class="n">seqtext_printer_evaluator</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">maxid</span><span class="p">,</span>
                                  <span class="n">id_input</span><span class="o">=</span><span class="n">sample_id</span><span class="p">,</span>
                                  <span class="n">dict_file</span><span class="o">=</span><span class="n">dict_file</span><span class="p">,</span>
                                  <span class="n">result_file</span><span class="o">=</span><span class="n">result_file</span><span class="p">)</span>
@@ -657,9 +625,9 @@ with maxid (when generating) as an input.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; Input Layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; Input Layer name.</li>
 <li><strong>result_file</strong> (<em>basestring</em>) &#8211; Path of the file to store the generated results.</li>
-<li><strong>id_input</strong> (<em>LayerOutput</em>) &#8211; Index of the input sequence, and the specified index will
+<li><strong>id_input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; Index of the input sequence, and the specified index will
 be prited in the gereated results. This an optional
 parameter.</li>
 <li><strong>dict_file</strong> (<em>basestring</em>) &#8211; Path of dictionary. This is an optional parameter.
@@ -684,11 +652,11 @@ Default is True. No space is added if set to False.</li>
 </dd></dl>
 
 </div>
-<div class="section" id="value-printer-evaluator">
-<h3>value_printer_evaluator<a class="headerlink" href="#value-printer-evaluator" title="永久链接至标题">¶</a></h3>
-<dl class="function">
+<div class="section" id="value-printer">
+<h3>value_printer<a class="headerlink" href="#value-printer" title="永久链接至标题">¶</a></h3>
+<dl class="class">
 <dt>
-<code class="descclassname">paddle.trainer_config_helpers.evaluators.</code><code class="descname">value_printer_evaluator</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.evaluator.</code><code class="descname">value_printer</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
 <dd><p>This Evaluator is used to print the values of input layers. It contains
 one or more input layers.</p>
 <p>The simple usage is:</p>
@@ -700,7 +668,7 @@ one or more input layers.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input</strong> (<em>LayerOutput|list</em>) &#8211; One or more input layers.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer|list</em>) &#8211; One or more input layers.</li>
 <li><strong>name</strong> (<em>None|basestring</em>) &#8211; Evaluator name.</li>
 </ul>
 </td>
@@ -718,6 +686,15 @@ one or more input layers.</p>
           </div>
           <footer>
   
+    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
+      
+        <a href="optimizer.html" class="btn btn-neutral float-right" title="Optimizer" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
+      
+      
+        <a href="layer.html" class="btn btn-neutral" title="Layers" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
+      
+    </div>
+  
 
   <hr/>
 
diff --git a/develop/doc_cn/api/v2/config/layer.html b/develop/doc_cn/api/v2/config/layer.html
index 96cd07503ea4f7d9bb442b3daa700f69890d51a5..001a68d567c85e139b9ce270ab3d40e060886be1 100644
--- a/develop/doc_cn/api/v2/config/layer.html
+++ b/develop/doc_cn/api/v2/config/layer.html
@@ -34,7 +34,7 @@
         <link rel="search" title="搜索" href="../../../search.html"/>
     <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/>
         <link rel="up" title="Model Configuration" href="../model_configs.html"/>
-        <link rel="next" title="Optimizer" href="optimizer.html"/>
+        <link rel="next" title="Evaluators" href="evaluators.html"/>
         <link rel="prev" title="Activation" href="activation.html"/> 
 
   <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
@@ -145,6 +145,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">模型配置</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
@@ -1460,10 +1461,10 @@ Each inputs is a projection or operator.</p>
 <li><strong>input</strong> &#8211; inputs layer. It is an optional parameter. If set,
 then this function will just return layer&#8217;s name.</li>
 <li><strong>act</strong> (<em>BaseActivation</em>) &#8211; Activation Type.</li>
-<li><strong>bias_attr</strong> (<a class="reference internal" href="../../v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute" title="paddle.trainer_config_helpers.attrs.ParameterAttribute"><em>ParameterAttribute</em></a><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
+<li><strong>bias_attr</strong> (<em>ParameterAttribute</em><em> or </em><em>None</em><em> or </em><em>bool</em>) &#8211; The Bias Attribute. If no bias, then pass False or
 something not type of ParameterAttribute. None will get a
 default Bias.</li>
-<li><strong>layer_attr</strong> (<a class="reference internal" href="../../v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute" title="paddle.trainer_config_helpers.attrs.ExtraLayerAttribute"><em>ExtraLayerAttribute</em></a>) &#8211; The extra layer config. Default is None.</li>
+<li><strong>layer_attr</strong> (<em>ExtraLayerAttribute</em>) &#8211; The extra layer config. Default is None.</li>
 </ul>
 </td>
 </tr>
@@ -3477,7 +3478,7 @@ It is used by recurrent layer group.</p>
   
     <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
       
-        <a href="optimizer.html" class="btn btn-neutral float-right" title="Optimizer" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
+        <a href="evaluators.html" class="btn btn-neutral float-right" title="Evaluators" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
       
       
         <a href="activation.html" class="btn btn-neutral" title="Activation" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
diff --git a/develop/doc_cn/api/v2/config/networks.html b/develop/doc_cn/api/v2/config/networks.html
index da2f72d51abac4773716d342e66d6108383a8346..af925f1c40f3b1fb95c17fe2b2023be3ab782e24 100644
--- a/develop/doc_cn/api/v2/config/networks.html
+++ b/develop/doc_cn/api/v2/config/networks.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">模型配置</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Networks</a></li>
@@ -416,6 +417,9 @@ False if no bias.</li>
 </table>
 </dd></dl>
 
+</div>
+<div class="section" id="small-vgg">
+<h3>small_vgg<a class="headerlink" href="#small-vgg" title="永久链接至标题">¶</a></h3>
 </div>
 <div class="section" id="vgg-16-network">
 <h3>vgg_16_network<a class="headerlink" href="#vgg-16-network" title="永久链接至标题">¶</a></h3>
@@ -798,6 +802,86 @@ gru_group, and gru_group is relatively better than simple_gru.</p>
 </table>
 </dd></dl>
 
+</div>
+<div class="section" id="simple-gru2">
+<h4>simple_gru2<a class="headerlink" href="#simple-gru2" title="永久链接至标题">¶</a></h4>
+<dl class="class">
+<dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.networks.</code><code class="descname">simple_gru2</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<dd><p>simple_gru2 is the same with simple_gru, but using grumemory instead
+Please see grumemory in layers.py for more detail about the maths.
+simple_gru2 is faster than simple_gru.</p>
+<p>The example usage is:</p>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">gru</span> <span class="o">=</span> <span class="n">simple_gru2</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">layer1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">256</span><span class="p">)</span>
+</pre></div>
+</div>
+<table class="docutils field-list" frame="void" rules="none">
+<col class="field-name" />
+<col class="field-body" />
+<tbody valign="top">
+<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; input layer name.</li>
+<li><strong>name</strong> (<em>basestring</em>) &#8211; name of the gru group.</li>
+<li><strong>size</strong> (<em>int</em>) &#8211; hidden size of the gru.</li>
+<li><strong>reverse</strong> (<em>bool</em>) &#8211; whether to process the input data in a reverse order</li>
+<li><strong>act</strong> (<em>paddle.v2.Activation.Base</em>) &#8211; type of the activiation</li>
+<li><strong>gate_act</strong> (<em>paddle.v2.Activation.Base</em>) &#8211; type of the gate activiation</li>
+<li><strong>gru_bias_attr</strong> (<em>paddle.v2.attr.ParameterAttribute|False</em>) &#8211; bias. False means no bias, None means default bias.</li>
+<li><strong>gru_attr</strong> (<em>paddle.v2.attr.ParameterAttribute|False</em>) &#8211; Extra parameter attribute of the gru layer.</li>
+</ul>
+</td>
+</tr>
+<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">the gru group.</p>
+</td>
+</tr>
+<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">paddle.v2.config_base.Layer</p>
+</td>
+</tr>
+</tbody>
+</table>
+</dd></dl>
+
+</div>
+<div class="section" id="bidirectional-gru">
+<h4>bidirectional_gru<a class="headerlink" href="#bidirectional-gru" title="永久链接至标题">¶</a></h4>
+<dl class="class">
+<dt>
+<em class="property">class </em><code class="descclassname">paddle.v2.networks.</code><code class="descname">bidirectional_gru</code><span class="sig-paren">(</span><em>*args</em>, <em>**kwargs</em><span class="sig-paren">)</span></dt>
+<dd><p>A bidirectional_gru is a recurrent unit that iterates over the input
+sequence both in forward and bardward orders, and then concatenate two
+outputs to form a final output. However, concatenation of two outputs
+is not the only way to form the final output, you can also, for example,
+just add them together.</p>
+<p>The example usage is:</p>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">bi_gru</span> <span class="o">=</span> <span class="n">bidirectional_gru</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="p">[</span><span class="n">input1</span><span class="p">],</span> <span class="n">size</span><span class="o">=</span><span class="mi">512</span><span class="p">)</span>
+</pre></div>
+</div>
+<table class="docutils field-list" frame="void" rules="none">
+<col class="field-name" />
+<col class="field-body" />
+<tbody valign="top">
+<tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
+<li><strong>name</strong> (<em>basestring</em>) &#8211; bidirectional gru layer name.</li>
+<li><strong>input</strong> (<em>paddle.v2.config_base.Layer</em>) &#8211; input layer.</li>
+<li><strong>size</strong> (<em>int</em>) &#8211; gru layer size.</li>
+<li><strong>return_seq</strong> (<em>bool</em>) &#8211; If set False, outputs of the last time step are
+concatenated and returned.
+If set True, the entire output sequences that are
+processed in forward and backward directions are
+concatenated and returned.</li>
+</ul>
+</td>
+</tr>
+<tr class="field-even field"><th class="field-name">返回:</th><td class="field-body"><p class="first">paddle.v2.config_base.Layer object.</p>
+</td>
+</tr>
+<tr class="field-odd field"><th class="field-name">返回类型:</th><td class="field-body"><p class="first last">paddle.v2.config_base.Layer</p>
+</td>
+</tr>
+</tbody>
+</table>
+</dd></dl>
+
 </div>
 </div>
 <div class="section" id="simple-attention">
diff --git a/develop/doc_cn/api/v2/config/optimizer.html b/develop/doc_cn/api/v2/config/optimizer.html
index 9d26df3c59e6154336047a84d515607defafbc96..e8c79fd18c871fbe9673c37dcadf7238a2121a7f 100644
--- a/develop/doc_cn/api/v2/config/optimizer.html
+++ b/develop/doc_cn/api/v2/config/optimizer.html
@@ -35,7 +35,7 @@
     <link rel="top" title="PaddlePaddle  文档" href="../../../index.html"/>
         <link rel="up" title="Model Configuration" href="../model_configs.html"/>
         <link rel="next" title="Pooling" href="pooling.html"/>
-        <link rel="prev" title="Layers" href="layer.html"/> 
+        <link rel="prev" title="Evaluators" href="evaluators.html"/> 
 
   <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
   <link rel="stylesheet" href="../../../_static/css/override.css" type="text/css" />
@@ -145,6 +145,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">模型配置</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
@@ -374,7 +375,7 @@ w &amp; = w - \frac{\eta} {\sqrt{v(w,t) + \epsilon}} \nabla Q_{i}(w)\end{split}\
         <a href="pooling.html" class="btn btn-neutral float-right" title="Pooling" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
       
       
-        <a href="layer.html" class="btn btn-neutral" title="Layers" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
+        <a href="evaluators.html" class="btn btn-neutral" title="Evaluators" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
       
     </div>
   
diff --git a/develop/doc_cn/api/v2/config/pooling.html b/develop/doc_cn/api/v2/config/pooling.html
index 62d8d2357550cb6244b867712ad14f396e31a49f..2aec9dc797f04093dce38afe7c6125b6b1d5ad14 100644
--- a/develop/doc_cn/api/v2/config/pooling.html
+++ b/develop/doc_cn/api/v2/config/pooling.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2 current"><a class="reference internal" href="../model_configs.html">模型配置</a><ul class="current">
 <li class="toctree-l3"><a class="reference internal" href="activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="optimizer.html">Optimizer</a></li>
 <li class="toctree-l3 current"><a class="current reference internal" href="#">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v2/data.html b/develop/doc_cn/api/v2/data.html
index efecb67a7960c1567b30bb7df0f2e6f9f6064997..68ef31a40882c1fb6a32252ceeb22c7d2521d217 100644
--- a/develop/doc_cn/api/v2/data.html
+++ b/develop/doc_cn/api/v2/data.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v2/model_configs.html b/develop/doc_cn/api/v2/model_configs.html
index 9107f8065424c291de2b28c9c4920074abec8d28..068d18ef794e1de3b53d1f6a89443141dfcd7cb6 100644
--- a/develop/doc_cn/api/v2/model_configs.html
+++ b/develop/doc_cn/api/v2/model_configs.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2 current"><a class="current reference internal" href="#">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/networks.html">Networks</a></li>
@@ -193,6 +194,7 @@
 <ul>
 <li class="toctree-l1"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l1"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l1"><a class="reference internal" href="config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/api/v2/run_logic.html b/develop/doc_cn/api/v2/run_logic.html
index 285e5b43501fb54a50970e7222ccdbab1075389f..67f8687c1ff6bc924500c4d46e11a0896615f7f1 100644
--- a/develop/doc_cn/api/v2/run_logic.html
+++ b/develop/doc_cn/api/v2/run_logic.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="config/networks.html">Networks</a></li>
@@ -501,8 +502,9 @@ index that reader returns.</li>
 <dd><p>Infer a neural network by given neural network output and parameters.  The
 user should pass either a batch of input data or reader method.</p>
 <p>Example usages:</p>
-<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">result</span> <span class="o">=</span> <span class="n">paddle</span><span class="o">.</span><span class="n">infer</span><span class="p">(</span><span class="n">prediction</span><span class="p">,</span> <span class="n">parameters</span><span class="p">,</span> <span class="nb">input</span><span class="o">=</span><span class="n">SomeData</span><span class="p">,</span>
-                      <span class="n">batch_size</span><span class="o">=</span><span class="mi">32</span><span class="p">)</span>
+<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">result</span> <span class="o">=</span> <span class="n">paddle</span><span class="o">.</span><span class="n">infer</span><span class="p">(</span><span class="n">outptut_layer</span><span class="o">=</span><span class="n">prediction</span><span class="p">,</span>
+                      <span class="n">parameters</span><span class="o">=</span><span class="n">parameters</span><span class="p">,</span>
+                      <span class="nb">input</span><span class="o">=</span><span class="n">SomeData</span><span class="p">)</span>
 <span class="k">print</span> <span class="n">result</span>
 </pre></div>
 </div>
diff --git a/develop/doc_cn/design/api.html b/develop/doc_cn/design/api.html
index 4389aa254711431b8b98acf016745387525c70a5..7412ca9529d435ba137bddcffc245f86f3902cc3 100644
--- a/develop/doc_cn/design/api.html
+++ b/develop/doc_cn/design/api.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/design/cluster_train/README.html b/develop/doc_cn/design/cluster_train/README.html
index 763a365b875d8561ee0fbaedd3632022caaeec04..f68554d2c92f0d20f0af9d618270eb31a0732c50 100644
--- a/develop/doc_cn/design/cluster_train/README.html
+++ b/develop/doc_cn/design/cluster_train/README.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/design/cluster_train/checkpointing.html b/develop/doc_cn/design/cluster_train/checkpointing.html
index 4990569e49cdd208828d1959585cffd24f686348..35ea191ba8e2fcc2dbecef2038164c0d6ee72d1b 100644
--- a/develop/doc_cn/design/cluster_train/checkpointing.html
+++ b/develop/doc_cn/design/cluster_train/checkpointing.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/design/cluster_train/data_dispatch.html b/develop/doc_cn/design/cluster_train/data_dispatch.html
index 4ba7272e0fed85e4d34a8d5cf14a26cece7d9c5a..8e28936760adedc71c8c158f62078ba2d624f8dd 100644
--- a/develop/doc_cn/design/cluster_train/data_dispatch.html
+++ b/develop/doc_cn/design/cluster_train/data_dispatch.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/design/multi_language_interface/00.why_plain_c.html b/develop/doc_cn/design/multi_language_interface/00.why_plain_c.html
index b22828637163f28da052e9297271d31377d005c0..481158360f0f3f0e8abdbe9509ba81262f6a51fc 100644
--- a/develop/doc_cn/design/multi_language_interface/00.why_plain_c.html
+++ b/develop/doc_cn/design/multi_language_interface/00.why_plain_c.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/design/multi_language_interface/01.inference_implementation.html b/develop/doc_cn/design/multi_language_interface/01.inference_implementation.html
index 8fc26f3255c266e1c8c0e4662ce2fb8c1cd0efba..ce4415572fc3ec809639931b074179d160a3efde 100644
--- a/develop/doc_cn/design/multi_language_interface/01.inference_implementation.html
+++ b/develop/doc_cn/design/multi_language_interface/01.inference_implementation.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/design/reader/README.html b/develop/doc_cn/design/reader/README.html
index 6ab3ec5910f6791a8926dd278729d6b1ca804c34..a5b7bc4b986dc3e490ba20662b0aeb41161ebd08 100644
--- a/develop/doc_cn/design/reader/README.html
+++ b/develop/doc_cn/design/reader/README.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/design/releasing_process.html b/develop/doc_cn/design/releasing_process.html
index 157b51a257987169eeb3abdd3edff5abe18dde06..89ca1994b3e74edf22c9e449627e934edb0deb37 100644
--- a/develop/doc_cn/design/releasing_process.html
+++ b/develop/doc_cn/design/releasing_process.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/faq/index_cn.html b/develop/doc_cn/faq/index_cn.html
index 52843289064cb07f5c8e3dea4c742f9f5f59bb25..a5e3b89038de098b33ab94c82ed8b4c4f2486c29 100644
--- a/develop/doc_cn/faq/index_cn.html
+++ b/develop/doc_cn/faq/index_cn.html
@@ -143,6 +143,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/genindex.html b/develop/doc_cn/genindex.html
index c1299f2b9ce3eadc00aff18663c46c2cb33dd002..9943b0fb35f31ba9445a57f944e3d0353e8afc51 100644
--- a/develop/doc_cn/genindex.html
+++ b/develop/doc_cn/genindex.html
@@ -143,6 +143,7 @@
 <li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
@@ -187,56 +188,8 @@
 <h1 id="index">索引</h1>
 
 <div class="genindex-jumpbox">
- <a href="#D"><strong>D</strong></a>
- | <a href="#E"><strong>E</strong></a>
- | <a href="#P"><strong>P</strong></a>
- | <a href="#S"><strong>S</strong></a>
  
 </div>
-<h2 id="D">D</h2>
-<table style="width: 100%" class="indextable genindextable"><tr>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/data_sources.html#paddle.trainer_config_helpers.data_sources.define_py_data_sources2">define_py_data_sources2() (在 paddle.trainer_config_helpers.data_sources 模块中)</a>
-</li>
-  </ul></td>
-</tr></table>
-
-<h2 id="E">E</h2>
-<table style="width: 100%" class="indextable genindextable"><tr>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ExtraAttr">ExtraAttr() (在 paddle.trainer_config_helpers.attrs 模块中)</a>
-</li>
-  </ul></td>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ExtraLayerAttribute">ExtraLayerAttribute (paddle.trainer_config_helpers.attrs 中的类)</a>
-</li>
-  </ul></td>
-</tr></table>
-
-<h2 id="P">P</h2>
-<table style="width: 100%" class="indextable genindextable"><tr>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#module-paddle.trainer_config_helpers.attrs">paddle.trainer_config_helpers.attrs (模块)</a>
-</li>
-      <li><a href="api/v1/trainer_config_helpers/data_sources.html#module-paddle.trainer_config_helpers.data_sources">paddle.trainer_config_helpers.data_sources (模块)</a>
-</li>
-  </ul></td>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParamAttr">ParamAttr() (在 paddle.trainer_config_helpers.attrs 模块中)</a>
-</li>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute">ParameterAttribute (paddle.trainer_config_helpers.attrs 中的类)</a>
-</li>
-  </ul></td>
-</tr></table>
-
-<h2 id="S">S</h2>
-<table style="width: 100%" class="indextable genindextable"><tr>
-  <td style="width: 33%; vertical-align: top;"><ul>
-      <li><a href="api/v1/trainer_config_helpers/attrs.html#paddle.trainer_config_helpers.attrs.ParameterAttribute.set_default_parameter_name">set_default_parameter_name() (paddle.trainer_config_helpers.attrs.ParameterAttribute 方法)</a>
-</li>
-  </ul></td>
-</tr></table>
-
 
 
            </div>
diff --git a/develop/doc_cn/getstarted/basic_usage/index_cn.html b/develop/doc_cn/getstarted/basic_usage/index_cn.html
index e7c58b94a730c39e7bd49bd6e6c734850dbd0907..101324f34c824551c281f8663220fae2d0f85cfa 100644
--- a/develop/doc_cn/getstarted/basic_usage/index_cn.html
+++ b/develop/doc_cn/getstarted/basic_usage/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/getstarted/build_and_install/cmake/build_from_source_cn.html b/develop/doc_cn/getstarted/build_and_install/cmake/build_from_source_cn.html
index 0cba4401dcd181dd17b2256d60a142f0c81e4dbe..5f03493ec84e2b38228c2cb9a458268d4e36f23d 100644
--- a/develop/doc_cn/getstarted/build_and_install/cmake/build_from_source_cn.html
+++ b/develop/doc_cn/getstarted/build_and_install/cmake/build_from_source_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html b/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html
index 18062713bd1c866031b04ccfc2e3e6c1f74be591..b796d486ea77da9ef4c6d8d0138650721f2b4f0f 100644
--- a/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html
+++ b/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/getstarted/build_and_install/index_cn.html b/develop/doc_cn/getstarted/build_and_install/index_cn.html
index fa306e564267091fcc63ed5dd4672a77d4c2cb03..8e6dcd7a8abb6d2903460de7aab3b8e040845e8f 100644
--- a/develop/doc_cn/getstarted/build_and_install/index_cn.html
+++ b/develop/doc_cn/getstarted/build_and_install/index_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/getstarted/build_and_install/ubuntu_install_cn.html b/develop/doc_cn/getstarted/build_and_install/ubuntu_install_cn.html
index 8798b6d307af0dea04ccf2bef1c2cc186b715ee1..702fe3c4a0b4ae0e0e4f5fe441d41a1aa177c817 100644
--- a/develop/doc_cn/getstarted/build_and_install/ubuntu_install_cn.html
+++ b/develop/doc_cn/getstarted/build_and_install/ubuntu_install_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/getstarted/index_cn.html b/develop/doc_cn/getstarted/index_cn.html
index f7c8225eb98ba45829fdd82c53d852016560c5a6..4e0d37982f6474e7dbebe958ac32690147028a36 100644
--- a/develop/doc_cn/getstarted/index_cn.html
+++ b/develop/doc_cn/getstarted/index_cn.html
@@ -144,6 +144,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/deep_model/rnn/hierarchical_layer_cn.html b/develop/doc_cn/howto/deep_model/rnn/hierarchical_layer_cn.html
index 908ac6c88676a8686e81e4d9dcbee3539ec08597..0a48366d8a522c91cf4e38e67a70ed16967752b0 100644
--- a/develop/doc_cn/howto/deep_model/rnn/hierarchical_layer_cn.html
+++ b/develop/doc_cn/howto/deep_model/rnn/hierarchical_layer_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html b/develop/doc_cn/howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html
index 5a5845408f30c407d0bb1676a2870ebcdcbba54f..7b3694ebcb2ebabbd1efa82e683b98b219031729 100644
--- a/develop/doc_cn/howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html
+++ b/develop/doc_cn/howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/deep_model/rnn/index_cn.html b/develop/doc_cn/howto/deep_model/rnn/index_cn.html
index 270a355093bdced7bc5e45bdf333dd2168d7401f..d5d42bbfcaa79a4b0347c79f204255f80bc550d0 100644
--- a/develop/doc_cn/howto/deep_model/rnn/index_cn.html
+++ b/develop/doc_cn/howto/deep_model/rnn/index_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/deep_model/rnn/recurrent_group_cn.html b/develop/doc_cn/howto/deep_model/rnn/recurrent_group_cn.html
index 6af24347010d9182708169d98fe75b2d05958051..9b31daa719e4780a5e8edb8c856909a28ecff803 100644
--- a/develop/doc_cn/howto/deep_model/rnn/recurrent_group_cn.html
+++ b/develop/doc_cn/howto/deep_model/rnn/recurrent_group_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/deep_model/rnn/rnn_config_cn.html b/develop/doc_cn/howto/deep_model/rnn/rnn_config_cn.html
index 1d9d90c9559dc009adcee869e6e5e228830324ab..c5ee29ad2d12fa58153a83a8108205903162bb70 100644
--- a/develop/doc_cn/howto/deep_model/rnn/rnn_config_cn.html
+++ b/develop/doc_cn/howto/deep_model/rnn/rnn_config_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
@@ -286,7 +287,7 @@ layer(引导层)</strong>，其输出被用作Memory的初始值。
 用于计算编码向量的加权和。加权和用来生成 <span class="math">\(y_t\)</span> 。</p>
 <p>模型的编码器部分如下所示。它叫做<code class="docutils literal"><span class="pre">grumemory</span></code>来表示门控循环神经网络。如果网络架构简单，那么推荐使用循环神经网络的方法，因为它比
 <code class="docutils literal"><span class="pre">recurrent_group</span></code>
-更快。我们已经实现了大多数常用的循环神经网络架构，可以参考 <a class="reference internal" href="../../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers"><span class="std std-ref">Layers</span></a> 了解更多细节。</p>
+更快。我们已经实现了大多数常用的循环神经网络架构，可以参考 <span class="xref std std-ref">api_trainer_config_helpers_layers</span> 了解更多细节。</p>
 <p>我们还将编码向量投射到 <code class="docutils literal"><span class="pre">decoder_size</span></code>
 维空间。这通过获得反向循环网络的第一个实例，并将其投射到
 <code class="docutils literal"><span class="pre">decoder_size</span></code> 维空间完成：</p>
diff --git a/develop/doc_cn/howto/dev/contribute_to_paddle_cn.html b/develop/doc_cn/howto/dev/contribute_to_paddle_cn.html
index c0f85b9c516cdc6a583f1f3dcc83b365d694d1bc..26d6af6e62e5d77e0342ef9698eb57fe2808727a 100644
--- a/develop/doc_cn/howto/dev/contribute_to_paddle_cn.html
+++ b/develop/doc_cn/howto/dev/contribute_to_paddle_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/dev/new_layer_cn.html b/develop/doc_cn/howto/dev/new_layer_cn.html
index 7fda20968f71dc6cc34f6f5ac2bfb95cc277e4ee..b4f89413c633cee81f32f3397038e4b957bc6ff3 100644
--- a/develop/doc_cn/howto/dev/new_layer_cn.html
+++ b/develop/doc_cn/howto/dev/new_layer_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/dev/write_docs_cn.html b/develop/doc_cn/howto/dev/write_docs_cn.html
index 5a24c34d2d135f72dcef7f32fc5fb622e7317efe..9975949e136c274318d65babc7d8c07ef08d9599 100644
--- a/develop/doc_cn/howto/dev/write_docs_cn.html
+++ b/develop/doc_cn/howto/dev/write_docs_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/index_cn.html b/develop/doc_cn/howto/index_cn.html
index f9b6837c3fed9e174ce38d2988f81f918636c12a..6459e5258e3eecca3b8c52ffd1fa4c62d454bfdd 100644
--- a/develop/doc_cn/howto/index_cn.html
+++ b/develop/doc_cn/howto/index_cn.html
@@ -144,6 +144,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/optimization/gpu_profiling_cn.html b/develop/doc_cn/howto/optimization/gpu_profiling_cn.html
index c986ce591b3b83936907338f48295cfc721af53e..c5e22d4d8f330b770db936a17ca76a77322223bc 100644
--- a/develop/doc_cn/howto/optimization/gpu_profiling_cn.html
+++ b/develop/doc_cn/howto/optimization/gpu_profiling_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/cluster/cluster_train_cn.html b/develop/doc_cn/howto/usage/cluster/cluster_train_cn.html
index 0dbaf4a743ac1f220ca04ab66f2cbc8efa98b873..815031235d3ac586083c630cadab497241f3df33 100644
--- a/develop/doc_cn/howto/usage/cluster/cluster_train_cn.html
+++ b/develop/doc_cn/howto/usage/cluster/cluster_train_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/cmd_parameter/arguments_cn.html b/develop/doc_cn/howto/usage/cmd_parameter/arguments_cn.html
index 0742dbaa3eceea49e886aa0453369ee51640ac47..6511d20c7bccbb63e470c9ab1403ddbc4e134aa3 100644
--- a/develop/doc_cn/howto/usage/cmd_parameter/arguments_cn.html
+++ b/develop/doc_cn/howto/usage/cmd_parameter/arguments_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/cmd_parameter/detail_introduction_cn.html b/develop/doc_cn/howto/usage/cmd_parameter/detail_introduction_cn.html
index 6add093d6e5c11993055dfd7bb03bc23d06dc9ad..7c8fe4aaa740b52355feb8f42ad34ee87ef06877 100644
--- a/develop/doc_cn/howto/usage/cmd_parameter/detail_introduction_cn.html
+++ b/develop/doc_cn/howto/usage/cmd_parameter/detail_introduction_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/cmd_parameter/index_cn.html b/develop/doc_cn/howto/usage/cmd_parameter/index_cn.html
index 1e6b6aa93d2fe66bedbc0b9b0bde8a74dd9221bb..3ef9a2f60f8f1cdb9ac87fbdd35c2b622e8f0c44 100644
--- a/develop/doc_cn/howto/usage/cmd_parameter/index_cn.html
+++ b/develop/doc_cn/howto/usage/cmd_parameter/index_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/cmd_parameter/use_case_cn.html b/develop/doc_cn/howto/usage/cmd_parameter/use_case_cn.html
index b34f7943b9aa2c65f742761d6f5292ac2c1a5e4f..7ab8f978daaa8888be0d768c08b5667b62bc9501 100644
--- a/develop/doc_cn/howto/usage/cmd_parameter/use_case_cn.html
+++ b/develop/doc_cn/howto/usage/cmd_parameter/use_case_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/concepts/use_concepts_cn.html b/develop/doc_cn/howto/usage/concepts/use_concepts_cn.html
index 7ce2e795bbb48bbd8dcbecbba5ef74a8aae55a43..66b04abbe64814bdc01c384c84eaf88a32b21604 100644
--- a/develop/doc_cn/howto/usage/concepts/use_concepts_cn.html
+++ b/develop/doc_cn/howto/usage/concepts/use_concepts_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
@@ -304,7 +305,7 @@
 </div>
 <div class="section" id="id6">
 <h3><a class="toc-backref" href="#id14">优化算法配置</a><a class="headerlink" href="#id6" title="永久链接至标题">¶</a></h3>
-<p>通过 <a class="reference internal" href="../../../api/v1/trainer_config_helpers/optimizers.html#api-trainer-config-helpers-optimizers-settings"><span class="std std-ref">settings</span></a> 接口设置神经网络所使用的训练参数和 <a class="reference internal" href="../../../api/v1/trainer_config_helpers/optimizers.html#api-trainer-config-helpers-optimizers"><span class="std std-ref">Optimizers</span></a> ，包括学习率、batch_size、优化算法、正则方法等，具体的使用方法请参考 <a class="reference internal" href="../../../api/v1/trainer_config_helpers/optimizers.html#api-trainer-config-helpers-optimizers-settings"><span class="std std-ref">settings</span></a> 文档。</p>
+<p>通过 <span class="xref std std-ref">api_trainer_config_helpers_optimizers_settings</span> 接口设置神经网络所使用的训练参数和 <span class="xref std std-ref">api_trainer_config_helpers_optimizers</span> ，包括学习率、batch_size、优化算法、正则方法等，具体的使用方法请参考 <span class="xref std std-ref">api_trainer_config_helpers_optimizers_settings</span> 文档。</p>
 </div>
 <div class="section" id="id7">
 <h3><a class="toc-backref" href="#id15">网络结构配置</a><a class="headerlink" href="#id7" title="永久链接至标题">¶</a></h3>
@@ -319,13 +320,13 @@
 </ul>
 <p>这个配置文件网络由 <code class="docutils literal"><span class="pre">data_layer</span></code> 、 <code class="docutils literal"><span class="pre">simple_img_conv_pool</span></code> 、 <code class="docutils literal"><span class="pre">fc_layer</span></code> 组成。</p>
 <ul class="simple">
-<li><a class="reference internal" href="../../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-data-layer"><span class="std std-ref">data_layer</span></a>  ： 通常每个配置文件都会包括 <code class="docutils literal"><span class="pre">data_layer</span></code> ，定义输入数据大小。</li>
+<li><span class="xref std std-ref">api_trainer_config_helpers_layers_data_layer</span>  ： 通常每个配置文件都会包括 <code class="docutils literal"><span class="pre">data_layer</span></code> ，定义输入数据大小。</li>
 <li><a class="reference internal" href="../../../api/v2/config/networks.html#api-trainer-config-helpers-network-simple-img-conv-pool"><span class="std std-ref">simple_img_conv_pool</span></a> ：是一个组合层，包括了图像的卷积 (convolution)和池化(pooling)。</li>
-<li><a class="reference internal" href="../../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-fc-layer"><span class="std std-ref">fc_layer</span></a> ：全连接层，激活函数为Softmax，这里也可叫分类层。</li>
+<li><span class="xref std std-ref">api_trainer_config_helpers_layers_fc_layer</span> ：全连接层，激活函数为Softmax，这里也可叫分类层。</li>
 </ul>
 </li>
 <li><p class="first">损失函数和评估器：损失函数即为网络的优化目标，评估器可以评价模型结果。</p>
-<p>PaddlePaddle包括很多损失函数和评估起，详细可以参考 <a class="reference internal" href="../../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-cost-layers"><span class="std std-ref">Cost Layers</span></a> 和 <a class="reference internal" href="../../../api/v1/trainer_config_helpers/evaluators.html#api-trainer-config-helpers-evaluators"><span class="std std-ref">Evaluators</span></a> 。这里 <code class="docutils literal"><span class="pre">classification_cost</span></code> 默认使用多类交叉熵损失函数和分类错误率统计评估器。</p>
+<p>PaddlePaddle包括很多损失函数和评估起，详细可以参考 <span class="xref std std-ref">api_trainer_config_helpers_layers_cost_layers</span> 和 <span class="xref std std-ref">api_trainer_config_helpers_evaluators</span> 。这里 <code class="docutils literal"><span class="pre">classification_cost</span></code> 默认使用多类交叉熵损失函数和分类错误率统计评估器。</p>
 </li>
 <li><p class="first"><code class="docutils literal"><span class="pre">outputs</span></code>: 标记网络输出的函数为 <code class="docutils literal"><span class="pre">outputs</span></code> 。</p>
 <p>训练阶段，网络的输出为神经网络的优化目标；预测阶段，网络的输出也可通过 <code class="docutils literal"><span class="pre">outputs</span></code> 标记。</p>
@@ -338,7 +339,7 @@
     <span class="n">out</span> <span class="o">+=</span> <span class="n">full_matrix_projection</span><span class="p">(</span><span class="nb">input</span><span class="o">=</span><span class="n">data</span><span class="p">)</span>
 </pre></div>
 </div>
-<p>PaddlePaddle 可以使用 <code class="docutils literal"><span class="pre">mixed</span> <span class="pre">layer</span></code> 配置出非常复杂的网络，甚至可以直接配置一个完整的LSTM。用户可以参考 <a class="reference internal" href="../../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-mixed-layer"><span class="std std-ref">mixed_layer</span></a> 的相关文档进行配置。</p>
+<p>PaddlePaddle 可以使用 <code class="docutils literal"><span class="pre">mixed</span> <span class="pre">layer</span></code> 配置出非常复杂的网络，甚至可以直接配置一个完整的LSTM。用户可以参考 <span class="xref std std-ref">api_trainer_config_helpers_layers_mixed_layer</span> 的相关文档进行配置。</p>
 </div>
 </div>
 <div class="section" id="id8">
diff --git a/develop/doc_cn/howto/usage/k8s/k8s_basis_cn.html b/develop/doc_cn/howto/usage/k8s/k8s_basis_cn.html
index ec52162f8df287a9dc933dc121ecbb36c086762d..1aaa7ddcad8f7e3963ff1cc8f79fd952adb9fa95 100644
--- a/develop/doc_cn/howto/usage/k8s/k8s_basis_cn.html
+++ b/develop/doc_cn/howto/usage/k8s/k8s_basis_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/k8s/k8s_cn.html b/develop/doc_cn/howto/usage/k8s/k8s_cn.html
index 92b4ac34a9b85215e9a5c1ee0c696556133f15e5..a906f2630ab649412cd3a3c41884d1e704425f38 100644
--- a/develop/doc_cn/howto/usage/k8s/k8s_cn.html
+++ b/develop/doc_cn/howto/usage/k8s/k8s_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/k8s/k8s_distributed_cn.html b/develop/doc_cn/howto/usage/k8s/k8s_distributed_cn.html
index dbae3cb25eaf0b7e069c6002240884c1ea7c9e80..dbca32984919fee012afd6f5de00575aa842b33c 100644
--- a/develop/doc_cn/howto/usage/k8s/k8s_distributed_cn.html
+++ b/develop/doc_cn/howto/usage/k8s/k8s_distributed_cn.html
@@ -145,6 +145,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/k8s/src/k8s_data/README.html b/develop/doc_cn/howto/usage/k8s/src/k8s_data/README.html
index d6b9d2c27b6221f0a5682e1a796be51d1079c04d..4f94b49e33619d43da2817dc29db84290a9eed05 100644
--- a/develop/doc_cn/howto/usage/k8s/src/k8s_data/README.html
+++ b/develop/doc_cn/howto/usage/k8s/src/k8s_data/README.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/howto/usage/k8s/src/k8s_train/README.html b/develop/doc_cn/howto/usage/k8s/src/k8s_train/README.html
index f07f50b1a17a62e05f5c616c42499dde83556ecd..276e87e7df48df430bfd9173bc79f4f9675508b6 100644
--- a/develop/doc_cn/howto/usage/k8s/src/k8s_train/README.html
+++ b/develop/doc_cn/howto/usage/k8s/src/k8s_train/README.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../../../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../../../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/index_cn.html b/develop/doc_cn/index_cn.html
index 470b5639a9823920f69b53c2eadadd174cf5b5ff..2c1623fd62c66581bc37780f42a69b1b98d7a8d7 100644
--- a/develop/doc_cn/index_cn.html
+++ b/develop/doc_cn/index_cn.html
@@ -143,6 +143,7 @@
 <li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/objects.inv b/develop/doc_cn/objects.inv
index f1d912ca4bb602ad1bd0f0194621302c8bc8e566..d69d70beb920445fab874079da0f565e874247bd 100644
Binary files a/develop/doc_cn/objects.inv and b/develop/doc_cn/objects.inv differ
diff --git a/develop/doc_cn/py-modindex.html b/develop/doc_cn/py-modindex.html
deleted file mode 100644
index ae568bb2b1ffd3e9eaf7160a9260f27433a8d047..0000000000000000000000000000000000000000
--- a/develop/doc_cn/py-modindex.html
+++ /dev/null
@@ -1,274 +0,0 @@
-
-
-<!DOCTYPE html>
-<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
-<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
-<head>
-  <meta charset="utf-8">
-  
-  <meta name="viewport" content="width=device-width, initial-scale=1.0">
-  
-  <title>Python 模块索引 &mdash; PaddlePaddle  文档</title>
-  
-
-  
-  
-
-  
-
-  
-  
-    
-
-  
-
-  
-  
-    <link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
-  
-
-  
-  
-        <link rel="index" title="索引"
-              href="genindex.html"/>
-        <link rel="search" title="搜索" href="search.html"/>
-    <link rel="top" title="PaddlePaddle  文档" href="index.html"/>
- 
-
-  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
-  <link rel="stylesheet" href="_static/css/override.css" type="text/css" />
-  <script>
-  var _hmt = _hmt || [];
-  (function() {
-    var hm = document.createElement("script");
-    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
-    var s = document.getElementsByTagName("script")[0]; 
-    s.parentNode.insertBefore(hm, s);
-  })();
-  </script>
-
-  
-
-
-
-  
-  <script src="_static/js/modernizr.min.js"></script>
-
-</head>
-
-<body class="wy-body-for-nav" role="document">
-
-  
-  <header class="site-header">
-    <div class="site-logo">
-      <a href="/"><img src="_static/images/PP_w.png"></a>
-    </div>
-    <div class="site-nav-links">
-      <div class="site-menu">
-        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Folk me on Github</a>
-        <div class="language-switcher dropdown">
-          <a type="button" data-toggle="dropdown">
-            <span>English</span>
-            <i class="fa fa-angle-up"></i>
-            <i class="fa fa-angle-down"></i>
-          </a>
-          <ul class="dropdown-menu">
-            <li><a href="/doc_cn">中文</a></li>
-            <li><a href="/doc">English</a></li>
-          </ul>
-        </div>
-        <ul class="site-page-links">
-          <li><a href="/">Home</a></li>
-        </ul>
-      </div>
-      <div class="doc-module">
-        
-        <ul>
-<li class="toctree-l1"><a class="reference internal" href="getstarted/index_cn.html">新手入门</a></li>
-<li class="toctree-l1"><a class="reference internal" href="howto/index_cn.html">进阶指南</a></li>
-<li class="toctree-l1"><a class="reference internal" href="api/index_cn.html">API</a></li>
-<li class="toctree-l1"><a class="reference internal" href="faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-<div role="search">
-  <form id="rtd-search-form" class="wy-form" action="search.html" method="get">
-    <input type="text" name="q" placeholder="Search docs" />
-    <input type="hidden" name="check_keywords" value="yes" />
-    <input type="hidden" name="area" value="default" />
-  </form>
-</div>        
-      </div>
-    </div>
-  </header>
-  
-  <div class="main-content-wrap">
-
-    
-    <nav class="doc-menu-vertical" role="navigation">
-        
-          
-          <ul>
-<li class="toctree-l1"><a class="reference internal" href="getstarted/index_cn.html">新手入门</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="getstarted/build_and_install/index_cn.html">安装与编译</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="getstarted/build_and_install/docker_install_cn.html">PaddlePaddle的Docker容器使用方式</a></li>
-<li class="toctree-l3"><a class="reference internal" href="getstarted/build_and_install/ubuntu_install_cn.html">Ubuntu部署PaddlePaddle</a></li>
-<li class="toctree-l3"><a class="reference internal" href="getstarted/build_and_install/cmake/build_from_source_cn.html">PaddlePaddle的编译选项</a></li>
-</ul>
-</li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="howto/index_cn.html">进阶指南</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/cmd_parameter/index_cn.html">设置命令行参数</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="howto/usage/cmd_parameter/use_case_cn.html">使用案例</a></li>
-<li class="toctree-l3"><a class="reference internal" href="howto/usage/cmd_parameter/arguments_cn.html">参数概述</a></li>
-<li class="toctree-l3"><a class="reference internal" href="howto/usage/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/concepts/use_concepts_cn.html">基本使用概念</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/cluster/cluster_train_cn.html">运行分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/k8s/k8s_basis_cn.html">Kubernetes 简介</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/k8s/k8s_cn.html">Kubernetes单机训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/usage/k8s/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/dev/write_docs_cn.html">如何贡献/修改文档</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
-<li class="toctree-l2"><a class="reference internal" href="howto/deep_model/rnn/index_cn.html">RNN相关模型</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="howto/deep_model/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
-<li class="toctree-l3"><a class="reference internal" href="howto/deep_model/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="howto/deep_model/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="howto/optimization/gpu_profiling_cn.html">GPU性能分析与调优</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="api/index_cn.html">API</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">模型配置</a><ul>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
-<li class="toctree-l3"><a class="reference internal" href="api/v2/config/attr.html">Parameter Attribute</a></li>
-</ul>
-</li>
-<li class="toctree-l2"><a class="reference internal" href="api/v2/data.html">数据访问</a></li>
-<li class="toctree-l2"><a class="reference internal" href="api/v2/run_logic.html">训练与应用</a></li>
-</ul>
-</li>
-<li class="toctree-l1"><a class="reference internal" href="faq/index_cn.html">FAQ</a></li>
-</ul>
-
-        
-    </nav>
-    
-    <section class="doc-content-wrap">
-
-      
-
-
-
-
-
-
-
-
-
-<div role="navigation" aria-label="breadcrumbs navigation">
-  <ul class="wy-breadcrumbs">
-      
-    <li>Python 模块索引</li>
-  </ul>
-</div>
-      
-      <div class="wy-nav-content" id="doc-content">
-        <div class="rst-content">
-          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
-           <div itemprop="articleBody">
-            
-
-   <h1>Python 模块索引</h1>
-
-   <div class="modindex-jumpbox">
-   <a href="#cap-p"><strong>p</strong></a>
-   </div>
-
-   <table class="indextable modindextable">
-     <tr class="pcap"><td></td><td>&#160;</td><td></td></tr>
-     <tr class="cap" id="cap-p"><td></td><td>
-       <strong>p</strong></td><td></td></tr>
-     <tr>
-       <td><img src="_static/minus.png" class="toggler"
-              id="toggle-1" style="display: none" alt="-" /></td>
-       <td>
-       <code class="xref">paddle</code></td><td>
-       <em></em></td></tr>
-     <tr class="cg-1">
-       <td></td>
-       <td>&#160;&#160;&#160;
-       <a href="api/v1/trainer_config_helpers/attrs.html#module-paddle.trainer_config_helpers.attrs"><code class="xref">paddle.trainer_config_helpers.attrs</code></a></td><td>
-       <em></em></td></tr>
-     <tr class="cg-1">
-       <td></td>
-       <td>&#160;&#160;&#160;
-       <a href="api/v1/trainer_config_helpers/data_sources.html#module-paddle.trainer_config_helpers.data_sources"><code class="xref">paddle.trainer_config_helpers.data_sources</code></a></td><td>
-       <em></em></td></tr>
-   </table>
-
-
-           </div>
-          </div>
-          <footer>
-  
-
-  <hr/>
-
-  <div role="contentinfo">
-    <p>
-        &copy; Copyright 2016, PaddlePaddle developers.
-
-    </p>
-  </div>
-  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
-
-</footer>
-
-        </div>
-      </div>
-
-    </section>
-
-  </div>
-  
-
-
-  
-
-    <script type="text/javascript">
-        var DOCUMENTATION_OPTIONS = {
-            URL_ROOT:'./',
-            VERSION:'',
-            COLLAPSE_INDEX:false,
-            FILE_SUFFIX:'.html',
-            HAS_SOURCE:  true,
-            SOURCELINK_SUFFIX: ".txt",
-        };
-    </script>
-      <script type="text/javascript" src="_static/jquery.js"></script>
-      <script type="text/javascript" src="_static/underscore.js"></script>
-      <script type="text/javascript" src="_static/doctools.js"></script>
-      <script type="text/javascript" src="_static/translations.js"></script>
-      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>
-       
-  
-
-  
-  
-    <script type="text/javascript" src="_static/js/theme.js"></script>
-  
-  
-  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
-  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
-  <script src="_static/js/paddle_doc_init.js"></script> 
-
-</body>
-</html>
\ No newline at end of file
diff --git a/develop/doc_cn/search.html b/develop/doc_cn/search.html
index 576fec53b1e38dd726d9d5b068852830b5454704..834425f69af24df995beb611b120a00d6a984540 100644
--- a/develop/doc_cn/search.html
+++ b/develop/doc_cn/search.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/searchindex.js b/develop/doc_cn/searchindex.js
index c0622b5719d869a64200dcd4c25e0632ba21f1da..2ea26474f6ce6225827b67001db459ac99bf273d 100644
--- a/develop/doc_cn/searchindex.js
+++ b/develop/doc_cn/searchindex.js
@@ -1 +1 @@
-Search.setIndex({docnames:["about/index_cn","api/index_cn","api/v1/data_provider/dataprovider_cn","api/v1/data_provider/pydataprovider2_cn","api/v1/index_cn","api/v1/predict/swig_py_paddle_cn","api/v1/trainer_config_helpers/activations","api/v1/trainer_config_helpers/attrs","api/v1/trainer_config_helpers/data_sources","api/v1/trainer_config_helpers/evaluators","api/v1/trainer_config_helpers/layers","api/v1/trainer_config_helpers/networks","api/v1/trainer_config_helpers/optimizers","api/v1/trainer_config_helpers/poolings","api/v2/config/activation","api/v2/config/attr","api/v2/config/layer","api/v2/config/networks","api/v2/config/optimizer","api/v2/config/pooling","api/v2/data","api/v2/model_configs","api/v2/run_logic","design/api","design/cluster_train/README","design/cluster_train/checkpointing","design/cluster_train/data_dispatch","design/multi_language_interface/00.why_plain_c","design/multi_language_interface/01.inference_implementation","design/reader/README","design/releasing_process","faq/index_cn","getstarted/basic_usage/index_cn","getstarted/build_and_install/cmake/build_from_source_cn","getstarted/build_and_install/docker_install_cn","getstarted/build_and_install/index_cn","getstarted/build_and_install/ubuntu_install_cn","getstarted/index_cn","howto/deep_model/rnn/hierarchical_layer_cn","howto/deep_model/rnn/hrnn_rnn_api_compare_cn","howto/deep_model/rnn/index_cn","howto/deep_model/rnn/recurrent_group_cn","howto/deep_model/rnn/rnn_config_cn","howto/dev/contribute_to_paddle_cn","howto/dev/new_layer_cn","howto/dev/write_docs_cn","howto/index_cn","howto/optimization/gpu_profiling_cn","howto/usage/cluster/cluster_train_cn","howto/usage/cmd_parameter/arguments_cn","howto/usage/cmd_parameter/detail_introduction_cn","howto/usage/cmd_parameter/index_cn","howto/usage/cmd_parameter/use_case_cn","howto/usage/concepts/use_concepts_cn","howto/usage/k8s/k8s_basis_cn","howto/usage/k8s/k8s_cn","howto/usage/k8s/k8s_distributed_cn","howto/usage/k8s/src/k8s_data/README","howto/usage/k8s/src/k8s_train/README","index_cn","tutorials/embedding_model/index_cn","tutorials/image_classification/index_cn","tutorials/imagenet_model/resnet_model_cn","tutorials/index_cn","tutorials/quick_start/index_cn","tutorials/rec/ml_dataset_cn","tutorials/rec/ml_regression_cn","tutorials/semantic_role_labeling/index_cn","tutorials/sentiment_analysis/index_cn","tutorials/text_generation/index_cn"],envversion:50,filenames:["about/index_cn.md","api/index_cn.rst","api/v1/data_provider/dataprovider_cn.rst","api/v1/data_provider/pydataprovider2_cn.rst","api/v1/index_cn.rst","api/v1/predict/swig_py_paddle_cn.rst","api/v1/trainer_config_helpers/activations.rst","api/v1/trainer_config_helpers/attrs.rst","api/v1/trainer_config_helpers/data_sources.rst","api/v1/trainer_config_helpers/evaluators.rst","api/v1/trainer_config_helpers/layers.rst","api/v1/trainer_config_helpers/networks.rst","api/v1/trainer_config_helpers/optimizers.rst","api/v1/trainer_config_helpers/poolings.rst","api/v2/config/activation.rst","api/v2/config/attr.rst","api/v2/config/layer.rst","api/v2/config/networks.rst","api/v2/config/optimizer.rst","api/v2/config/pooling.rst","api/v2/data.rst","api/v2/model_configs.rst","api/v2/run_logic.rst","design/api.md","design/cluster_train/README.md","design/cluster_train/checkpointing.md","design/cluster_train/data_dispatch.md","design/multi_language_interface/00.why_plain_c.md","design/multi_language_interface/01.inference_implementation.md","design/reader/README.md","design/releasing_process.md","faq/index_cn.rst","getstarted/basic_usage/index_cn.rst","getstarted/build_and_install/cmake/build_from_source_cn.rst","getstarted/build_and_install/docker_install_cn.rst","getstarted/build_and_install/index_cn.rst","getstarted/build_and_install/ubuntu_install_cn.rst","getstarted/index_cn.rst","howto/deep_model/rnn/hierarchical_layer_cn.rst","howto/deep_model/rnn/hrnn_rnn_api_compare_cn.rst","howto/deep_model/rnn/index_cn.rst","howto/deep_model/rnn/recurrent_group_cn.md","howto/deep_model/rnn/rnn_config_cn.rst","howto/dev/contribute_to_paddle_cn.md","howto/dev/new_layer_cn.rst","howto/dev/write_docs_cn.rst","howto/index_cn.rst","howto/optimization/gpu_profiling_cn.rst","howto/usage/cluster/cluster_train_cn.md","howto/usage/cmd_parameter/arguments_cn.md","howto/usage/cmd_parameter/detail_introduction_cn.md","howto/usage/cmd_parameter/index_cn.rst","howto/usage/cmd_parameter/use_case_cn.md","howto/usage/concepts/use_concepts_cn.rst","howto/usage/k8s/k8s_basis_cn.md","howto/usage/k8s/k8s_cn.md","howto/usage/k8s/k8s_distributed_cn.md","howto/usage/k8s/src/k8s_data/README.md","howto/usage/k8s/src/k8s_train/README.md","index_cn.rst","tutorials/embedding_model/index_cn.md","tutorials/image_classification/index_cn.md","tutorials/imagenet_model/resnet_model_cn.md","tutorials/index_cn.md","tutorials/quick_start/index_cn.rst","tutorials/rec/ml_dataset_cn.md","tutorials/rec/ml_regression_cn.rst","tutorials/semantic_role_labeling/index_cn.md","tutorials/sentiment_analysis/index_cn.md","tutorials/text_generation/index_cn.md"],objects:{"paddle.trainer_config_helpers":{attrs:[7,0,0,"-"],data_sources:[8,0,0,"-"]},"paddle.trainer_config_helpers.attrs":{ExtraAttr:[7,1,1,""],ExtraLayerAttribute:[7,2,1,""],ParamAttr:[7,1,1,""],ParameterAttribute:[7,2,1,""]},"paddle.trainer_config_helpers.attrs.ParameterAttribute":{set_default_parameter_name:[7,3,1,""]},"paddle.trainer_config_helpers.data_sources":{define_py_data_sources2:[8,4,1,""]}},objnames:{"0":["py","module","Python \u6a21\u5757"],"1":["py","attribute","Python \u5c5e\u6027"],"2":["py","class","Python \u7c7b"],"3":["py","method","Python \u65b9\u6cd5"],"4":["py","function","Python \u51fd\u6570"]},objtypes:{"0":"py:module","1":"py:attribute","2":"py:class","3":"py:method","4":"py:function"},terms:{"00012\u7684\u6a21\u578b\u6709\u7740\u6700\u9ad8\u7684bleu\u503c27":69,"0005\u4e58\u4ee5batch":61,"000\u4e2a\u5df2\u6807\u6ce8\u8fc7\u7684\u9ad8\u6781\u6027\u7535\u5f71\u8bc4\u8bba\u7528\u4e8e\u8bad\u7ec3":68,"000\u4e2a\u7528\u4e8e\u6d4b\u8bd5":68,"000\u4e2atxt\u6587\u4ef6":68,"000\u4f4d\u7528\u6237\u5bf94":65,"000\u5e45\u56fe\u50cf\u4e0a\u6d4b\u8bd5\u4e86\u6a21\u578b\u7684\u5206\u7c7b\u9519\u8bef\u7387":62,"000\u5f20\u7070\u5ea6\u56fe\u7247\u7684\u6570\u5b57\u5206\u7c7b\u6570\u636e\u96c6":3,"000\u6761\u8bc4\u4ef7":65,"000\u90e8\u7535\u5f71\u76841":65,"00186201e":5,"00m":47,"02595v1":[10,16],"03m":47,"0424m":47,"0473v3":[11,17],"05d":61,"0630u":47,"06u":47,"0810u":47,"08823112e":5,"0957m":47,"0\u53f7\u8bad\u7ec3\u8282\u70b9\u662f\u4e3b\u8bad\u7ec3\u8282\u70b9":50,"0\u5c42\u5e8f\u5217":38,"0\u8868\u793a\u7b2c\u4e00\u6b21\u7ecf\u8fc7\u8bad\u7ec3\u96c6":68,"0ab":[10,16],"0b1":36,"0rc1":[30,34],"0rc2":30,"10000\u5f20\u4f5c\u4e3a\u6d4b\u8bd5\u96c6":61,"10007_10":68,"10014_7":68,"100m":31,"101\u5c42\u548c152\u5c42\u7684\u7f51\u7edc\u7ed3\u6784\u4e2d":62,"101\u5c42\u548c152\u5c42\u7684\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6\u53ef\u53c2\u7167":62,"101\u5c42\u7f51\u7edc\u6a21\u578b":62,"10\u4e2d\u7684\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6":61,"10\u6570\u636e\u96c6":61,"10\u6570\u636e\u96c6\u5305\u542b60000\u5f2032x32\u7684\u5f69\u8272\u56fe\u7247":61,"10\u6570\u636e\u96c6\u7684\u5b98\u65b9\u7f51\u5740":61,"10\u6570\u636e\u96c6\u8bad\u7ec3\u4e00\u4e2a\u5377\u79ef\u795e\u7ecf\u7f51\u7edc":61,"1150u":47,"11\u5b9e\u73b0\u4e86c":28,"11e6":55,"12194102e":5,"124n":47,"128\u7ef4\u548c256\u7ef4":60,"13m":55,"1490u":47,"14\u6570\u636e\u96c6":69,"14\u6570\u636e\u96c6\u4e0a\u5f97\u5230\u826f\u597d\u8868\u73b0\u7684\u8bad\u7ec3\u8fc7\u7a0b":69,"14\u8fd9\u79cd\u5199\u6cd5\u5c06\u4f1a\u6d4b\u8bd5\u6a21\u578b":52,"152\u5c42\u7f51\u7edc\u6a21\u578b":62,"15501715e":5,"1550u":47,"15\u884c":39,"1636k":69,"16u":47,"173m":62,"173n":47,"1770u":47,"18\u5c81\u4ee5\u4e0b":65,"18e457ce3d362ff5f3febf8e7f85ffec852f70f3b629add10aed84f930a68750":55,"197u":47,"1\u7684\u5c42\u4e4b\u5916":52,"1\u7a00\u758f\u6570\u636e":44,"1\u8f6e\u5b58\u50a8\u7684\u6240\u6709\u6a21\u578b":52,"1\u9664\u4ee5batch":61,"1m\u6570\u636e\u96c6\u4e2d":66,"1m\u7684\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6\u5728\u76ee\u5f55":66,"200\u6570\u636e\u96c6\u4e0a\u4f7f\u7528vgg\u6a21\u578b\u8bad\u7ec3\u4e00\u4e2a\u9e1f\u7c7b\u56fe\u7247\u5206\u7c7b\u6a21\u578b":61,"210u":47,"211839e770f7b538e2d8":[11,17],"215n":47,"228u":47,"234m":62,"24\u5c81":65,"2520u":47,"25639710e":5,"25k":64,"2680u":47,"26\u884c":39,"27787406e":5,"279n":47,"27m":47,"285m":47,"2863m":47,"28\u7684\u56fe\u7247\u50cf\u7d20\u7070\u5ea6\u503c":3,"28\u7ef4\u7684\u7a20\u5bc6\u6d6e\u70b9\u6570\u5411\u91cf\u548c\u4e00\u4e2a":3,"28m":47,"2977m":47,"29997\u4e2a\u6700\u9ad8\u9891\u5355\u8bcd\u548c3\u4e2a\u7279\u6b8a\u7b26\u53f7":69,"2\u4e09\u7c7b\u7684\u6bd4\u4f8b\u4e3a":31,"2\u4e2a\u6d6e\u70b9\u6570":32,"2\u5206\u522b\u4ee3\u88683\u4e2a\u8282\u70b9\u7684trainer":56,"2\u610f\u5473\u77400\u53f7\u548c1\u53f7gpu\u5c06\u4f1a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97fc1\u548cfc2\u5c42":52,"2\u8fd9\u51e0\u4e2a\u76ee\u5f55\u8868\u793apaddlepaddle\u8282\u70b9\u4e0etrain":56,"2nd":[10,16],"302n":47,"30u":47,"3206326\u4e2a\u8bcd\u548c4\u4e2a\u7279\u6b8a\u6807\u8bb0":60,"32777140e":5,"328n":47,"32\u7ef4":60,"32u":47,"32x32":20,"331n":47,"3320u":47,"34\u5c81":65,"35\u65f6":69,"36540484e":5,"36u":47,"3710m":47,"3768m":47,"387u":47,"38u":47,"3920u":47,"39u":47,"3\u4e2a\u7279\u6b8a\u7b26\u53f7":69,"3\u53f7gpu":31,"4035m":47,"4090u":47,"4096mb":50,"4279m":47,"43630644e":5,"43u":47,"448a5b355b84":55,"44\u5c81":65,"4560u":47,"4563m":47,"45u":47,"4650u":47,"4726m":47,"473m":55,"48565123e":5,"48684503e":5,"49316648e":5,"49\u5c81":65,"4gb":50,"500\u4e2atxt\u6587\u4ef6":68,"500m":31,"50\u5c42":62,"50\u5c42\u7f51\u7edc\u6a21\u578b":62,"51111044e":5,"514u":47,"525n":47,"526u":47,"53018653e":5,"536u":47,"5460u":47,"5470u":47,"54u":47,"55\u5c81":65,"55g":69,"5690m":47,"573u":47,"578n":47,"5798m":47,"586u":47,"58s":55,"5969m":47,"5\u4e2a\u6d4b\u8bd5\u6837\u4f8b\u548c2\u4e2a\u751f\u6210\u5f0f\u6837\u4f8b":60,"5\u5230\u672c\u5730\u73af\u5883\u4e2d":36,"6080u":47,"6082v4":[10,16],"6140u":47,"6305m":47,"639u":47,"64\u7ef4":60,"655u":47,"6780u":47,"6810u":47,"682u":47,"6970u":47,"6\u4e07\u4ebf\u6b21\u6d6e\u70b9\u8fd0\u7b97\u6bcf\u79d2":47,"6\u4e2a\u8282\u70b9":48,"6\u5143\u4e0a\u4e0b\u6587\u4f5c\u4e3a\u8f93\u5165\u5c42":60,"704u":47,"70634608e":5,"7090u":47,"72296313e":5,"72u":47,"73u":47,"75u":47,"760u":47,"767u":47,"783n":47,"784u":47,"78m":47,"7eamaa":20,"7kb":55,"8250u":47,"8300u":47,"830n":47,"849m":47,"85625684e":5,"861u":47,"864k":69,"8661m":47,"877\u4e2a\u88ab\u5411\u91cf\u5316\u7684\u8bcd":60,"877\u884c":60,"892m":47,"901n":47,"90u":47,"918u":47,"9247m":47,"924n":47,"9261m":47,"93137714e":5,"9330m":47,"94u":47,"9530m":47,"96644767e":5,"983m":47,"988u":47,"997u":47,"99982715e":5,"99m":62,"99u":47,"9\u4e2d\u7684\u4e00\u4e2a\u6570\u5b57":3,"9f18":55,"\u0233":32,"\u03b5":32,"\u4e00":39,"\u4e00\u4e2a":53,"\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a0\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u4f8b\u5b50\u662f\u623f\u4ea7\u4f30\u503c":32,"\u4e00\u4e2a\u5178\u578b\u7684\u795e\u7ecf\u7f51\u7edc\u5982\u4e0b\u56fe\u6240\u793a":61,"\u4e00\u4e2a\u5206\u5e03\u5f0f\u4f5c\u4e1a\u91cc\u5305\u62ec\u82e5\u5e72trainer\u8fdb\u7a0b\u548c\u82e5\u5e72paramet":53,"\u4e00\u4e2a\u5206\u5e03\u5f0f\u7684\u5b58\u50a8\u7cfb\u7edf":54,"\u4e00\u4e2a\u5206\u5e03\u5f0fpaddle\u8bad\u7ec3\u4efb\u52a1\u4e2d\u7684\u6bcf\u4e2a\u8fdb\u7a0b\u90fd\u53ef\u4ee5\u4ececeph\u8bfb\u53d6\u6570\u636e":55,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8fdb\u5165":41,"\u4e00\u4e2a\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u5305\u542b\u5982\u4e0b\u5c42":61,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8fdb\u5165":41,"\u4e00\u4e2a\u53cc\u5c42rnn\u7531\u591a\u4e2a\u5355\u5c42rnn\u7ec4\u6210":41,"\u4e00\u4e2a\u53ef\u8c03\u7528\u7684\u51fd\u6570":41,"\u4e00\u4e2a\u57fa\u672c\u7684\u5e94\u7528\u573a\u666f\u662f\u533a\u5206\u7ed9\u5b9a\u6587\u672c\u7684\u8912\u8d2c\u4e24\u6781\u6027":68,"\u4e00\u4e2a\u6216\u591a\u4e2a":54,"\u4e00\u4e2a\u6570\u636e\u96c6\u5927\u90e8\u5206\u5e8f\u5217\u957f\u5ea6\u662f100":31,"\u4e00\u4e2a\u6587\u4ef6":3,"\u4e00\u4e2a\u662f\u6d6e\u70b9\u8ba1\u7b97\u91cf":47,"\u4e00\u4e2a\u662f\u76f4\u63a5\u628a\u4e00\u4e2a\u5bb9\u5668\u8f6c\u6362\u6210\u955c\u50cf":34,"\u4e00\u4e2a\u72ec\u7acb\u7684\u5143\u7d20":38,"\u4e00\u4e2a\u72ec\u7acb\u7684\u8bcd\u8bed":38,"\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u6a21\u578b\u7531\u5927\u91cf\u7684\u53c2\u6570\u7ec4\u6210":25,"\u4e00\u4e2a\u7b80\u5355\u7684\u4f8b\u5b50\u5982":68,"\u4e00\u4e2a\u7b80\u5355\u7684\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6\u4e3a":53,"\u4e00\u4e2a\u7ec8\u7aef\u8fd0\u884cvi":34,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad\u90e8\u5206\u628a\u8f93\u5165\u8f6c\u5316\u4e3a\u76f8\u5e94\u7684\u8f93\u51fa":44,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u53c2\u6570\u662f\u5728":44,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684c":44,"\u4e00\u4e2a\u91cd\u8981\u7684\u95ee\u9898\u662f\u9009\u62e9\u6b63\u786e\u7684learning_r":31,"\u4e00\u4e2agpu\u8bbe\u5907\u4e0a\u4e0d\u5141\u8bb8\u914d\u7f6e\u591a\u4e2a\u6a21\u578b":50,"\u4e00\u4e2alabel":39,"\u4e00\u4e2alogging\u5bf9\u8c61":3,"\u4e00\u4e2amemory\u5305\u542b":42,"\u4e00\u4e2apass\u610f\u5473\u7740paddlepaddle\u8bad\u7ec3\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u6837\u672c\u88ab\u904d\u5386\u4e00\u6b21":67,"\u4e00\u4e2apass\u8868\u793a\u8fc7\u4e00\u904d\u6240\u6709\u8bad\u7ec3\u6837\u672c":64,"\u4e00\u4e2apod\u4e2d\u7684\u6240\u6709\u5bb9\u5668\u4f1a\u88ab\u8c03\u5ea6\u5230\u540c\u4e00\u4e2anode\u4e0a":54,"\u4e00\u4e2apserver\u8fdb\u7a0b\u5171\u7ed1\u5b9a\u591a\u5c11\u7aef\u53e3\u7528\u6765\u505a\u7a00\u758f\u66f4\u65b0":53,"\u4e00\u4e9b\u60c5\u51b5\u4e0b":53,"\u4e00\u4e9b\u968f\u673a\u5316\u566a\u58f0\u6dfb\u52a0\u90fd\u5e94\u8be5\u5728dataprovider\u4e2d\u5b8c\u6210":53,"\u4e00\u4eba":39,"\u4e00\u53e5\u8bdd\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":41,"\u4e00\u53f0\u673a\u5668\u4e0a\u9762\u7684\u7ebf\u7a0b\u6570\u91cf":66,"\u4e00\u65e9":39,"\u4e00\u662fbatch":31,"\u4e00\u6761\u6837\u672c":3,"\u4e00\u6837\u8bbe\u7f6e":48,"\u4e00\u6b21\u4f5c\u4e1a\u79f0\u4e3a\u4e00\u4e2ajob":54,"\u4e00\u6b21\u6027\u676f\u5b50":39,"\u4e00\u6b21yield\u8c03\u7528":3,"\u4e00\u79cd\u5e38\u7528\u7684\u505a\u6cd5\u662f\u7528\u5b66\u4e60\u7684\u6a21\u578b\u5bf9\u53e6\u5916\u4e00\u7ec4\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u9884\u6d4b":32,"\u4e00\u7bc7\u8bba\u6587":69,"\u4e00\u7ea7\u76ee\u5f55":[68,69],"\u4e00\u81f4":[38,39],"\u4e00\u822c\u4e0d\u5141\u8bb8\u518d\u4ece":30,"\u4e00\u822c\u4ece":43,"\u4e00\u822c\u5728paddlepaddle\u4e2d":39,"\u4e00\u822c\u60c5\u51b5\u4e0b":[2,32],"\u4e00\u822c\u63a8\u8350\u8bbe\u7f6e\u6210true":3,"\u4e00\u822c\u662f\u5c01\u88c5\u4e86\u8bb8\u591a\u590d\u6742\u64cd\u4f5c\u7684\u96c6\u5408":53,"\u4e00\u822c\u662f\u7531\u4e8e\u76f4\u63a5\u4f20\u9012\u5927\u5b57\u5178\u5bfc\u81f4\u7684":31,"\u4e00\u822c\u6765\u8bf4":42,"\u4e00\u822c\u800c\u8a00":69,"\u4e00\u822c\u8868\u793a":39,"\u4e00\u884c\u4e3a\u4e00\u4e2a\u6837\u672c":64,"\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":3,"\u4e09\u7ea7\u76ee\u5f55":[68,69],"\u4e0a":43,"\u4e0a\u4e0b\u6587\u5927\u5c0f\u8bbe\u7f6e\u4e3a1\u7684\u4e00\u4e2a\u6837\u672c\u7684\u7279\u5f81\u5982\u4e0b":67,"\u4e0a\u53d1\u8868\u7684\u8bc4\u8bba\u5206\u6210\u6b63\u9762\u8bc4\u8bba\u548c\u8d1f\u9762\u8bc4\u8bba\u4e24\u7c7b":68,"\u4e0a\u56fe\u4e2d\u865a\u7ebf\u7684\u8fde\u63a5":39,"\u4e0a\u56fe\u63cf\u8ff0\u4e86\u4e00\u4e2a3\u8282\u70b9\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3\u573a\u666f":56,"\u4e0a\u7f51":39,"\u4e0a\u8fd0\u884c":34,"\u4e0a\u8ff0\u4ee3\u7801\u5c06bias\u5168\u90e8\u521d\u59cb\u5316\u4e3a1":31,"\u4e0a\u8ff0\u7684\u4ee3\u7801\u7247\u6bb5\u5305\u542b\u4e86\u4e24\u79cd\u65b9\u6cd5":47,"\u4e0a\u8ff0\u811a\u672c\u4f7f\u7528":48,"\u4e0b":61,"\u4e0b\u56fe\u4e2d\u5c31\u5c55\u793a\u4e86\u4e00\u4e9b\u5173\u4e8e\u5185\u5b58\u6570\u636e\u8fc1\u5f99\u548c\u8ba1\u7b97\u8d44\u6e90\u5229\u7528\u7387\u7684\u5efa\u8bae":47,"\u4e0b\u56fe\u5c55\u793a\u4e86\u6240\u6709\u7684\u56fe\u7247\u7c7b\u522b":61,"\u4e0b\u56fe\u5c55\u793a\u4e86\u65f6\u95f4\u6269\u5c55\u76842\u5c42":67,"\u4e0b\u56fe\u5c55\u793a\u7684\u662f\u57fa\u4e8e\u6b8b\u5dee\u7684\u8fde\u63a5\u65b9\u5f0f":62,"\u4e0b\u56fe\u63cf\u8ff0\u4e86\u7528\u6237\u4f7f\u7528\u6846\u56fe":53,"\u4e0b\u56fe\u662f\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u793a\u610f\u56fe":44,"\u4e0b\u6587\u4ee5nlp\u4efb\u52a1\u4e3a\u4f8b":41,"\u4e0b\u6587\u4f7f\u7528":56,"\u4e0b\u6587\u5c31\u662f\u7528job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u8fdb\u884c\u8bad\u7ec3":55,"\u4e0b\u6b21":39,"\u4e0b\u7684":56,"\u4e0b\u8868\u5c55\u793a\u4e86batch":62,"\u4e0b\u8f7d\u5b8c\u6570\u636e\u540e":55,"\u4e0b\u8f7d\u5b8c\u76f8\u5173\u5b89\u88c5\u5305\u540e":36,"\u4e0b\u8f7d\u6570\u636e\u96c6":61,"\u4e0b\u8f7dwmt":69,"\u4e0b\u9762\u4e3e\u4e2a\u7b80\u5355\u7684\u4f8b\u5b50":47,"\u4e0b\u9762\u4ecb\u7ecd\u9884\u5904\u7406\u8fc7\u7a0b\u5177\u4f53\u7684\u6b65\u9aa4":66,"\u4e0b\u9762\u5148\u7b80\u8981\u4ecb\u7ecd\u4e00\u4e0b\u672c\u6587\u7528\u5230\u7684\u51e0\u4e2akubernetes\u6982\u5ff5":54,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u6570\u636e\u6e90\u914d\u7f6e":53,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u67d0\u4e00\u7c7b\u6587\u4ef6\u7684\u5b9e\u73b0\u65b9\u5f0f":28,"\u4e0b\u9762\u5217\u51fa\u4e86":42,"\u4e0b\u9762\u5217\u51fa\u4e86\u5168\u8fde\u63a5\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":44,"\u4e0b\u9762\u5c06\u5206\u522b\u4ecb\u7ecd\u8fd9\u4e24\u90e8\u5206":53,"\u4e0b\u9762\u5c31\u6839\u636e\u8fd9\u51e0\u4e2a\u6b65\u9aa4\u5206\u522b\u4ecb\u7ecd":56,"\u4e0b\u9762\u6211\u4eec\u7ed9\u51fa\u4e86\u4e00\u4e2a\u914d\u7f6e\u793a\u4f8b":61,"\u4e0b\u9762\u662f\u4e00\u4e2a\u8bef\u5dee\u66f2\u7ebf\u56fe\u7684\u793a\u4f8b":61,"\u4e0b\u9762\u662fcifar":61,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u5b9e\u73b0\u4e86":44,"\u4e0b\u9762\u7684\u4f8b\u5b50\u4f7f\u7528\u4e86":62,"\u4e0b\u9762\u7684\u4f8b\u5b50\u540c\u6837\u4f7f\u7528\u4e86":62,"\u4e0b\u9762\u7ed9\u51fa\u4e86\u4e00\u4e2a\u4f8b\u5b50":44,"\u4e0b\u9762\u811a\u672c\u7b26\u5408paddlepaddle\u671f\u5f85\u7684\u8bfb\u53d6\u6570\u636e\u7684python\u7a0b\u5e8f\u7684\u6a21\u5f0f":32,"\u4e0b\u9762\u8fd9\u4e9blayer\u80fd\u591f\u63a5\u53d7\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165":38,"\u4e0d":39,"\u4e0d\u4e00\u5b9a\u548c\u65f6\u95f4\u6709\u5173\u7cfb":3,"\u4e0d\u4f1a\u518d\u4ece":31,"\u4e0d\u4f7f\u7528\u9759\u6001\u5e93":27,"\u4e0d\u4f7f\u7528\u989d\u5916\u7a7a\u95f4":44,"\u4e0d\u4f7f\u7528c":27,"\u4e0d\u4f7f\u7528swig":27,"\u4e0d\u5305\u542b\u5728\u5b57\u5178\u4e2d\u7684\u5355\u8bcd":69,"\u4e0d\u540c":67,"\u4e0d\u540c\u4e3b\u673a":54,"\u4e0d\u540c\u4ea7\u54c1":68,"\u4e0d\u540c\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u6570\u636e\u5927\u5c0f\u7684\u6700\u5927\u503c\u4e0e\u6700\u5c0f\u503c\u7684\u6bd4\u7387":50,"\u4e0d\u540c\u5c42\u7684\u7279\u5f81\u7531\u5206\u53f7":62,"\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u662f\u4e0d\u540c\u7684":42,"\u4e0d\u540c\u7248\u672c\u7684\u7f16\u8bd1\u5668\u4e4b\u95f4":27,"\u4e0d\u540c\u7684\u4f18\u5316\u7b97\u6cd5\u9700\u8981\u4f7f\u7528\u4e0d\u540c\u5927\u5c0f\u7684\u5185\u5b58":31,"\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":56,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":3,"\u4e0d\u540c\u7a7a\u95f4\u7684\u8d44\u6e90\u540d\u53ef\u4ee5\u91cd\u590d":54,"\u4e0d\u540c\u8bed\u8a00\u7684\u63a5\u53e3\u9002\u5e94\u4e0d\u540c\u8bed\u8a00\u7684\u7279\u6027":27,"\u4e0d\u540c\u8f93\u5165\u542b\u6709\u7684\u5b50\u53e5":41,"\u4e0d\u540c\u8f93\u5165\u5e8f\u5217\u542b\u6709\u7684\u8bcd\u8bed\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":41,"\u4e0d\u540cdataprovider\u5bf9\u6bd4\u5982\u4e0b":39,"\u4e0d\u540cpod\u4e4b\u95f4\u53ef\u4ee5\u901a\u8fc7ip\u5730\u5740\u8bbf\u95ee":54,"\u4e0d\u542b\u53ef\u5b66\u4e60\u53c2\u6570":53,"\u4e0d\u5728":28,"\u4e0d\u5c11":39,"\u4e0d\u5d4c\u5165\u5176\u4ed6\u8bed\u8a00\u89e3\u91ca\u5668":27,"\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":27,"\u4e0d\u5e94\u8be5\u88ab\u62c6\u89e3":41,"\u4e0d\u5fc5\u518d\u5c06\u4efb\u610f\u957f\u5ea6\u6e90\u8bed\u53e5\u4e2d\u7684\u6240\u6709\u4fe1\u606f\u538b\u7f29\u81f3\u4e00\u4e2a\u5b9a\u957f\u7684\u5411\u91cf\u4e2d":69,"\u4e0d\u6307\u5b9a\u65f6":41,"\u4e0d\u63d0\u4f9b\u5206\u5e03\u5f0f\u5b58\u50a8":54,"\u4e0d\u652f\u6301":67,"\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":3,"\u4e0d\u663e\u793a\u7684\u5199\u6bcf\u4e2a\u7c7b\u5177\u4f53\u5305\u542b\u4ec0\u4e48":27,"\u4e0d\u6ee1\u8db3\u94a9\u5b50\u7684":43,"\u4e0d\u7f13\u5b58\u4efb\u4f55\u6570\u636e":3,"\u4e0d\u80fd\u88ab\u63d0\u4ea4\u5230":43,"\u4e0d\u8fc7":39,"\u4e0d\u8fdc":39,"\u4e0d\u9519":39,"\u4e0d\u9700\u8981\u5bf9\u5e8f\u5217\u6570\u636e\u8fdb\u884c\u4efb\u4f55\u9884\u5904\u7406":42,"\u4e0e":[56,60,69],"\u4e0e\u529f\u80fd\u5206\u652f\u4e0d\u540c\u7684\u662f":30,"\u4e0e\u5355\u5c42rnn\u7684\u914d\u7f6e\u7c7b\u4f3c":39,"\u4e0e\u53ef\u80fd\u6709\u7684":30,"\u4e0e\u5728":67,"\u4e0e\u672c\u5730\u8bad\u7ec3\u76f8\u540c":48,"\u4e0e\u6b64\u4e0d\u540c\u7684\u662f":56,"\u4e0e\u7ffb\u8bd1":69,"\u4e0e\u8bad\u7ec3\u4e0d\u540c":68,"\u4e0e\u8bad\u7ec3\u6a21\u578b\u4e0d\u540c\u7684\u662f":69,"\u4e0e\u8fd9\u4e2a\u8bad\u7ec3\u6570\u636e\u4ea4\u4e92\u7684layer":31,"\u4e0eimdb\u7f51\u7ad9\u63d0\u4f9b\u7684\u4e00\u81f4":65,"\u4e0ejob":56,"\u4e0etime":65,"\u4e14":39,"\u4e14\u589e\u52a0\u4e00\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00":27,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":3,"\u4e14\u652f\u6301\u90e8\u7f72\u5230":54,"\u4e14\u6bcf\u4e2a\u53e5\u5b50\u8868\u793a\u4e3a\u5bf9\u5e94\u7684\u8bcd\u8868\u7d22\u5f15\u6570\u7ec4":39,"\u4e14\u8c03\u7528\u65f6\u4e0d\u80fd\u629b\u51fa\u5f02\u5e38\u6216\u51fa\u73b0\u8fd0\u884c\u65f6\u9519\u8bef":28,"\u4e14\u9ed8\u8ba4\u5728\u8bad\u7ec3\u96c6\u4e0a\u6784\u5efa\u5b57\u5178":68,"\u4e14c99\u652f\u6301bool\u7c7b\u578b\u548c\u5b9a\u957f\u6574\u6570":27,"\u4e14c99\u76f8\u5bf9\u4e8ec11\u4f7f\u7528\u66f4\u52a0\u5e7f\u6cdb":27,"\u4e24":39,"\u4e24\u4e2a\u5217\u8868\u6587\u4ef6":48,"\u4e24\u4e2a\u5b50\u76ee\u5f55\u4e0b":45,"\u4e24\u4e2a\u5d4c\u5957\u7684":41,"\u4e24\u4e2a\u64cd\u4f5c":47,"\u4e24\u4e2a\u6587\u4ef6\u5939":61,"\u4e24\u4e2a\u8f93\u5165\u7279\u5f81\u5728\u8fd9\u4e2a\u6d41\u7a0b\u4e2d\u8d77\u7740\u81f3\u5173\u91cd\u8981\u7684\u4f5c\u7528":67,"\u4e24\u4e2a\u8f93\u5165\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u4e5f\u5e76\u4e0d\u76f8\u540c":39,"\u4e24\u4e2a\u90e8\u5206":45,"\u4e24\u79cd\u7c7b\u522b":64,"\u4e24\u8005\u5747\u4e3a\u7eaf\u6587\u672c\u6587\u4ef6":2,"\u4e2a":64,"\u4e2a\u5185\u5b58\u6c60\u5b9e\u9645\u4e0a\u51b3\u5b9a\u4e86shuffle\u7684\u7c92\u5ea6":31,"\u4e2a\u5355\u8bcd":69,"\u4e2a\u6027\u5316\u63a8\u8350":[30,63],"\u4e2a\u6279\u6b21\u540e\u6253\u5370\u4e00\u4e2a":66,"\u4e2a\u6279\u6b21\u7684\u53c2\u6570\u5e73\u5747\u503c\u8fdb\u884c\u6d4b\u8bd5":50,"\u4e2a\u6a21\u578b\u6d4b\u8bd5\u6570\u636e":50,"\u4e2d":[27,28,31,44,53,56,61,64,66,67,68],"\u4e2d\u4e0d\u8981\u6dfb\u52a0\u5927\u6587\u4ef6\u7b49":43,"\u4e2d\u4ecb\u7ecd\u7684\u65b9\u6cd5":60,"\u4e2d\u4efb\u610f\u7b2ci\u884c\u7684\u53e5\u5b50\u4e4b\u95f4\u90fd\u5fc5\u987b\u6709\u7740\u4e00\u4e00\u5bf9\u5e94\u7684\u5173\u7cfb":69,"\u4e2d\u4efb\u610f\u7b2ci\u884c\u7684\u53e5\u5b50\u4e4b\u95f4\u90fd\u6709\u7740\u4e00\u4e00\u5bf9\u5e94\u7684\u5173\u7cfb":69,"\u4e2d\u5173\u4e8e\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc\u7684\u4ecb\u7ecd":39,"\u4e2d\u5199\u5165json\u5185\u5bb9":25,"\u4e2d\u5305\u542b\u4e86\u8bad\u7ec3\u6a21\u578b\u7684\u57fa\u672c\u547d\u4ee4":64,"\u4e2d\u5305\u542b\u5982\u4e0b\u8868\u6240\u793a\u76843\u4e2a\u6587\u4ef6\u5939":69,"\u4e2d\u5355\u5143\u6d4b\u8bd5\u7684\u4e00\u90e8\u5206":43,"\u4e2d\u5b89\u88c5":48,"\u4e2d\u5b8c\u5168\u4e00\u81f4":27,"\u4e2d\u5b8c\u6210":68,"\u4e2d\u5b9a\u4e49":42,"\u4e2d\u5b9a\u4e49\u4f7f\u7528\u54ea\u79cddataprovid":2,"\u4e2d\u5b9a\u4e49\u548c\u4f7f\u7528":41,"\u4e2d\u5b9e\u73b0\u7684\u7ed3\u6784\u4f53":28,"\u4e2d\u5bfc\u51fa\u9884\u5b9a\u4e49\u7684\u7f51\u7edc":68,"\u4e2d\u5c55\u793a\u4e86\u5982\u4f55\u4f7f\u7528python\u6765\u63d0\u53d6\u7279\u5f81":62,"\u4e2d\u6307\u5b9a":50,"\u4e2d\u6307\u5b9a\u7684\u540d\u5b57":52,"\u4e2d\u6307\u5b9a\u7684\u5c42\u987a\u5e8f\u4e00\u81f4":62,"\u4e2d\u63d0\u51fa\u7684resnet\u7f51\u7edc\u7ed3\u6784\u57282015\u5e74imagenet\u5927\u89c4\u6a21\u89c6\u89c9\u8bc6\u522b\u7ade\u8d5b":62,"\u4e2d\u641c\u7d22\u8fd9\u51e0\u4e2a\u5e93":33,"\u4e2d\u6587\u6587\u6863":45,"\u4e2d\u6587\u6587\u6863\u76ee\u5f55":45,"\u4e2d\u6587\u7ef4\u57fa\u767e\u79d1\u9875\u9762":39,"\u4e2d\u6709\u8bb8\u591a\u7684\u7279\u5f81":65,"\u4e2d\u6bcf\u4e2apod\u7684ip\u5730\u5740":56,"\u4e2d\u6bcf\u5c42\u7684\u6570\u503c\u7edf\u8ba1":50,"\u4e2d\u7684":62,"\u4e2d\u7684\u4e00\u884c":[3,43],"\u4e2d\u7684\u5185\u5bb9":67,"\u4e2d\u7684\u5bf9\u5e94\u5206\u652f\u5373\u53ef":43,"\u4e2d\u7684\u63a5\u53e3":66,"\u4e2d\u7684\u6570\u636e":62,"\u4e2d\u7684\u6570\u636e\u662f\u5426\u4e3a\u5e8f\u5217\u6a21\u5f0f":66,"\u4e2d\u7684\u6570\u636e\u8fdb\u884c\u9884\u6d4b":62,"\u4e2d\u7684\u6570\u6910\u96c6\u7684\u7ed3\u6784\u5982\u4e0b":68,"\u4e2d\u7684\u6bcf\u4e00\u884c\u547d\u4ee4":66,"\u4e2d\u7684\u751f\u6210\u7ed3\u679c\u5982\u4e0b\u6240\u793a":69,"\u4e2d\u7684\u7528\u6237\u8bc1\u4e66":54,"\u4e2d\u7684\u7b2ci\u884c":69,"\u4e2d\u7684\u8bf4\u660e":3,"\u4e2d\u7684\u8fd9\u4e9b\u6570\u636e\u6587\u4ef6":65,"\u4e2d\u770b\u5230\u4e0b\u9762\u7684\u6587\u4ef6":68,"\u4e2d\u83b7\u53d6":56,"\u4e2d\u8ba4\u771f\u8bbe\u7f6e":48,"\u4e2d\u8bbe\u7f6e":48,"\u4e2d\u8bbe\u7f6e\u7684\u6240\u6709\u8282\u70b9":48,"\u4e2d\u8be6\u7ec6\u4ecb\u7ecd":44,"\u4e2d\u8bfb\u53d6":3,"\u4e2d\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":42,"\u4e2d\u914d\u7f6e\u7684\u6548\u679c\u4e00\u81f4":3,"\u4e34\u65f6\u53d8\u91cf\u7b49\u7b49":31,"\u4e3a":[3,42],"\u4e3a0":3,"\u4e3a\u4e86\u4f7f\u7528\u63d0\u524d\u7f16\u5199\u7684\u811a\u672c":68,"\u4e3a\u4e86\u4fdd\u8bc1\u6548\u7387":44,"\u4e3a\u4e86\u4fdd\u8bc1gpu\u9a71\u52a8\u80fd\u591f\u5728\u955c\u50cf\u91cc\u9762\u6b63\u5e38\u8fd0\u884c":34,"\u4e3a\u4e86\u5145\u5206\u7684\u968f\u673a\u6253\u4e71\u8bad\u7ec3\u96c6":68,"\u4e3a\u4e86\u5b8c\u6210\u5206\u5e03\u5f0f\u673a\u5668\u5b66\u4e60\u8bad\u7ec3\u4efb\u52a1":54,"\u4e3a\u4e86\u5c01\u88c5\u80fd\u591f\u6b63\u786e\u5de5\u4f5c":44,"\u4e3a\u4e86\u63cf\u8ff0\u65b9\u4fbf":41,"\u4e3a\u4e86\u65b9\u4fbf\u5728\u56fd":34,"\u4e3a\u4e86\u65b9\u4fbf\u5927\u5bb6":43,"\u4e3a\u4e86\u65b9\u4fbf\u8d77\u89c1":48,"\u4e3a\u4e86\u66b4\u9732\u7684\u63a5\u53e3\u5c3d\u91cf\u7b80\u5355":28,"\u4e3a\u4e86\u66f4\u7075\u6d3b\u7684\u914d\u7f6e":53,"\u4e3a\u4e86\u6ee1\u8db3\u8bad\u7ec3":48,"\u4e3a\u4e86\u7528\u6237\u80fd\u591f\u7075\u6d3b\u7684\u5904\u7406\u6570\u636e":53,"\u4e3a\u4e86\u8fbe\u5230\u6027\u80fd\u6700\u4f18":47,"\u4e3a\u4e86\u8fd8\u539f":32,"\u4e3a\u4e86\u907f\u514d\u7528\u6237\u76f4\u63a5\u5199\u590d\u6742\u7684protobuf":53,"\u4e3a\u4f8b":64,"\u4e3a\u4f8b\u521b\u5efa\u5206\u5e03\u5f0f\u7684\u5355\u8fdb\u7a0b\u8bad\u7ec3":48,"\u4e3a\u4f8b\u8fdb\u884c\u9884\u6d4b":64,"\u4e3a\u53c2\u6570\u77e9\u9635\u7684\u5bbd\u5ea6":31,"\u4e3a\u60a8\u505a\u6027\u80fd\u8c03\u4f18\u63d0\u4f9b\u4e86\u65b9\u5411":47,"\u4e3a\u60f3\u4fee\u6b63\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u7528\u6237\u63d0\u4f9b\u4e86\u5c06\u6587\u672c\u8bcd\u5411\u91cf\u6a21\u578b\u8f6c\u6362\u4e3a\u4e8c\u8fdb\u5236\u6a21\u578b\u7684\u547d\u4ee4":60,"\u4e3a\u65b9\u4fbf\u4f5c\u4e1a\u542f\u52a8\u63d0\u4f9b\u4e86\u4e24\u4e2a\u72ec\u7279\u7684\u547d\u4ee4\u9009\u9879":48,"\u4e3a\u6b64":55,"\u4e3a\u8f93\u51fa\u5206\u914d\u5185\u5b58":44,"\u4e3a\u96c6\u7fa4\u4f5c\u4e1a\u8bbe\u7f6e\u989d\u5916\u7684":48,"\u4e3ajson\u6216yaml\u683c\u5f0f":66,"\u4e3aoutput_\u7533\u8bf7\u5185\u5b58":44,"\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":50,"\u4e3b\u8981\u5305\u62ec\u4ee5\u4e0b\u4e94\u4e2a\u6b65\u9aa4":5,"\u4e3b\u8981\u539f\u56e0":39,"\u4e3b\u8981\u539f\u56e0\u5305\u62ec\u4e24\u4e2a\u65b9\u9762":31,"\u4e3b\u8981\u539f\u56e0\u662f\u589e\u52a0\u4e86\u521d\u59cb\u5316\u673a\u5236":3,"\u4e3b\u8981\u6765\u81ea\u5317\u7f8e\u6d32":61,"\u4e3b\u8981\u7531layer\u7ec4\u6210":53,"\u4e3b\u8981\u7684\u89e3\u51b3\u529e\u6cd5\u662f\u51cf\u5c0f\u5b66\u4e60\u5f8b\u6216\u8005\u5bf9\u6570\u636e\u8fdb\u884c\u5f52\u4e00\u5316\u5904\u7406":31,"\u4e3b\u8981\u804c\u8d23\u5728\u4e8e\u5c06\u8bad\u7ec3\u6570\u636e\u4f20\u5165\u5185\u5b58\u6216\u8005\u663e\u5b58":64,"\u4e3e\u4e00\u4e2a\u4f8b\u5b50":31,"\u4e3e\u4f8b":31,"\u4e3e\u4f8b\u8bf4\u660e":39,"\u4e4b\u524d":43,"\u4e4b\u524d\u914d\u7f6e\u6587\u4ef6\u4e2d":64,"\u4e4b\u540e":[32,44],"\u4e4b\u540e\u4f60\u4f1a\u5f97\u5230\u8bad\u7ec3":48,"\u4e4b\u540e\u4f7f\u7528":44,"\u4e4b\u540e\u4f7f\u7528\u77e9\u9635\u8fd0\u7b97\u51fd\u6570\u6765\u8ba1\u7b97":44,"\u4e4b\u540e\u521d\u59cb\u5316\u6240\u6709\u7684\u6743\u91cd\u77e9\u9635":44,"\u4e4b\u540e\u5b9a\u4e49\u7684":61,"\u4e4b\u5916\u7684\u6240\u6709\u5934\u6587\u4ef6":28,"\u4e4b\u95f4\u7684\u8ddd\u79bb":32,"\u4e4b\u95f4\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":41,"\u4e58\u4e0a\u8f93\u51fa\u7684\u68af\u5ea6":44,"\u4e58\u9664\u7b49\u65f6\u5019":31,"\u4e5d\u4e2a":67,"\u4e5f":39,"\u4e5f\u4e0d\u4f7f\u7528\u5176\u4ed6\u52a8\u6001\u5e93":27,"\u4e5f\u4e0d\u5b58\u5728\u4e00\u4e2asubseq\u76f4\u63a5\u751f\u6210\u4e0b\u4e00\u4e2asubseq\u7684\u60c5\u51b5":41,"\u4e5f\u4e0d\u5e94\u8be5\u62a5\u9519":28,"\u4e5f\u4e0d\u751f\u6210":28,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":43,"\u4e5f\u53ef\u4ee5\u53bb\u6389\u8fd9\u4e9b\u8bc1\u4e66\u7684\u914d\u7f6e":54,"\u4e5f\u53ef\u4ee5\u5728\u5f00\u53d1\u955c\u50cf\u4e2d\u542f\u52a8\u4e00\u4e2asshd\u670d\u52a1":34,"\u4e5f\u53ef\u4ee5\u662f\u4e00\u4e2a\u8bcd\u8bed":41,"\u4e5f\u53ef\u4ee5\u8bf4\u662f\u67d0\u4e9b\u7279\u5b9a\u6307\u4ee4\u7684\u4f7f\u7528\u60c5\u51b5":47,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539":56,"\u4e5f\u53ef\u4ee5\u901a\u8fc7saving_period_by_batches\u8bbe\u7f6e\u6bcf\u9694\u591a\u5c11batch\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":64,"\u4e5f\u53ef\u4ee5\u914d\u7f6e\u4e0d\u540c\u7684\u91cd\u8bd5\u673a\u5236":54,"\u4e5f\u5c31\u662f":43,"\u4e5f\u5c31\u662f\u5c06\u8bcd\u5411\u91cf\u6a21\u578b\u8fdb\u4e00\u6b65\u6f14\u5316\u4e3a\u4e09\u4e2a\u65b0\u6b65\u9aa4":64,"\u4e5f\u5c31\u662f\u81ea\u5df1\u7528\u6237\u540d\u4e0b\u7684":43,"\u4e5f\u5c31\u662f\u8bf4":[50,52,60],"\u4e5f\u5f97\u5230\u4e00\u4e2a\u7528\u6237\u7279\u5f81":66,"\u4e5f\u63cf\u8ff0\u4e86\u5bb9\u5668\u9700\u8981\u4f7f\u7528\u7684\u5b58\u50a8\u5377\u6302\u8f7d\u7684\u60c5\u51b5":56,"\u4e5f\u652f\u6301cpu\u7684\u6027\u80fd\u5206\u6790":47,"\u4e5f\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":39,"\u4e5f\u662f\u5bb9\u5668\u4e0enode\u4e4b\u95f4\u5171\u4eab\u6587\u4ef6\u7684\u65b9\u5f0f":54,"\u4e5f\u662fdecoder\u5faa\u73af\u5c55\u5f00\u7684\u4f9d\u636e":41,"\u4e5f\u662fpaddlepaddle\u6240\u80fd\u591f\u4fdd\u8bc1\u7684shuffle\u7c92\u5ea6":3,"\u4e5f\u6ca1\u7528":31,"\u4e5f\u79f0\u4e3arnn\u6a21\u578b":64,"\u4e5f\u79f0\u4f5c":53,"\u4e5f\u8bb8\u662f\u56e0\u4e3a\u9700\u8981\u5b89\u88c5":61,"\u4e5f\u9700\u8981\u4e24\u6b21\u968f\u673a\u9009\u62e9\u5230\u76f8\u540cgenerator\u7684\u65f6\u5019":3,"\u4e66\u5199":27,"\u4e7e":39,"\u4e86":39,"\u4e86\u89e3\u60a8\u7684\u786c\u4ef6":47,"\u4e86\u89e3\u66f4\u591a\u7ec6\u8282":42,"\u4e86\u89e3\u66f4\u591a\u8be6\u7ec6\u4fe1\u606f":42,"\u4e8c\u7ea7\u76ee\u5f55":[68,69],"\u4e8c\u7ef4\u77e9\u9635":62,"\u4e8c\u8005\u8bed\u610f\u4e0a\u5b8c\u5168\u4e00\u81f4":39,"\u4e8c\u8fdb\u5236":60,"\u4e92\u76f8\u901a\u4fe1":54,"\u4e92\u8054\u7f51\u7535\u5f71\u6570\u636e\u5e93":68,"\u4e94\u661f\u7ea7":39,"\u4e9a\u9a6c\u900a":68,"\u4ea4\u901a":39,"\u4ea4\u901a\u4fbf\u5229":39,"\u4eab\u53d7\u60a8\u7684\u65c5\u7a0b":34,"\u4eba\u8138\u8bc6\u522b":26,"\u4ec0\u4e48":66,"\u4ec5\u4ec5\u4f7f\u7528":27,"\u4ec5\u4ec5\u662f\u4e00\u4e9b\u5173\u952e\u8bcd":68,"\u4ec5\u4ec5\u662f\u4e24\u4e2a\u5168\u8fde\u63a5\u5c42":66,"\u4ec5\u4ec5\u662f\u7b80\u5355\u7684\u5d4c\u5165":66,"\u4ec5\u5305\u542b\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6570\u6910\u96c6":68,"\u4ec5\u5728\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u65f6\u6709\u6548":44,"\u4ec5\u5bf9\u7a00\u758f\u6570\u636e\u6709\u6548":44,"\u4ec5\u9700\u8981\u77e5\u9053\u5982\u4f55\u4ece":3,"\u4ecb\u7ecd\u4e86\u4e00\u79cd\u901a\u8fc7ssh\u8fdc\u7a0b\u5206\u53d1\u4efb\u52a1":56,"\u4ecb\u7ecd\u5206\u5e03\u5f0f\u8bad\u7ec3\u4e4b\u524d":54,"\u4ecb\u7ecdpaddlepaddle\u7684\u57fa\u672c\u4f7f\u7528\u65b9\u6cd5":64,"\u4ece":[30,47,67],"\u4ece0\u5230num":50,"\u4ece\u4e00\u4e2aword\u751f\u6210\u4e0b\u4e00\u4e2aword":41,"\u4ece\u5185\u6838\u51fd\u6570\u7684\u89d2\u5ea6":47,"\u4ece\u56fe\u4e2d\u53ef\u4ee5\u770b\u5230":32,"\u4ece\u5916\u90e8\u7f51\u7ad9\u4e0a\u4e0b\u8f7d\u7684\u539f\u59cb\u6570\u6910\u96c6":68,"\u4ece\u5927\u5230\u5c0f":69,"\u4ece\u6570\u636e\u63d0\u4f9b\u7a0b\u5e8f\u52a0\u8f7d\u5b9e\u4f8b":67,"\u4ece\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u770b":39,"\u4ece\u6bcf\u4e2a\u5355\u8bcd\u5de6\u53f3\u4e24\u7aef\u5206\u522b\u83b7\u53d6k\u4e2a\u76f8\u90bb\u7684\u5355\u8bcd":64,"\u4ece\u78c1\u76d8\u6587\u4ef6\u4e2d\u52a0\u8f7duuid\u6587\u4ef6\u540d\u7684\u68c0\u67e5\u70b9\u5feb\u7167\u6587\u4ef6":25,"\u4ece\u7b2c0\u4e2a\u8bc4\u4f30\u5230\u5f53\u524d\u8bc4\u4f30\u4e2d":69,"\u4ece\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u7684\u5e73\u5747\u635f\u5931":68,"\u4ece\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u7684\u5e73\u5747cost":69,"\u4ece\u800c\u53ef\u4ee5\u505a\u4e00\u4e9b\u4e0e\u8ba1\u7b97\u91cd\u53e0\u7684\u5de5\u4f5c":44,"\u4ece\u800c\u53ef\u4ee5\u62df\u5408\u4efb\u610f\u7684\u51fd\u6570\u6765\u5b66\u4e60\u590d\u6742\u7684\u6570\u636e\u5173\u7cfb":32,"\u4ece\u800c\u751f\u6210\u591a\u4e2agener":3,"\u4ece\u800c\u80fd\u591f\u88abpaddlepaddl":64,"\u4ece\u800c\u9632\u6b62\u8fc7\u62df\u5408":2,"\u4ece\u8be5\u94fe\u63a5":69,"\u4ece\u8bed\u4e49\u4e0a\u770b":41,"\u4ece\u8f93\u5165\u6570\u636e\u4e0a\u770b":39,"\u4ece\u8f93\u51fa\u65e5\u5fd7\u53ef\u4ee5\u770b\u5230":32,"\u4ece\u9884\u8bad\u7ec3\u6a21\u578b\u4e2d":60,"\u4eceetcd\u4e2d\u8bfb\u53d6\u8282\u70b9":25,"\u4ecestart":50,"\u4ecetest":69,"\u4ed3\u5e93\u7684\u8fdc\u7a0b\u4e3b\u673a":43,"\u4ed4\u7ec6\u89c2\u5bdf":62,"\u4ed6\u4e3b\u8981\u5305\u542b\u4e86\u5b9e\u9645\u66b4\u9732\u7684\u7c7b\u578b\u7ed3\u6784":28,"\u4ed6\u4eec\u5206\u522b\u662f":39,"\u4ed6\u4eec\u5728paddle\u7684\u6587\u6863\u548capi\u4e2d\u662f\u4e00\u4e2a\u6982\u5ff5":39,"\u4ed6\u4eec\u63d0\u51fa\u6b8b\u5dee\u5b66\u4e60\u7684\u6846\u67b6\u6765\u7b80\u5316\u7f51\u7edc\u7684\u8bad\u7ec3":62,"\u4ed6\u662f\u5c06":28,"\u4ed6\u7684\u76ee\u6807\u662f\u4f7f\u7528c":27,"\u4ee3\u66ff":56,"\u4ee3\u7801":66,"\u4ee3\u7801\u4e2d9":39,"\u4ee3\u7801\u5982\u4e0b":42,"\u4ee3\u7801\u6ce8\u91ca\u8bf7\u9075\u5b88":43,"\u4ee3\u7801\u751f\u6210\u7684\u7b26\u53f7\u53ef\u80fd\u4e0d\u4e00\u81f4":27,"\u4ee3\u8868\u5bbf\u4e3b\u673a\u76ee\u5f55":56,"\u4ee3\u8868\u7f16\u53f7":66,"\u4ee3\u8868\u8fd9\u4e2ashard\u7684\u6700\u5927index":26,"\u4ee3\u8868shard\u7684index":26,"\u4ee5\u4e0a":43,"\u4ee5\u4e0a\u4ee3\u7801\u7684reader\u8f93\u51fa\u7684data":26,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u4e0b\u751f\u6210100\u4e2a\u6587\u4ef6":26,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u6309\u7167\u751f\u6210\u7684dockerfile\u628a\u751f\u6210\u7684\u7a0b\u5e8f\u62f7\u8d1d\u5230\u751f\u4ea7\u955c\u50cf\u4e2d\u5e76\u505a\u76f8\u5e94\u7684\u914d\u7f6e":34,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u7f16\u8bd1paddlepaddl":34,"\u4ee5\u4e0a\u65b9\u6cd5\u5728gpu\u955c\u50cf\u91cc\u4e5f\u80fd\u7528":34,"\u4ee5\u4e0b":[26,66],"\u4ee5\u4e0b\u4ee3\u7801\u6bb5\u5b9a\u4e49\u4e86\u4e09\u4e2a\u8f93\u5165":42,"\u4ee5\u4e0b\u4ee3\u7801\u7247\u6bb5\u5b9a\u4e49":42,"\u4ee5\u4e0b\u6211\u4eec\u7ffb\u8bd1\u6570\u636e\u96c6\u7f51\u7ad9\u4e2dreadme\u6587\u4ef6\u7684\u63cf\u8ff0":65,"\u4ee5\u4e0b\u6307\u4ee4\u80fd\u68c0\u67e5linux\u7535\u8111\u662f\u5426\u652f\u6301avx":34,"\u4ee5\u4e0b\u6559\u7a0b\u5c06\u6307\u5bfc\u60a8\u63d0\u4ea4\u4ee3\u7801":43,"\u4ee5\u4e0b\u662f\u5bf9\u4e0a\u8ff0\u6570\u636e\u52a0\u8f7d\u7684\u89e3\u91ca":64,"\u4ee5\u4e0b\u6b65\u9aa4\u57fa\u4e8e":48,"\u4ee5\u4e0b\u793a\u8303\u5982\u4f55\u4f7f\u7528\u9884\u8bad\u7ec3\u7684\u4e2d\u6587\u5b57\u5178\u548c\u8bcd\u5411\u91cf\u8fdb\u884c\u77ed\u8bed\u6539\u5199":60,"\u4ee5\u4e0b\u9009\u9879\u5fc5\u987b\u5728":48,"\u4ee5\u4ea4\u4e92\u5bb9\u5668\u65b9\u5f0f\u8fd0\u884c\u5f00\u53d1\u955c\u50cf":34,"\u4ee5\u4fbf\u7528\u6237":48,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u7684\u6b63\u786e\u8ba1\u7b97":44,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u8ba1\u7b97\u7684\u6b63\u786e\u6027":44,"\u4ee5\u5206\u7c7b\u6765\u81ea":68,"\u4ee5\u53ca":44,"\u4ee5\u53ca\u4f7f\u7528\u5b50\u5e8f\u5217\u6765\u5b9a\u4e49\u5206\u7ea7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":42,"\u4ee5\u53ca\u53cc\u5c42\u5e8f\u5217":38,"\u4ee5\u53ca\u5728wmt":69,"\u4ee5\u53ca\u5982\u4f55\u5728\u5c42\u4e4b\u95f4\u8fdb\u884c\u8fde\u63a5":61,"\u4ee5\u53ca\u6570\u636e\u8bfb\u53d6\u51fd\u6570":53,"\u4ee5\u53ca\u751f\u6210\u521b\u5efa\u751f\u4ea7\u955c\u50cf\u7684dockerfil":34,"\u4ee5\u53ca\u8ba1\u7b97\u903b\u8f91\u5728\u5e8f\u5217\u4e0a\u7684\u5faa\u73af\u5c55\u5f00":41,"\u4ee5\u53ca\u8f93\u5165\u7684\u68af\u5ea6":44,"\u4ee5\u53canumpi":26,"\u4ee5\u53capaddle\u5982\u4f55\u5904\u7406\u591a\u79cd\u7c7b\u578b\u7684\u8f93\u5165":66,"\u4ee5\u53carelu":44,"\u4ee5\u76f8\u5bf9\u8def\u5f84\u5f15\u7528":2,"\u4ee5\u83b7\u5f97\u66f4\u597d\u7684\u7f51\u7edc\u6027\u80fd":48,"\u4ee5\u9017\u53f7":60,"\u4ee5\u9017\u53f7\u95f4\u9694":50,"\u4ef7\u683c":39,"\u4efb\u52a1":66,"\u4efb\u52a1\u6765\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":48,"\u4efb\u610f\u5c06\u4e00\u4e9b\u6570\u636e\u7ec4\u5408\u6210\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":39,"\u4efb\u610f\u65f6\u523b\u53ea\u53ef\u80fd\u540c\u65f6\u6709\u4e00\u53f0\u670d\u52a1\u5668\u6545\u969c":25,"\u4f18\u5316":68,"\u4f18\u5316\u5668\u5219\u7528\u94fe\u5f0f\u6cd5\u5219\u6765\u5bf9\u6bcf\u4e2a\u53c2\u6570\u8ba1\u7b97\u635f\u5931\u51fd\u6570\u7684\u68af\u5ea6":44,"\u4f18\u5316\u7b97\u6cd5":53,"\u4f1a\u5148\u8fdb\u884c\u53c2\u6570\u7684\u521d\u59cb\u5316\u4e0e\u89e3\u6790":56,"\u4f1a\u5171\u4eab\u53c2\u6570":31,"\u4f1a\u52a0\u8f7d\u4e0a\u4e00\u8f6e\u7684\u53c2\u6570":50,"\u4f1a\u53d8\u6210\u8bcd\u8868\u4e2d\u7684\u4f4d\u7f6e":39,"\u4f1a\u542f\u52a8pserver\u4e0etrainer\u8fdb\u7a0b":56,"\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u751f\u6210\u4e24\u4e2a\u5b50\u76ee\u5f55":45,"\u4f1a\u5bf9\u6bcf\u4e00\u4e2a\u6fc0\u6d3b\u6682\u5b58\u4e00\u4e9b\u6570\u636e":31,"\u4f1a\u5bf9\u8fd9\u7c7b\u8f93\u5165\u8fdb\u884c\u62c6\u89e3":41,"\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7248\u672cpython\u5728\u4e00\u4e2a\u8fdb\u7a0b\u91cc\u7684bug":27,"\u4f1a\u5c06\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u62fc\u63a5":41,"\u4f1a\u5c06\u7b2c\u4e00\u4e2a":31,"\u4f1a\u6210\u4e3astep\u51fd\u6570\u7684\u8f93\u5165":41,"\u4f1a\u6254\u5230\u8fd9\u6761\u6570\u636e":3,"\u4f1a\u62a5\u9519":41,"\u4f1a\u6839\u636e\u547d\u4ee4\u884c\u53c2\u6570\u6307\u5b9a\u7684\u6d4b\u8bd5\u65b9\u5f0f":2,"\u4f1a\u6839\u636einput_types\u68c0\u67e5\u6570\u636e\u7684\u5408\u6cd5\u6027":3,"\u4f1a\u76f4\u63a5\u62a5\u9519\u9000\u51fa":27,"\u4f1a\u76f8\u5e94\u5730\u6539\u53d8\u8f93\u51fa\u7684\u5c3a\u5bf8":44,"\u4f1a\u81ea\u9002\u5e94\u5730\u4ece\u8fd9\u4e9b\u5411\u91cf\u4e2d\u9009\u62e9\u4e00\u4e2a\u5b50\u96c6\u51fa\u6765":69,"\u4f1a\u83b7\u53d6\u5f53\u524dnamespace\u4e0b\u7684\u6240\u6709pod":56,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":41,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u975e\u5e8f\u5217":41,"\u4f1a\u88abpickle\u5e8f\u5217\u5316\u6210\u5b57\u7b26\u4e32":26,"\u4f20\u5165":[3,26],"\u4f20\u5165\u4e0a\u4e00\u6b65\u89e3\u6790\u51fa\u6765\u7684\u6a21\u578b\u914d\u7f6e\u5c31\u53ef\u4ee5\u521b\u5efa\u4e00\u4e2a":5,"\u4f20\u5165\u9884\u6d4b\u6570\u636e":5,"\u4f20\u7ed9dataprovider\u7684\u67d0\u4e00\u4e2aargs\u8fc7\u5927":31,"\u4f20\u9012\u7ed9\u914d\u7f6e\u6587\u4ef6\u7684\u53c2\u6570":50,"\u4f46":28,"\u4f46\u4e0d\u66b4\u9732":28,"\u4f46\u4e0d\u7528\u4e8e\u8ba1\u7b97\u68af\u5ea6":44,"\u4f46\u4e0d\u9700\u8981\u63d0\u524d\u521b\u5efa":50,"\u4f46\u4e8e\u53cc\u5c42\u5e8f\u5217\u7684lstm\u6765\u8bf4":39,"\u4f46\u548c\u5355\u5c42rnn\u4e0d\u540c":39,"\u4f46\u5728\u8d77\u521d\u7684\u51e0\u8f6e\u8bad\u7ec3\u4e2d\u5b83\u4eec\u90fd\u5728\u5feb\u901f\u903c\u8fd1\u771f\u5b9e\u503c":32,"\u4f46\u5b50\u53e5\u542b\u6709\u7684\u8bcd\u8bed\u6570\u53ef\u4ee5\u4e0d\u76f8\u7b49":41,"\u4f46\u5c3d\u91cf\u8bf7\u4fdd\u6301\u7f16\u8bd1\u548c\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u662f\u540c\u4e00\u4e2a\u7248\u672c":33,"\u4f46\u5e76\u6ca1\u6709\u7ecf\u8fc7\u56de\u5f52\u6d4b\u8bd5":30,"\u4f46\u5e8f\u5217\u8f93\u51fa\u65f6":39,"\u4f46\u5f53\u8c03\u7528\u8fc7\u4e00\u6b21\u540e":3,"\u4f46\u6240\u6709fork\u7684\u7248\u672c\u5e93\u7684\u6240\u6709\u5206\u652f\u90fd\u76f8\u5f53\u4e8e\u7279\u6027\u5206\u652f":30,"\u4f46\u662f":[31,39],"\u4f46\u662f2008\u5e74\u4e4b\u524d\u751f\u4ea7\u7684\u65e7\u7535\u8111\u4e0d\u652f\u6301avx":34,"\u4f46\u662f\u4e5f\u6ca1\u6709\u5fc5\u8981\u5220\u9664\u65e0\u7528\u7684\u6587\u4ef6":48,"\u4f46\u662f\u53c8\u8fc7\u4e8e\u7410\u788e":28,"\u4f46\u662f\u5927\u90e8\u5206\u53c2\u6570\u662f\u4e3a\u5f00\u53d1\u8005\u63d0\u4f9b\u7684":49,"\u4f46\u662f\u5b50\u5e8f\u5217\u7684\u6570\u76ee\u5fc5\u987b\u4e00\u6837":39,"\u4f46\u662f\u6211\u4eec\u5e76\u4e0d\u63a8\u8350\u8fd9\u79cd\u65b9\u6cd5":34,"\u4f46\u662f\u652f\u6301avx\u6307\u4ee4\u96c6":43,"\u4f46\u662f\u65b9\u4fbf\u8c03\u8bd5\u548c\u6d4bbenchmark":33,"\u4f46\u662f\u6bcf\u4e2a\u6837\u672c\u4ec5\u5305\u542b\u51e0\u4e2a\u8bcd":52,"\u4f46\u662f\u7a81\u7136\u6709\u4e00\u4e2a10000\u957f\u7684\u5e8f\u5217":31,"\u4f46\u662f\u89e3\u91ca\u6027\u8bed\u8a00":27,"\u4f46\u662f\u8fd9\u4e2a\u503c\u4e0d\u53ef\u4ee5\u8c03\u7684\u8fc7\u5927":53,"\u4f46\u662f\u8fd9\u79cd\u65b9\u6cd5\u5728\u6bcf\u5c42\u53ea\u4fdd\u5b58\u9884\u8bbe\u6570\u91cf\u7684\u6700\u4f18\u72b6\u6001":69,"\u4f46\u662f\u8fdc\u672a\u5b8c\u5584":0,"\u4f46\u662f\u9690\u85cf\u5c42\u4e2d\u7684\u6bcf\u4e2a\u666e\u901a\u8282\u70b9\u88ab\u4e00\u4e2a\u8bb0\u5fc6\u5355\u5143\u66ff\u6362":68,"\u4f46\u662fbatch":31,"\u4f46\u6709\u503c\u7684\u5730\u65b9\u5fc5\u987b\u4e3a1":3,"\u4f46\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":3,"\u4f46\u8fd9\u4e2a\u5173\u7cfb\u53ef\u80fd\u4e0d\u6b63\u786e":3,"\u4f4d\u7f6e":39,"\u4f4f":39,"\u4f53\u88c1\u5b57\u5178":66,"\u4f53\u88c1\u5b57\u6bb5":66,"\u4f59\u5f26\u76f8\u4f3c\u5ea6\u56de\u5f52":66,"\u4f59\u5f26\u76f8\u4f3c\u5ea6\u5c42":66,"\u4f5c\u4e3a\u4e0b\u4e00\u4e2a\u5b50\u53e5memory\u7684\u521d\u59cb\u72b6\u6001":39,"\u4f5c\u4e3a\u4f8b\u5b50\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":42,"\u4f5c\u4e3a\u53c2\u6570\u7684id":31,"\u4f5c\u4e3a\u5f53\u524d\u65f6\u523b\u8f93\u5165":41,"\u4f5c\u4e3a\u6d88\u606f\u957f\u5ea6":53,"\u4f5c\u4e3a\u793a\u4f8b\u6570\u636e":65,"\u4f5c\u4e3a\u7c7b\u53e5\u67c4":27,"\u4f5c\u4e3a\u7ebf\u6027\u56de\u5f52\u7684\u8f93\u5165":32,"\u4f5c\u4e3a\u8f93\u51fa":42,"\u4f5c\u4e3a\u96c6\u7fa4\u8bad\u7ec3\u7684\u5de5\u4f5c\u7a7a\u95f4":48,"\u4f5c\u4e3aboot_layer\u4f20\u7ed9\u4e0b\u4e00\u4e2a\u5b50\u53e5\u7684memori":39,"\u4f5c\u5bb6":65,"\u4f5c\u7528":38,"\u4f60\u4e5f\u53ef\u4ee5\u4f7f\u7528\u8fd9\u4e09\u4e2a\u503c":62,"\u4f60\u4e5f\u53ef\u4ee5\u5148\u8df3\u8fc7\u672c\u6587\u7684\u89e3\u91ca\u73af\u8282":64,"\u4f60\u4e5f\u53ef\u4ee5\u7b80\u5355\u7684\u8fd0\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4":60,"\u4f60\u4e5f\u53ef\u4ee5\u901a\u8fc7\u5728\u547d\u4ee4\u884c\u53c2\u6570\u4e2d\u589e\u52a0\u4e00\u4e2a\u53c2\u6570\u5982":62,"\u4f60\u4e5f\u8bb8\u53ef\u4ee5\u5c1d\u8bd5\u66f4\u8001\u7684\u65b9\u6cd5":34,"\u4f60\u53ea\u9700\u5b8c\u6210":48,"\u4f60\u53ea\u9700\u8981\u5728\u547d\u4ee4\u884c\u8f93\u5165\u4ee5\u4e0b\u547d\u4ee4":64,"\u4f60\u53ea\u9700\u8981\u6309\u7167\u5982\u4e0b\u65b9\u5f0f\u7ec4\u7ec7\u6570\u636e":69,"\u4f60\u53ef\u4ee5\u4f7f\u7528":62,"\u4f60\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u811a\u672c\u4e0b\u8f7d":68,"\u4f60\u53ef\u4ee5\u4f7f\u7528\u5b83\u6765\u8bbe\u7f6e":48,"\u4f60\u53ef\u4ee5\u4f7f\u7528\u672c\u5730\u8bad\u7ec3\u4e2d\u7684\u76f8\u540c\u6a21\u578b\u6587\u4ef6\u8fdb\u884c\u96c6\u7fa4\u8bad\u7ec3":48,"\u4f60\u53ef\u4ee5\u5728\u4efb\u4f55\u65f6\u5019\u7528":66,"\u4f60\u53ef\u4ee5\u5bf9\u6a21\u578b\u8fdb\u884c\u8bc4\u4f30":66,"\u4f60\u53ef\u4ee5\u5c06\u7f51\u7edc\u914d\u7f6e\u6210\u67d0\u4e9b\u5c42\u4f7f\u7528gpu\u8ba1\u7b97":52,"\u4f60\u53ef\u4ee5\u6267\u884c\u4e0a\u8ff0\u547d\u4ee4\u6765\u4e0b\u8f7d\u6240\u6709\u7684\u6a21\u578b\u548c\u5747\u503c\u6587\u4ef6":62,"\u4f60\u53ef\u4ee5\u901a\u8fc7\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u5f97\u5230resnet\u7f51\u7edc\u7684\u7ed3\u6784\u53ef\u89c6\u5316\u56fe":62,"\u4f60\u53ef\u4ee5\u9884\u6d4b\u4efb\u4f55\u7528\u6237\u5bf9\u4e8e\u4efb\u4f55\u4e00\u90e8\u7535\u5f71\u7684\u8bc4\u4ef7":66,"\u4f60\u5c06\u4f1a\u770b\u5230\u4ee5\u4e0b\u7684\u6a21\u578b\u7ed3\u6784":60,"\u4f60\u5c06\u4f1a\u770b\u5230\u5982\u4e0b\u6d88\u606f":69,"\u4f60\u5c06\u4f1a\u770b\u5230\u5982\u4e0b\u7ed3\u679c":62,"\u4f60\u5c06\u4f1a\u770b\u5230\u7279\u5f81\u5b58\u50a8\u5728":62,"\u4f60\u5c06\u4f1a\u770b\u5230\u8fd9\u6837\u7684\u6d88\u606f":69,"\u4f60\u5c06\u5728\u76ee\u5f55":68,"\u4f60\u5c06\u770b\u5230\u5982\u4e0b\u7684\u4fe1\u606f":66,"\u4f60\u7684\u5de5\u4f5c\u7a7a\u95f4\u5e94\u5982\u4e0b\u6240\u793a":48,"\u4f60\u8fd8\u53ef\u4ee5\u5c06\u7528\u6237\u548c":48,"\u4f60\u9700\u8981\u4e00\u4e9b\u66f4\u590d\u6742\u7684\u5355\u5143\u6d4b\u8bd5\u6765\u4fdd\u8bc1\u4f60\u5b9e\u73b0\u7684\u7f51\u7edc\u5c42\u662f\u6b63\u786e\u7684":44,"\u4f60\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a\u8bbe\u5907\u7684id\u53f7":52,"\u4f60\u9700\u8981\u5728\u914d\u7f6ecmake\u65f6\u5c06":44,"\u4f60\u9700\u8981\u5b89\u88c5python\u7684\u7b2c\u4e09\u65b9\u5e93":66,"\u4f60\u9700\u8981\u628a\u8be5\u6587\u4ef6\u52a0\u5165":44,"\u4f7f\u5176\u8f6c\u53d8\u4e3a\u7ef4\u5ea6\u4e3ahidden_dim\u7684\u65b0\u5411\u91cf":64,"\u4f7f\u5f97":32,"\u4f7f\u5f97\u4e24\u4e2a\u5b57\u5178\u6709\u76f8\u540c\u7684\u4e0a\u4e0b\u6587":69,"\u4f7f\u5f97\u5355\u5143\u6d4b\u8bd5\u6709\u4e00\u4e2a\u5e72\u51c0\u7684\u73af\u5883":31,"\u4f7f\u5f97\u642d\u6a21\u578b\u65f6\u66f4\u65b9\u4fbf":44,"\u4f7f\u5f97\u6700\u7ec8\u5f97\u5230\u7684\u6a21\u578b\u51e0\u4e4e\u4e0e\u771f\u5b9e\u6a21\u578b\u4e00\u81f4":32,"\u4f7f\u7528":[28,30,31,34,39,41,42,44,47,50,53,64,67,68],"\u4f7f\u75280\u53f7\u548c1\u53f7gpu\u8ba1\u7b97fc2\u5c42":52,"\u4f7f\u75280\u53f7gpu\u8ba1\u7b97fc2\u5c42":52,"\u4f7f\u752810\u4e2a\u88c1\u526a\u56fe\u50cf\u5757":62,"\u4f7f\u75281\u53f7gpu\u8ba1\u7b97fc3\u5c42":52,"\u4f7f\u75282\u53f7\u548c3\u53f7gpu\u8ba1\u7b97fc3\u5c42":52,"\u4f7f\u7528\u4e00\u4e2a\u5c3a\u5ea6\u4e3a":44,"\u4f7f\u7528\u4e00\u4e2a\u8bcd\u524d\u4e24\u4e2a\u8bcd\u548c\u540e\u4e24\u4e2a\u8bcd":31,"\u4f7f\u7528\u4e0a\u6587\u521b\u5efa\u7684yaml\u6587\u4ef6\u521b\u5efakubernet":55,"\u4f7f\u7528\u4e0b\u9762\u547d\u4ee4":26,"\u4f7f\u7528\u4e86":53,"\u4f7f\u7528\u4e86\u540c\u6837\u7684parameter\u548cbia":31,"\u4f7f\u7528\u4e86\u57fa\u4e8e\u53e5\u6cd5\u7ed3\u6784\u7684\u9884\u5b9a\u4e49\u7279\u5f81\u6a21\u677f":67,"\u4f7f\u7528\u4e86avx\u6307\u4ee4\u96c6":36,"\u4f7f\u7528\u4ee5\u4e0b\u547d\u4ee4\u8fdb\u884c\u6a21\u578b\u8bad\u7ec3":60,"\u4f7f\u7528\u5177\u6709softmax\u6fc0\u6d3b\u7684\u5168\u8fde\u63a5\u524d\u9988\u5c42\u6765\u6267\u884c\u5206\u7c7b\u4efb\u52a1":68,"\u4f7f\u7528\u52a8\u6001\u5e93":27,"\u4f7f\u7528\u540c\u6837\u7684\u8bad\u7ec3\u6570\u636eblock":25,"\u4f7f\u7528\u591a\u5757\u663e\u5361\u8bad\u7ec3":31,"\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bad\u7ec3":31,"\u4f7f\u7528\u5982\u4e0b\u53c2\u6570":61,"\u4f7f\u7528\u5982\u4e0b\u547d\u4ee4":60,"\u4f7f\u7528\u5b66\u4e60\u5b8c\u6210\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u751f\u6210\u5e8f\u5217":42,"\u4f7f\u7528\u5bb9\u5668\u65b9\u5f0f\u8fd0\u884c\u8bad\u7ec3\u4efb\u52a1\u7684kubernet":56,"\u4f7f\u7528\u6211\u4eec\u4e4b\u524d\u6784\u9020\u7684\u955c\u50cf":55,"\u4f7f\u7528\u624b\u5de5\u6307\u5b9a\u7aef\u53e3\u6570\u91cf":53,"\u4f7f\u7528\u663e\u5361\u8bad\u7ec3":31,"\u4f7f\u7528\u667a\u80fd\u6307\u9488\u7684\u539f\u56e0\u662f":28,"\u4f7f\u7528\u6848\u4f8b":51,"\u4f7f\u7528\u7684":31,"\u4f7f\u7528\u76f8\u5bf9\u8def\u5f84\u7684\u5f15\u7528\u65b9\u5f0f":28,"\u4f7f\u7528\u8005\u4e0d\u9700\u8981\u5173\u5fc3":50,"\u4f7f\u7528\u8005\u53ea\u9700\u8981\u5173\u6ce8\u4e8e\u8bbe\u8ba1rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":41,"\u4f7f\u7528\u8005\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684python\u811a\u672c\u6765\u8bfb\u53d6\u53c2\u6570\u503c":62,"\u4f7f\u7528\u8005\u65e0\u9700\u5173\u5fc3\u8fd9\u4e2a\u53c2\u6570":50,"\u4f7f\u7528\u8005\u901a\u5e38\u65e0\u9700\u5173\u5fc3":50,"\u4f7f\u7528\u81ea\u52a8\u7684\u66ff\u8865\u6765\u66ff\u4ee3\u7ecf\u9a8c\u4e30\u5bcc\u7684\u4eba\u5de5\u8bc4\u5224":69,"\u4f7f\u7528\u8c13\u8bcd\u4e0a\u4e0b\u6587":67,"\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u5b8c\u6210\u5bf9\u65b0\u6570\u636e\u7684\u9884\u6d4b":25,"\u4f7f\u7528\u8fd9\u4e2a\u811a\u672c\u524d\u8bf7\u786e\u8ba4\u5df2\u7ecf\u5b89\u88c5\u4e86pillow\u53ca\u76f8\u5173\u4f9d\u8d56\u6a21\u5757":61,"\u4f7f\u7528\u8fd9\u79cd\u65b9\u5f0f":39,"\u4f7f\u7528\u8fdc\u7a0b\u7a00\u758f\u65b9\u5f0f\u8bad\u7ec3\u65f6":44,"\u4f7f\u7528\u968f\u673a\u68af\u5ea6\u4e0b\u964d":68,"\u4f7f\u7528\u9759\u6001\u5e93\u548c\u52a8\u6001\u5e93\u96be\u5ea6\u5dee\u4e0d\u591a":27,"\u4f7f\u7528\u9884\u8bad\u7ec3\u7684\u6807\u51c6\u683c\u5f0f\u8bcd\u5411\u91cf\u6a21\u578b":60,"\u4f7f\u7528args\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u8bbe\u7f6e":3,"\u4f7f\u7528c":28,"\u4f7f\u7528c99\u505a\u63a5\u53e3":27,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c11\u7684\u539f\u56e0\u662f":27,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c89":27,"\u4f7f\u7528checkgrad\u6a21\u5f0f\u65f6\u7684\u53c2\u6570\u53d8\u5316\u5927\u5c0f":50,"\u4f7f\u7528cpu\u4e24\u7ebf\u7a0b\u8ba1\u7b97fc4\u5c42":52,"\u4f7f\u7528cpu\u8ba1\u7b97fc4\u5c42":52,"\u4f7f\u7528cpu\u8bad\u7ec3":68,"\u4f7f\u7528init":52,"\u4f7f\u7528lstm\u4f5c\u4e3aencod":39,"\u4f7f\u7528max":61,"\u4f7f\u7528memory\u7684rnn\u5b9e\u73b0\u4fbf\u5982\u4e0b\u56fe\u6240\u793a":39,"\u4f7f\u7528model":52,"\u4f7f\u7528paddlepaddl":64,"\u4f7f\u7528python\u6570\u636e\u63d0\u4f9b\u5668":61,"\u4f7f\u7528rdma\u8fd8\u662ftcp\u4f20\u8f93\u534f\u8bae":50,"\u4f7f\u7528regress":30,"\u4f7f\u7528swig\u53ea\u652f\u6301cpython\u89e3\u91ca\u5668":27,"\u4f7f\u7528swig\u9700\u8981\u591a\u8bed\u8a00\u7ed1\u5b9a\u7684\u5f00\u53d1\u4eba\u5458\u719f\u7ec3\u638c\u63e1swig\u914d\u7f6e":27,"\u4f7f\u7528void":27,"\u4f7f\u8f93\u5165\u5c42\u5230\u9690\u85cf\u5c42\u7684\u795e\u7ecf\u5143\u662f\u5168\u90e8\u8fde\u63a5\u7684":61,"\u4f86":39,"\u4f8b\u5982":[3,27,28,30,31,33,39,42,44,47,48,49,50,52,53,56,62,64,66,68],"\u4f8b\u5982\u4e0a\u6587\u7684pod":54,"\u4f8b\u5982\u4e0a\u9762\u7684":32,"\u4f8b\u5982\u4ee5\u592a\u7f51\u7684":48,"\u4f8b\u5982\u4f7f\u7528":31,"\u4f8b\u5982\u586b\u5145":42,"\u4f8b\u5982\u5bf9\u4e8ejava\u6216\u8005python":27,"\u4f8b\u5982\u5bf9\u4e8ejava\u6765\u8bf4":27,"\u4f8b\u5982\u5bf9\u4e8epython":27,"\u4f8b\u5982\u5c06\u7b2c\u4e00\u6761\u6570\u636e\u8f6c\u5316\u4e3a":39,"\u4f8b\u5982\u6587\u672c\u5206\u7c7b\u4e2d":39,"\u4f8b\u5982\u672c\u4f8b\u4e2d\u7684\u4e24\u4e2a\u7279\u5f81":39,"\u4f8b\u5982\u673a\u5668\u4e0a\u67094\u5757gpu":31,"\u4f8b\u5982\u7b2c300\u4e2apass\u7684\u6a21\u578b\u4f1a\u88ab\u4fdd\u5b58\u5728":61,"\u4f8b\u5982c":27,"\u4f8b\u5982hostpath":54,"\u4f8b\u5982java\u4e0epython\u7684\u9519\u8bef\u5904\u7406\u662f\u76f4\u63a5\u6254\u51fa\u6765except":27,"\u4f8b\u5982output\u76ee\u5f55\u4e0b\u5c31\u5b58\u653e\u4e86\u8f93\u51fa\u7ed3\u679c":56,"\u4f8b\u5982python\u53ef\u4ee5\u4f7f\u7528":27,"\u4f8b\u5982python\u7684":27,"\u4f8b\u5982sigmoid":44,"\u4f8b\u5982sigmoid\u53d8\u6362":64,"\u4f8b\u5b50\u4e2d\u662f":44,"\u4f8b\u5b50\u4e2d\u662f0":44,"\u4f8b\u5b50\u4e2d\u662f100":44,"\u4f8b\u5b50\u4e2d\u662f4096":44,"\u4f8b\u5b50\u4e2d\u662f8192":44,"\u4f8b\u5b50\u4e2d\u662ffc":44,"\u4f8b\u5b50\u4e2d\u662fsoftmax":44,"\u4f8b\u5b50\u4f7f\u7528":54,"\u4f9bpaddlepaddle\u52a0\u8f7d":50,"\u4f9d\u636e\u5206\u7c7b\u9519\u8bef\u7387\u83b7\u5f97\u6700\u4f73\u6a21\u578b\u8fdb\u884c\u6d4b\u8bd5":68,"\u4f9d\u6b21\u7c7b\u63a8":30,"\u4f9d\u8d56\u4e8epython\u7684":61,"\u4fbf\u4e8e\u5b58\u50a8\u8d44\u6e90\u7ba1\u7406\u548cpod\u5f15\u7528":54,"\u4fbf\u4e8e\u672c\u5730\u9a8c\u8bc1\u548c\u6d4b\u8bd5":54,"\u4fbf\u4e8e\u7528\u6237\u6d4f\u89c8c":34,"\u4fbf\u5229":39,"\u4fbf\u548c\u5355\u5c42rnn\u914d\u7f6e\u4e2d\u7684":39,"\u4fbf\u5b9c":39,"\u4fbf\u662f\u5c06\u9759\u6001\u5e93\u52a0\u5165jvm\u4e2d":27,"\u4fdd\u5b58\u6a21\u578b\u53c2\u6570\u7684\u76ee\u5f55":50,"\u4fdd\u5b58\u751f\u6210\u7ed3\u679c\u7684\u6587\u4ef6":69,"\u4fdd\u5b58\u7f51\u7edc\u5c42\u8f93\u51fa\u7ed3\u679c\u7684\u76ee\u5f55":50,"\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c\u7684\u6587\u4ef6\u540d":50,"\u4fdd\u6301\u5bbd\u9ad8\u6bd4\u7f29\u653e\u5230\u77ed\u8fb9\u4e3a256":62,"\u4fe1\u53f7\u6765\u81ea\u52a8\u7ec8\u6b62\u5b83\u542f\u52a8\u7684\u6240\u6709\u8fdb\u7a0b":48,"\u4fee\u590d\u6240\u6709bug\u540e":30,"\u4fee\u590ddocker\u7f16\u8bd1\u955c\u50cf\u95ee\u9898":30,"\u4fee\u590dubuntu":30,"\u4fee\u6539":[54,55],"\u4fee\u6539\u542f\u52a8\u811a\u672c\u540e":55,"\u4fee\u6539\u6210\u66f4\u5feb\u7684\u7248\u672c":47,"\u4fee\u6539\u6587\u6863":46,"\u503c\u5f97\u6ce8\u610f\u7684\u662f":[39,43],"\u503c\u5f97\u6df1\u5165\u5206\u6790":47,"\u503c\u7c7b\u578b":52,"\u5047\u5982\u6211\u4eec\u662f\u4e09\u5206\u7c7b\u95ee\u9898":31,"\u5047\u8bbe":44,"\u5047\u8bbe\u53d8\u91cf":32,"\u5047\u8bbe\u60a8\u5df2\u7ecf\u5b8c\u6210\u4e86\u4e00\u4e2aai\u8bad\u7ec3\u7684python\u7a0b\u5e8f":34,"\u5047\u8bbe\u635f\u5931\u51fd\u6570\u662f":44,"\u5047\u8bbe\u8bcd\u5411\u91cf\u7ef4\u5ea6\u4e3a32":60,"\u504f\u7f6e\u53c2\u6570":62,"\u504f\u7f6e\u53c2\u6570\u7684\u5927\u5c0f":44,"\u505a\u5982\u4e0b\u51e0\u4e2a\u64cd\u4f5c":30,"\u505a\u63a5\u53e3":27,"\u505c\u6b62\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":25,"\u505c\u6b62\u52a0\u8f7d\u6570\u636e":50,"\u505c\u7535":39,"\u50cf":34,"\u513f\u7ae5\u7247":65,"\u5143\u7d20":38,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684\u8f93\u5165\u4fe1\u606f":38,"\u5148\u4f7f\u7528\u547d\u4ee4":53,"\u5148\u5b9e\u73b0\u6a21\u578b\u63a8\u65ad\u7684api":28,"\u5148\u8c03\u7528initializer\u51fd\u6570":64,"\u5168\u5bb6":39,"\u5168\u8fde\u63a5\u5c42":[32,53,60,61,66],"\u5168\u8fde\u63a5\u5c42\u4ee5\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":44,"\u5168\u8fde\u63a5\u5c42\u5c06\u7535\u5f71\u7684\u6bcf\u4e2a\u7279\u5f81\u7ed3\u5408\u6210\u4e00\u4e2a\u7535\u5f71\u7279\u5f81":66,"\u5168\u8fde\u63a5\u5c42\u6743\u91cd":62,"\u5168\u8fde\u63a5\u5c42\u6ca1\u6709\u7f51\u7edc\u5c42\u914d\u7f6e\u7684\u8d85\u53c2\u6570":44,"\u5168\u8fde\u63a5\u5c42\u7684\u5b9e\u73b0\u4f4d\u4e8e":44,"\u5168\u8fde\u63a5\u5c42\u7684\u6bcf\u4e2a\u8f93\u51fa\u90fd\u8fde\u63a5\u5230\u4e0a\u4e00\u5c42\u7684\u6240\u6709\u7684\u795e\u7ecf\u5143\u4e0a":44,"\u5168\u8fde\u63a5\u5c42python\u5c01\u88c5\u7684\u4f8b\u5b50\u4e2d\u5305\u542b\u4e0b\u9762\u51e0\u6b65":44,"\u516b\u4e2a\u7279\u5f81\u5206\u522b\u8f6c\u6362\u4e3a\u5411\u91cf":67,"\u516c\u5f0f":34,"\u516c\u94a5\u5199\u5165":48,"\u516d\u4e2a\u7279\u5f81\u548c\u6807\u7b7e\u90fd\u662f\u7d22\u5f15\u69fd":67,"\u5171\u4eab\u4efb\u52a1\u4e2d\u8bbe\u7f6e\u7684\u6570\u636e\u8fdb\u884c\u8bad\u7ec3\u548c\u6d4b\u8bd5":67,"\u5171\u4eab\u5b58\u50a8\u6302\u5728\u7684\u8def\u5f84":56,"\u5171\u670932":60,"\u5173\u4e8e\u5982\u4f55\u5b9a\u4e49\u7f51\u7edc\u4e2d\u7684\u5c42":61,"\u5173\u4e8e\u65f6\u95f4\u5e8f\u5217":39,"\u5173\u4e8e\u6784\u5efa\u548c\u6d4b\u8bd5\u7684\u66f4\u591a\u4fe1\u606f":43,"\u5173\u4e8epaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":56,"\u5173\u4e8eunbound":41,"\u5173\u4e8evgg\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u7684\u63cf\u8ff0\u53ef\u4ee5\u53c2\u8003":61,"\u5176\u4e0b\u5b50\u6587\u4ef6\u5939\u7684\u7ed3\u6784\u5982\u4e0b":61,"\u5176\u4e2d":[3,27,30,31,32,42,44,53,60,61,62],"\u5176\u4e2d156\u548c285\u662f\u8fd9\u4e9b\u56fe\u50cf\u7684\u5206\u7c7b\u6807\u7b7e":62,"\u5176\u4e2d50000\u5f20\u56fe\u7247\u4f5c\u4e3a\u8bad\u7ec3\u96c6":61,"\u5176\u4e2d\u5206\u522b\u5305\u542b\u4e86cifar":61,"\u5176\u4e2d\u5305\u542b6":65,"\u5176\u4e2d\u5305\u542b\u4e86200\u79cd\u9e1f\u7c7b\u7684\u7167\u7247":61,"\u5176\u4e2d\u5305\u542b\u7b97\u6cd5\u548c\u7f51\u7edc\u914d\u7f6e":68,"\u5176\u4e2d\u5305\u62ec\u51fd\u6570":67,"\u5176\u4e2d\u5b9a\u4e49\u4e86\u6a21\u578b\u67b6\u6784\u548csolver\u914d\u7f6e":69,"\u5176\u4e2d\u6570\u636e\u6e90\u914d\u7f6e\u4e0edataprovider\u7684\u5173\u7cfb\u662f":53,"\u5176\u4e2d\u6587\u672c\u8f93\u5165\u7c7b\u578b\u5b9a\u4e49\u4e3a\u6574\u6570\u65f6\u5e8f\u7c7b\u578binteger_value_sequ":64,"\u5176\u4e2d\u6bcf\u4e00\u884c\u5bf9\u5e94\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u5730\u5740":2,"\u5176\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u53cc\u5c42\u5e8f\u5217\u4e2d\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a":38,"\u5176\u4e2d\u6bcf\u4e2a\u5411\u91cf\u5bf9\u5e94\u8f93\u5165\u8bed\u53e5\u4e2d\u7684\u4e00\u4e2a\u5143\u7d20":69,"\u5176\u4e2d\u6bcf\u6761pass\u82b1\u8d39\u4e867\u4e2a\u5c0f\u65f6":69,"\u5176\u4e2d\u6bcf\u884c\u6570\u636e\u4ee3\u8868\u4e00\u5f20\u56fe\u7247":3,"\u5176\u4e2d\u8be6\u7ec6\u8bf4\u660e\u4e86\u6a21\u578b\u67b6\u6784":69,"\u5176\u4e2d\u8f93\u5165\u56fe\u50cf\u7684\u989c\u8272\u901a\u9053\u987a\u5e8f\u4e3a":62,"\u5176\u4e2dbeam":69,"\u5176\u4e2dcheckgrad\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":50,"\u5176\u4e2dmean\u548cstd\u662f\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u53c2\u6570":50,"\u5176\u4e2dvalue\u5373\u4e3asoftmax\u5c42\u7684\u8f93\u51fa":5,"\u5176\u4ed6":65,"\u5176\u4ed6\u4eba\u5f88\u5bb9\u6613\u770b\u61c2\u955c\u50cf\u751f\u6210\u8fc7\u7a0b":34,"\u5176\u4ed6\u516d\u884c\u5217\u51fa\u4e86\u96c6\u675f\u641c\u7d22\u7684\u7ed3\u679c":69,"\u5176\u4ed6\u5185\u5b58\u6742\u9879":31,"\u5176\u4ed6\u5185\u5b58\u6742\u9879\u662f\u6307paddlepaddle\u672c\u8eab\u6240\u7528\u7684\u4e00\u4e9b\u5185\u5b58":31,"\u5176\u4ed6\u51fd\u6570\u5747\u8fd4\u56de":28,"\u5176\u4ed6\u53c2\u6570\u4f7f\u7528":3,"\u5176\u4ed6\u53c2\u6570\u8bf7\u53c2\u8003":64,"\u5176\u4ed6\u6240\u6709\u5c42\u90fd\u4f1a\u4f7f\u7528gpu\u8ba1\u7b97":52,"\u5176\u4ed6\u7528\u6237\u7684fork\u7248\u672c\u5e93\u5e76\u4e0d\u9700\u8981\u4e25\u683c\u9075\u5b88":30,"\u5176\u4ed6\u9ad8\u7ea7\u529f\u80fd\u5305\u62ec\u5b9a\u4e49\u591a\u4e2amemori":42,"\u5176\u4f1a\u81ea\u52a8\u88ab\u52a0\u5165\u7f16\u8bd1\u5217\u8868":44,"\u5176\u4f59\u884c\u662f":60,"\u5176\u4f5c\u7528\u662f\u5c06\u6570\u636e\u4f20\u5165\u5185\u5b58\u6216\u663e\u5b58":2,"\u5176\u5177\u4f53\u8bf4\u660e\u4e86\u5b57\u6bb5\u7c7b\u578b\u548c\u6587\u4ef6\u540d\u79f0":66,"\u5176\u5185\u90e8\u7684\u6587\u4ef6\u4e5f\u4f1a\u968f\u4e4b\u6d88\u5931":54,"\u5176\u5305\u62ec\u4e24\u4e2a\u51fd\u6570":64,"\u5176\u53c2\u6570\u5982\u4e0b":3,"\u5176\u5b83\u90e8\u5206\u548c\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u7ed3\u6784\u4e00\u81f4":64,"\u5176\u5b83layer\u7684\u8f93\u51fa":41,"\u5176\u5b9e\u4e5f\u662f\u548c\u6bcf\u4e2amini":31,"\u5176\u63d0\u4f9b\u5e94\u7528\u90e8\u7f72":54,"\u5176\u6b21":[3,39,64],"\u5176\u76ee\u7684\u662f\u5728\u7ed9\u5b9a\u7684\u8f93\u5165\u53e5\u5b50\u4e2d\u53d1\u73b0\u6bcf\u4e2a\u8c13\u8bcd\u7684\u8c13\u8bcd\u8bba\u5143\u7ed3\u6784":67,"\u5176\u8bf4\u660e\u5982\u4e0b":39,"\u5176\u8f93\u5165\u53c2\u6570\u5982\u4e0b":69,"\u5176\u8f93\u51fa\u88ab\u7528\u4f5cmemory\u7684\u521d\u59cb\u503c":42,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":[28,31],"\u5177\u4f53\u539f\u56e0\u53c2\u8003":28,"\u5177\u4f53\u53ef\u4ee5\u53c2\u8003":[3,44],"\u5177\u4f53\u53ef\u53c2\u8003\u6587\u6863":41,"\u5177\u4f53\u5982\u4e0b":34,"\u5177\u4f53\u60c5\u51b5\u56e0\u4eba\u800c\u5f02":47,"\u5177\u4f53\u64cd\u4f5c\u5982\u4e0b":31,"\u5177\u4f53\u6d41\u7a0b\u5982\u4e0b":64,"\u5177\u4f53\u7684\u4f7f\u7528\u65b9\u6cd5\u8bf7\u53c2\u8003":53,"\u5177\u4f53\u7684\u683c\u5f0f\u8bf4\u660e":3,"\u5177\u4f53\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":31,"\u5177\u4f53\u8ba1\u7b97\u662f\u901a\u8fc7\u5185\u90e8\u7684":53,"\u5177\u4f53\u8bf7\u53c2\u7167\u793a\u4f8b":62,"\u5177\u4f53\u8bf7\u53c2\u8003":[3,28,43],"\u5177\u4f53\u8bf7\u89c1":43,"\u5177\u6709\u76f8\u540c\u7684\u7ed3\u679c\u4e86":39,"\u5177\u6709\u81ea\u5faa\u73af\u8fde\u63a5\u7684\u795e\u7ecf\u5143":68,"\u517c\u5907\u6613\u7528\u6027":0,"\u5185":42,"\u5185\u5b58":47,"\u5185\u5b58\u5bb9\u9650\u9608\u503c":50,"\u5185\u5bb9":64,"\u5185\u5bb9\u5982\u4e0b":55,"\u5185\u5c42inner_step\u7684recurrent_group\u548c\u5355\u5c42\u5e8f\u5217\u7684\u51e0\u4e4e\u4e00\u6837":39,"\u5185\u5df2\u7ecf\u5305\u542bpaddlepaddle\u7684\u6267\u884c\u7a0b\u5e8f\u4f46\u662f\u8fd8\u6ca1\u4e0a\u8ff0\u529f\u80fd":56,"\u5185\u7684\u5f00\u53d1\u8005\u4e0b\u8f7ddocker\u955c\u50cf":34,"\u5185\u90e8":56,"\u5185\u90e8\u9a71\u52a8python\u89e3\u91ca\u5668\u8fdb\u884c\u6a21\u578b\u914d\u7f6e\u89e3\u6790\u548c\u6570\u636e\u8bfb\u53d6":27,"\u518d\u4e3apaddle\u7684\u8bad\u7ec3\u8fc7\u7a0b\u63d0\u4f9b\u6587\u4ef6\u5217\u8868":66,"\u518d\u4f20\u5165\u7ed9train":53,"\u518d\u5728\u6bcf\u4e00\u4e2aapi\u4e2d\u81ea\u5df1\u68c0\u67e5\u7c7b\u578b":27,"\u518d\u57fa\u4e8e":30,"\u518d\u5bf9\u6bcf\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u8fdb\u884c\u5904\u7406":39,"\u518d\u5bf9\u6bcf\u4e00\u53e5\u8bdd\u7684\u7f16\u7801\u5411\u91cf\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u6bb5\u843d\u7684\u5411\u91cf":39,"\u518d\u5bf9\u8fd9\u4e2a\u6bb5\u843d\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":39,"\u518d\u6307\u5b9a":33,"\u518d\u6b21\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":47,"\u518d\u7528\u8fd9\u4e2a\u68af\u5ea6\u53bb\u548c":44,"\u518d\u901a\u8fc7\u51fd\u6570":56,"\u5192\u9669\u7247":65,"\u5197\u4f59\u7b49\u529f\u80fd":54,"\u5199\u4ee3\u7801":27,"\u5199\u5165\u5feb\u7167\u6570\u636e":25,"\u5199\u5b8c\u6a21\u578b\u914d\u7f6e\u4e4b\u540e":69,"\u5199\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5\u662f\u4e00\u4e2a\u9a8c\u8bc1\u65b0\u5b9e\u73b0\u7684\u5c42\u662f\u5426\u6b63\u786e\u7684\u76f8\u5bf9\u7b80\u5355\u7684\u529e\u6cd5":44,"\u519c\u6c11":65,"\u51c6\u5907":39,"\u51c6\u5907\u597d\u6570\u636e":66,"\u51c6\u5907\u7528\u6765\u5b66\u4e60\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u5e8f\u5217\u6570\u636e":42,"\u51c6\u5907\u9884\u6d4b\u6570\u636e":5,"\u51cf\u5c0f\u5e8f\u5217\u7684\u957f\u5ea6":31,"\u51cf\u5c0f\u8fd9\u4e2a\u5185\u5b58\u6c60\u5373\u53ef\u51cf\u5c0f\u5185\u5b58\u5360\u7528":31,"\u51cf\u5c0fbatch":31,"\u51fa\u53bb\u73a9":39,"\u51fa\u5dee":39,"\u51fa\u6765":39,"\u51fa\u73b0\u4ee5\u4e0b\u9519\u8bef":31,"\u51fa\u73b0\u8fd9\u4e2a\u95ee\u9898\u7684\u4e3b\u8981\u539f\u56e0\u662f":31,"\u51fd\u6570":[3,32,42,44,47,67,68],"\u51fd\u6570\u4e2d":42,"\u51fd\u6570\u4e2d\u4f7f\u7528":3,"\u51fd\u6570\u4e2d\u8bbe\u7f6e\u7684":48,"\u51fd\u6570\u5047\u8bbe":42,"\u51fd\u6570\u52a0\u5230\u4ee3\u7801\u4e2d":47,"\u51fd\u6570\u5373\u53ef\u5b8c\u6210\u8f6c\u6362":26,"\u51fd\u6570\u53ea\u5173\u6ce8\u4e8ernn\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8ba1\u7b97":41,"\u51fd\u6570\u540d\u4e3a":28,"\u51fd\u6570\u547d\u540d":27,"\u51fd\u6570\u5c06\u8fd4\u56de\u4e09\u4e2a\u6574\u6570\u5217\u8868":42,"\u51fd\u6570\u5c31\u662f\u6839\u636e\u8be5\u673a\u5236\u914d\u7f6e\u7684":3,"\u51fd\u6570\u5f97\u5230\u7684\u68af\u5ea6\u53bb\u5bf9\u6bd4":44,"\u51fd\u6570\u5fc5\u987b\u5148\u8c03\u7528\u57fa\u7c7b\u4e2d\u7684\u51fd\u6570":44,"\u51fd\u6570\u5fc5\u987b\u8fd4\u56de\u4e00\u4e2a\u6216\u591a\u4e2alayer\u7684\u8f93\u51fa":41,"\u51fd\u6570\u6307\u51fa\u4e86\u5728\u8bad\u7ec3\u65f6\u9700\u8981\u4ece\u53c2\u6570\u670d\u52a1\u5668\u53d6\u51fa\u7684\u884c":44,"\u51fd\u6570\u6765\u5c06\u4fe1\u606f\u8f93\u51fa\u5230\u754c\u9762\u4e2d":47,"\u51fd\u6570\u67e5\u8be2\u8f6f\u4ef6\u5305\u76f8\u5173api\u8bf4\u660e":5,"\u51fd\u6570\u7684":3,"\u51fd\u6570\u7684\u5b9e\u73b0\u662f\u6b63\u786e\u7684":44,"\u51fd\u6570\u7684\u5f00\u5934\u5fc5\u987b\u8c03\u7528":44,"\u5206\u4e3a\u597d\u8bc4":64,"\u5206\u522b\u4e3a":60,"\u5206\u522b\u4e3atrain":69,"\u5206\u522b\u4ece\u8bcd\u8bed\u548c\u53e5\u5b50\u7ea7\u522b\u7f16\u7801\u8f93\u5165\u6570\u636e":41,"\u5206\u522b\u4f7f\u7528\u5355\u53cc\u5c42rnn\u4f5c\u4e3a\u7f51\u7edc\u914d\u7f6e\u7684\u6a21\u578b":39,"\u5206\u522b\u5305\u542b\u4e86\u6cd5\u8bed\u5230\u82f1\u8bed\u7684\u5e73\u884c\u8bed\u6599\u5e93\u7684\u8bad\u7ec3\u6570\u636e":69,"\u5206\u522b\u5b9a\u4e49\u5b50\u53e5\u7ea7\u522b\u548c\u8bcd\u8bed\u7ea7\u522b\u4e0a\u9700\u8981\u5b8c\u6210\u7684\u8fd0\u7b97":41,"\u5206\u522b\u5bf9\u5e94\u4e8e\u53d8\u91cf":32,"\u5206\u522b\u662f":38,"\u5206\u522b\u662frnn\u72b6\u6001\u548c\u8f93\u5165\u7684\u53d8\u6362\u77e9\u9635":42,"\u5206\u522b\u662fsentences\u548clabel":39,"\u5206\u522b\u662fwords\u548clabel":39,"\u5206\u522b\u8ba1\u7b97\u6bcf\u4e2a\u53c2\u6570\u7684\u68af\u5ea6":44,"\u5206\u522b\u8fdb\u884c\u5e8f\u5217\u64cd\u4f5c":39,"\u5206\u5272":[65,67],"\u5206\u5272\u6587\u4ef6\u7684\u65b9\u6cd5\u662f":66,"\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":25,"\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":54,"\u5206\u6210\u4e24\u90e8\u5206":3,"\u5206\u652f":[30,43],"\u5206\u652f\u4e00\u65e6\u5efa\u7acb":30,"\u5206\u652f\u4e0a":43,"\u5206\u652f\u4e0a\u521b\u5efa\u65b0\u5206\u652f":43,"\u5206\u652f\u4e2d":30,"\u5206\u652f\u4e3a\u5f00\u53d1":30,"\u5206\u652f\u4e3a\u6bcf\u4e00\u6b21release\u65f6\u5efa\u7acb\u7684\u4e34\u65f6\u5206\u652f":30,"\u5206\u652f\u4e3a\u7a33\u5b9a":30,"\u5206\u652f\u529f\u80fd\u7684\u5c01\u95ed":30,"\u5206\u652f\u5408\u5165":30,"\u5206\u652f\u5408\u5165master\u5206\u652f":30,"\u5206\u652f\u540c\u6b65\u4e3b\u7248\u672c\u5e93\u7684":30,"\u5206\u652f\u540d":43,"\u5206\u652f\u540d\u4e3a":30,"\u5206\u652f\u5b58\u5728\u7684\u65f6\u5019":30,"\u5206\u652f\u6d3e\u751f\u51fa\u65b0\u7684\u5206\u652f":30,"\u5206\u652f\u7684\u7248\u672c\u90fd\u662f\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5\u548c\u56de\u5f52\u6d4b\u8bd5\u7684\u7248\u672c":30,"\u5206\u652f\u7684\u7248\u672c\u90fd\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5":30,"\u5206\u652f\u89c4\u8303":43,"\u5206\u6790\u5f97\u5230\u7684\u4fe1\u606f\u7528\u4e8e\u534f\u52a9\u8fdb\u884c\u7a0b\u5e8f\u7684\u4f18\u5316":47,"\u5206\u7247":25,"\u5206\u7c7b\u6210\u6b63\u9762\u60c5\u7eea\u548c\u8d1f\u9762\u60c5\u7eea\u4e24\u7c7b":3,"\u5206\u7c7b\u8bef\u5dee\u662f0":68,"\u5206\u7c7b\u9519\u8bef\u7387\u548c\u6a21\u578b\u5927\u5c0f\u7531\u4e0b\u8868\u7ed9\u51fa":62,"\u5206\u8bcd\u5e8f\u5217\u7684\u5f00\u59cb":60,"\u5206\u8bcd\u5e8f\u5217\u7684\u7ed3\u675f":60,"\u5206\u8bcd\u98ce\u683c\u5982\u4e0b":60,"\u5206\u914d\u5230\u5f53\u524d\u6570\u636e\u5757\u6837\u672c\u6570\u7684\u56db\u5206\u4e4b\u4e00":50,"\u5206\u9694":[60,66],"\u5206\u9694\u7b26\u4e3a":65,"\u5207\u6362\u5230":43,"\u5207\u6362\u5230\u6240\u5efa\u5206\u652f":43,"\u5217\u8868":66,"\u5217\u8868\u5982\u4e0b":3,"\u5219\u4e0d\u5728\u4e4e\u5185\u5b58\u6682\u5b58\u591a\u5c11\u6761\u6570\u636e":3,"\u5219\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":44,"\u5219\u4f1a\u9884\u5148\u8bfb\u53d6\u5168\u90e8\u6570\u636e\u5230\u5185\u5b58\u4e2d":3,"\u5219\u4f1a\u9ed8\u8ba4\u751f\u6210\u4e00\u4e2alist\u6587\u4ef6":53,"\u5219\u4f7f\u7528\u533a\u57df\u6807\u8bb0":67,"\u5219\u4f7f\u7528\u540c\u6b65\u8bad\u7ec3":50,"\u5219\u4f7f\u7528\u542f\u52a8\u53c2\u6570\u5b9a\u4e49\u7684\u521d\u59cb\u5316\u65b9\u6cd5\u521d\u59cb\u5316\u53c2\u6570":25,"\u5219\u4f7f\u7528\u8be5\u53c2\u6570\u4f5c\u4e3a\u9ed8\u8ba4\u503c":50,"\u5219\u5148\u505a\u5d4c\u5165":66,"\u5219\u53ef\u4ee5\u50cf":48,"\u5219\u5b57\u4e0e\u5b57\u4e4b\u95f4\u7528\u7a7a\u683c\u5206\u9694":64,"\u5219\u5ffd\u7565":25,"\u5219\u603b\u4f1a\u663e\u793a\u963b\u9694\u6458\u8981\u4fe1\u606f":50,"\u5219\u628a\u53e6\u4e00\u4e2a\u6162\u901f\u7684kill\u6389":25,"\u5219\u63a8\u8350\u5927\u4e8e\u8bad\u7ec3\u65f6batch":3,"\u5219\u662f\u5e26gui\u7684nvidia\u53ef\u89c6\u5316\u6027\u80fd\u5206\u6790\u5de5\u5177":47,"\u5219\u663e\u793a\u963b\u9694\u6027\u80fd\u7684\u6458\u8981\u4fe1\u606f":50,"\u5219\u76f4\u63a5\u5f15\u5165\u53e6\u4e00\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5219\u9700\u8981\u4f7f\u7528\u7b49\u4e8e\u6743\u91cd\u53c2\u6570\u89c4\u6a21\u5927\u7ea65\u500d\u7684\u5185\u5b58":31,"\u5219\u9700\u8981\u56de\u6eda\u5230\u4e0a\u4e00\u4e2a\u68c0\u67e5\u70b9":25,"\u5219\u9700\u8981\u914d\u7f6e":54,"\u521b\u5efa\u4e00\u4e2akubernet":56,"\u521b\u5efa\u548c\u53d1\u5e03\u81ea\u5df1\u7684ai\u7a0b\u5e8f\u955c\u50cf":34,"\u521b\u5efa\u5e76\u5207\u6362\u5230\u65b0\u5206\u652f":43,"\u521b\u5efa\u6210\u529f\u540e":56,"\u521b\u5efa\u8bad\u7ec3\u6570\u636e\u7684":69,"\u521b\u5efagener":3,"\u521d\u59cb\u5316\u4e4b\u540e":5,"\u521d\u59cb\u5316\u504f\u7f6e\u5411\u91cf":44,"\u521d\u59cb\u5316\u65f6\u8c03\u7528\u7684\u51fd\u6570":3,"\u521d\u59cb\u5316\u6743\u91cd\u8868":44,"\u521d\u59cb\u5316\u6a21\u578b\u7684\u8def\u5f84":50,"\u521d\u59cb\u5316\u6a21\u578b\u7684\u8def\u5f84\u914d\u7f6e\u4e3a":60,"\u521d\u59cb\u5316\u7236\u7c7b":44,"\u521d\u59cb\u5316biases_":44,"\u521d\u59cb\u5316paddlepaddle\u73af\u5883":5,"\u521d\u59cb\u72b6\u6001":41,"\u5220\u9664":43,"\u5220\u9664\u78c1\u76d8\u76ee\u5f55\u4e2d\u4e0d\u662f\u5f53\u524duuid\u7684\u5feb\u7167\u6587\u4ef6":25,"\u5229\u7528\u5206\u5e03\u5f0f\u8bad\u7ec3\u9a7e\u9a6d\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":31,"\u5229\u7528\u5355\u8bcdid\u67e5\u627e\u8be5\u5355\u8bcd\u5bf9\u5e94\u7684\u8fde\u7eed\u5411\u91cf":64,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90\u53ef\u4ee5\u5206\u4e3a\u4e00\u4e0b\u51e0\u4e2a\u65b9\u5f0f\u6765\u8fdb\u884c":31,"\u5229\u7528\u6b63\u5219\u8868\u8fbe\u5f0f\u6765\u89e3\u6790\u8be5\u7279\u5f81":66,"\u5229\u7528\u8bad\u7ec3\u96c6\u751f\u6210\u7684\u5b57\u5178":68,"\u5229\u7528\u8fd9\u79cd\u7279\u6027":41,"\u5229\u7528\u903b\u8f91\u56de\u5f52\u6a21\u578b\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":64,"\u5229\u7528kubernetes\u80fd\u65b9\u4fbf\u5730\u7ba1\u7406\u8de8\u673a\u5668\u8fd0\u884c\u5bb9\u5668\u5316\u7684\u5e94\u7528":54,"\u5229\u843d":39,"\u5230":[25,31,42],"\u5230\u6240\u6709\u8282\u70b9\u800c\u4e0d\u7528\u5bc6\u7801":48,"\u5230\u672c\u5730":43,"\u5230\u6b64":34,"\u5230\u76ee\u524d\u4e3a\u6b62":67,"\u5236\u4f5c\u65b0\u955c\u50cf\u6765\u5b8c\u6210\u4ee5\u4e0a\u7684\u5de5\u4f5c":56,"\u5236\u4f5cpaddlepaddle\u5f00\u53d1\u955c\u50cf":34,"\u5236\u4f5cpaddlepaddle\u751f\u4ea7\u955c\u50cf":34,"\u5236\u4f5cpaddlepaddle\u955c\u50cf":56,"\u5237\u7259":39,"\u524d\u4e00\u7bc7\u6587\u7ae0\u4ecb\u7ecd\u4e86\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673apaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":56,"\u524d\u4e09\u884cimport\u4e86\u5b9a\u4e49network":69,"\u524d\u53f0":39,"\u524d\u5411\u4f20\u64ad":44,"\u524d\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u5165":44,"\u524d\u5411\u548c\u540e\u5411":44,"\u524d\u81ea\u52a8\u68c0\u67e5\u4e00\u4e9b\u57fa\u672c\u4e8b\u5b9c":43,"\u5269\u4e0b\u7684pass\u4f1a\u76f4\u63a5\u4ece\u5185\u5b58\u91cc":3,"\u529f\u80fd\u7684\u6b63\u786e\u6027\u5305\u62ec\u9a8c\u8bc1paddle\u76ee\u524d\u7684":30,"\u52a0\u4e0a\u504f\u7f6e\u5411\u91cf":44,"\u52a0\u4e86l2\u6b63\u5219\u548c\u68af\u5ea6\u622a\u65ad":64,"\u52a0\u5165":47,"\u52a0\u6743\u548c\u7528\u6765\u751f\u6210":42,"\u52a0\u6743\u7f16\u7801\u5411\u91cf":42,"\u52a0\u8f7d\u6570\u636e":67,"\u52a0\u8f7d\u6a21\u578b":67,"\u52a0\u8f7d\u6a21\u578b\u53c2\u6570":69,"\u52a0\u8f7dtest":50,"\u52a0\u901fpaddlepaddle\u8bad\u7ec3\u53ef\u4ee5\u8003\u8651\u4ece\u4ee5\u4e0b\u51e0\u4e2a\u65b9\u9762":31,"\u52a8\u4f5c\u7247":65,"\u52a8\u6001\u5e93":27,"\u52a8\u753b\u7247":65,"\u52a8\u8bcd":67,"\u52a9\u624b":44,"\u5305\u542b12":68,"\u5305\u542b20\u4e2a\u8bad\u7ec3\u6837\u4f8b":60,"\u5305\u542b3\u4e2a\u5c5e\u6027":60,"\u5305\u542b50":68,"\u5305\u542b\u4e86\u67d0\u79cd\u7c7b\u578b\u7684\u7c7b\u578b\u5b9a\u4e49\u548c\u66b4\u9732\u7684\u5168\u90e8\u51fd\u6570":28,"\u5305\u542b\u7684\u6240\u6709\u4f9d\u8d56\u5047\u8bbe\u90fd\u53ef\u4ee5\u5728paddle\u7684\u8fd0\u884c\u5bb9\u5668\u4e2d":34,"\u5305\u5e76\u91cd\u65b0\u7f16\u8bd1paddlepaddl":31,"\u5305\u62ec":[26,50,64,67,69],"\u5305\u62ec\u4e86\u56fe\u50cf\u7684\u5377\u79ef":53,"\u5305\u62ec\u4ee5\u4e0b\u4e24\u79cd":3,"\u5305\u62ec\u53d1\u884c\u65f6\u95f4":65,"\u5305\u62ec\u5b57\u7b26\u4e32\u5206\u914d":31,"\u5305\u62ec\u5b66\u4e60\u7387":53,"\u5305\u62ec\u6570\u636e\u8f93\u5165":32,"\u5305\u62ec\u6743\u91cdw\u548c\u504f\u7f6eb":25,"\u5305\u62ec\u751f\u6210cpu":33,"\u5305\u62ec\u7b80\u5355\u7684":64,"\u5305\u62ecbool":52,"\u5305\u62ecdocker\u955c\u50cf":35,"\u5305\u62eclinux":34,"\u5305\u662f\u6700\u65b0\u7684":31,"\u5305\u6bd4\u8f83\u8001":31,"\u5305\u7684\u65b9\u6cd5\u662f":31,"\u533a\u522b\u662f\u540c\u65f6\u5904\u7406\u4e86\u4e24\u4e2a\u8f93\u5165":39,"\u533a\u522b\u662frnn\u4f7f\u7528\u4e24\u5c42\u5e8f\u5217\u6a21\u578b":39,"\u533b\u751f":65,"\u533b\u7597\u4fdd\u5065":65,"\u5341\u4e00":39,"\u5347\u5e8f\u6392\u5217":69,"\u534e\u6da6\u4e07\u5bb6":39,"\u534f\u540c\u5b8c\u6210releas":30,"\u5355\u4e2a\u503c":26,"\u5355\u4f4d\u662fmb":50,"\u5355\u5143\u6d4b\u8bd5\u4f1a\u5f15\u7528site":31,"\u5355\u5143\u6d4b\u8bd5checkgrad_ep":49,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u53e5\u5b50\u662f\u4e00\u6837\u7684":39,"\u5355\u53cc\u5c42rnn":40,"\u5355\u53d8\u91cf\u7684\u7ebf\u6027\u56de\u5f52":32,"\u5355\u5c42":41,"\u5355\u5c42\u4e0d\u7b49\u957frnn":39,"\u5355\u5c42\u548c\u53cc\u5c42\u5e8f\u5217\u7684\u4f7f\u7528\u548c\u793a\u4f8b2\u4e2d\u7684\u793a\u4f8b\u7c7b\u4f3c":39,"\u5355\u5c42\u5e8f\u5217":38,"\u5355\u5c42\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20":38,"\u5355\u5c42\u5e8f\u5217\u7b2ci\u4e2a\u5143\u7d20":38,"\u5355\u5c42\u6216\u53cc\u5c42":38,"\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":39,"\u5355\u5c42rnn":[39,41],"\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":39,"\u5355\u673a\u6a21\u5f0f\u7528\u547d\u4ee4":53,"\u5355\u673a\u8bad\u7ec3\u901a\u5e38\u53ea\u5305\u62ec\u4e00\u4e2atrainer\u8fdb\u7a0b":53,"\u5355\u673acpu\u8bad\u7ec3":31,"\u5355\u673agpu\u8bad\u7ec3":31,"\u5355\u6b65\u51fd\u6570":42,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u5728":42,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u90fd\u975e\u5e38\u7b80\u5355":42,"\u5355\u6b65\u51fd\u6570\u7684\u5b9e\u73b0\u5982\u4e0b\u6240\u793a":42,"\u5355\u70b9\u6545\u969c":25,"\u5355\u8fdb\u5355\u51fa":41,"\u5360\u4f4d\u7b26":60,"\u536b\u751f":39,"\u5373":[28,31,32,45,56,64,68],"\u5373\u4e00\u4e2a\u5c06\u5355\u8bcd\u5b57\u7b26\u4e32\u6620\u5c04\u5230\u5355\u8bcdid\u7684\u5b57\u5178":3,"\u5373\u4e0a\u8ff0\u4ee3\u7801\u4e2d\u7684\u7b2c19\u884c":39,"\u5373\u4e0d\u8981\u5c06\u6bcf\u4e00\u4e2a\u6837\u672c\u90fd\u653e\u5165train":3,"\u5373\u4e0d\u9700\u8981\u4f7f\u7528memori":39,"\u5373\u4e3a\u4e00\u4e2a\u65f6\u95f4\u6b65":39,"\u5373\u4e3a\u5355\u5c42rnn\u5e8f\u5217\u7684\u4f7f\u7528\u4ee3\u7801":39,"\u5373\u4e3a\u65f6\u95f4\u5e8f\u5217\u7684\u8f93\u5165":39,"\u5373\u4e3a\u8fd9\u4e2a\u53cc\u5c42rnn\u7684\u7f51\u7edc\u7ed3\u6784":39,"\u5373\u4e3a\u8fd9\u4e2a\u6570\u636e\u6587\u4ef6\u7684\u540d\u5b57":3,"\u5373\u4e8c\u7ef4\u6570\u7ec4":39,"\u5373\u4f7f\u7528":28,"\u5373\u4f7f\u7528\u6237\u76f4\u63a5\u5f15\u7528\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5373\u4f7f\u95f4\u9694\u5f88\u5c0f":50,"\u5373\u4f7fc":28,"\u5373\u4f7fprocess\u51fd\u6570\u91cc\u9762\u53ea\u6709\u4e00\u4e2ayield":3,"\u5373\u4f8b\u5982":28,"\u5373\u4fbf\u8bbe\u7f6e":31,"\u5373\u4fbfpaddl":28,"\u5373\u521d\u59cb\u72b6\u6001\u4e3a0":41,"\u5373\u5305\u542b\u65f6\u95f4\u6b65\u4fe1\u606f":3,"\u5373\u5355\u65f6\u95f4\u6b65\u6267\u884c\u7684\u51fd\u6570":42,"\u5373\u53cc\u5411lstm\u548c\u4e09\u5c42\u5806\u53e0lstm":68,"\u5373\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":39,"\u5373\u53cc\u5c42rnn\u7684\u6bcf\u4e2a\u72b6\u6001":41,"\u5373\u53ef":32,"\u5373\u53ef\u4ee5\u6781\u5927\u7684\u52a0\u901f\u6570\u636e\u8f7d\u5165\u6d41\u7a0b":31,"\u5373\u53ef\u8fdb\u884c\u4e0b\u4e00\u6b65\u64cd\u4f5c":45,"\u5373\u5728\u53cc\u5c42\u5e8f\u5217\u7684\u539f\u59cb\u6570\u636e\u4e2d":39,"\u5373\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":31,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":3,"\u5373\u5b8c\u6210\u67d0\u4e00\u4e2a\u4efb\u52a1\u7684\u6700\u5c11\u51fd\u6570":28,"\u5373\u5bf9\u7b2c\u4e09\u6b65\u8fdb\u884c\u66ff\u6362":64,"\u5373\u5c06\u4e00\u6bb5\u82f1\u6587\u6587\u672c\u6570\u636e":3,"\u5373\u5c06\u4e00\u6bb5\u8bdd\u8fdb\u884c\u5206\u7c7b":39,"\u5373\u5f53\u524d\u65f6\u95f4\u6b65\u4e0b\u7684\u795e\u7ecf\u7f51\u7edc\u4f9d\u8d56\u524d\u4e00\u4e2a\u65f6\u95f4\u6b65\u795e\u7ecf\u7f51\u7edc\u4e2d\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u8f93\u51fa":39,"\u5373\u6211\u4eec\u7684\u8bad\u7ec3\u76ee\u6807":32,"\u5373\u628a\u5355\u5c42rnn\u751f\u6210\u540e\u7684subseq\u7ed9\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42seq":41,"\u5373\u6574\u4e2a\u53cc\u5c42group\u662f\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":39,"\u5373\u6574\u4e2a\u8f93\u5165\u5e8f\u5217":38,"\u5373\u6574\u6570\u6570\u7ec4":39,"\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":39,"\u5373\u662f\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684\u7f51\u7edc\u8fde\u63a5":39,"\u5373\u66b4\u9732":28,"\u5373\u6b63\u9762\u548c\u8d1f\u9762":68,"\u5373\u6b63\u9762\u8bc4\u4ef7\u6807\u7b7e\u548c\u8d1f\u9762\u8bc4\u4ef7\u6807\u7b7e":68,"\u5373\u7279\u5f81\u7684\u6570\u7ec4":39,"\u5373\u7f51\u5361\u540d":56,"\u5373\u82e5\u5e72\u6570\u636e\u6587\u4ef6\u8def\u5f84\u7684\u67d0\u4e00\u4e2a":3,"\u5373\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0nan\u6216\u8005inf":31,"\u5373\u8bbe\u7f6e":31,"\u5373\u8fd9\u4e2a\u52a8\u6001\u5e93\u662f\u4e0d\u4f9d\u8d56\u4e8e\u5176\u4ed6\u4efb\u4f55\u6587\u4ef6\u7684":27,"\u5373define_py_data_sources2\u5e94\u6539\u4e3a":31,"\u5373input":41,"\u5373rnn\u4e4b\u95f4\u6709\u4e00\u6b21\u5d4c\u5957\u5173\u7cfb":39,"\u5377\u79ef\u5c42":61,"\u5377\u79ef\u5c42\u6743\u91cd":62,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fa8\u8bc6\u56fe\u7247\u4e2d\u7684\u4e3b\u4f53":61,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u5728\u56fe\u7247\u5206\u7c7b\u4e0a\u6709\u7740\u60ca\u4eba\u7684\u6027\u80fd":61,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u662f\u4e00\u79cd\u4f7f\u7528\u5377\u79ef\u5c42\u7684\u524d\u5411\u795e\u7ecf\u7f51\u7edc":61,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u80fd\u591f\u5f88\u597d\u7684\u8868\u793a\u8fd9\u4e24\u7c7b\u4fe1\u606f":61,"\u5377\u79ef\u7f51\u7edc\u662f\u4e00\u79cd\u7279\u6b8a\u7684\u4ece\u8bcd\u5411\u91cf\u8868\u793a\u5230\u53e5\u5b50\u8868\u793a\u7684\u65b9\u6cd5":64,"\u5378\u8f7dpaddlepaddle\u5305":31,"\u538b\u7f29\u6210\u4e00\u4e2a\u5411\u91cf":39,"\u539f\u56e0\u5728\u4e8e\u6ca1\u6709\u628a\u673a\u5668\u4e0acuda\u76f8\u5173\u7684\u9a71\u52a8\u548c\u5e93\u6620\u5c04\u5230\u5bb9\u5668\u5185\u90e8":31,"\u539f\u56e0\u662f\u672a\u8bbe\u7f6ecuda\u8fd0\u884c\u65f6\u73af\u5883\u53d8\u91cf":36,"\u53bb\u8fc7":39,"\u53c2\u6570":[3,7,8,9,10,11,12,15,16,17,18,19,20,22,27,44,49,56,60,62,68],"\u53c2\u6570\u5171\u4eab\u7684\u914d\u7f6e\u793a\u4f8b\u4e3a":31,"\u53c2\u6570\u521d\u59cb\u5316\u8def\u5f84":67,"\u53c2\u6570\u5373\u53ef":68,"\u53c2\u6570\u540d":62,"\u53c2\u6570\u6570\u91cf":64,"\u53c2\u6570\u670d\u52a1\u5668":49,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":50,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u76d1\u542c\u7aef\u53e3":50,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u7f51\u7edc\u8bbe\u5907\u540d\u79f0":50,"\u53c2\u6570\u670d\u52a1\u5668\u7684ip\u5730\u5740":50,"\u53c2\u6570\u670d\u52a1\u5668\u7a00\u758f\u66f4\u65b0\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":50,"\u53c2\u6570\u6765\u63a7\u5236\u7f13\u5b58\u65b9\u6cd5":31,"\u53c2\u6570\u6982\u8ff0":51,"\u53c2\u6570\u7684\u89e3\u6790":56,"\u53c2\u6570\u7ef4\u5ea6":60,"\u53c2\u6570\u884c":60,"\u53c2\u6570\u8bbe\u7f6e\u4e86\u5916\u5c42":39,"\u53c2\u6570\u9700\u8981\u5b9e\u73b0":42,"\u53c2\u7167\u4e0a\u8ff0\u6b65\u9aa4\u66f4\u65b0":43,"\u53c2\u8003":[27,54],"\u53c2\u8003\u5f3a\u8c03\u90e8\u5206":47,"\u53c2\u8003\u6587\u732e":69,"\u53c2\u8003\u65f6\u95f4\u5e8f\u5217":39,"\u53c2\u8003\u955c\u50cf\u7684":56,"\u53c8":39,"\u53c8\u662f\u4e00\u4e2a\u5355\u5c42\u7684\u5e8f\u5217":38,"\u53c8\u8981\u4fdd\u8bc1\u6570\u636e\u662f\u968f\u673a\u7684":31,"\u53ca":44,"\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u9690\u85cf\u72b6\u6001":42,"\u53cc\u5c42":41,"\u53cc\u5c42\u4e0d\u7b49\u957frnn":39,"\u53cc\u5c42\u5e8f\u5217":38,"\u53cc\u5c42\u5e8f\u5217\u6216\u5355\u5c42\u5e8f\u5217":38,"\u53cc\u5c42\u5e8f\u5217\u6570\u636e\u4e00\u5171\u67094\u4e2a\u6837\u672c":39,"\u53cc\u5c42\u5e8f\u5217\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u5e8f\u5217":38,"\u53cc\u5c42\u5e8f\u5217\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":41,"\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u4e2d\u6bcf\u4e2a\u5143\u7d20":38,"\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"\u53cc\u5c42\u6216\u8005\u5355\u5c42":38,"\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u7684\u4ee3\u7801":39,"\u53cc\u5c42rnn":41,"\u53cc\u5c42rnn\u6570\u636e\u968f\u610f\u52a0\u4e86\u4e00\u4e9b\u9694\u65ad":39,"\u53cc\u5c42rnn\u987e\u540d\u601d\u4e49":39,"\u53cc\u7f13\u51b2":53,"\u53cc\u8fdb\u5355\u51fa":41,"\u53cc\u8fdb\u53cc\u51fa":41,"\u53cd\u4e4b\u5219":67,"\u53cd\u5411\u4f20\u64ad":44,"\u53cd\u5411\u4f20\u64ad\u6839\u636e\u8f93\u51fa\u7684\u68af\u5ea6":44,"\u53d1\u5e03":34,"\u53d1\u5e03\u5230dockerhub":30,"\u53d1\u5e03\u5230github":30,"\u53d1\u6563\u5230\u4e86\u4e00\u4e2a\u6570\u503c\u7279\u522b\u5927\u7684\u5730\u65b9":31,"\u53d1\u884c\u548c\u7ef4\u62a4":43,"\u53d1\u9001\u53c2\u6570\u7684\u7aef\u53e3\u53f7":50,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230cuda\u5de5\u5177\u94fe":33,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gtest":33,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230swig":33,"\u53d8\u6362\u77e9\u9635":44,"\u53d8\u91cf\u6765\u8bbe\u7f6e\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761":3,"\u53e3\u5934":39,"\u53e3\u7edf\u8ba1\u5b66\u4fe1\u606f\u7684\u7528\u6237\u624d\u88ab\u5305\u542b\u5728\u6570\u636e\u96c6\u4e2d":65,"\u53e5\u5b50":68,"\u53e5\u5b50\u4e2d\u7684\u7ec4\u5757\u5c06\u4f1a\u626e\u6f14\u67d0\u4e9b\u8bed\u4e49\u89d2\u8272":67,"\u53e5\u5b50\u8868\u793a\u7684\u8ba1\u7b97\u66f4\u65b0\u4e3a\u4e24\u6b65":64,"\u53e6\u4e00\u4e2a\u4f8b\u5b50\u662f\u901a\u8fc7\u5206\u6790\u6bcf\u65e5twitter\u535a\u5ba2\u7684\u6587\u672c\u5185\u5bb9\u6765\u9884\u6d4b\u80a1\u7968\u53d8\u52a8":68,"\u53e6\u4e00\u4e2a\u597d\u5904\u662f\u6211\u4eec\u53ef\u4ee5\u628apaddlepaddle\u5bb9\u5668\u8fd0\u884c\u5728\u8fdc\u7a0b\u670d\u52a1\u5668\u4e0a":34,"\u53e6\u4e00\u4e2a\u662f\u5185\u5b58\u64cd\u4f5c\u91cf":47,"\u53e6\u4e00\u4e2a\u662f\u521b\u5efadockerfile\u5e76\u8fd0\u884cdock":34,"\u53e6\u4e00\u4e2a\u662f\u6bcf\u6761\u5e8f\u5217":31,"\u53e6\u4e00\u4e2a\u7ec8\u7aef\u8fd0\u884cpython":34,"\u53e6\u4e00\u65b9\u9762":68,"\u53e6\u4e00\u79cd\u65b9\u5f0f\u662f\u5c06\u7f51\u7edc\u5c42\u5212\u5206\u5230\u4e0d\u540c\u7684gpu\u4e0a\u53bb\u8ba1\u7b97":52,"\u53e6\u5916":[39,53],"\u53e6\u5916\u4e24\u4e2a\u5206\u522b\u662f\u6ed1\u52a8\u5747\u503c\u548c\u65b9\u5dee":62,"\u53e6\u5916\u7a00\u758f\u66f4\u65b0\u7684\u7aef\u53e3\u5982\u679c\u592a\u5927\u7684\u8bdd":53,"\u53ea\u4f5c\u4e3aread":41,"\u53ea\u4fdd\u5b58\u6700\u540e\u4e00\u8f6e\u7684\u53c2\u6570":50,"\u53ea\u5141\u8bb8\u6574\u6570\u7684\u661f\u7ea7":65,"\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":33,"\u53ea\u5bf9\u7279\u6b8a\u5728\u7ebf\u7cfb\u7edf\u8003\u8651\u4e24\u53f0\u4ee5\u4e0a\u540c\u65f6\u6545\u969c\u7684\u5bb9\u707e":25,"\u53ea\u622a\u53d6\u4e2d\u5fc3\u65b9\u5f62\u7684\u56fe\u50cf\u533a\u57df":62,"\u53ea\u662f\u53cc\u5c42\u5e8f\u5217\u5c06\u5176\u53c8\u505a\u4e86\u5b50\u5e8f\u5217\u5212\u5206":39,"\u53ea\u662f\u5c06\u53e5\u5b50\u7528\u8fde\u7eed\u5411\u91cf\u8868\u793a\u66ff\u6362\u4e3a\u7528\u7a00\u758f\u5411\u91cf\u8868\u793a":64,"\u53ea\u662f\u8bf4\u660e\u6570\u636e\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684":3,"\u53ea\u662f\u8bf7\u4e0d\u8981\u5fd8\u8bb0\u63d0\u524d\u5728\u7269\u7406\u673a\u4e0a\u5b89\u88c5gpu\u6700\u65b0\u9a71\u52a8":34,"\u53ea\u66b4\u9732\u6982\u5ff5\u7684\u63a5\u53e3":28,"\u53ea\u6709":39,"\u53ea\u67092\u4e2a\u914d\u7f6e\u4e0d\u4e00\u6837":60,"\u53ea\u6709\u542b\u6709\u4eba":65,"\u53ea\u6709\u5f53\u8bbe\u7f6e\u4e86spars":50,"\u53ea\u7528\u4e8e\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d\u6307\u5b9a\u8f93\u5165\u6570\u636e":41,"\u53ea\u80fd\u6d4b\u8bd5\u5355\u4e2a\u6a21\u578b":52,"\u53ea\u80fd\u8c03\u7528paddle\u7684\u52a8\u6001\u5e93":27,"\u53ea\u8981\u4e00\u7cfb\u5217\u7279\u5f81\u6570\u636e\u4e2d\u7684":39,"\u53ea\u8981\u51fa\u73b0\u6d6e\u70b9\u6570\u5f02\u5e38":31,"\u53ea\u8981\u5728docker\u91cc\u542f\u52a8paddlepaddle\u7684\u65f6\u5019\u7ed9\u5b83\u4e00\u4e2a\u540d\u5b57":34,"\u53ea\u8bfbmemory\u8f93\u5165":41,"\u53ea\u9488\u5bf9\u5185\u5b58":31,"\u53ea\u9700\u4e2d\u65ad":48,"\u53ea\u9700\u4f7f\u7528":48,"\u53ea\u9700\u5220\u9664\u6700\u540e\u4e00\u884c\u4e2d\u7684\u6ce8\u91ca\u5e76\u628a":68,"\u53ea\u9700\u5728linux\u4e0b\u8fd0\u884c\u5982\u4e0b\u547d\u4ee4":69,"\u53ea\u9700\u7528\u4f60\u5b9a\u4e49\u7684\u76ee\u5f55\u4fee\u6539":48,"\u53ea\u9700\u77e5\u9053\u8fd9\u662f\u4e00\u4e2a\u6807\u8bb0\u5c5e\u6027\u7684\u65b9\u6cd5\u5c31\u53ef\u4ee5\u4e86":3,"\u53ea\u9700\u8981":42,"\u53ea\u9700\u8981\u4e00\u884c\u4ee3\u7801\u5c31\u53ef\u4ee5\u8c03\u7528\u8fd9\u4e2apydataprovider2":3,"\u53ea\u9700\u8981\u5728\u51fd\u6570\u4e2d\u8c03\u7528\u591a\u6b21yield\u5373\u53ef":3,"\u53ea\u9700\u8981\u6062\u590d\u8fd9\u53f0\u8282\u70b9":25,"\u53ea\u9700\u8981\u7b80\u5355\u5730\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4":69,"\u53ea\u9700\u8981\u7b80\u5355\u7684\u8fd0\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5373\u53ef":66,"\u53ea\u9700\u8981\u8fd0\u884c":66,"\u53ef\u4ee5":[39,43,48],"\u53ef\u4ee5\u4ee5\u540e\u53f0\u8fdb\u7a0b\u65b9\u5f0f\u8fd0\u884c\u5bb9\u5668":34,"\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":3,"\u53ef\u4ee5\u4f30\u8ba1\u51fa\u5982\u679c\u6a21\u578b\u91c7\u7528\u4e0d\u53d8\u7684\u8f93\u51fa\u6700\u5c0f\u7684cost0\u662f\u591a\u5c11":31,"\u53ef\u4ee5\u4f7f\u7528":[31,53],"\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4":36,"\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u4ee3\u7801":31,"\u53ef\u4ee5\u4f7f\u7528\u8be5\u53c2\u6570":50,"\u53ef\u4ee5\u4f7f\u7528kubernetes\u7684\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":56,"\u53ef\u4ee5\u4f7f\u7528python\u7684":5,"\u53ef\u4ee5\u51cf\u5c0f\u7cfb\u7edf\u590d\u6742\u6027":25,"\u53ef\u4ee5\u51cf\u5c11\u7f13\u5b58\u6c60\u7684\u5927\u5c0f":31,"\u53ef\u4ee5\u521b\u5efa\u4e00\u4e2a":55,"\u53ef\u4ee5\u53c2\u7167\u4e0b\u9762\u7684\u547d\u4ee4\u8fdb\u884c\u5b89\u88c5":61,"\u53ef\u4ee5\u53c2\u8003":[39,42,53,54,56,69],"\u53ef\u4ee5\u53c2\u8003\u4fdd\u5b58\u5728":60,"\u53ef\u4ee5\u542f\u52a8":53,"\u53ef\u4ee5\u542f\u52a8\u4e00\u4e2atrainer\u8fdb\u7a0b":53,"\u53ef\u4ee5\u542f\u52a8\u5206\u5e03\u5f0f\u4f5c\u4e1a":53,"\u53ef\u4ee5\u544a\u8bc9\u60a8\u67d0\u4e2a\u64cd\u4f5c\u5230\u5e95\u82b1\u4e86\u591a\u957f\u65f6\u95f4":47,"\u53ef\u4ee5\u5728":34,"\u53ef\u4ee5\u5728\u4efb\u4f55\u673a\u5668\u4e0a\u6267\u884c\u7684":27,"\u53ef\u4ee5\u5728\u5171\u4eab\u5b58\u50a8\u4e0a\u67e5\u770b\u8f93\u51fa\u7684\u65e5\u5fd7\u548c\u6a21\u578b":56,"\u53ef\u4ee5\u5728\u5f88\u5927\u7a0b\u5ea6\u4e0a\u6d88\u9664\u6b67\u4e49":67,"\u53ef\u4ee5\u5728\u7f51\u7ad9\u4e0a\u627e\u5230":67,"\u53ef\u4ee5\u5728\u8fd9\u4e2a":43,"\u53ef\u4ee5\u5728kubernetes\u4e2d\u6309\u7167":54,"\u53ef\u4ee5\u5c06\u67d0\u4e00\u4e2a\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovider2":3,"\u53ef\u4ee5\u5c06\u78c1\u76d8\u4e0a\u67d0\u4e2a\u76ee\u5f55\u5171\u4eab\u7ed9\u7f51\u7edc\u4e2d\u5176\u4ed6\u673a\u5668\u8bbf\u95ee":54,"\u53ef\u4ee5\u5c06memory\u7406\u89e3\u4e3a\u4e00\u4e2a\u65f6\u5ef6\u64cd\u4f5c":41,"\u53ef\u4ee5\u5e2e\u60a8\u63d0\u4f9b\u4e00\u4e9b\u5b9a\u4f4d\u6027\u80fd\u74f6\u9888\u7684\u5efa\u8bae":47,"\u53ef\u4ee5\u5feb\u901f\u8fed\u4ee3":34,"\u53ef\u4ee5\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u7f16\u8bd1\u751f\u6210\u6587\u6863":45,"\u53ef\u4ee5\u628a\u672c\u5730\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5b58\u50a8\u96c6\u7fa4\u4e2d":26,"\u53ef\u4ee5\u6307\u5b9a\u54ea\u4e00\u4e2a\u8f93\u5165\u548c\u8f93\u51fa\u5e8f\u5217\u4fe1\u606f\u4e00\u81f4":39,"\u53ef\u4ee5\u6309\u5982\u4e0b\u7684\u7ed3\u6784\u6765\u51c6\u5907\u6570\u6910":68,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":[38,41],"\u53ef\u4ee5\u662f\u4e00\u4e2a\u975e\u5e8f\u5217":41,"\u53ef\u4ee5\u662f\u4ee5\u4e0b\u51e0\u79cd":44,"\u53ef\u4ee5\u663e\u793a\u5730\u6307\u5b9a\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7528\u4e8e\u521d\u59cb\u5316memori":41,"\u53ef\u4ee5\u6709\u4ee5\u4e0b\u4e24\u79cd":41,"\u53ef\u4ee5\u6709\u53ef\u5b66\u4e60\u7684\u53c2\u6570":53,"\u53ef\u4ee5\u6709\u6548\u51cf\u5c0f\u7f51\u7edc\u7684\u963b\u585e":50,"\u53ef\u4ee5\u6709\u6548\u7684\u907f\u514dparamet":25,"\u53ef\u4ee5\u67e5\u770b":56,"\u53ef\u4ee5\u67e5\u770b\u6b64pod\u8fd0\u884c\u7684\u5bbf\u4e3b\u673a":55,"\u53ef\u4ee5\u6d4b\u8bd5\u591a\u4e2a\u6a21\u578b":52,"\u53ef\u4ee5\u7528\u4e8e\u4ece\u5b98\u65b9\u7f51\u7ad9\u4e0a\u4e0b\u8f7dcifar":61,"\u53ef\u4ee5\u7528\u4e8e\u5c0f\u91cf\u6570\u636e\u7684\u9a8c\u8bc1":54,"\u53ef\u4ee5\u7528\u4e8e\u63a5\u6536\u548cpydataprovider2\u4e00\u6837\u7684\u8f93\u5165\u6570\u636e\u5e76\u8f6c\u6362\u6210\u9884\u6d4b\u63a5\u53e3\u6240\u9700\u7684\u6570\u636e\u7c7b\u578b":5,"\u53ef\u4ee5\u7528\u4ee5\u4e0b\u6307\u4ee4":26,"\u53ef\u4ee5\u7528\u5982\u4e0b\u547d\u4ee4":43,"\u53ef\u4ee5\u7528\u6765\u8ba1\u7b97cpu\u51fd\u6570\u6216cuda\u5185\u6838\u7684\u65f6\u95f4\u6d88\u8017":47,"\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u975e\u5e8f\u5217\u8f93\u5165":38,"\u53ef\u4ee5\u7cbe\u786e\u8bf4\u660e\u4e00\u4e2a\u957f\u8017\u65f6\u64cd\u4f5c\u7684\u5177\u4f53\u539f\u56e0":47,"\u53ef\u4ee5\u7ee7\u7eed\u5728\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f\u63d0\u4ea4\u4ee3\u7801":30,"\u53ef\u4ee5\u7f16\u5199":34,"\u53ef\u4ee5\u8003\u8651\u4f7f\u7528\u4e00\u4e9b\u4f18\u5316\u7b97\u6cd5":31,"\u53ef\u4ee5\u8bbe\u7f6e":61,"\u53ef\u4ee5\u8fd0\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u751f\u6210":66,"\u53ef\u4ee5\u8fd0\u884c\u811a\u672ctrain":61,"\u53ef\u4ee5\u9009\u62e9\u662f\u5426\u4f7f\u7528\u53c2\u6570":52,"\u53ef\u4ee5\u901a\u8fc7":54,"\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539\u8fd9\u4e24\u4e2a\u51fd\u6570\u6765\u5b9e\u73b0\u590d\u6742\u7684\u7f51\u7edc\u914d\u7f6e":42,"\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528":5,"\u53ef\u4ee5\u901a\u8fc7\u9636\u6bb5\u6027\u7684\u4fdd\u5b58\u6bcf\u4e2aparamet":25,"\u53ef\u4ee5\u901a\u8fc7show_parameter_stats_period\u8bbe\u7f6e\u6253\u5370\u53c2\u6570\u4fe1\u606f\u7b49":64,"\u53ef\u4ee5\u901a\u8fc7volume\u6216persistentvolume\u6302\u8f7d\u5b58\u50a8\u7a7a\u95f4\u5230\u6bcf\u4e2a\u5bb9\u5668\u4e2d":26,"\u53ef\u7528\u4e8e\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u89e3\u6790\u8fd9\u4e9b\u53c2\u6570":52,"\u53ef\u7528\u5728\u6d4b\u8bd5\u6216\u8bad\u7ec3\u65f6\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b":64,"\u53ef\u80fd\u4f1a\u5bfc\u81f4\u51fa\u9519":56,"\u53ef\u80fd\u4f1a\u9020\u6210\u7f51\u7edc\u62e5\u585e":25,"\u53ef\u80fd\u7684\u4ee3\u7801\u4e3a":31,"\u53ef\u80fd\u7684\u539f\u56e0\u662f":31,"\u53ef\u80fd\u7684\u53c2\u6570\u662f":53,"\u53ef\u80fd\u7684\u60c5\u51b5\u4e0b":47,"\u53ef\u9009":[3,44],"\u53f3\u8fb9\u662f":62,"\u5403":39,"\u5403\u996d":39,"\u5404\u65b9\u9762":39,"\u5404\u9879\u53c2\u6570\u7684\u8be6\u7ec6\u8bf4\u660e\u53ef\u4ee5\u5728\u547d\u4ee4\u884c\u53c2\u6570\u76f8\u5173\u6587\u6863\u4e2d\u627e\u5230":61,"\u5408":39,"\u5408\u5e76":69,"\u5408\u5e76\u6bcf\u4e2a":69,"\u5408\u7406":39,"\u540c\u65f6":[31,47],"\u540c\u65f6\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":33,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u52a0\u901f\u5f00\u59cb\u8bad\u7ec3\u524d\u6570\u636e\u8f7d\u5165\u7684\u8fc7\u7a0b":31,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u901a\u8fc7":43,"\u540c\u65f6\u4e5f\u80fd\u591f\u5f15\u5165\u66f4\u52a0\u590d\u6742\u7684\u8bb0\u5fc6\u673a\u5236":41,"\u540c\u65f6\u4f1a\u8ba1\u7b97\u5206\u7c7b\u51c6\u786e\u7387":64,"\u540c\u65f6\u4f60\u53ef\u4ee5\u4f7f\u7528":62,"\u540c\u65f6\u4f7f\u7528\u4e86l2\u6b63\u5219":64,"\u540c\u65f6\u5176\u5185\u90e8\u5b9e\u73b0\u53ef\u4ee5\u907f\u514d\u7eafcpu\u7248\u672cpaddlepaddle\u5728\u6267\u884c\u672c\u8bed\u53e5\u65f6\u53d1\u751f\u5d29\u6e83":47,"\u540c\u65f6\u518d\u5c06":30,"\u540c\u65f6\u53ef\u4ee5\u4f7f\u7528\u6237\u53ea\u5173\u6ce8\u5982\u4f55\u4ece\u6587\u4ef6\u4e2d\u8bfb\u53d6\u6bcf\u4e00\u6761\u6570\u636e":3,"\u540c\u65f6\u5728\u5185\u5b58\u91cc\u76f4\u63a5\u968f\u5373\u9009\u53d6\u6570\u636e\u6765\u505ashuffl":31,"\u540c\u65f6\u5c06\u53c2\u6570\u521d\u59cb\u5316\u4e3a":31,"\u540c\u65f6\u6211\u4eec\u5e0c\u671b\u5e7f\u5927\u5f00\u53d1\u8005\u79ef\u6781\u63d0\u4f9b\u53cd\u9988\u548c\u8d21\u732e\u6e90\u4ee3\u7801":0,"\u540c\u65f6\u63d0\u8d77":30,"\u540c\u65f6\u6b22\u8fce\u8d21\u732e\u66f4\u591a\u7684\u5b89\u88c5\u5305":35,"\u540c\u65f6\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":52,"\u540c\u65f6\u8bbe\u7f6e\u5185\u5b58\u7f13\u5b58\u529f\u80fd":31,"\u540c\u65f6\u8bbe\u7f6e\u5b83\u7684input_types\u5c5e\u6027":3,"\u540c\u65f6\u9884\u6d4b\u7f51\u7edc\u901a\u5e38\u76f4\u63a5\u8f93\u51fa\u6700\u540e\u4e00\u5c42\u7684\u7ed3\u679c\u800c\u4e0d\u662f\u50cf\u8bad\u7ec3\u7f51\u7edc\u4e00\u6837\u518d\u63a5\u4e00\u5c42cost":5,"\u540c\u6837\u4e5f\u53ef\u4ee5\u5728\u6d4b\u8bd5\u6a21\u5f0f\u4e2d\u6307\u5b9a\u6a21\u578b\u8def\u5f84":50,"\u540c\u6837\u529f\u80fd\u7684":53,"\u540c\u6837\u53ef\u4ee5\u6269\u5c55\u5230\u53cc\u5c42\u5e8f\u5217\u7684\u5904\u7406\u4e0a":41,"\u540c\u6b65\u6267\u884c\u64cd\u4f5c\u7684\u7ebf\u7a0b\u6570":50,"\u540d\u5b57\u4fee\u9970":27,"\u540d\u79f0":64,"\u540e":[31,33,43,56,68],"\u540e\u5411\u4f20\u64ad":44,"\u540e\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u51fa\u7684\u68af\u5ea6":44,"\u540e\u9762\u8fde\u5168\u8fde\u63a5\u5c42\u548csoftmax\u5c42":68,"\u5411\u6307\u5b9a\u7684\u76ee\u5f55\u4e2d\u4e00\u4e2a\u65b0\u7684\u6587\u4ef6":25,"\u5411\u91cfenable_parallel_vector":49,"\u5411paddle\u7684\u4e3b\u7248\u672c\u5e93\u63d0\u4ea4":30,"\u5426":33,"\u5426\u5219":[2,48,66],"\u5426\u5219\u4f1a\u628a":43,"\u5426\u5219\u4f60\u9700\u8981\u81ea\u5df1\u4e0b\u8f7d":69,"\u5426\u5219\u4f7f\u7528\u591a\u673a\u8bad\u7ec3":50,"\u5426\u5219\u4f7f\u7528cpu\u6a21\u5f0f":50,"\u5426\u5219\u4f7f\u7528gpu":52,"\u5426\u5219\u5b83\u4ee5\u4e00\u4e2a\u5e8f\u5217\u8f93\u5165":42,"\u5426\u5219\u5f97\u628apaddle\u9759\u6001\u5e93\u94fe\u63a5\u5230\u89e3\u91ca\u5668\u91cc":27,"\u5426\u5219\u9891\u7e41\u7684\u591a\u8282\u70b9\u5de5\u4f5c\u7a7a\u95f4\u90e8\u7f72\u53ef\u80fd\u4f1a\u5f88\u9ebb\u70e6":48,"\u5426\u5b9a":67,"\u542b\u4e49":[62,68],"\u542b\u53ef\u5b66\u4e60\u53c2\u6570":53,"\u542b\u6709":65,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u548c\u5b50\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":44,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u6574\u6570":44,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":44,"\u542f\u52a8\u4e00\u4e2a\u65b0\u7684\u7ebf\u7a0b\u5f00\u59cb\u4fdd\u5b58\u68c0\u67e5\u70b9":25,"\u542f\u52a8\u4e00\u4e2apserver\u8fdb\u7a0b":53,"\u542f\u52a8\u4e4b\u540e":53,"\u542f\u52a8\u5bb9\u5668\u5f00\u59cb\u8bad\u7ec3":56,"\u542f\u52a8\u5e76\u884c\u5411\u91cf\u7684\u9608\u503c":50,"\u542f\u52a8\u5feb\u901f\u5e94\u7b54":50,"\u542f\u7528\u68af\u5ea6\u53c2\u6570\u7684\u9608\u503c":50,"\u5440":39,"\u544a\u8bc9paddle\u54ea\u4e2a\u6587\u4ef6\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u914d\u7f6e\u6587\u4ef6":66,"\u544a\u8bc9paddle\u5c06\u6a21\u578b\u4fdd\u5b58\u5728":66,"\u5468\u56f4":39,"\u547d\u4ee4":48,"\u547d\u4ee4\u4e3a":55,"\u547d\u4ee4\u521b\u5efa\u65b0\u955c\u50cf":55,"\u547d\u4ee4\u53ef\u4ee5\u8bbe\u7f6e":33,"\u547d\u4ee4\u6307\u5b9a\u7684\u53c2\u6570\u4f1a\u4f20\u5165\u7f51\u7edc\u914d\u7f6e\u4e2d":64,"\u547d\u4ee4\u884c\u53c2\u6570\u6587\u6863":64,"\u547d\u4ee4\u8bbe\u7f6e\u8be5\u7c7b\u7f16\u8bd1\u9009\u9879":33,"\u547d\u4ee4\u9009\u9879\u5e76\u4e14":48,"\u547d\u540d\u4e3a":43,"\u547d\u540d\u7a7a\u95f4":54,"\u547d\u540d\u7a7a\u95f4\u4e3b\u8981\u4e3a\u4e86\u5bf9\u8c61\u8fdb\u884c\u903b\u8f91\u4e0a\u7684\u5206\u7ec4\u4fbf\u4e8e\u7ba1\u7406":54,"\u548c":[26,27,28,30,31,32,33,39,42,43,44,45,47,48,52,53,54,60,61,64,66,69],"\u548c\u4e00\u4e2a\u5df2\u7ecf\u5206\u8bcd\u540e\u7684\u53e5\u5b50":39,"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":3,"\u548c\u4e2d\u6587\u6587\u6863":45,"\u548c\u4e4b\u524d\u51cf\u5c0f\u901a\u8fc7\u51cf\u5c0f\u7f13\u5b58\u6c60\u6765\u51cf\u5c0f\u5185\u5b58\u5360\u7528\u7684\u539f\u7406\u4e00\u81f4":31,"\u548c\u504f\u7f6e\u5411\u91cf":44,"\u548c\u533a\u57df\u6807\u8bb0":67,"\u548c\u53cc\u5c42\u5e8f\u5217\u542b\u6709subseq":38,"\u548c\u5728":3,"\u548c\u5bf9\u8c61\u5b58\u50a8api":54,"\u548c\u5dee\u8bc4":64,"\u548c\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":38,"\u548c\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165":42,"\u548c\u68af\u5ea6\u622a\u65ad":64,"\u548c\u6a21\u578b\u8def\u5f84":68,"\u548c\u6c60\u5316":53,"\u548c\u771f\u5b9e":32,"\u548c\u793a\u4f8b2\u4e2d\u7684\u914d\u7f6e\u7c7b\u4f3c":39,"\u548c\u79bb\u7ebf\u6570\u636e\u7684\u65b9\u5f0f":26,"\u548c\u7b2c6\u884c\u7684":69,"\u548c\u90e8\u5206layer":41,"\u548cadam\u5b66\u4e60\u65b9\u6cd5":69,"\u548cargument":67,"\u548ccudnn":36,"\u548cpool":38,"\u548cpython\u63a5\u53e3\u6765\u63d0\u53d6\u7279\u5f81":62,"\u54c1\u8d28":39,"\u54ea\u4e2atrainer\u5148\u5b8c\u6210block\u7684\u8bad\u7ec3":25,"\u54ea\u4e9b\u4e0d\u662f":39,"\u552f\u4e00\u9700\u8981\u505a\u7684\u662f\u5c06\u76f8\u5e94\u7c7b\u578b\u8bbe\u7f6e\u4e3a\u8f93\u5165":42,"\u5546\u52a1":39,"\u554a":39,"\u559c\u5267\u7247":65,"\u5668":64,"\u56db\u79cd\u6570\u636e\u7c7b\u578b":3,"\u56de\u5f52\u8bef\u5dee\u4ee3\u4ef7\u5c42":32,"\u56e0\u4e3a\u5168\u8fde\u63a5\u5c42\u7684\u6fc0\u6d3b\u53ef\u4ee5\u662fsoftmax":44,"\u56e0\u4e3a\u5176\u4e3a\u8d1f\u8d23\u63d0\u4f9bgradient":53,"\u56e0\u4e3a\u5355\u4e2a\u8c13\u8bcd\u4e0d\u80fd\u7cbe\u786e\u5730\u63cf\u8ff0\u8c13\u8bcd\u4fe1\u606f":67,"\u56e0\u4e3a\u53c2\u6570":52,"\u56e0\u4e3a\u589e\u52a0\u8fd9\u4e2a\u503c":53,"\u56e0\u4e3a\u5b83\u4eec\u7684\u8ba1\u7b97\u6548\u7387\u6bd4":42,"\u56e0\u4e3a\u5b83\u6bd4":42,"\u56e0\u4e3a\u5b98\u65b9\u955c\u50cf":56,"\u56e0\u4e3a\u5bb9\u5668\u5185\u7684\u6587\u4ef6\u90fd\u662f\u6682\u65f6\u5b58\u5728\u7684":54,"\u56e0\u4e3a\u8be5\u6587\u4ef6\u53ef\u9002\u7528\u4e8e\u9884\u6d4b":61,"\u56e0\u4e3a\u8fd9\u6837\u505a\u4e5f\u6ca1\u6cd5\u4fdd\u8bc1\u6d88\u9664\u968f\u673a\u6027":25,"\u56e0\u4e3adocker\u80fd\u5728\u6240\u6709\u4e3b\u8981\u64cd\u4f5c\u7cfb\u7edf":34,"\u56e0\u4e3apaddlepaddle\u7684v2":45,"\u56e0\u4e3apython\u7684\u641c\u7d22\u8def\u5f84\u662f\u4f18\u5148\u5df2\u7ecf\u5b89\u88c5\u7684python\u5305":31,"\u56e0\u4e3aswig\u5728\u7b2c\u4e09\u65b9\u8bed\u8a00\u4e2d\u66b4\u9732\u7684\u51fd\u6570\u540d":27,"\u56e0\u6b64":[2,3,39,41,44,53],"\u56e0\u6b64\u4f7f\u7528":3,"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":39,"\u56e0\u6b64\u53ef\u4ee5\u4f7f\u7528\u8be5\u9009\u9879":60,"\u56e0\u6b64\u53ef\u80fd\u4f1a\u6709\u4e00\u4e9b\u9519\u8bef\u548c\u4e0d\u4e00\u81f4\u53d1\u751f":65,"\u56e0\u6b64\u5982\u679c\u8fd9\u4e2a\u811a\u672c\u8fd0\u884c\u5931\u8d25":61,"\u56e0\u6b64\u5b83\u662finteger_value_sub_sequ":39,"\u56e0\u6b64\u6211\u4eec\u91c7\u7528\u8f93\u51fa\u7684\u52a0\u6743\u548c":44,"\u56e0\u6b64\u6709\u4e24\u79cd\u89e3\u51b3\u65b9\u6848":3,"\u56e0\u6b64\u7528\u6237\u5e76\u4e0d\u9700\u8981\u5173\u5fc3\u5b83\u4eec":49,"\u56e0\u6b64\u8be5\u5c42\u4e2d\u6ca1\u6709\u504f\u7f6e":62,"\u56e0\u6b64\u9519\u8bef\u7684\u4f7f\u7528\u4e8c\u8fdb\u5236\u53d1\u884c\u7248\u53ef\u80fd\u4f1a\u5bfc\u81f4\u8fd9\u79cd\u9519\u8bef":31,"\u56e0\u6b64init_hook\u5c3d\u91cf\u4f7f\u7528":3,"\u56e2\u8d2d\u7f51\u7ad9":68,"\u56fe":[62,68],"\u56fe2\u662f\u53cc\u5411lstm\u7f51\u7edc":68,"\u56fe3\u662f\u4e09\u5c42lstm\u7ed3\u6784":68,"\u56fe\u4e2d\u6bcf\u4e2a\u7070\u8272\u65b9\u5757\u662f\u4e00\u53f0\u673a\u5668":53,"\u56fe\u50cf\u5206\u7c7b":[30,63],"\u56fe\u50cf\u5927\u5c0f\u4e3a3":62,"\u56fe\u50cf\u63cf\u8ff0":69,"\u56fe\u7247\u5206\u4e3a10\u7c7b":61,"\u56fe\u7684\u5e95\u90e8\u662fword":68,"\u56fe\u8868":[34,68],"\u5728":[3,28,30,36,38,39,42,43,48,53,62,64,65,67],"\u5728\u4e00\u4e2a\u4e0d\u53ef\u4e2d\u65ad\u5e76\u7f3a\u5c11\u5907\u4efd\u7684\u8bad\u7ec3\u4efb\u52a1\u4e2d":25,"\u5728\u4e00\u4e2a\u529f\u80fd\u9f50\u5168\u7684kubernetes\u673a\u7fa4\u91cc":55,"\u5728\u4e00\u4e2a\u53c2\u6570\u7684\u68af\u5ea6\u88ab\u66f4\u65b0\u540e":44,"\u5728\u4e00\u4e2a\u5468\u671f\u5185\u6d4b\u8bd5\u6240\u6709\u6570\u636e":67,"\u5728\u4e00\u8f6e\u4e2d\u6bcfsave":50,"\u5728\u4e0a\u56fe\u4e2d\u663e\u793a\u4e86\u5728\u4e00\u4e2a\u5b9e\u9645\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u5e94\u7528":26,"\u5728\u4e0a\u9762\u4ee3\u7801\u4e2d":39,"\u5728\u4e0b\u4e00\u7bc7\u4e2d":55,"\u5728\u4e0b\u9762\u4f8b\u5b50\u91cc":64,"\u5728\u4e0b\u9762\u7684\u4f8b\u5b50\u4e2d":61,"\u5728\u4e0d\u540c\u64cd\u4f5c\u7cfb\u7edf":54,"\u5728\u4e0d\u540c\u7684\u5e94\u7528\u91cc":53,"\u5728\u4e4b\u540e\u7684":31,"\u5728\u4efb\u610f\u957f\u5ea6\u8bed\u53e5\u7ffb\u8bd1\u7684\u573a\u666f\u4e0b\u90fd\u53ef\u4ee5\u89c2\u5bdf\u5230\u5176\u6548\u679c\u7684\u63d0\u5347":69,"\u5728\u4f7f\u7528\u5b83\u4e4b\u524d\u8bf7\u5b89\u88c5paddlepaddle\u7684python":68,"\u5728\u5168\u8fde\u63a5\u5c42\u4e2d":44,"\u5728\u51c6\u5907\u53d1\u8d77":43,"\u5728\u51fa\u73b0\u5355\u70b9\u6545\u969c\u65f6":25,"\u5728\u51fd\u6570":56,"\u5728\u5206\u5e03\u5f0f\u73af\u5883\u4e2d\u6d4b\u8bd5":50,"\u5728\u5206\u5e03\u5f0f\u8bad\u7ec3\u4e2d":50,"\u5728\u5355\u5c42\u6570\u636e\u7684\u57fa\u7840\u4e0a":39,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u52a0\u8f7d\u548c\u4fdd\u5b58\u53c2\u6570":50,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u7ec8\u7aef\u6bcflog":50,"\u5728\u53cc\u5c42rnn\u4e2d\u7684\u7ecf\u5178\u60c5\u51b5\u662f\u5c06\u5185\u5c42\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u6570\u636e":39,"\u5728\u53cd\u5411\u4f20\u9012\u7684\u65f6\u5019":31,"\u5728\u53d8\u6362\u65f6\u9700\u8981\u5c06\u8f93\u5165\u5e8f\u5217\u4f20\u5165":39,"\u5728\u5404\u4e2a\u673a\u5668\u4e0a\u8fd0\u884c\u5982\u4e0b\u547d\u4ee4":53,"\u5728\u540c\u4e00\u4e2a\u547d\u540d\u7a7a\u95f4\u4e2d":54,"\u5728\u542f\u52a8job\u4e4b\u524d":56,"\u5728\u56fd\u5185":34,"\u5728\u58f0\u660edataprovider\u7684\u65f6\u5019\u4f20\u5165dictionary\u4f5c\u4e3a\u53c2\u6570":3,"\u5728\u591acpu\u8bad\u7ec3\u65f6\u5171\u4eab\u8be5\u53c2\u6570":50,"\u5728\u5b9e\u73b0\u8fc7\u7a0b\u4e2d":28,"\u5728\u5bb9\u5668\u521b\u5efa\u540e":56,"\u5728\u5bf9\u5bb9\u5668\u7684\u63cf\u8ff0":56,"\u5728\u5c42\u4e2d\u6307\u5b9a":52,"\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d":41,"\u5728\u5f00\u59cb\u8bad\u7ec3\u4e4b\u524d":61,"\u5728\u5f02\u6784\u96c6\u7fa4\u4e2d":25,"\u5728\u5f15\u5165\u5176\u4ed6\u7c7b\u578b\u7684\u5934\u6587\u4ef6\u65f6":28,"\u5728\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":44,"\u5728\u5f97\u5230":56,"\u5728\u5feb\u7167\u5199\u5165\u5b8c\u6210\u540e":25,"\u5728\u60a8\u7684\u5b9e\u9645\u73af\u5883\u4e2d":25,"\u5728\u6211\u4eec\u7684\u4f8b\u5b50\u4e2d":42,"\u5728\u6211\u4eec\u7684\u6d4b\u8bd5\u4e2d":68,"\u5728\u63d0\u4ea4":43,"\u5728\u6570\u636e\u52a0\u8f7d\u548c\u7f51\u7edc\u914d\u7f6e\u5b8c\u6210\u4e4b\u540e":64,"\u5728\u6570\u636e\u96c6\u53ef\u4ee5\u88ab\u8bad\u7ec3\u4e4b\u524d":26,"\u5728\u6587\u4ef6":66,"\u5728\u6587\u4ef6\u7684\u5f00\u59cb":53,"\u5728\u6709\u65b0\u7684\u5355\u8bcd\u6765\u4e34\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u9aa4\u5185":68,"\u5728\u672c\u4f8b\u4e2d":[39,43,52],"\u5728\u672c\u4f8b\u4e2d\u6ca1\u6709\u4f7f\u7528":3,"\u5728\u672c\u6559\u7a0b\u4e2d":[42,61],"\u5728\u672c\u6587\u4e2d":48,"\u5728\u672c\u6587\u4e2d\u4f7f\u7528\u7684":48,"\u5728\u672c\u6f14\u793a\u4e2d":68,"\u5728\u672c\u793a\u4f8b\u4e2d":[39,68],"\u5728\u672c\u8282\u4e2d":42,"\u5728\u6811\u7684\u6bcf\u4e00\u5c42\u4e0a":50,"\u5728\u6837\u4f8b\u4e2d":28,"\u5728\u6a21\u578b\u6587\u4ef6\u7684":48,"\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u901a\u8fc7":64,"\u5728\u6b64":[0,49,52],"\u5728\u6b64\u4e3a\u65b9\u4fbf\u5bf9\u6bd4\u4e0d\u540c\u7f51\u7edc\u7ed3\u6784":64,"\u5728\u6b64\u611f\u8c22":60,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":42,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u76f8\u7b49":39,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u957f":42,"\u5728\u6bcf\u4e2a\u673a\u5668\u4e2d":53,"\u5728\u6bcf\u4e2apod\u4e0a\u90fd\u901a\u8fc7volume\u65b9\u5f0f\u6302\u8f7d\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u7684\u4e00\u4e2a\u76ee\u5f55\u7528\u4e8e\u4fdd\u5b58\u8bad\u7ec3\u6570\u636e\u548c\u8f93\u51fa\u7ed3\u679c":56,"\u5728\u6bcf\u8bad\u7ec3":66,"\u5728\u6d4b\u8bd5\u9636\u6bb5":50,"\u5728\u6d4b\u8bd5\u9636\u6bb5\u5b83\u4eec\u5c06\u4f1a\u88ab\u52a0\u8f7d\u5230\u6a21\u578b\u4e2d":62,"\u5728\u6e90\u7801\u76ee\u5f55\u6811\u7684\u6839\u76ee\u5f55\u4e2d\u8fd0\u884c":43,"\u5728\u6f14\u793a\u4e2d":67,"\u5728\u7269\u7406\u673a\u4e0a\u624b\u52a8\u90e8\u7f72":54,"\u5728\u751f\u6210\u65f6":42,"\u5728\u751f\u6210\u8fc7\u7a0b\u4e2d":69,"\u5728\u751f\u6210\u8fc7\u7a0b\u4e2d\u6211\u4eec\u4f7f\u7528sgd\u8bad\u7ec3\u7b97\u6cd5":69,"\u5728\u7528\u6237\u4f7f\u7528c":28,"\u5728\u7528\u6237\u6587\u4ef6user":66,"\u5728\u7535\u5f71\u6587\u4ef6movi":66,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u4f7f\u7528attention\u7248\u672c\u7684gru\u7f16\u89e3\u7801\u7f51\u7edc":69,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u4f7f\u7528sgd\u8bad\u7ec3\u7b97\u6cd5":69,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u5b9a\u4e49\u4e86\u4e00\u4e2a\u5e8f\u5217\u5230\u5e8f\u5217\u7684\u751f\u6210\u6570\u636e":69,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u5b9a\u4e49\u4e86\u4e00\u4e2a\u5e8f\u5217\u5230\u5e8f\u5217\u7684\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6570\u636e":69,"\u5728\u7a0b\u5e8f\u5f00\u59cb\u9636\u6bb5":5,"\u5728\u7b2c\u4e00\u884c\u4e2d\u6211\u4eec\u8f7d\u5165\u7528\u4e8e\u5b9a\u4e49\u7f51\u7edc\u7684\u51fd\u6570":61,"\u5728\u7ebf\u6a21\u578b\u9884\u6d4b\u670d\u52a1":26,"\u5728\u7f16\u8bd1\u5b89\u88c5\u6b63\u786e\u4e4b\u540e":45,"\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":44,"\u5728\u7f51\u7edc\u914d\u7f6e\u91cc":3,"\u5728\u7ffb\u8bd1\u6cd5\u8bed\u53e5\u5b50\u4e4b\u524d":69,"\u5728\u811a\u672c":66,"\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d":38,"\u5728\u8bad\u7ec3\u4e2d":42,"\u5728\u8bad\u7ec3\u4e4b\u524d":56,"\u5728\u8bad\u7ec3\u4e86":66,"\u5728\u8bad\u7ec3\u4e86\u51e0\u4e2a\u8f6e\u6b21\u4ee5\u540e":66,"\u5728\u8bad\u7ec3\u5b8c\u6210\u540e":61,"\u5728\u8bad\u7ec3\u6570\u96c6\u4e0a\u8bad\u7ec3\u751f\u6210\u8bcd\u5411\u91cf\u5b57\u5178":60,"\u5728\u8bad\u7ec3\u65f6":55,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":[56,69],"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6bcfshow":50,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u8fdb\u884c\u6d4b\u8bd5":2,"\u5728\u8bc4\u5ba1\u8fc7\u7a0b\u4e2d":30,"\u5728\u8be5\u914d\u7f6e\u76847":39,"\u5728\u8bed\u8a00\u751f\u6210\u9886\u57df\u4e2d":69,"\u5728\u8d2d\u7269\u7f51\u7ad9\u4e0a":64,"\u5728\u8f6f\u4ef6\u5de5\u7a0b\u7684\u8303\u7574\u91cc":47,"\u5728\u8f93\u51fa\u7684\u8fc7\u7a0b\u4e2d":41,"\u5728\u8fd0\u884c":68,"\u5728\u8fd9\u4e2a":30,"\u5728\u8fd9\u4e2a\u4efb\u52a1\u4e2d":69,"\u5728\u8fd9\u4e2a\u4f8b\u5b50\u4e2d":[32,68],"\u5728\u8fd9\u4e2a\u4f8b\u5b50\u91cc":[44,55],"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":39,"\u5728\u8fd9\u4e2a\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":27,"\u5728\u8fd9\u4e2a\u6559\u7a0b\u4e2d":47,"\u5728\u8fd9\u4e2a\u6a21\u578b\u4e2d":42,"\u5728\u8fd9\u4e2a\u793a\u4f8b\u4e2d":67,"\u5728\u8fd9\u4e2a\u9636\u6bb5\u7684\u4ee3\u7801\u6b63\u5728\u7ecf\u5386\u56de\u5f52\u6d4b\u8bd5":30,"\u5728\u8fd9\u4e9b\u5934\u6587\u4ef6\u4e2d":28,"\u5728\u8fd9\u4e9b\u6587\u4ef6\u4e2d":28,"\u5728\u8fd9\u4e9b\u7f51\u7edc\u4e2d":66,"\u5728\u8fd9\u4e9blayer\u4e2d":39,"\u5728\u8fd9\u6b65\u4efb\u52a1\u4e2d":68,"\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b":[42,44],"\u5728\u8fd9\u79cd\u7ed3\u6784\u4e2d":41,"\u5728\u8fd9\u7bc7\u6587\u6863\u91cc":55,"\u5728\u8fd9\u7bc7\u6587\u7ae0\u91cc":56,"\u5728\u8fd9\u91cc":41,"\u5728\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u5168\u8fde\u63a5\u5c42\u4f5c\u4e3a\u4f8b\u5b50\u6765\u5c55\u793a\u5b9e\u73b0\u65b0\u7f51\u7edc\u5c42\u6240\u9700\u8981\u7684\u56db\u4e2a\u6b65\u9aa4":44,"\u5728\u914d\u7f6e\u4e2d\u9700\u8981\u8bfb\u53d6\u5916\u90e8\u5b57\u5178":3,"\u5728\u914d\u7f6e\u6587\u4ef6\u4e2d\u7684":62,"\u5728\u91c7\u7528sgd":31,"\u5728\u96c6\u7fa4\u4e0a\u8bad\u7ec3\u4e00\u4e2a\u7a00\u758f\u6a21\u578b\u9700\u8981\u52a0\u4e0a\u4e0b\u9762\u7684\u53c2\u6570":52,"\u5728\u96c6\u7fa4\u8fd0\u884cmapreduc":26,"\u5728\u9884\u5904\u7406\u542b\u6709\u591a\u884c\u6570\u6910\u7684\u6587\u672c\u6587\u4ef6\u65f6\u53c2\u6570\u8bbe\u7f6e\u7a0d\u6709\u4e0d\u540c":68,"\u5728\u9884\u6d4b\u5e8f\u5217\u6216\u6bb5\u843d\u7684\u60c5\u611f\u4e2d\u8d77\u4e3b\u8981\u4f5c\u7528":68,"\u5728aws\u4e0a\u5feb\u901f\u90e8\u7f72\u96c6\u7fa4":54,"\u5728c":27,"\u5728c\u7684\u5934\u6587\u4ef6":27,"\u5728cephfs\u5b58\u50a8\u7cfb\u7edf\u4e2d\u7684\u516c\u5f00\u76ee\u5f55":26,"\u5728cub":61,"\u5728generator\u7684\u4e0a\u4e0b\u6587\u4e2d\u5c3d\u91cf\u7559\u4e0b\u975e\u5e38\u5c11\u7684\u53d8\u91cf\u5f15\u7528":3,"\u5728kubernetes\u4e0a\u8fd0\u884c\u7684\u4e0d\u540c\u7684\u8ba1\u7b97\u6846\u67b6":26,"\u5728kubernetes\u4e2d\u521b\u5efa\u7684\u6240\u6709\u8d44\u6e90\u5bf9\u8c61":54,"\u5728linux\u4e0b":69,"\u5728meta\u6587\u4ef6\u4e2d\u6709\u4e24\u79cd\u7279\u5f81":66,"\u5728movielen":66,"\u5728paddl":56,"\u5728paddle\u4e2d":52,"\u5728paddle\u4e4b\u4e0a\u8fd0\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u8f93\u51fa\u7684\u6a21\u578b\u4f1a\u63d0\u4f9b\u7ed9\u5728\u7ebf\u4eba\u8138\u8bc6\u522b\u7684\u5e94\u7528\u4f7f\u7528":26,"\u5728paddlepaddle\u4e2d":41,"\u5728paddlepaddle\u7684\u6587\u6863\u4e2d":39,"\u5728paddlepaddle\u91cc":32,"\u5728paramet":25,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49":41,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49memori":41,"\u5728trainer":52,"\u5730\u5740\u4e5f\u53ef\u4ee5\u4e3ahdfs\u6587\u4ef6\u8def\u5f84":2,"\u5730\u6bb5":39,"\u5730\u7406\u4f4d\u7f6e":39,"\u5730\u94c1\u7ad9":39,"\u5747\u4f1a\u88ab\u5b89\u88c5\u5230includ":28,"\u5747\u503c\u56fe\u50cf\u6587\u4ef6":62,"\u5747\u5300\u5206\u5e03":31,"\u5747\u5300\u5206\u5e03\u7684\u8303\u56f4\u662f":50,"\u5747\u662f\u5728":28,"\u5747\u6709\u4e09\u4e2a\u5b50\u5e8f\u5217":39,"\u5747\u6709\u4e24\u7ec4\u7279\u5f81":39,"\u57fa\u4e8e\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":41,"\u57fa\u4e8e\u5b57\u6bcd\u7684\u8bcd\u5d4c\u5165\u7279\u5f81":66,"\u57fa\u4e8epython\u7684\u6a21\u578b\u9884\u6d4b":5,"\u57fa\u4e8epython\u7684\u9884\u6d4b":[4,64],"\u57fa\u672c\u4e0a\u548cmnist\u6837\u4f8b\u4e00\u81f4":3,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":46,"\u57fa\u672c\u76f8\u540c":60,"\u586b\u5199":43,"\u589e\u52a0\u4e86\u4e00\u6761cd\u547d\u4ee4":55,"\u589e\u52a0\u5982\u4e0b\u53c2\u6570":52,"\u589e\u52a0\u68af\u5ea6\u68c0\u6d4b\u7684\u5355\u5143\u6d4b\u8bd5":44,"\u58f0\u660epython\u6570\u636e\u6e90":66,"\u5904\u7406\u5668\u6709\u4e24\u4e2a\u5173\u952e\u6027\u80fd\u9650\u5236":47,"\u5904\u7406\u6570\u636e\u7684python\u811a\u672c\u6587\u4ef6":64,"\u5904\u7406\u7684\u8f93\u5165\u5e8f\u5217\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u4e09\u79cd\u7c7b\u578b":41,"\u5904\u7406\u76f8\u4f3c\u5ea6\u56de\u5f52":66,"\u5904\u7406\u8fc7\u7a0b\u4e2d\u6570\u636e\u5b58\u50a8\u683c\u5f0f":61,"\u5904\u7406batch":53,"\u5907\u6ce8":47,"\u590d\u6742\u5ea6\u6216\u65f6\u95f4\u590d\u6742\u5ea6":47,"\u5916\u5c42memory\u662f\u4e00\u4e2a\u5143\u7d20":39,"\u5916\u5c42outer_step\u4e2d":39,"\u591a\u4e2a\u503c":26,"\u591a\u4e2ainput\u4ee5list\u65b9\u5f0f\u8f93\u5165":64,"\u591a\u4e2aparamet":25,"\u591a\u53e5\u8bdd\u8fdb\u4e00\u6b65\u6784\u6210\u4e86\u6bb5\u843d":41,"\u591a\u673a\u8bad\u7ec3":31,"\u591a\u673a\u8bad\u7ec3\u7684\u7ecf\u5178\u62d3\u6251\u7ed3\u6784\u5982\u4e0b":53,"\u591a\u7ebf\u7a0b\u7684\u6570\u636e\u8bfb\u53d6":3,"\u591a\u8f6e\u5bf9\u8bdd\u7b49\u66f4\u4e3a\u590d\u6742\u7684\u8bed\u8a00\u6570\u636e":41,"\u5927\u578b\u7535\u5f71\u8bc4\u8bba\u6570\u636e\u96c6":68,"\u5927\u591a\u6570\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u51fd\u6570":44,"\u5927\u591a\u6570\u5c42\u9700\u8981\u8bbe\u7f6e\u4e3a":44,"\u5927\u591a\u6570\u6210\u529f\u7684srl\u7cfb\u7edf\u662f\u5efa\u7acb\u5728\u67d0\u79cd\u5f62\u5f0f\u7684\u53e5\u6cd5\u5206\u6790\u7ed3\u679c\u4e4b\u4e0a\u7684":67,"\u5927\u591a\u6570\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u652f\u6301\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":44,"\u5927\u591a\u6570\u8bed\u8a00\u90fd\u652f\u6301\u4f7f\u7528c\u8bed\u8a00api":27,"\u5927\u5b66\u751f":65,"\u5927\u5bb6\u53ef\u4ee5\u901a\u8fc7\u5b83\u5236\u4f5c\u548c\u5206\u4eab\u5e26\u6709\u4ee3\u7801":34,"\u5927\u5c0f":48,"\u5929":39,"\u5929\u4e00\u5e7f\u573a":39,"\u5929\u4e00\u9601":39,"\u5929\u732b":68,"\u5934\u6587\u4ef6\u4e2d\u628a\u53c2\u6570\u5b9a\u4e49\u4e3a\u7c7b\u7684\u6210\u5458\u53d8\u91cf":44,"\u5934\u6587\u4ef6\u5982\u4e0b":44,"\u5947\u5e7b\u7247":65,"\u597d":39,"\u597d\u5403":39,"\u597d\u8bc4":64,"\u5982":[3,42,43,48,52,53],"\u59822":48,"\u5982\u4e0b":[3,66,68],"\u5982\u4e0b\u56fe\u6240\u793a":[39,47,61],"\u5982\u4e0b\u6240\u793a":[52,62,66],"\u5982\u4e0b\u662f\u4e00\u6bb5\u4f7f\u7528mnist":5,"\u5982\u4e0b\u8868\u683c":64,"\u5982\u4f55":66,"\u5982\u4f55\u5b58\u50a8\u7b49\u7b49":3,"\u5982\u4f55\u89e3\u6790\u8be5\u5730\u5740\u4e5f\u662f\u7528\u6237\u81ea\u5b9a\u4e49dataprovider\u65f6\u9700\u8981\u8003\u8651\u7684\u5730\u65b9":2,"\u5982\u4f55\u8d21\u732e":46,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":46,"\u5982\u4f55\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":64,"\u5982\u4fe1\u606f\u63d0\u53d6":67,"\u5982\u56fe2\u6240\u793a":68,"\u5982\u56fe\u4e2dtrainer":25,"\u5982\u5f62\u5bb9\u8bcd\u548c\u526f\u8bcd":68,"\u5982\u60f3\u4e86\u89e3\u66f4\u591a\u8be6\u7ec6\u7684\u89e3\u91ca":69,"\u5982\u672c\u4f8b\u4e2d":3,"\u5982\u672c\u4f8b\u7684":3,"\u5982\u679c\u4e00\u4e2a\u7f51\u7edc\u5c42\u9700\u8981\u914d\u7f6e\u7684\u8bdd":44,"\u5982\u679c\u4e0a\u9762\u4e24\u6b65\u51fa\u73b0\u9519\u8bef":25,"\u5982\u679c\u4e0b\u8f7d\u6210\u529f":62,"\u5982\u679c\u4e0d\u4e3a0":50,"\u5982\u679c\u4e0d\u4e86\u89e3":3,"\u5982\u679c\u4e0d\u5207\u8bcd":64,"\u5982\u679c\u4e0d\u6536\u655b":31,"\u5982\u679c\u4e0d\u662f\u5e8f\u5217":66,"\u5982\u679c\u4e3a":3,"\u5982\u679c\u4e3a0":50,"\u5982\u679c\u4e3afals":50,"\u5982\u679c\u4e3atrue":[3,50],"\u5982\u679c\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe\u7f6e":33,"\u5982\u679c\u4ed4\u7ec6\u8bbe\u7f6e\u7684\u8bdd":50,"\u5982\u679c\u4f20\u5165\u4e00\u4e2alist\u7684\u8bdd":53,"\u5982\u679c\u4f20\u5165\u5b57\u7b26\u4e32\u7684\u8bdd":53,"\u5982\u679c\u4f60\u4e0d\u9700\u8981\u8fd9\u4e2a\u64cd\u4f5c":68,"\u5982\u679c\u4f60\u53ea\u9700\u8981\u4f7f\u7528\u7b80\u5355\u7684rnn":42,"\u5982\u679c\u4f60\u5b89\u88c5gpu\u7248\u672c\u7684paddlepaddl":68,"\u5982\u679c\u4f60\u60f3\u4f7f\u7528\u8fd9\u4e9b\u7279\u6027":52,"\u5982\u679c\u4f60\u60f3\u8981\u4fdd\u5b58\u67d0\u4e9b\u5c42\u7684\u7279\u5f81\u56fe":50,"\u5982\u679c\u4f60\u60f3\u8fdb\u884c\u8bf8\u5982\u8bed\u4e49\u8f6c\u8ff0":69,"\u5982\u679c\u4f60\u6267\u884c\u5176\u5b83\u7684\u7528\u60c5\u611f\u5206\u6790\u6765\u5206\u7c7b\u6587\u672c\u7684\u4efb\u52a1":68,"\u5982\u679c\u4f60\u6b63\u5728\u5904\u7406\u5e8f\u5217\u6807\u8bb0\u4efb\u52a1":42,"\u5982\u679c\u4f60\u6ca1\u6709gpu\u73af\u5883":61,"\u5982\u679c\u4f60\u8981\u4e3a\u4e86\u6d4b\u8bd5\u800c\u589e\u52a0\u65b0\u7684\u6587\u4ef6":44,"\u5982\u679c\u4f7f\u7528":[48,60],"\u5982\u679c\u4f7f\u7528gpu\u7248\u672c\u7684paddlepaddl":36,"\u5982\u679c\u4f7f\u7528nvidia":34,"\u5982\u679c\u4f7f\u7528ssl\u8ba4\u8bc1":54,"\u5982\u679c\u4f7f\u7528swig\u6211\u4eec\u9700\u8981\u5c06\u5728interface\u6587\u4ef6\u91cc":27,"\u5982\u679c\u51fa\u73b0\u4ee5\u4e0bpython\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86\u7684\u60c5\u51b5":31,"\u5982\u679c\u53c2\u6570\u4fdd\u5b58\u4e0b\u6765\u7684\u6a21\u578b\u76ee\u5f55":31,"\u5982\u679c\u53c2\u6570\u6a21\u578b\u6587\u4ef6\u7f3a\u5931":60,"\u5982\u679c\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u672a\u8bbe\u7f6easync":50,"\u5982\u679c\u5728\u8bad\u7ec3\u671f\u95f4\u540c\u65f6\u53d1\u8d77\u53e6\u5916\u4e00\u4e2a\u8fdb\u7a0b\u8fdb\u884c\u6d4b\u8bd5":50,"\u5982\u679c\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6ebatch":50,"\u5982\u679c\u5728\u8bad\u7ec3nlp\u76f8\u5173\u6a21\u578b\u65f6":31,"\u5982\u679c\u5931\u8d25":30,"\u5982\u679c\u5b58\u5728\u67d0\u4e9btrainer\u6267\u884c\u901f\u5ea6\u8fc7\u6162\u4f1a\u5f71\u54cd\u6574\u4f53\u96c6\u7fa4\u7684\u901f\u5ea6":25,"\u5982\u679c\u5b83\u4f4d\u4e8e\u8c13\u8bcd\u4e0a\u4e0b\u6587\u533a\u57df\u4e2d":67,"\u5982\u679c\u5c06\u8fd9\u4e2a\u5185\u5b58\u6c60\u51cf\u5c0f":31,"\u5982\u679c\u5df2\u5b89\u88c5":67,"\u5982\u679c\u5df2\u7ecf\u6709pod\u8fd0\u884c":56,"\u5982\u679c\u5df2\u7ecf\u6b63\u5728\u6267\u884c\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":25,"\u5982\u679c\u5f00\u542f\u4f1a\u5bfc\u81f4\u8fd0\u884c\u7565\u6162":33,"\u5982\u679c\u60a8":34,"\u5982\u679c\u60a8\u60f3\u8981\u66f4\u6df1\u5165\u4e86\u89e3deep":34,"\u5982\u679c\u60a8\u6709\u597d\u7684\u5efa\u8bae\u6765":66,"\u5982\u679c\u60a8\u7684gpu\u7406\u8bba\u53ef\u4ee5\u8fbe\u52306":47,"\u5982\u679c\u60f3\u4e3a\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u8fd4\u56de\u591a\u6761\u6837\u672c":3,"\u5982\u679c\u60f3\u4f7f\u7528\u53ef\u89c6\u5316\u7684\u5206\u6790\u5668":47,"\u5982\u679c\u60f3\u5f88\u597d\u7684\u7406\u89e3\u7a0b\u5e8f\u7684\u884c\u4e3a":47,"\u5982\u679c\u60f3\u8981\u4e86\u89e3\u53cc\u5c42rnn\u5728\u5177\u4f53\u95ee\u9898\u4e2d\u7684\u4f7f\u7528":39,"\u5982\u679c\u60f3\u8981\u542f\u7528paddlepaddle\u7684\u5185\u7f6e\u5b9a\u65f6\u5668":47,"\u5982\u679c\u6211\u77e5\u9053\u5185\u6838\u82b1\u4e8610ms\u6765\u79fb\u52a81gb\u6570\u636e":47,"\u5982\u679c\u6267\u884c\u5931\u8d25":54,"\u5982\u679c\u6267\u884c\u6210\u529f":62,"\u5982\u679c\u63d0\u793a\u6b63\u786e":45,"\u5982\u679c\u63d0\u793a\u9519\u8bef":45,"\u5982\u679c\u6570\u636e\u6587\u4ef6\u5b58\u4e8e\u672c\u5730\u78c1\u76d8":2,"\u5982\u679c\u6570\u636e\u89c4\u6a21\u6bd4\u8f83\u5927":53,"\u5982\u679c\u6570\u6910\u83b7\u53d6\u6210\u529f":68,"\u5982\u679c\u662f\u4f7f\u7528\u975essl\u65b9\u5f0f\u8bbf\u95ee":54,"\u5982\u679c\u662f\u5176\u5b83\u7c7b\u578b":26,"\u5982\u679c\u662f\u5e8f\u5217":66,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165":41,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":41,"\u5982\u679c\u6709\u66f4\u590d\u6742\u7684\u4f7f\u7528":2,"\u5982\u679c\u6709\u9700\u8981\u4fee\u6539\u7684\u5730\u65b9":43,"\u5982\u679c\u6709bugfix\u7684\u884c\u4e3a":30,"\u5982\u679c\u672a\u8bbe\u7f6e":50,"\u5982\u679c\u672a\u8bbe\u7f6egpu":52,"\u5982\u679c\u67d0\u4e00\u4e2a\u7c7b\u578b\u9700\u8981\u5f15\u7528\u53e6\u4e00\u4e2a\u7c7b\u578b":28,"\u5982\u679c\u67d0\u4e00\u4e2apaddl":28,"\u5982\u679c\u67d0\u4e00\u4e2apaddle\u6982\u5ff5\u5fc5\u987b\u8981\u66b4\u9732":28,"\u5982\u679c\u67d0\u4e00\u5757\u6839\u672c\u5c31\u4e0d\u600e\u4e48\u8017\u65f6":47,"\u5982\u679c\u68c0\u67e5\u5230\u5206\u914d\u5728\u4e0d\u540c\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u53c2\u6570\u7684\u5206\u5e03\u4e0d\u5747\u5300\u6b21\u6570\u5927\u4e8echeck":50,"\u5982\u679c\u6ca1\u6709\u5b9a\u4e49memori":41,"\u5982\u679c\u6ca1\u6709\u8bbe\u7f6e":69,"\u5982\u679c\u6ca1\u6709\u8bbe\u7f6etest":2,"\u5982\u679c\u6d88\u606f\u6570\u636e\u592a\u5c0f":50,"\u5982\u679c\u6ee1\u8db3\u6761\u4ef6":25,"\u5982\u679c\u7528\u6237\u4e0d\u663e\u793a\u6307\u5b9a\u8fd4\u56de\u6570\u636e\u7684\u5bf9\u5e94\u5173\u7cfb":3,"\u5982\u679c\u7528\u6237\u60f3\u8981\u4e86\u89e3\u8be6\u7ec6\u7684\u6570\u636e\u96c6\u7684\u683c\u5f0f":60,"\u5982\u679c\u7528\u6237\u60f3\u8981\u81ea\u5b9a\u4e49\u521d\u59cb\u5316\u65b9\u5f0f":31,"\u5982\u679c\u7528\u6237\u8981\u628apaddle\u7684\u9759\u6001\u5e93":27,"\u5982\u679c\u771f\u60f3\u6316\u6398\u5185\u6838\u6df1\u5904\u7684\u67d0\u4e2a\u79d8\u5bc6":47,"\u5982\u679c\u7a0b\u5e8f\u5d29\u6e83\u4f60\u4e5f\u53ef\u4ee5\u624b\u52a8\u7ec8\u6b62":48,"\u5982\u679c\u7cfb\u7edf\u5b89\u88c5\u4e86\u591a\u4e2apython\u7248\u672c":31,"\u5982\u679c\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":44,"\u5982\u679c\u7f51\u7edc\u67b6\u6784\u7b80\u5355":42,"\u5982\u679c\u8981\u4f7f\u7528\u53cc\u5411lstm":68,"\u5982\u679c\u8981\u4f7f\u7528gpu":34,"\u5982\u679c\u8981\u542f\u7528gpu":48,"\u5982\u679c\u8981\u8fd0\u884c\u6240\u6709\u7684\u5355\u5143\u6d4b\u8bd5":43,"\u5982\u679c\u8bad\u7ec3\u4e00\u4e2apass":31,"\u5982\u679c\u8bad\u7ec3\u8fc7\u7a0b\u542f\u52a8\u6210\u529f\u7684\u8bdd":66,"\u5982\u679c\u8bad\u7ec3\u8fc7\u7a0b\u7684\u7684cost\u660e\u663e\u9ad8\u4e8e\u8fd9\u4e2a\u5e38\u6570\u8f93\u51fa\u7684cost":31,"\u5982\u679c\u8bbe\u7f6e":3,"\u5982\u679c\u8bbe\u7f6e\u8be5\u53c2\u6570":50,"\u5982\u679c\u8c03\u7528\u9759\u6001\u5e93\u53ea\u80fd\u5c06\u9759\u6001\u5e93\u4e0e\u89e3\u91ca\u5668\u94fe\u63a5":27,"\u5982\u679c\u8f93\u51fa\u662fno":34,"\u5982\u679c\u8fd0\u884c\u6210\u529f":[62,68],"\u5982\u679c\u96c6\u7fa4\u8282\u70b9\u6570\u91cf\u5c11":48,"\u5982\u679c\u9700\u8981\u5305\u542b\u66f4\u591a\u7684\u4f9d\u8d56":34,"\u5982\u679c\u9700\u8981\u6269\u5927\u77e9\u9635":44,"\u5982\u679c\u9700\u8981\u7f29\u51cf\u77e9\u9635":44,"\u5982\u679clearning_rate\u592a\u5927":31,"\u5982\u679clearning_rate\u592a\u5c0f":31,"\u5982\u679cpaddlepaddle\u5305\u5df2\u7ecf\u5728python\u7684sit":31,"\u5982\u679cparamet":25,"\u5982\u6bcf\u4e2a\u6587\u4ef6\u53ea\u6709\u4e00\u4e2a":43,"\u5982\u795e\u7ecf\u5143\u6fc0\u6d3b\u503c\u7b49":31,"\u5982\u8981build\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf":43,"\u5982\u9ad8\u4eae\u90e8\u5206":47,"\u5b50":39,"\u5b50\u53e5":41,"\u5b50\u53e5\u7684\u5355\u8bcd\u6570\u548c\u6307\u5b9a\u7684\u4e00\u4e2a\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":41,"\u5b57\u5178":69,"\u5b57\u5178\u4f1a\u5305\u542b\u8f93\u5165\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u5355\u8bcd":69,"\u5b57\u5178\u5171\u5305\u542b":60,"\u5b57\u5178\u6587\u4ef6":[67,68],"\u5b57\u5178\u91c7\u7528utf8\u7f16\u7801":60,"\u5b57\u5178imdb":68,"\u5b57\u6bb5\u4e2d":56,"\u5b57\u6bb5\u8868\u793a\u5bb9\u5668\u7684\u73af\u5883\u53d8\u91cf":56,"\u5b57\u6bb5\u8868\u793a\u8fd9\u4e2ajob\u4f1a\u540c\u65f6\u5f00\u542f3\u4e2apaddlepaddle\u8282\u70b9":56,"\u5b57\u7b26\u4e32":26,"\u5b58\u50a8":26,"\u5b58\u50a8\u5377":54,"\u5b58\u50a8\u5728\u8bb0\u5fc6\u5355\u5143\u533a\u5757\u7684\u5386\u53f2\u4fe1\u606f\u88ab\u66f4\u65b0\u7528\u6765\u8fed\u4ee3\u7684\u5b66\u4e60\u5355\u8bcd\u4ee5\u5408\u7406\u7684\u5e8f\u5217\u7a0b\u73b0":68,"\u5b58\u50a8\u6a21\u578b\u7684\u8def\u5f84":69,"\u5b58\u50a8\u7740\u7535\u5f71\u6216\u7528\u6237\u4fe1\u606f":66,"\u5b58\u5165settings\u5bf9\u8c61":3,"\u5b58\u5728\u6216\u66f4\u6539\u4e3a\u5176\u5b83\u6a21\u578b\u8def\u5f84":68,"\u5b66\u4e60\u6210\u672c\u9ad8":27,"\u5b66\u4e60\u7b97\u6cd5":32,"\u5b66\u672f":65,"\u5b81\u6ce2":39,"\u5b83\u4e0d\u4ec5\u80fd\u591f\u5904\u7406imdb\u6570\u636e":68,"\u5b83\u4eec\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4f5c\u4e3a\u7f51\u7edc\u7684\u51fa\u53e3":32,"\u5b83\u4eec\u7684\u5927\u5c0f\u662f":42,"\u5b83\u4eec\u7684\u6587\u4ef6\u540d\u662f":26,"\u5b83\u4eec\u8fd8\u53ef\u4ee5\u4f9b\u90a3\u4e9b\u8fd0\u884c\u66f4\u590d\u6742\u7684\u96c6\u7fa4\u7ba1\u7406\u7cfb\u7edf":48,"\u5b83\u4eec\u90fd\u662f\u5e8f\u5217":42,"\u5b83\u4f1a\u5728dataprovider\u521b\u5efa\u7684\u65f6\u5019\u6267\u884c":3,"\u5b83\u4f7f\u752850\u5c42\u7684resnet\u6a21\u578b\u6765\u5bf9":62,"\u5b83\u5305\u542b\u4ee5\u4e0b\u51e0\u6b65":44,"\u5b83\u5305\u542b\u4ee5\u4e0b\u53c2\u6570":44,"\u5b83\u5305\u542b\u56db\u4e2a\u7248\u672c":36,"\u5b83\u5305\u542b\u7684\u5c5e\u6027\u53c2\u6570\u5982\u4e0b":3,"\u5b83\u5305\u62ec\u4e86\u4e00\u4e2a\u53cc\u5411\u7684gru\u4f5c\u4e3a\u7f16\u7801\u5668\u548c\u89e3\u7801\u5668":69,"\u5b83\u53eb\u505a":42,"\u5b83\u53ef\u4ee5\u5728\u53e5\u5b50\u7ea7\u522b\u5229\u7528\u53ef\u6269\u5c55\u7684\u4e0a\u4e0b\u6587":68,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u51cf\u5c11\u5206\u53d1\u5ef6\u8fdf":48,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u6211\u4eec\u683c\u5f0f\u5316\u6e90\u4ee3\u7801":43,"\u5b83\u53ef\u4ee5\u6307\u6d4b\u91cf\u4e00\u4e2a\u7a0b\u5e8f\u7684\u7a7a\u95f4":47,"\u5b83\u53ef\u4ee5\u88ab\u5e94\u7528\u4e8e\u8fdb\u884c\u673a\u5668\u7ffb\u8bd1":69,"\u5b83\u53ef\u80fd\u6709\u4e0d\u6b62\u4e00\u4e2a\u6743\u91cd":44,"\u5b83\u540c\u65f6\u5b66\u4e60\u6392\u5217":69,"\u5b83\u548c\u6570\u636e\u4f20\u5165\u51fd\u6570\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570":3,"\u5b83\u5b58\u50a8\u5f53\u524d\u8282\u70b9\u6240\u6709\u8bad\u7ec3":48,"\u5b83\u5b9a\u4e49\u4e86":42,"\u5b83\u5b9a\u4e49\u4e86\u6a21\u578b\u53c2\u6570\u6539\u53d8\u7684\u89c4\u5219":32,"\u5b83\u5b9a\u4e49\u89e3\u7801\u7f51\u7edc\u7684":42,"\u5b83\u5c06\u88ab\u5206\u53d1\u5230":48,"\u5b83\u5c06\u8f93\u5165\u8bed\u53e5\u7f16\u7801\u4e3a\u5411\u91cf\u7684\u5e8f\u5217":69,"\u5b83\u5c06\u8fd4\u56de\u5982\u4e0b\u7684\u5b57\u5178":62,"\u5b83\u5c31\u4f1a\u5728\u6e90\u8bed\u53e5\u4e2d\u641c\u7d22\u51fa\u6700\u76f8\u5173\u4fe1\u606f\u7684\u4f4d\u7f6e\u7684\u96c6\u5408":69,"\u5b83\u652f\u6301\u591a\u7ebf\u7a0b\u66f4\u65b0":44,"\u5b83\u662finteger_value\u7c7b\u578b\u7684":39,"\u5b83\u662finteger_value_sequence\u7c7b\u578b\u7684":39,"\u5b83\u6709\u52a9\u4e8e\u5e2e\u52a9\u9891\u7e41\u4fee\u6539\u548c\u8bbf\u95ee\u5de5\u4f5c\u533a\u6587\u4ef6\u7684\u7528\u6237\u51cf\u5c11\u8d1f\u62c5":48,"\u5b83\u6a21\u62df\u4e86\u89e3\u7801\u7ffb\u8bd1\u8fc7\u7a0b\u4e2d\u5728\u6e90\u8bed\u53e5\u4e2d\u7684\u641c\u7d22":69,"\u5b83\u7684":42,"\u5b83\u7684\u6536\u655b\u901f\u5ea6\u6bd4":68,"\u5b83\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20":38,"\u5b83\u7684\u76ee\u7684\u662f\u9884\u6d4b\u5728\u4e00\u4e2a\u5e8f\u5217\u4e2d\u8868\u8fbe\u7684\u60c5\u611f\u6001\u5ea6":68,"\u5b83\u7684\u8f93\u5165\u4e0e\u7ecf\u8fc7\u5b66\u4e60\u7684\u53c2\u6570\u505a\u5185\u79ef\u5e76\u52a0\u4e0a\u504f\u7f6e":44,"\u5b83\u76f4\u63a5\u5b66\u4e60\u6bb5\u843d\u8868\u793a":68,"\u5b83\u80fd\u591f\u4ece\u8bcd\u7ea7\u5230\u5177\u6709\u53ef\u53d8\u4e0a\u4e0b\u6587\u957f\u5ea6\u7684\u4e0a\u4e0b\u6587\u7ea7\u522b\u6765\u603b\u7ed3\u8868\u793a":68,"\u5b83\u8bfb\u5165\u6570\u636e\u5e76\u5c06\u5b83\u4eec\u4f20\u8f93\u5230\u63a5\u4e0b\u6765\u7684\u7f51\u7edc\u5c42":32,"\u5b83\u8fd4\u56degen":69,"\u5b83\u8fd4\u56detrain":69,"\u5b83\u9700\u8981\u5728\u8fd9\u91cc\u6307\u5b9a":68,"\u5b83\u9996\u5148\u8c03\u7528\u57fa\u6784\u9020\u51fd\u6570":44,"\u5b89\u6392":39,"\u5b89\u88c5\u4e0e\u7f16\u8bd1":37,"\u5b89\u88c5\u5305\u7684\u4e0b\u8f7d\u5730\u5740\u662f":36,"\u5b89\u88c5\u540e\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":28,"\u5b89\u88c5\u597ddocker\u4e4b\u540e\u53ef\u4ee5\u4f7f\u7528\u6e90\u7801\u76ee\u5f55\u4e0b\u7684\u811a\u672c\u6784\u5efa\u6587\u6863":45,"\u5b89\u88c5\u5b8c\u6210\u540e":36,"\u5b89\u88c5\u6d41\u7a0b":64,"\u5b89\u88c5\u8be5\u8f6f\u4ef6\u5305\u5c31\u53ef\u4ee5\u5728python\u73af\u5883\u4e0b\u5b9e\u73b0\u6a21\u578b\u9884\u6d4b":5,"\u5b89\u88c5paddlepaddl":64,"\u5b89\u88c5pillow":61,"\u5b89\u9759":39,"\u5b8c\u6210":[43,53],"\u5b8c\u6210\u4efb\u610f\u7684\u8fd0\u7b97\u903b\u8f91":41,"\u5b8c\u6210\u540evolume\u4e2d\u7684\u6587\u4ef6\u5185\u5bb9\u5927\u81f4\u5982\u4e0b":56,"\u5b8c\u6210\u5f00\u53d1":34,"\u5b8c\u6210\u6570\u636e\u7684\u9884\u5904\u7406":26,"\u5b8c\u6210\u76f8\u5e94\u7684\u8ba1\u7b97":38,"\u5b8c\u6574\u6e90\u7801\u53ef\u53c2\u8003":31,"\u5b8c\u6574\u7684\u4ee3\u7801\u89c1":5,"\u5b8c\u6574\u7684\u53c2\u6570\u77e9\u9635\u88ab\u5206\u5e03\u5728\u4e0d\u540c\u7684\u53c2\u6570\u670d\u52a1\u5668\u4e0a":44,"\u5b8c\u6574\u7684\u6570\u636e\u63d0\u4f9b\u6587\u4ef6\u5728":42,"\u5b8c\u6574\u7684\u914d\u7f6e\u6587\u4ef6\u5728":42,"\u5b98\u65b9\u6587\u6863":33,"\u5b9a\u4e49\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185rnn\u5355\u5143\u5b8c\u6210\u7684\u8ba1\u7b97":41,"\u5b9a\u4e49\u4e00\u4e2apython\u7684":3,"\u5b9a\u4e49\u4e86\u4e00\u4e2a\u53ea\u8bfb\u7684memori":41,"\u5b9a\u4e49\u4e86\u7f51\u7edc\u7684\u6570\u636e\u69fd":67,"\u5b9a\u4e49\u4e86\u7f51\u7edc\u7ed3\u6784":61,"\u5b9a\u4e49\u4e86\u7f51\u7edc\u7ed3\u6784\u5e76\u4fdd\u5b58\u4e3a":32,"\u5b9a\u4e49\u5728\u5916\u5c42":41,"\u5b9a\u4e49\u5f02\u6b65\u8bad\u7ec3\u7684\u957f\u5ea6":50,"\u5b9a\u4e49\u6570\u636e\u6765\u6e90":32,"\u5b9a\u4e49\u6e90\u8bed\u53e5\u7684\u6570\u636e\u5c42":42,"\u5b9a\u4e49\u89e3\u7801\u5668\u7684memori":42,"\u5b9a\u4e49\u8bad\u7ec3\u6570\u6910\u548c\u6d4b\u8bd5\u6570\u6910\u63d0\u4f9b\u8005":68,"\u5b9a\u4e49\u8f93\u5165\u6570\u636e\u5927\u5c0f":53,"\u5b9a\u4e49\u8f93\u5165\u6570\u636e\u7684\u7c7b\u578b":32,"\u5b9a\u4e49\u8f93\u51fa\u51fd\u6570":42,"\u5b9a\u4e49\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5355\u6b65\u51fd\u6570":42,"\u5b9e\u4f8b\u5982\u4e0b":67,"\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":39,"\u5b9e\u73b0\u524d\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":44,"\u5b9e\u73b0\u5355\u6b65\u51fd\u6570":42,"\u5b9e\u73b0\u540e\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":44,"\u5b9e\u73b0\u6570\u636e\u8f93\u5165\u51fd\u6570":3,"\u5b9e\u73b0\u6784\u9020\u51fd\u6570":44,"\u5b9e\u73b0\u7b80\u5355":27,"\u5b9e\u73b0\u7ec6\u8282":44,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad":44,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u540e\u5411\u4f20\u64ad":44,"\u5b9e\u73b0\u8bcd\u8bed\u548c\u53e5\u5b50\u4e24\u4e2a\u7ea7\u522b\u7684\u53cc\u5c42rnn\u7ed3\u6784":41,"\u5b9e\u73b0\u8be5\u5c42\u7684c":44,"\u5b9e\u9645\u4e0a\u53ea\u6709":62,"\u5b9e\u9645\u4e0a\u662fcsv\u6587\u4ef6":65,"\u5ba2\u6237":39,"\u5ba2\u6237\u670d\u52a1":65,"\u5bb6":39,"\u5bb9\u5668":54,"\u5bb9\u5668\u4e0d\u4f1a\u4fdd\u7559\u5728\u8fd0\u884c\u65f6\u751f\u6210\u7684\u6570\u636e":54,"\u5bb9\u5668\u8fd0\u884c\u90fd\u8fd0\u884c":56,"\u5bbf\u4e3b\u673a\u76ee\u5f55":54,"\u5bc4\u5b58\u5668\u4f7f\u7528\u60c5\u51b5\u548c\u5171\u4eab\u5185\u5b58\u4f7f\u7528\u60c5\u51b5\u80fd\u8ba9\u6211\u4eec\u5bf9gpu\u7684\u6574\u4f53\u4f7f\u7528\u6709\u66f4\u597d\u7684\u7406\u89e3":47,"\u5bf9":39,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f01\u5411\u91cf":3,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f\u6d6e\u70b9\u5411\u91cf":3,"\u5bf9\u4e8e":42,"\u5bf9\u4e8e\u4e0d\u540c\u8bed\u8a00":27,"\u5bf9\u4e8e\u4e24\u79cd\u4e0d\u540c\u7684\u8f93\u5165\u6570\u636e\u7c7b\u578b":39,"\u5bf9\u4e8e\u5185\u5b58\u8f83\u5c0f\u7684\u673a\u5668":3,"\u5bf9\u4e8e\u5355\u5c42rnn":39,"\u5bf9\u4e8e\u5355\u5c42rnn\u7684\u6570\u636e\u4e00\u5171\u6709\u4e24\u4e2a\u6837\u672c":39,"\u5bf9\u4e8e\u53cc\u5c42rnn":39,"\u5bf9\u4e8e\u540c\u4e00\u6bb5c":27,"\u5bf9\u4e8e\u540c\u6837\u7684\u6570\u636e":39,"\u5bf9\u4e8e\u591a\u8bed\u8a00\u63a5\u53e3":27,"\u5bf9\u4e8e\u5927\u591a\u6570\u8bed\u8a00":27,"\u5bf9\u4e8e\u6211\u4eec\u652f\u6301\u7684\u5168\u90e8\u77e9\u9635\u64cd\u4f5c":44,"\u5bf9\u4e8e\u6811\u7684\u6bcf\u4e00\u5c42":69,"\u5bf9\u4e8e\u6bb5\u843d\u7684\u6587\u672c\u5206\u7c7b":39,"\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5355\u5c42rnn\u7684\u6570\u636e":39,"\u5bf9\u4e8e\u6bcf\u4f4d\u7528\u6237":66,"\u5bf9\u4e8e\u6bcf\u79cd\u7c7b\u578b":28,"\u5bf9\u4e8e\u6bcf\u79cdc":28,"\u5bf9\u4e8e\u7b80\u5355\u7684\u591a\u673a\u534f\u540c\u8bad\u7ec3\u4f7f\u7528\u4e0a\u8ff0\u65b9\u5f0f\u5373\u53ef":53,"\u5bf9\u4e8e\u7ed9\u5b9a\u7684\u4e00\u6761\u6587\u672c":64,"\u5bf9\u4e8e\u914d\u5907\u6709\u6ce8\u610f\u529b\u673a\u5236\u7684\u89e3\u7801\u5668":42,"\u5bf9\u4e8eamazon":64,"\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":47,"\u5bf9\u5168\u8fde\u63a5\u5c42\u6765\u8bf4":44,"\u5bf9\u56fe\u7247\u8fdb\u884c\u9884\u5904\u7406":61,"\u5bf9\u5e94\u4e00\u4e2a\u5b50\u53e5":41,"\u5bf9\u5e94\u4e00\u4e2a\u8bcd":41,"\u5bf9\u5e94\u4e8e\u5b57\u5178":60,"\u5bf9\u5e94\u7684":3,"\u5bf9\u6027\u80fd\u5c24\u5176\u662f\u5185\u5b58\u5360\u7528\u6709\u4e00\u5b9a\u7684\u5f00\u9500":53,"\u5bf9\u6570\u636e\u96c6\u8fdb\u884c\u9884\u5904\u7406\u7684\u57fa\u672c\u547d\u4ee4\u662f":69,"\u5bf9\u6574\u4e2a\u65b0\u5411\u91cf\u96c6\u5408\u7684\u6bcf\u4e00\u4e2a\u7ef4\u5ea6\u53d6\u6700\u5927\u503c\u6765\u8868\u793a\u6700\u540e\u7684\u53e5\u5b50":64,"\u5bf9\u6587\u6863\u5904\u7406\u540e\u5f62\u6210\u7684\u5355\u8bcd\u5411\u91cf":68,"\u5bf9\u673a\u5668\u7ffb\u8bd1\u7684\u4eba\u5de5\u8bc4\u4f30\u5de5\u4f5c\u5f88\u5e7f\u6cdb\u4f46\u4e5f\u5f88\u6602\u8d35":69,"\u5bf9\u6bcf\u4e2a\u8f93\u5165":44,"\u5bf9\u6bcf\u4e2a\u8f93\u5165\u4e58\u4e0a\u53d8\u6362\u77e9\u9635":44,"\u5bf9\u6bd4":27,"\u5bf9\u6fc0\u6d3b\u6c42\u5bfc":44,"\u5bf9\u7528\u6237\u6765\u8bf4":3,"\u5bf9\u8bad\u7ec3\u6570\u636e\u8fdb\u884cshuffl":3,"\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u975e\u7ebf\u6027\u53d8\u6362":64,"\u5bf9\u8c61":[31,53],"\u5bf9\u8c61\u5b58\u50a8\u4e3a\u6587\u4ef6":66,"\u5bf9\u8f93\u5165\u53c2\u6570\u7684\u5b89\u5168\u6027\u8fdb\u884c\u4e86\u5fc5\u8981\u7684\u5224\u65ad":28,"\u5bf9\u8f93\u51fa\u7684\u5408\u5e76":41,"\u5bf9\u8fd9\u4e2a\u7248\u672c\u7684\u63d0\u4ea4":30,"\u5bf9\u9762":39,"\u5bf9check":3,"\u5bf9sparse_binary_vector\u548csparse_float_vector":3,"\u5bfc\u51fa\u8fd9\u4e9b\u63a5\u53e3":28,"\u5bfc\u81f4\u4e86\u6d6e\u70b9\u6570\u6ea2\u51fa":31,"\u5bfc\u81f4\u53c2\u6570\u6536\u655b\u5230\u4e86\u4e00\u4e9b\u5947\u5f02\u7684\u60c5\u51b5":31,"\u5bfc\u81f4\u53c2\u6570\u7d2f\u52a0":31,"\u5bfc\u81f4\u7f16\u8bd1paddlepaddle\u5931\u8d25":31,"\u5bfc\u81f4\u8bad\u7ec3\u65f6\u95f4\u8fc7\u957f":31,"\u5c01\u88c5\u4e86":47,"\u5c01\u88c5\u8be5\u5c42\u7684python\u63a5\u53e3":44,"\u5c06":[3,30,31,47,53,64],"\u5c06\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u62c6\u5206\u6210\u591a\u4efd":25,"\u5c06\u4e0a\u4e00\u65f6\u95f4\u6b65\u6240\u751f\u6210\u7684\u8bcd\u7684\u5411\u91cf\u6765\u4f5c\u4e3a\u5f53\u524d\u65f6\u95f4\u6b65\u7684\u8f93\u5165":42,"\u5c06\u4ed6\u4eec\u79fb\u52a8\u5230\u76ee\u5f55":66,"\u5c06\u4f1a\u81ea\u52a8\u8ba1\u7b97\u51fa\u4e00\u4e2a\u5408\u9002\u7684\u503c":50,"\u5c06\u5176\u8bbe\u7f6e\u6210":31,"\u5c06\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e\u5148\u53d8\u6362\u6210\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e":39,"\u5c06\u542b\u6709\u5b50\u53e5":41,"\u5c06\u542b\u6709\u8bcd\u8bed\u7684\u53e5\u5b50\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":41,"\u5c06\u56fe\u7247\u6309\u7167\u4e0a\u8ff0\u7ed3\u6784\u5b58\u50a8\u597d\u4e4b\u540e":61,"\u5c06\u5728":61,"\u5c06\u5728\u8fd0\u884c\u65f6\u62a5\u9519":48,"\u5c06\u5916\u90e8\u7684\u5b58\u50a8\u670d\u52a1\u5728kubernetes\u4e2d\u63cf\u8ff0\u6210\u4e3a\u7edf\u4e00\u7684\u8d44\u6e90\u5f62\u5f0f":54,"\u5c06\u591a\u53e5\u8bdd\u770b\u6210\u4e00\u4e2a\u6574\u4f53\u540c\u65f6\u4f7f\u7528encoder\u538b\u7f29":39,"\u5c06\u591a\u53f0\u673a\u5668\u7684\u6d4b\u8bd5\u7ed3\u679c\u5408\u5e76":50,"\u5c06\u5927\u91cf\u7684":27,"\u5c06\u5b57\u5178\u7684\u5730\u5740\u4f5c\u4e3aargs\u4f20\u7ed9dataprovid":31,"\u5c06\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u524d\u5411\u548c\u53cd\u5411\u90e8\u5206\u6df7\u5408\u5728\u4e00\u8d77":42,"\u5c06\u6570\u636e\u5904\u7406\u6210\u89c4\u8303\u683c\u5f0f":60,"\u5c06\u6570\u636e\u7ec4\u5408\u6210batch\u8fdb\u884c\u8bad\u7ec3":3,"\u5c06\u6570\u636e\u8f6c\u6362\u4e3apaddle\u7684\u683c\u5f0f":61,"\u5c06\u65b0\u5206\u652f\u7684\u7248\u672c\u6253\u4e0atag":30,"\u5c06\u65b0\u5efa\u7684\u6743\u91cd\u52a0\u5165\u6743\u91cd\u8868":44,"\u5c06\u65e5\u5fd7\u5199\u5165\u6587\u4ef6":66,"\u5c06\u672c\u5730\u7684\u4fee\u6539\u63a8\u9001\u5230":43,"\u5c06\u6837\u672c\u7684\u5730\u5740\u653e\u5165\u53e6\u4e00\u4e2a\u6587\u672c\u6587\u4ef6":3,"\u5c06\u6b64\u76ee\u5f55\u6302\u8f7d\u4e3a\u5bb9\u5668\u7684":56,"\u5c06\u6bcf\u4e2a\u6e90\u8bed\u8a00\u5230\u76ee\u6807\u8bed\u8a00\u7684\u5e73\u884c\u8bed\u6599\u5e93\u6587\u4ef6\u5408\u5e76\u4e3a\u4e00\u4e2a\u6587\u4ef6":69,"\u5c06\u73af\u5883\u53d8\u91cf\u8f6c\u6362\u6210paddle\u7684\u547d\u4ee4\u884c\u53c2\u6570":56,"\u5c06\u7528\u6237\u7684\u539f\u59cb\u6570\u636e\u8f6c\u6362\u6210\u7cfb\u7edf\u53ef\u4ee5\u8bc6\u522b\u7684\u6570\u636e\u7c7b\u578b":53,"\u5c06\u7ed3\u679c\u4fdd\u5b58\u5230\u6b64\u76ee\u5f55\u91cc":56,"\u5c06\u884c\u4e2d\u7684\u6570\u636e\u8f6c\u6362\u6210\u4e0einput_types\u4e00\u81f4\u7684\u683c\u5f0f":3,"\u5c06\u88ab\u5206\u4e3a":60,"\u5c06\u8bad\u7ec3\u6587\u4ef6\u4e0e\u5207\u5206\u597d\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5171\u4eab\u5b58\u50a8":56,"\u5c06\u8be5\u53e5\u8bdd\u5305\u542b\u7684\u6240\u6709\u5355\u8bcd\u5411\u91cf\u6c42\u5e73\u5747":64,"\u5c06\u8df3\u8fc7\u5206\u53d1\u9636\u6bb5\u76f4\u63a5\u542f\u52a8\u6240\u6709\u8282\u70b9\u7684\u96c6\u7fa4\u4f5c\u4e1a":48,"\u5c06\u8fd9\u79cd\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684\u8fde\u63a5\u7528\u4e00\u4e2a\u7279\u6b8a\u7684\u795e\u7ecf\u7f51\u7edc\u5355\u5143\u5b9e\u73b0":39,"\u5c06\u8fdc\u7a0b\u4ed3\u5e93":43,"\u5c06\u900f\u660e":48,"\u5c06ip\u6392\u5e8f\u751f\u6210\u7684\u5e8f\u53f7\u4f5c\u4e3atrain":56,"\u5c06master\u5206\u652f\u7684\u5408\u5165commit\u6253\u4e0atag":30,"\u5c11\u4e8e5":48,"\u5c1a\u53ef":39,"\u5c31":39,"\u5c31\u4f1a\u751f\u6210\u975e\u5e38\u591a\u7684gener":3,"\u5c31\u53ef\u4ee5\u518d\u8fd0\u884c\u53e6\u4e00\u4e2anginx":34,"\u5c31\u53ef\u4ee5\u5c06\u6570\u636e\u4f20\u9001\u7ed9paddlepaddle\u4e86":3,"\u5c31\u53ef\u4ee5\u5c06\u8fd9\u4e9b\u6587\u4ef6\u6301\u4e45\u5316\u5b58\u50a8":54,"\u5c31\u5f88\u5bb9\u6613\u5bfc\u81f4\u5185\u5b58\u8d85\u9650":31,"\u5c31\u662f":39,"\u5c31\u662f\u6a21\u578b\u7684\u53c2\u6570":32,"\u5c31\u662f\u7528\u4e8e\u5c55\u793a\u4e0a\u8ff0\u5206\u6790\u5de5\u5177\u7684\u7528\u6cd5":47,"\u5c31\u80fd\u591f\u5f88\u65b9\u4fbf\u7684\u5b8c\u6210\u6570\u636e\u4e0b\u8f7d\u548c\u76f8\u5e94\u7684\u9884\u5904\u7406\u5de5\u4f5c":64,"\u5c31\u8fd9\u4e48\u7b80\u5355":34,"\u5c31\u901a\u5e38\u7684gpu\u6027\u80fd\u5206\u6790\u6765\u8bf4":47,"\u5c31\u9700\u8981\u5bf9\u8fd9\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00\u589e\u52a0\u4e00\u4e9b\u5b9a\u4e49":27,"\u5c31\u9700\u8981\u9009\u62e9\u4f7f\u7528no":34,"\u5c3a\u5bf8":62,"\u5c40\u90e8\u5173\u8054\u6027\u8d28\u548c\u7a7a\u95f4\u4e0d\u53d8\u6027\u8d28":61,"\u5c42\u540e\u5f97\u5230\u6df1\u5ea6":67,"\u5c42\u548c\u8f93\u5165\u7684\u914d\u7f6e":44,"\u5c42\u6743\u91cd":62,"\u5c42\u6b21\u5316\u7684rnn":41,"\u5c42\u7279\u5f81":62,"\u5c42\u7684\u540d\u79f0\u4e0e":42,"\u5c42\u7684\u5927\u5c0f":44,"\u5c42\u7684\u7279\u5f81":62,"\u5c42\u7684\u7c7b\u578b":44,"\u5c42\u7684\u8f93\u5165":67,"\u5c42\u7684\u8f93\u5165\u548c\u8f93\u51fa\u4f5c\u4e3a\u4e0b\u4e00\u4e2a":67,"\u5c42\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684":42,"\u5c42\u7ec4\u6210\u4e00\u5bf9":67,"\u5c45\u7136":39,"\u5c55\u793a\u4e86\u4e00\u79cd\u65b9\u6cd5":69,"\u5c55\u793a\u4e86\u4e0a\u8ff0\u7f51\u7edc\u6a21\u578b\u7684\u8bad\u7ec3\u6548\u679c":64,"\u5c55\u793a\u4e86\u5982\u4f55\u5c06\u6bcf\u4e2a\u7279\u5f81\u6620\u5c04\u5230\u4e00\u4e2a\u5411\u91cf":66,"\u5c5e\u6027":67,"\u5d4c\u5165\u5c42":66,"\u5d4c\u5165\u7279\u5f81\u5b57\u5178":66,"\u5d4c\u5165\u7f16\u53f7\u4f1a\u6839\u636e\u5355\u8bcd\u6392\u5e8f":66,"\u5de5\u4f5c\u6a21\u5f0f":50,"\u5de5\u4f5c\u7a7a\u95f4":48,"\u5de5\u4f5c\u7a7a\u95f4\u4e2d\u7684":48,"\u5de5\u4f5c\u7a7a\u95f4\u6839\u76ee\u5f55":48,"\u5de5\u4f5c\u7a7a\u95f4\u76ee\u5f55\u7684\u5de5\u4f5c\u7a7a\u95f4":48,"\u5de5\u4f5c\u7a7a\u95f4\u914d\u7f6e":48,"\u5de5\u5177":68,"\u5de5\u5177\u4e2d\u7684\u811a\u672c":68,"\u5de5\u5177\u6765\u7ba1\u7406":43,"\u5de5\u7a0b\u5e08":65,"\u5de6\u56fe\u6784\u9020\u7f51\u7edc\u6a21\u5757\u7684\u65b9\u5f0f\u88ab\u7528\u4e8e34\u5c42\u7684\u7f51\u7edc\u4e2d":62,"\u5de6\u8fb9\u662f":62,"\u5dee\u8bc4":64,"\u5df2\u6253\u5f00":43,"\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":50,"\u5df2\u7ecf\u63d0\u4f9b\u4e86\u811a\u672c\u6765\u5e2e\u52a9\u60a8\u521b\u5efa\u8fd9\u4e24\u4e2a\u6587\u4ef6":48,"\u5e02\u573a":65,"\u5e02\u9762\u4e0a\u5df2\u7ecf\u6709nvidia\u6216\u7b2c\u4e09\u65b9\u63d0\u4f9b\u7684\u4f17\u591a\u5de5\u5177":47,"\u5e0c\u671b\u52a0\u901f\u8bad\u7ec3":53,"\u5e0c\u671b\u80fd\u8ba9\u6211\u4eec\u77e5\u6653":66,"\u5e2e\u52a9\u6211\u4eec\u5b8c\u6210\u5bf9\u8f93\u5165\u5e8f\u5217\u7684\u62c6\u5206":41,"\u5e2e\u52a9\u6211\u4eec\u66f4\u597d\u5730\u63cf\u8ff0\u6bb5\u843d":41,"\u5e2e\u52a9\u6211\u4eec\u6784\u9020\u4e00\u4e9b\u590d\u6742\u7684\u8f93\u5165\u4fe1\u606f":38,"\u5e38\u5e38\u51fa\u73b0":31,"\u5e38\u7528\u4f18\u5316\u7b97\u6cd5\u5305\u62ecmomentum":64,"\u5e38\u89c1\u7684\u53ef\u9009\u5b58\u50a8\u670d\u52a1\u5305\u62ec":54,"\u5e72\u51c0":39,"\u5e73\u53f0\u4e3a\u60f3\u89c2\u6d4b\u8bcd\u5411\u91cf\u7684\u7528\u6237\u63d0\u4f9b\u4e86\u5c06\u4e8c\u8fdb\u5236\u8bcd\u5411\u91cf\u6a21\u578b\u8f6c\u6362\u4e3a\u6587\u672c\u6a21\u578b\u7684\u529f\u80fd":60,"\u5e73\u5747\u6545\u969c\u4fee\u590d\u65f6\u95f4":25,"\u5e73\u5747\u6545\u969c\u7387":25,"\u5e73\u5747\u7279\u5f81\u56fe\u7684\u9ad8\u5ea6\u53ca\u5bbd\u5ea6":61,"\u5e74\u9f84":65,"\u5e74\u9f84\u4ece\u4e0b\u5217\u5217\u8868\u8303\u56f4\u4e2d\u9009\u53d6":65,"\u5e74\u9f84\u548c\u804c\u4e1a":66,"\u5e76\u4e0d\u4fdd\u8bc1":44,"\u5e76\u4e0d\u662f\u4f7f\u7528\u53cc\u5c42rnn\u89e3\u51b3\u5b9e\u9645\u7684\u95ee\u9898":39,"\u5e76\u4e0d\u662fkubernetes\u4e2d\u7684node\u6982\u5ff5":56,"\u5e76\u4e0d\u771f\u6b63\u7684\u548c":39,"\u5e76\u4e14":[3,42],"\u5e76\u4e14\u4f7f\u7528":28,"\u5e76\u4e14\u5185\u5c42\u7684\u5e8f\u5217\u64cd\u4f5c\u4e4b\u95f4\u72ec\u7acb\u65e0\u4f9d\u8d56":39,"\u5e76\u4e14\u5206\u522b\u91cd\u547d\u540d\u6587\u4ef6\u540e\u7f00":69,"\u5e76\u4e14\u52a0\u4e0a\u4e0b\u9762\u7684\u547d\u4ee4\u884c\u53c2\u6570":52,"\u5e76\u4e14\u5305\u62ecunit":43,"\u5e76\u4e14\u53ea\u6709\u4e00\u4e2a\u6743\u91cd":62,"\u5e76\u4e14\u53ef\u4ee5\u88ab\u63d0\u4ea4\u7684job\u76f4\u63a5\u4f7f\u7528":26,"\u5e76\u4e14\u53ef\u80fd\u4f1a\u52a0\u901f\u8bad\u7ec3\u8fc7\u7a0b":31,"\u5e76\u4e14\u540e\u7eed\u4ecd\u5728\u4e0d\u65ad\u6539\u8fdb":32,"\u5e76\u4e14\u542f\u52a8\u8bad\u7ec3":56,"\u5e76\u4e14\u5728\u5185\u5b58\u8db3\u591f\u7684\u60c5\u51b5\u4e0b\u8d8a\u5927\u8d8a\u597d":3,"\u5e76\u4e14\u5728\u5e38\u89c1\u7684\u5e73\u53f0\u4e0a":27,"\u5e76\u4e14\u5728\u968f\u540e\u7684\u8bfb\u53d6\u6570\u636e\u8fc7\u7a0b\u4e2d\u586b\u5145\u8bcd\u8868":64,"\u5e76\u4e14\u5728dataprovider\u4e2d\u5b9e\u73b0\u5982\u4f55\u8bbf\u95ee\u8bad\u7ec3\u6587\u4ef6\u5217\u8868":2,"\u5e76\u4e14\u5b83\u4eec\u7684\u987a\u5e8f\u4e0e":62,"\u5e76\u4e14\u5bf9\u7528\u6237\u7684\u7279\u5f81\u505a\u540c\u6837\u7684\u64cd\u4f5c":66,"\u5e76\u4e14\u5c06\u9884\u5904\u7406\u597d\u7684\u6570\u636e\u96c6\u5b58\u653e\u5728":69,"\u5e76\u4e14\u67e5\u8be2paddlepaddle\u5355\u5143\u6d4b\u8bd5\u7684\u65e5\u5fd7":31,"\u5e76\u4e14\u7b2c\u4e8c\u4e2a\u662f\u53cd\u5411lstm":68,"\u5e76\u4e14\u7cfb\u7edf\u6bcf\u4e00\u8f6e\u8bad\u7ec3\u5f00\u59cb\u65f6\u4f1a\u91cd\u7f6edataprovid":53,"\u5e76\u4e14\u7f16\u8bd1\u80fd\u901a\u8fc7\u4ee3\u7801\u6837\u5f0f\u68c0\u67e5":43,"\u5e76\u4e14\u8ba9\u63a5\u53e3\u8131\u79bb\u5b9e\u73b0\u7ec6\u8282":27,"\u5e76\u4e14\u8f93\u51fa\u4e00\u4e2a":43,"\u5e76\u4e14\u901a\u8fc7\u7ed9\u51fa\u5f53\u524d\u76ee\u6807\u5355\u8bcd\u6765\u9884\u6d4b\u4e0b\u4e00\u4e2a\u76ee\u6807\u5355\u8bcd":69,"\u5e76\u4e14\u96c6\u7fa4\u4f5c\u4e1a\u4e2d\u7684\u6240\u6709\u8282\u70b9\u5c06\u5728\u6b63\u5e38\u60c5\u51b5\u4e0b\u5904\u7406\u5177\u6709\u76f8\u540c\u903b\u8f91\u4ee3\u7801\u7684\u6587\u4ef6":48,"\u5e76\u4e14\u9700\u8981\u91cd\u5199\u57fa\u7c7b\u4e2d\u7684\u4ee5\u4e0b\u51e0\u4e2a\u865a\u51fd\u6570":44,"\u5e76\u4e14softmax\u5c42\u7684\u4e24\u4e2a\u8f93\u5165\u4e5f\u4f7f\u7528\u4e86\u540c\u6837\u7684\u53c2\u6570":31,"\u5e76\u4f20\u5165\u76f8\u5e94\u7684\u547d\u4ee4\u884c\u53c2\u6570\u521d\u59cb\u5316paddlepaddl":5,"\u5e76\u4f7f\u7528":67,"\u5e76\u4f7f\u7528\u4e86dropout":64,"\u5e76\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u6765\u5bf9\u56fe\u7247\u8fdb\u884c\u5206\u7c7b":61,"\u5e76\u521b\u5efa\u4e86\u4e00\u4e2a\u65b0\u6587\u4ef6":43,"\u5e76\u5220\u9664":30,"\u5e76\u5220\u9664\u66f4\u65e9\u7684\u5feb\u7167":25,"\u5e76\u52a0\u8f7d\u5176\u4e2d\u7684\u53c2\u6570":25,"\u5e76\u5728\u4e58\u79ef\u7ed3\u679c\u4e0a\u518d\u52a0\u4e0a\u7ef4\u5ea6\u4e3a":44,"\u5e76\u5728\u6700\u5f00\u59cb\u521d\u59cb\u5316\u4e3a\u8d77\u59cb\u8bcd":42,"\u5e76\u5728\u7b14\u8bb0\u672c\u4e0a\u901a\u8fc7ssh\u4e0e\u5176\u8fde\u63a5":34,"\u5e76\u5728\u7c7b\u6784\u5efa\u51fd\u6570\u4e2d\u628a\u5b83\u653e\u5165\u4e00\u4e2a\u7c7b\u6210\u5458\u53d8\u91cf\u91cc":44,"\u5e76\u5728\u96c6\u7fa4\u4e2d\u8fd0\u884c\u591a\u4e2a\u5206\u5e03\u5f0f\u6570\u636e\u5904\u7406\u4efb\u52a1":26,"\u5e76\u5bf9\u76f8\u5e94\u7684\u53c2\u6570\u8c03\u7528":44,"\u5e76\u5c06\u5176\u6295\u5c04\u5230":42,"\u5e76\u5c06\u5b83\u4eec\u6309\u7167\u542f\u53d1\u4ee3\u4ef7":69,"\u5e76\u5c06\u5b83\u4eec\u653e\u5728":69,"\u5e76\u5c06\u6bcf\u8f6e\u7684\u6a21\u578b\u7ed3\u679c\u5b58\u653e\u5728":32,"\u5e76\u5c06c":28,"\u5e76\u5c06develop\u548ctest\u6570\u636e\u5206\u522b\u653e\u5165\u4e0d\u540c\u7684\u6587\u4ef6\u5939":69,"\u5e76\u60f3\u4f7f\u7528gpu\u6765\u8bad\u7ec3\u8bbe\u7f6e\u4e3atru":68,"\u5e76\u628a\u5feb\u7167\u4fdd\u5b58\u5230\u8fd9\u4e2a\u76ee\u5f55\u4e0b":25,"\u5e76\u6307\u5b9a\u7aef\u53e3\u53f7":53,"\u5e76\u6307\u5b9abatch":69,"\u5e76\u63d0\u4f9b\u4e86\u7b80\u5355\u7684cache\u529f\u80fd":3,"\u5e76\u6b22\u8fce\u60a8\u6765\u53c2\u4e0e\u8d21\u732e":68,"\u5e76\u6ca1\u6709paddle\u7279\u522b\u9700\u8981\u7684\u7279\u6027":27,"\u5e76\u7531":67,"\u5e76\u7ed9\u51fa\u5206\u7c7b\u7ed3\u679c":61,"\u5e76\u7ed9\u51fa\u7684\u76f8\u5173\u6a21\u578b\u683c\u5f0f\u7684\u5b9a\u4e49":60,"\u5e76\u88ab\u53cd\u5411\u5904\u7406":67,"\u5e76\u88ab\u5b58\u50a8\u5728\u8bf8\u5982hadoop":26,"\u5e76\u89c2\u5bdf\u7ed3\u679c":47,"\u5e76\u8bb0\u5f55\u5b83\u7684\u7f16\u53f7":43,"\u5e76\u8bbe\u7f6e":[36,48],"\u5e76\u9002\u5e94github\u7684\u7279\u6027\u505a\u4e86\u4e00\u4e9b\u533a\u522b":30,"\u5e76\u9010\u6e10\u5c55\u793a\u66f4\u52a0\u6df1\u5165\u7684\u529f\u80fd":64,"\u5e8a\u4e0a\u7528\u54c1":39,"\u5e8a\u57ab":39,"\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":38,"\u5e8f\u5217\u5230\u5e8f\u5217":69,"\u5e8f\u5217\u6570\u636e\u662f\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u9762\u5bf9\u7684\u4e00\u79cd\u4e3b\u8981\u8f93\u5165\u6570\u636e\u7c7b\u578b":41,"\u5e8f\u5217\u662f\u4e00\u79cd\u5e38\u89c1\u7684\u6570\u636e\u7c7b\u578b":38,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u5927\u591a\u9075\u5faaencod":41,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u7684\u8f93\u5165":41,"\u5e8f\u5217\u7684\u5f00\u59cb":69,"\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u539f\u6765\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u5143\u7d20\u7684\u5e73\u5747\u503c":38,"\u5e8f\u5217\u7684\u7ed3\u5c3e":69,"\u5e8f\u5217\u7684\u7ed3\u675f":69,"\u5e93":48,"\u5e93\u7684\u8def\u5f84":48,"\u5e94\u7528\u524d\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":42,"\u5e94\u7528\u53cd\u5411\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":42,"\u5e94\u7528\u6a21\u578b":64,"\u5e94\u8be5":39,"\u5e94\u8be5\u4e0e\u5b83\u7684memory\u540d\u5b57\u76f8\u540c":42,"\u5e94\u8be5\u964d\u4f4e\u5b66\u4e60\u7387":31,"\u5e95\u5c42\u8fdb\u7a0b":48,"\u5efa\u7acb\u4e00\u4e2a":43,"\u5efa\u7acb\u4e00\u4e2a\u6d3b\u8dc3\u7684\u5f00\u6e90\u793e\u533a":0,"\u5efa\u8bae":30,"\u5efa\u8bae\u5c06\u5176\u8bbe\u7f6e\u4e3a\u8f83\u5927":48,"\u5efa\u8bae\u5c06\u8be5\u53c2\u6570\u8bbe\u4e3atrue":50,"\u5f00\u53d1\u4e86\u6a21\u578b\u9884\u6d4b\u7684\u6837\u4f8b\u4ee3\u7801":28,"\u5f00\u53d1\u4eba\u5458\u4f7f\u7528":43,"\u5f00\u53d1\u4eba\u5458\u53ef\u4ee5\u5728docker\u5f00\u53d1\u955c\u50cf\u4e2d\u5f00\u53d1paddlepaddl":34,"\u5f00\u53d1\u8005\u4fee\u6539\u81ea\u5df1\u7684\u4ee3\u7801":30,"\u5f00\u53d1\u8005\u7ed9paddlepaddle\u4ee3\u7801\u589e\u52a0\u7684\u6ce8\u91ca\u4ee5pr\u7684\u5f62\u5f0f\u63d0\u4ea4\u5230github\u4e2d":45,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4e2d":30,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4f7f\u7528":30,"\u5f00\u53d1\u955c\u50cf":[34,43],"\u5f00\u53d1\u955c\u50cf\u5305\u542b\u4e86\u4ee5\u4e0b\u5de5\u5177":34,"\u5f00\u59cb":32,"\u5f00\u59cb\u63d0\u4f9b\u670d\u52a1":25,"\u5f00\u59cb\u6807\u8bb0":42,"\u5f00\u59cb\u8bad\u7ec3\u6a21\u578b":64,"\u5f00\u59cb\u9636\u6bb5":47,"\u5f02\u6b65\u8bfb\u53d6\u7b49\u95ee\u9898":53,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":49,"\u5f15\u5165\u4e86\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5f15\u5165lstm\u6a21\u578b\u4e3b\u8981\u662f\u4e3a\u4e86\u514b\u670d\u6d88\u5931\u68af\u5ea6\u7684\u95ee\u9898":68,"\u5f15\u5165paddlepaddle\u7684pydataprovider2\u5305":3,"\u5f15\u53d1":20,"\u5f15\u5bfc\u5c42":42,"\u5f15\u7528":48,"\u5f15\u7528memory\u5f97\u5230\u8fd9layer\u4e0a\u4e00\u65f6\u523b\u8f93\u51fa":41,"\u5f3a\u70c8\u63a8\u8350":39,"\u5f3a\u70c8\u63a8\u8350\u4f7f\u7528virtualenv\u6765\u521b\u9020\u4e00\u4e2a\u5e72\u51c0\u7684python\u73af\u5883":66,"\u5f52\u4e00\u5316\u6982\u7387\u5411\u91cf":42,"\u5f53":52,"\u5f53\u4f20\u9012\u76f8\u540c\u7684\u6837\u672c\u6570\u65f6":68,"\u5f53\u4f60\u6267\u884c\u547d\u4ee4":44,"\u5f53\u51fd\u6570\u8fd4\u56de\u7684\u65f6\u5019":3,"\u5f53\u524d\u5355\u8bcd\u5728\u76f8\u6bd4\u4e4b\u4e0b\u603b\u662f\u88ab\u5f53\u4f5c\u771f\u503c":69,"\u5f53\u524d\u5355\u8bcd\u662f\u89e3\u7801\u5668\u6700\u540e\u4e00\u6b65\u7684\u8f93\u51fa":69,"\u5f53\u524d\u65f6\u95f4\u6b65\u5904\u7684memory\u7684\u8f93\u51fa\u4f5c\u4e3a\u4e0b\u4e00\u65f6\u95f4\u6b65memory\u7684\u8f93\u5165":42,"\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":44,"\u5f53\u524d\u7684\u8f93\u5165y\u548c\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51farnn_state\u505a\u4e86\u4e00\u4e2a\u5168\u94fe\u63a5":39,"\u5f53\u524d\u8bc4\u4f30\u4e2d":69,"\u5f53\u524dbatch\u7684cost":69,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":64,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":64,"\u5f53\u529f\u80fd\u5206\u652f\u5f00\u53d1\u5b8c\u6bd5\u540e":30,"\u5f53\u5728\u7f51\u7edc\u5c42\u914d\u7f6e\u4e2d\u8bbe\u7f6e":50,"\u5f53\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":50,"\u5f53\u5bb9\u5668\u56e0\u4e3a\u5404\u79cd\u539f\u56e0\u88ab\u9500\u6bc1\u65f6":54,"\u5f53\u6240\u6709\u6570\u636e\u8bfb\u53d6\u5b8c\u4e00\u8f6e\u540e":53,"\u5f53\u6240\u6709pod\u90fd\u5904\u4e8erunning\u72b6\u6001":56,"\u5f53\u6a21\u578b\u53c2\u6570\u4e0d\u5b58\u5728\u65f6":50,"\u5f53\u6a21\u578b\u8bad\u7ec3\u597d\u4e86\u4e4b\u540e":64,"\u5f53\u6a21\u5f0f\u4e3a":50,"\u5f53\u7136":47,"\u5f53\u7528\u6237\u4f7f\u7528\u5b8c\u8fd9\u4e2a\u53c2\u6570\u540e":28,"\u5f53\u7f51\u7edc\u5c42\u7528\u4e00\u4e2a\u6279\u6b21\u505a\u8bad\u7ec3\u65f6":44,"\u5f53\u89e3\u8bfb\u6bcf\u4e00\u4e2a":42,"\u5f53\u8bad\u7ec3\u6570\u636e\u975e\u5e38\u591a\u65f6":3,"\u5f53\u8d85\u8fc7\u8be5\u9608\u503c\u65f6":50,"\u5f53\u8f93\u5165\u662f\u7ef4\u5ea6\u5f88\u9ad8\u7684\u7a00\u758f\u6570\u636e\u65f6":52,"\u5f53\u9700\u8981\u5feb\u901f\u6216\u8005\u9891\u7e41\u7684\u8bc4\u4f30\u65f6":69,"\u5f53classif":69,"\u5f62\u6210recurr":41,"\u5f62\u6210recurrent\u8fde\u63a5":41,"\u5f62\u72b6":62,"\u5f88":[39,64],"\u5f88\u591a":39,"\u5f88\u591a\u5f00\u53d1\u8005\u4f1a\u4f7f\u7528\u8fdc\u7a0b\u7684\u5b89\u88c5\u6709gpu\u7684\u670d\u52a1\u5668\u5de5\u4f5c":34,"\u5f88\u5b89\u9759":39,"\u5f88\u5bb9\u6613\u5bfc\u81f4\u67d0\u4e00\u4e2a\u53c2\u6570\u670d\u52a1\u5668\u6ca1\u6709\u5206\u914d\u5230\u4efb\u4f55\u53c2\u6570":53,"\u5f88\u5e72\u51c0":39,"\u5f88\u65b9\u4fbf":39,"\u5f88\u6709\u53ef\u80fd\u5b9e\u9645\u5e94\u7528\u5c31\u662f\u6ca1\u6709\u6309\u7167\u60a8\u7684\u9884\u671f\u60c5\u51b5\u8fd0\u884c":47,"\u5f88\u9002\u5408\u6784\u5efa\u7528\u4e8e\u7406\u89e3\u56fe\u7247\u5185\u5bb9\u7684\u6a21\u578b":61,"\u5f88\u96be\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":27,"\u5f88\u96be\u6574\u4f53\u4fee\u6b63":3,"\u5f8b\u5e08":65,"\u5f97":39,"\u5f97\u4f7f\u7528":27,"\u5f97\u5230\u53e5\u5b50\u7684\u8868\u793a":64,"\u5f97\u5230\u6700\u597d\u8f6e\u6b21\u4e0b\u7684\u6a21\u578b":66,"\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e2a\u65f6\u95f4\u6b65\u603b\u662f\u80fd\u591f\u5f15\u7528\u6240\u6709\u8f93\u5165":41,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4e2d":42,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4f5c\u4e3a\u4f7f\u7528":42,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u548c":42,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u9aa4\u987a\u5e8f\u5730\u5904\u7406\u5e8f\u5217":42,"\u5faa\u73af\u7f51\u7edc\u4ece":42,"\u5fc5\u8981":28,"\u5fc5\u987b":44,"\u5fc5\u987b\u4e00\u81f4":3,"\u5fc5\u987b\u4f7f\u7528python\u5173\u952e\u8bcd":3,"\u5fc5\u987b\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":39,"\u5fc5\u987b\u6307\u5411\u4e00\u4e2apaddlepaddle\u5b9a\u4e49\u7684lay":41,"\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":41,"\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":38,"\u5fc5\u987b\u7531\u53ea\u8bfbmemory\u7684":42,"\u5fd8\u8bb0\u95e8\u548c\u8f93\u51fa\u95e8":68,"\u5feb":[39,68],"\u5feb\u901f\u5165\u95e8":63,"\u5feb\u901f\u5728\u672c\u5730\u542f\u52a8\u4e00\u4e2a\u5355\u673a\u7684kubernetes\u670d\u52a1\u5668":54,"\u5feb\u901f\u90e8\u7f72\u96c6\u7fa4":54,"\u6027\u4ef7\u6bd4":39,"\u6027\u522b":[65,66],"\u6027\u80fd\u5206\u6790":47,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u662f\u7528\u4e8e\u7ed9\u5e94\u7528\u7a0b\u5e8f\u7684\u6027\u80fd\u505a\u5b9a\u91cf\u5206\u6790\u7684":47,"\u6027\u80fd\u5206\u6790\u662f\u6027\u80fd\u4f18\u5316\u7684\u5173\u952e\u4e00\u6b65":47,"\u6027\u80fd\u8c03\u4f18":49,"\u603b\u4f53\u6765\u8bf4":39,"\u603b\u8ba1\u7684\u53c2\u6570\u4e2a\u6570":60,"\u603b\u8bc4\u520610\u5206":68,"\u6050\u6016\u7247":65,"\u60a8\u4f1a\u5728\u63a5\u4e0b\u6765\u7684\u90e8\u5206\u4e2d\u83b7\u5f97\u66f4\u591a\u7684\u7ec6\u8282\u4ecb\u7ecd":47,"\u60a8\u53ef\u4ee5\u4efb\u610f\u4f7f\u7528\u4e00\u4e2a\u6216\u4e24\u4e2a\u6765\u5bf9\u611f\u5174\u8da3\u7684\u4ee3\u7801\u6bb5\u505a\u6027\u80fd\u5206\u6790":47,"\u60a8\u53ef\u4ee5\u5bfc\u5165":47,"\u60a8\u53ef\u4ee5\u91c7\u7528\u4e0b\u9762\u4e94\u4e2a\u6b65\u9aa4":47,"\u60a8\u5c06\u4e86\u89e3\u5982\u4f55":42,"\u60a8\u5c31\u80fd\u83b7\u5f97\u5982\u4e0b\u7684\u5206\u6790\u7ed3\u679c":47,"\u60a8\u6309\u5982\u4e0b\u6b65\u9aa4\u64cd\u4f5c\u5373\u53ef":47,"\u60a8\u6700\u597d\u5148\u786e\u8ba4":47,"\u60a8\u9700\u8981\u66f4\u6539":34,"\u60a8\u9996\u5148\u9700\u8981\u5728\u76f8\u5173\u4ee3\u7801\u6bb5\u4e2d\u52a0\u5165":47,"\u60ac\u7591\u7247":65,"\u60c5\u6001\u52a8\u8bcd":67,"\u60c5\u611f\u5206\u6790":[30,63],"\u60c5\u611f\u5206\u6790\u4e5f\u5e38\u7528\u4e8e\u57fa\u4e8e\u5927\u91cf\u8bc4\u8bba\u548c\u4e2a\u4eba\u535a\u5ba2\u6765\u76d1\u63a7\u793e\u4f1a\u5a92\u4f53":68,"\u60c5\u611f\u5206\u6790\u662f\u81ea\u7136\u8bed\u8a00\u7406\u89e3\u4e2d\u6700\u5178\u578b\u7684\u95ee\u9898\u4e4b\u4e00":68,"\u60c5\u611f\u5206\u6790\u6709\u8bb8\u591a\u5e94\u7528\u573a\u666f":68,"\u60ca\u9669\u7535\u5f71":65,"\u60f3\u4e86\u89e3\u66f4\u591a\u7ec6\u8282\u53ef\u4ee5\u53c2\u8003pydataprovider\u90e8\u5206\u7684\u6587\u6863":68,"\u610f\u5473\u7740\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u90fd\u662f\u76f8\u540c\u7684\u503c":42,"\u610f\u601d\u662f\u4e0d\u4f7f\u7528\u5e73\u5747\u53c2\u6570\u6267\u884c\u6d4b\u8bd5":50,"\u610f\u601d\u662f\u4e0d\u4fdd\u5b58\u7ed3\u679c":50,"\u610f\u601d\u662f\u4f7f\u7528\u7b2ctest":50,"\u610f\u601d\u662f\u5728gpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2agpu":50,"\u611f\u89c9":39,"\u620f\u5267":65,"\u6210\u529f\u8bad\u7ec3\u4e14\u9000\u51fa\u7684pod\u6570\u76ee\u4e3a3\u65f6":56,"\u6211\u4eec\u4e0d\u80fd\u901a\u8fc7\u5e38\u89c4\u7684\u68af\u5ea6\u68c0\u67e5\u7684\u65b9\u5f0f\u6765\u8ba1\u7b97\u68af\u5ea6":44,"\u6211\u4eec\u4e3a\u7528\u6237\u5b9a\u4ee5python\u63a5\u53e3\u6765\u914d\u7f6e\u7f51\u7edc":53,"\u6211\u4eec\u4e3b\u8981\u4f1a\u4ecb\u7ecdnvprof\u548cnvvp":47,"\u6211\u4eec\u4e5f\u53ef\u4ee5\u786e\u5b9a\u6bcf\u4e00\u4e2a\u53c2\u6570\u7684\u7c7b\u578b":28,"\u6211\u4eec\u4ec5\u4ec5\u662f\u5c06\u6bcf\u4e2a\u7279\u5f81\u79cd\u7c7b\u6620\u5c04\u5230\u4e00\u4e2a\u7279\u5f81\u5411\u91cf\u4e2d":66,"\u6211\u4eec\u4ec5\u6709\u4e00\u4e2a\u8f93\u5165":44,"\u6211\u4eec\u4ec5\u7528":66,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728":55,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884c\u5206\u5e03\u5f0fpaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":56,"\u6211\u4eec\u4ece\u63d0\u524d\u7ed9\u5b9a\u7684\u7c7b\u522b\u96c6\u5408\u4e2d\u9009\u62e9\u5176\u6240\u5c5e\u7c7b\u522b":64,"\u6211\u4eec\u4ee5mnist\u624b\u5199\u8bc6\u522b\u4e3a\u4f8b":3,"\u6211\u4eec\u4f1a\u53d1\u73b0\u6570\u636e\u96c6":69,"\u6211\u4eec\u4f1a\u5728":34,"\u6211\u4eec\u4f1a\u7ee7\u7eed\u4f7f\u7528\u73b0\u6709\u7684\u5185\u5b58\u5757":44,"\u6211\u4eec\u4f1a\u91cd\u65b0\u5206\u914d\u5185\u5b58":44,"\u6211\u4eec\u4f7f\u7528":[44,48,68],"\u6211\u4eec\u4f7f\u7528\u4e86":39,"\u6211\u4eec\u4f7f\u7528\u4e86\u4e00\u4e2a\u7f16\u89e3\u7801\u6a21\u578b\u7684\u6269\u5c55":69,"\u6211\u4eec\u4f7f\u7528\u4e86\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":68,"\u6211\u4eec\u4f7f\u7528\u5176\u4e2d\u7684":69,"\u6211\u4eec\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":27,"\u6211\u4eec\u4f7f\u7528\u6700\u5927\u6982\u7387\u7684\u6807\u7b7e\u4f5c\u4e3a\u7ed3\u679c":67,"\u6211\u4eec\u4f7f\u7528\u96c6\u675f\u641c\u7d22":69,"\u6211\u4eec\u4f7f\u7528paddlepaddle\u5728ilsvrc\u7684\u9a8c\u8bc1\u96c6\u517150":62,"\u6211\u4eec\u5047\u8bbe\u4e00\u53f0\u673a\u5668\u4e0a\u67094\u4e2agpu":52,"\u6211\u4eec\u5047\u8bbe\u623f\u4ea7\u7684\u4ef7\u683c":32,"\u6211\u4eec\u5148\u4ece\u4e00\u6761\u968f\u673a\u7684\u76f4\u7ebf":32,"\u6211\u4eec\u51c6\u5907\u4e86\u4e00\u4e2a\u811a\u672c":61,"\u6211\u4eec\u53ea\u4f7f\u7528\u5df2\u7ecf\u6807\u6ce8\u8fc7\u7684\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6":68,"\u6211\u4eec\u53ea\u6240\u4ee5\u4f7f\u7528lstm\u6765\u6267\u884c\u8fd9\u4e2a\u4efb\u52a1\u662f\u56e0\u4e3a\u5176\u6539\u8fdb\u7684\u8bbe\u8ba1\u5e76\u4e14\u5177\u6709\u95e8\u673a\u5236":68,"\u6211\u4eec\u53ea\u6f14\u793a\u4e00\u4e2a\u5355\u673a\u4f5c\u4e1a":55,"\u6211\u4eec\u53ea\u9700\u8981\u4f7f\u7528lstm":39,"\u6211\u4eec\u53ea\u9700\u8981\u8fd0\u884c":64,"\u6211\u4eec\u53ea\u9700\u8981\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u628a\u7f16\u8bd1\u597d\u7684paddlepaddle\u6253\u5305\u6210\u4e00\u4e2a":43,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528":61,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5b83\u6765\u751f\u6210\u5e8f\u5217":42,"\u6211\u4eec\u53ef\u4ee5\u5728":43,"\u6211\u4eec\u53ef\u4ee5\u5c06":48,"\u6211\u4eec\u53ef\u4ee5\u6309\u7167\u5982\u4e0b\u5c42\u6b21\u5b9a\u4e49\u975e\u5e8f\u5217":38,"\u6211\u4eec\u53ef\u4ee5\u751f\u6210":66,"\u6211\u4eec\u53ef\u4ee5\u76f4\u63a5\u89c2\u5bdf\u6a21\u578b\u7684\u53c2\u6570\u662f\u5426\u7b26\u5408\u9884\u671f\u6765\u8fdb\u884c\u68c0\u9a8c":32,"\u6211\u4eec\u53ef\u4ee5\u770b\u5230\u5728\u76ee\u5f55":68,"\u6211\u4eec\u53ef\u4ee5\u8bbe\u8ba1\u642d\u5efa\u4e00\u4e2a\u7075\u6d3b\u7684":41,"\u6211\u4eec\u53ef\u4ee5\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4\u6765\u505ableu\u8bc4\u4f30":69,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u4ee5\u4e0b\u547d\u4ee4\u6765\u8bad\u7ec3\u6a21\u578b":69,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u4ee5\u4e0b\u547d\u4ee4\u6765\u8fdb\u884c\u4ece\u6cd5\u8bed\u5230\u82f1\u8bed\u7684\u6587\u672c\u7ffb\u8bd1":69,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u5982\u4e0b\u547d\u4ee4\u8fdb\u884c\u9884\u5904\u7406\u5de5\u4f5c":61,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u65e5\u5fd7\u67e5\u770b\u5bb9\u5668\u8bad\u7ec3\u7684\u60c5\u51b5":56,"\u6211\u4eec\u57285\u5929\u91cc\u8bad\u7ec3\u4e8616\u4e2apass":69,"\u6211\u4eec\u5728\u51fd\u6570\u7684\u7ed3\u5c3e\u8fd4\u56de":42,"\u6211\u4eec\u5728\u62e5\u670950\u4e2a\u8282\u70b9\u7684\u96c6\u7fa4\u4e2d\u8bad\u7ec3\u6a21\u578b":69,"\u6211\u4eec\u5728\u8bad\u7ec3\u4e4b\u524d\u9700\u8981\u5e38\u89c1\u4e00\u4e2a\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":69,"\u6211\u4eec\u5728initialzier\u51fd\u6570\u91cc\u521d\u59cb\u5316\u8bcd\u8868":64,"\u6211\u4eec\u5bf9\u6a21\u578b\u8fdb\u884c\u4e86\u4ee5\u4e0b\u66f4\u6539":42,"\u6211\u4eec\u5c06":[56,66],"\u6211\u4eec\u5c06\u4e00\u6bb5\u8bdd\u770b\u6210\u53e5\u5b50\u7684\u6570\u7ec4":39,"\u6211\u4eec\u5c06\u4ecb\u7ecd\u5982\u4f55\u542f\u52a8\u5206\u5e03\u5f0f\u8bad\u7ec3\u4f5c\u4e1a":55,"\u6211\u4eec\u5c06\u4ee5":[48,64],"\u6211\u4eec\u5c06\u4ee5\u6700\u57fa\u672c\u7684\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u4f5c\u4e3a\u8d77\u70b9":64,"\u6211\u4eec\u5c06\u4f7f\u7528":42,"\u6211\u4eec\u5c06\u4f7f\u7528\u7b80\u5355\u7684":42,"\u6211\u4eec\u5c06\u4f7f\u7528cifar":61,"\u6211\u4eec\u5c06\u539f\u59cb\u6570\u636e\u7684\u6bcf\u4e00\u7ec4":39,"\u6211\u4eec\u5c06\u5728\u540e\u9762\u4ecb\u7ecd\u8bad\u7ec3\u548c\u9884\u6d4b\u6d41\u7a0b\u7684\u811a\u672c":64,"\u6211\u4eec\u5c06\u5b83\u4eec\u5212\u5206\u4e3a\u4e0d\u540c\u7684\u7c7b\u522b":49,"\u6211\u4eec\u5c06\u5bf9\u8fd9\u4e24\u4e2a\u6b65\u9aa4\u7ed9\u51fa\u4e86\u8be6\u7ec6\u7684\u89e3\u91ca":64,"\u6211\u4eec\u5c06\u653e\u7f6e\u4f9d\u8d56\u5e93":48,"\u6211\u4eec\u5c06\u8bc4\u5206\u5206\u6210\u4e24\u90e8\u5206":66,"\u6211\u4eec\u5c06\u9610\u91ca\u5982\u4f55\u5728\u96c6\u7fa4\u4e0a\u8fd0\u884c\u5206\u5e03\u5f0f":48,"\u6211\u4eec\u5c31\u53ef\u4ee5\u7740\u624b\u5bf9\u5206\u7c7b\u6a21\u578b\u8fdb\u884c\u8bad\u7ec3\u4e86":61,"\u6211\u4eec\u5c31\u53ef\u4ee5\u8bad\u7ec3\u6a21\u578b\u4e86":64,"\u6211\u4eec\u5c31\u53ef\u4ee5\u8fdb\u884c\u9884\u6d4b\u4e86":64,"\u6211\u4eec\u5c31\u5b8c\u6210\u4e86\u4e00\u6b21\u4ee3\u7801\u8d21\u732e\u7684\u8fc7\u7a0b":43,"\u6211\u4eec\u5c55\u793a\u5982\u4f55\u7528paddlepaddle\u89e3\u51b3":32,"\u6211\u4eec\u5df2\u7ecf\u5b9e\u73b0\u4e86\u5927\u591a\u6570\u5e38\u7528\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":42,"\u6211\u4eec\u5e0c\u671b\u80fd\u5728\u8fd9\u4e2a\u57fa\u7840\u4e0a\u4e0d\u65ad\u7684\u6539\u8fdb":0,"\u6211\u4eec\u5e0c\u671b\u80fd\u591f\u68c0\u9a8c\u6a21\u578b\u7684\u597d\u574f":32,"\u6211\u4eec\u5e94\u5f53\u4f1a\u5f97\u5230\u4e00\u4e2a\u540d\u4e3acifar":61,"\u6211\u4eec\u5efa\u8bae\u4f60\u4e3a\u4f60\u7684python\u5c01\u88c5\u5b9e\u73b0\u4e00\u4e2a":44,"\u6211\u4eec\u5efa\u8bae\u4f60\u5728\u5199\u65b0\u7f51\u7edc\u5c42\u65f6\u628a\u6d4b\u8bd5\u4ee3\u7801\u653e\u5165\u65b0\u7684\u6587\u4ef6\u4e2d":44,"\u6211\u4eec\u603b\u7ed3\u4e86\u5404\u4e2a\u7f51\u7edc\u7684\u590d\u6742\u5ea6\u548c\u6548\u679c":64,"\u6211\u4eec\u611f\u8c22":69,"\u6211\u4eec\u628apaddlepaddle\u7684\u7f16\u8bd1\u73af\u5883\u6253\u5305\u6210\u4e00\u4e2a\u955c\u50cf":34,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528":34,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":33,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528docker\u955c\u50cf\u6765\u90e8\u7f72\u73af\u5883":35,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u7f51\u7edc":68,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u8f6c\u6362\u65b9\u5f0f":26,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u6570\u636e\u9884\u5904\u7406\u811a\u672c":68,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u6784\u5efa\u811a\u672cbuild_doc":45,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u793a\u4f8b\u811a\u672c":62,"\u6211\u4eec\u63d0\u4f9b\u4e86\u56fd\u5185\u7684\u955c\u50cf\u670d\u52a1\u5668\u4f9b\u5927\u5bb6\u4f7f\u7528":34,"\u6211\u4eec\u63d0\u4f9b\u4e86\u811a\u672c\u6765\u6784\u5efa\u5b57\u5178\u548c\u9884\u5904\u7406\u6570\u6910":68,"\u6211\u4eec\u63d0\u4f9b\u4e86c":62,"\u6211\u4eec\u63d0\u4f9b\u4e86python\u5904\u7406\u6570\u636e\u7684\u63a5\u53e3":53,"\u6211\u4eec\u63d0\u4f9b\u53ef\u4ee5\u76f4\u63a5\u8fd0\u884cpaddlepaddl":34,"\u6211\u4eec\u662f\u5bf9\u6bcf\u4e00\u4e2a\u5b50\u5e8f\u5217\u53d6\u6700\u540e\u4e00\u4e2a\u5143\u7d20":39,"\u6211\u4eec\u6700\u7ec8\u7684\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165python\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u8bed\u8a00\u7684\u89e3\u91ca\u5668":27,"\u6211\u4eec\u6709\u4e00\u4e2a\u5e8f\u5217\u4f5c\u4e3a\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u72b6\u6001":42,"\u6211\u4eec\u7528":65,"\u6211\u4eec\u7528\u4ee5\u4e0b\u7684\u4e00\u4e9b":66,"\u6211\u4eec\u7528\u7f16\u53f7\u4f5c\u4e3akei":66,"\u6211\u4eec\u7528paddlepaddle\u89e3\u51b3\u4e86\u5355\u53d8\u91cf\u7ebf\u6027\u56de\u5f52\u95ee\u9898":32,"\u6211\u4eec\u7684\u5b57\u5178\u4f7f\u7528\u5185\u90e8\u7684\u5206\u8bcd\u5de5\u5177\u5bf9\u767e\u5ea6\u77e5\u9053\u548c\u767e\u5ea6\u767e\u79d1\u7684\u8bed\u6599\u8fdb\u884c\u5206\u8bcd\u540e\u4ea7\u751f":60,"\u6211\u4eec\u7684\u6807\u51c6\u5f00\u53d1\u6d41\u7a0b\u662f\u628a\u8fd9\u4e9b\u5de5\u5177\u90fd\u88c5\u8fdb\u4e00\u4e2adocker":43,"\u6211\u4eec\u7684\u8bad\u7ec3\u66f2\u7ebf\u5982\u4e0b":67,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u5c42rnn\u7684\u914d\u7f6e":39,"\u6211\u4eec\u770b\u4e00\u4e0b\u8bed\u4e49\u76f8\u540c\u7684\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":39,"\u6211\u4eec\u771f\u8bda\u5730\u611f\u8c22\u60a8\u7684\u5173\u6ce8":68,"\u6211\u4eec\u771f\u8bda\u5730\u611f\u8c22\u60a8\u7684\u8d21\u732e":43,"\u6211\u4eec\u79f0\u4e4b\u4e3a\u4e00\u4e2a0\u5c42\u7684\u5e8f\u5217":38,"\u6211\u4eec\u8fd8\u53ef\u4ee5\u767b\u5f55\u5230\u5bbf\u4e3b\u673a\u4e0a\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":55,"\u6211\u4eec\u8fd8\u5c06\u7f16\u7801\u5411\u91cf\u6295\u5c04\u5230":42,"\u6211\u4eec\u9009\u53d6\u5355\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e\u4e2d\u7684\u4e0d\u540c\u90e8\u5206":39,"\u6211\u4eec\u901a\u5e38\u5728\u6240\u6709\u8282\u70b9\u4e0a\u521b\u5efa\u4e00\u4e2a":48,"\u6211\u4eec\u901a\u5e38\u5c06\u4e00\u53e5\u8bdd\u7406\u89e3\u6210\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":39,"\u6211\u4eec\u901a\u8fc7\u8bfb\u53d6":56,"\u6211\u4eec\u9075\u5faa":69,"\u6211\u4eec\u90fd\u63d0\u4f9bpython\u7684\u8f6c\u6362\u5e93":26,"\u6211\u4eec\u91c7\u7528\u4e0a\u9762\u7684\u7279\u5f81\u6a21\u677f":67,"\u6211\u4eec\u91c7\u7528\u5355\u5c42lstm\u6a21\u578b":64,"\u6211\u4eec\u91c7\u7528\u7b2c\u4e8c\u4e2a\u65b9\u6cd5":34,"\u6211\u4eec\u9700\u8981\u5148\u521b\u5efa\u4e00\u4e2a\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":61,"\u6211\u4eec\u9700\u8981\u521b\u5efa\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":69,"\u6211\u4eec\u9700\u8981\u5236\u4f5c\u4e00\u4e2a\u5305\u542b\u8bad\u7ec3\u6570\u636e\u7684paddle\u955c\u50cf":55,"\u6211\u4eec\u9700\u8981\u5728\u96c6\u7fa4\u7684\u6240\u6709\u8282\u70b9\u4e0a\u5b89\u88c5":48,"\u6211\u4eec\u9700\u8981\u8ba1\u7b97":44,"\u6211\u4eec\u9700\u8981\u8bbe\u7f6e":66,"\u6211\u4eec\u9700\u8981\u9884\u5904\u7406\u6570\u6910\u5e76\u6784\u5efa\u4e00\u4e2a\u5b57\u5178":68,"\u6211\u4eec\u975e\u5e38\u6b22\u8fce\u60a8\u7528paddlepaddle\u6784\u5efa\u66f4\u597d\u7684\u793a\u4f8b":66,"\u6211\u4eec\u9884\u8bad\u7ec3\u5f97\u52304\u79cd\u4e0d\u540c\u7ef4\u5ea6\u7684\u8bcd\u5411\u91cf":60,"\u6211\u4eec\u9996\u5148\u5904\u7406\u7535\u5f71\u6216\u7528\u6237\u7684\u7279\u5f81\u6587\u4ef6":66,"\u6211\u4eec\u9ed8\u8ba4\u4f7f\u7528imdb\u7684\u6d4b\u8bd5\u6570\u636e\u96c6\u4f5c\u4e3a\u9a8c\u8bc1":68,"\u6211\u5220\u9664\u4e86":43,"\u6216":[3,47,53,61,67],"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"\u6216\u4e00\u4e2a\u5411\u91cf":41,"\u6216\u4e0d\u786e\u5b9a":65,"\u6216\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"\u6216\u662f\u624b\u52a8\u7f16\u8f91\u751f\u6210":66,"\u6216\u6700\u5927\u503c":38,"\u6216\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":2,"\u6216\u79f0pserver":53,"\u6216\u7b2c\u4e00\u4e2a":38,"\u6216\u7b2c\u4e00\u4e2a\u5143\u7d20":38,"\u6216\u8005":[27,28,31,34,36,38,39,47,53],"\u6216\u8005\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":41,"\u6216\u8005\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":[38,41],"\u6216\u8005\u4ece\u5de5\u5177\u7684\u754c\u9762\u91cc\u8fd0\u884c\u60a8\u7684\u5e94\u7528":47,"\u6216\u8005\u53cd\u5411\u5730\u4ece":42,"\u6216\u8005\u5728cpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2a\u7ebf\u7a0b":50,"\u6216\u8005\u5c06\u8fd9\u53f0\u8282\u70b9\u8fc1\u79fb\u5230\u53e6\u4e00\u4e2a\u8282\u70b9\u5e76\u542f\u52a8\u5373\u53ef\u6062\u590d\u8bad\u7ec3\u4efb\u52a1":25,"\u6216\u8005\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":49,"\u6216\u8005\u6570\u636e\u5e93\u8fde\u63a5\u8def\u5f84\u7b49":2,"\u6216\u8005\u6570\u7ec4\u7684\u6570\u7ec4\u8fd9\u4e2a\u6982\u5ff5":39,"\u6216\u8005\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":38,"\u6216\u8005\u662f\u4e00\u4e2a\u5c0f\u7684\u6587\u672c\u7247\u6bb5":68,"\u6216\u8005\u662f\u51fd\u6570\u8c03\u7528\u7684\u9891\u7387\u548c\u8017\u65f6\u7b49":47,"\u6216\u8005\u66f4\u65e9":31,"\u6216\u8005\u6bcf\u4e00\u4e2a\u7cfb\u5217\u91cc\u7684\u7279\u5f81\u6570\u636e":39,"\u6216\u8005\u7528tuple\u8868\u793a\u7684\u591a\u4e2a\u503c":26,"\u6216\u8005\u7531\u5b83\u4eec\u7ec4\u6210\u7684list":26,"\u6216\u8005\u76f4\u63a5\u4f7f\u7528\u4e0b\u9762\u7684shell\u547d\u4ee4":62,"\u6216\u8005\u76f4\u63a5\u6254\u6389\u975e\u5e38\u957f\u7684\u5e8f\u5217":31,"\u6216\u8005\u8f93\u5165\u6570\u636e\u5c3a\u5ea6\u8fc7\u5927":31,"\u6216\u8005\u91c7\u7528\u5e76\u884c\u8ba1\u7b97\u6765\u52a0\u901f\u67d0\u4e9b\u5c42\u7684\u66f4\u65b0":52,"\u6216\u8005\u9700\u8981\u53d1\u5e03\u60a8\u7684\u5e94\u7528\u7684\u955c\u50cf":34,"\u6216\u8005\u9700\u8981\u66f4\u9ad8\u7684\u6548\u7387":2,"\u6216\u8bbe\u7f6e\u4e3anone":2,"\u6216gpu":50,"\u6216gpu\u4e2a\u6570":68,"\u6218\u4e89\u7247":65,"\u623f":39,"\u623f\u95f4":39,"\u6240\u4ee5":[3,31,53],"\u6240\u4ee5\u4e00\u822c\u9700\u8981\u5bf9\u8bad\u7ec3\u7528\u7684\u6a21\u578b\u914d\u7f6e\u6587\u4ef6\u7a0d\u4f5c\u76f8\u5e94\u4fee\u6539\u624d\u80fd\u5728\u9884\u6d4b\u65f6\u4f7f\u7528":5,"\u6240\u4ee5\u4f60\u53ea\u7528\u6309\u4e0b\u9762\u7684\u7ed3\u6784\u6765\u7ec4\u7ec7\u6570\u6910\u5c31\u884c\u4e86":68,"\u6240\u4ee5\u505a\u6cd5\u53ef\u4ee5\u6709\u4e24\u79cd":31,"\u6240\u4ee5\u53ef\u4ee5\u5229\u7528\u5982\u4e0b\u65b9\u6cd5\u8bfb\u53d6\u6a21\u578b\u7684\u53c2\u6570":32,"\u6240\u4ee5\u53ef\u4ee5\u7b80\u5316\u5bf9\u73af\u5883\u7684\u8981\u6c42":55,"\u6240\u4ee5\u5728\u5199\u5165\u5feb\u7167\u7684\u8fc7\u7a0b\u4e2d":25,"\u6240\u4ee5\u5916\u5c42\u8f93\u51fa\u7684\u5e8f\u5217\u5f62\u72b6":39,"\u6240\u4ee5\u5982\u679c\u9700\u8981\u81ea\u884c\u914d\u7f6e\u5f00\u53d1\u73af\u5883\u9700\u8981\u8003\u8651\u7248\u672c\u7684\u56e0\u7d20":34,"\u6240\u4ee5\u5b83\u4eec\u4f7f\u7528\u540c\u4e00\u4e2aip\u5730\u5740":54,"\u6240\u4ee5\u5bf9":39,"\u6240\u4ee5\u5f88\u591a\u65f6\u5019\u4f60\u9700\u8981\u505a\u7684\u53ea\u662f\u5b9a\u4e49\u6b63\u786e\u7684\u7f51\u7edc\u5c42\u5e76\u628a\u5b83\u4eec\u8fde\u63a5\u8d77\u6765":32,"\u6240\u4ee5\u6027\u80fd\u4e5f\u5c31\u9010\u6b65\u53d8\u6210\u4e86\u6df1\u5ea6\u5b66\u4e60\u9886\u57df\u6700\u91cd\u8981\u7684\u6307\u6807":47,"\u6240\u4ee5\u6211\u4eec\u4f7f\u7528\u8fd9\u4e2a\u955c\u50cf\u6765\u4e0b\u8f7d\u8bad\u7ec3\u6570\u636e\u5230docker":55,"\u6240\u4ee5\u6211\u4eec\u53ef\u4ee5\u5728\u8fd9\u4e2a\u57fa\u7840\u4e0a":56,"\u6240\u4ee5\u6211\u4eec\u63a8\u8350\u4f7f\u7528\u57fa\u4e8edocker\u6765\u6784\u5efapaddlepaddle\u7684\u6587\u6863":45,"\u6240\u4ee5\u6211\u4eec\u9700\u8981\u5c06\u8f93\u5165\u6570\u636e\u6807\u8bb0\u6210":39,"\u6240\u4ee5\u63a8\u8350\u4f7f\u7528\u663e\u5f0f\u6307\u5b9a\u7684\u65b9\u5f0f\u6765\u8bbe\u7f6einput_typ":3,"\u6240\u4ee5\u653e\u4e00\u4e2a\u7a7a\u5217\u8868":32,"\u6240\u4ee5\u8bad\u7ec3":48,"\u6240\u4ee5\u8f93\u51fa\u7684value\u5305\u542b\u4e24\u4e2a\u5411\u91cf":5,"\u6240\u4ee5\u8fd9\u4e00\u6b65\u662f\u5fc5\u8981\u7684":44,"\u6240\u5bf9\u5e94\u7684\u8bcd\u8868index\u6570\u7ec4":39,"\u6240\u6709\u4e0e\u7c7b\u578b\u76f8\u5173\u7684\u51fd\u6570":28,"\u6240\u6709\u4ee3\u7801\u5fc5\u987b\u5177\u6709\u5355\u5143\u6d4b\u8bd5":43,"\u6240\u6709\u53c2\u6570\u7f6e\u4e3a\u96f6":50,"\u6240\u6709\u540c\u76ee\u5f55\u4e0b\u7684\u6587\u672c\u5b9e\u4f8b\u6587\u4ef6\u90fd\u662f\u540c\u7ea7\u522b\u7684":68,"\u6240\u6709\u547d\u4ee4\u884c\u9009\u9879\u53ef\u4ee5\u8bbe\u7f6e\u4e3a":48,"\u6240\u6709\u6587\u4ef6\u5217\u8868":3,"\u6240\u6709\u672c\u5730\u8bad\u7ec3":48,"\u6240\u6709\u6807\u8bb0\u7684\u6d4b\u8bd5\u96c6\u548c\u8bad\u7ec3\u96c6":68,"\u6240\u6709\u751f\u6210\u7684\u7684\u6587\u4ef6\u90fd\u5728build\u76ee\u5f55\u4e0b":34,"\u6240\u6709\u7684":[43,44],"\u6240\u6709\u7684\u4eba\u53e3\u7edf\u8ba1\u5b66\u4fe1\u606f\u7531\u7528\u6237\u81ea\u613f\u63d0\u4f9b":65,"\u6240\u6709\u7684\u5355\u6d4b\u90fd\u4f1a\u88ab\u6267\u884c\u4e00\u6b21":44,"\u6240\u6709\u7684\u63a5\u53e3\u5747\u4e3ac\u63a5\u53e3":28,"\u6240\u6709\u7684\u64cd\u4f5c\u90fd\u662f\u9488\u5bf9\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u8fdb\u884c\u7684":39,"\u6240\u6709\u7684\u7528\u6237\u4fe1\u606f\u90fd\u5305\u542b\u5728":65,"\u6240\u6709\u7684\u7535\u5f71\u4fe1\u606f\u90fd\u5305\u542b\u5728":65,"\u6240\u6709\u7684\u8bc4\u5206\u6570\u636e\u90fd\u5305\u542b\u5728":65,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u4f7f\u7528":44,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u5728":44,"\u6240\u6709\u7c7b\u578b\u540d\u4e3a":28,"\u6240\u6709\u7f51\u7edc\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u90fd\u4f4d\u4e8e":44,"\u6240\u6709\u8282\u70b9\u8fd0\u884c\u96c6\u7fa4\u4f5c\u4e1a\u7684\u4e3b\u673a\u540d\u6216":48,"\u6240\u6709\u8d21\u732e\u8005":0,"\u6240\u6709\u8f93\u5165\u5e8f\u5217\u5e94\u8be5\u6709\u76f8\u540c\u7684\u957f\u5ea6":42,"\u6240\u6709\u914d\u7f6e\u90fd\u80fd\u5728":64,"\u6240\u6784\u5efa\u7f51\u7edc\u7ed3\u6784\u7684\u7684\u6df1\u5ea6\u6bd4\u4e4b\u524d\u4f7f\u7528\u7684\u7f51\u7edc\u6709\u5927\u5e45\u5ea6\u7684\u63d0\u9ad8":62,"\u6240\u793a":67,"\u6240\u8c13\u65f6\u95f4\u6b65\u4fe1\u606f":3,"\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":27,"\u624b\u5de5\u827a\u8005":65,"\u624d\u4f1a\u91ca\u653e\u8be5\u6bb5\u5185\u5b58":3,"\u624d\u4f1astop":3,"\u624d\u80fd\u4fdd\u8bc1\u548c\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":39,"\u624d\u80fd\u5145\u5206\u5229\u7528mac":34,"\u624d\u80fd\u53d1\u6325\u5176\u5168\u90e8\u80fd\u529b":47,"\u6253\u5370\u5728\u5c4f\u5e55\u4e0a":66,"\u6253\u5370\u7684\u65e5\u5fd7\u53d8\u591a":33,"\u6253\u5f00":47,"\u6253\u5f00\u6587\u672c\u6587\u4ef6":3,"\u6253\u5f00\u6d4f\u89c8\u5668\u8bbf\u95ee\u5bf9\u5e94\u76ee\u5f55\u4e0b\u7684index":45,"\u6253\u5f00\u8fd9\u4e2a\u7f16\u8bd1\u9009\u9879":28,"\u6267\u884c":[36,67,68],"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5c31\u53ef\u4ee5\u9884\u5904\u7406\u6570\u6910":68,"\u6267\u884c\u4ee5\u4e0b\u64cd\u4f5c":42,"\u6267\u884c\u60a8\u7684\u4ee3\u7801":47,"\u6267\u884c\u7684\u547d\u4ee4\u5982\u4e0b":62,"\u6269\u5c55\u548c\u5ef6\u4f38":0,"\u6269\u5c55\u673a\u5236\u7b49\u529f\u80fd":54,"\u6279\u6b21\u540e\u6253\u5370\u65e5\u5fd7":66,"\u6279\u6b21\u5bf9\u5e73\u5747\u53c2\u6570\u8fdb\u884c\u6d4b\u8bd5":67,"\u6279\u6b21\u7684\u6570\u636e":66,"\u627e\u5230":42,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u539f\u56e0":47,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u90e8\u5206":47,"\u6280\u672f\u5458":65,"\u628a":[26,44],"\u628a\u4e4b\u524d\u793a\u4f8b\u4e2d\u8f6c\u6362\u5b8c\u6bd5\u7684random":26,"\u628a\u7528\u6237\u5728\u8d2d\u7269\u7f51\u7ad9":68,"\u628a\u7f16\u8bd1\u51fa\u6765\u7684paddlepaddle\u4e5f\u6253\u5305\u6210\u4e00\u4e2a\u955c":34,"\u628a\u8bad\u7ec3\u6570\u636e\u76f4\u63a5\u653e\u5728":55,"\u6293\u53d6\u4ea7\u54c1\u7684\u7528\u6237\u8bc4\u8bba\u5e76\u5206\u6790\u4ed6\u4eec\u7684\u60c5\u611f":68,"\u6295\u5c04\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b\u5230":42,"\u6295\u5c04\u7f16\u7801\u5411\u91cf\u5230":42,"\u62a5\u9519":36,"\u62bd\u53d6\u51fa\u7684\u65b0\u8bcd\u8868\u7684\u4fdd\u5b58\u8def\u5f84":60,"\u62bd\u53d6\u5bf9\u5e94\u7684\u8bcd\u5411\u91cf\u6784\u6210\u65b0\u7684\u8bcd\u8868":60,"\u62c6\u5206\u5230\u4e0d\u540c\u6587\u4ef6\u5939":69,"\u62c6\u89e3":41,"\u62c6\u89e3\u6210\u7684\u6bcf\u4e00\u53e5\u8bdd\u518d\u901a\u8fc7\u4e00\u4e2alstm\u7f51\u7edc":39,"\u62f7\u8d1d\u8bad\u7ec3\u6587\u4ef6\u5230\u5bb9\u5668\u5185":56,"\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5411\u91cf":64,"\u6301\u7eed\u96c6\u6210\u7cfb\u7edf\u4e5f\u53ef\u4ee5\u7b80\u5355\u5730\u590d\u73b0\u8fd9\u4e2a\u8fc7\u7a0b":34,"\u6307\u5411\u4e00\u4e2alayer":41,"\u6307\u5b9a":[31,41,42,53],"\u6307\u5b9a\u4e00\u53f0\u673a\u5668\u4e0a\u4f7f\u7528\u7684\u7ebf\u7a0b\u6570":50,"\u6307\u5b9a\u4e86dataprovider\u7684\u6587\u4ef6\u540d\u548c\u8fd4\u56de\u6570\u636e\u7684\u51fd\u6570\u540d":53,"\u6307\u5b9a\u4ee5\u592a\u7f51\u7c7b\u578b\u4e3atcp\u7f51\u7edc":53,"\u6307\u5b9a\u4f7f\u75282":31,"\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b\u8def\u5f84":64,"\u6307\u5b9a\u52a0\u8f7d\u7684\u65b9\u5f0f":50,"\u6307\u5b9a\u5de5\u4f5c\u6a21\u578b\u8fdb\u884c\u9884\u6d4b":62,"\u6307\u5b9a\u5de5\u4f5c\u6a21\u5f0f\u6765\u63d0\u53d6\u7279\u5f81":62,"\u6307\u5b9a\u63d0\u53d6\u7279\u5f81\u7684\u5c42":62,"\u6307\u5b9a\u662f\u5426\u4f7f\u7528gpu":62,"\u6307\u5b9a\u751f\u6210\u6570\u636e\u7684\u51fd\u6570":64,"\u6307\u5b9a\u7684\u5b57\u5178\u5355\u8bcd\u6570":69,"\u6307\u5b9a\u7684\u6570\u636e\u5c06\u4f1a\u88ab\u6d4b\u8bd5":64,"\u6307\u5b9a\u7684\u8f93\u5165\u4e0d\u4f1a\u88ab":41,"\u6307\u5b9a\u7f51\u7edc\u63a5\u53e3\u540d\u5b57\u4e3aeth0":53,"\u6307\u5b9a\u8bad\u7ec3\u6570\u636e\u548c\u6d4b\u8bd5\u6570\u636e":64,"\u6307\u5b9abatch":69,"\u6307\u5b9acudnn\u7684\u6700\u5927\u5de5\u4f5c\u7a7a\u95f4\u5bb9\u9650":50,"\u6307\u5bf9\u4e8e\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u8f93\u5165\u6570\u636e":39,"\u6307\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u4e4b\u540e\u5f97\u5230\u7684\u6240\u6709\u53c2\u6570":25,"\u6307\u793a\u4f7f\u7528\u54ea\u4e2agpu\u6838":50,"\u6307\u793a\u5728\u7b80\u5355\u7684recurrentlayer\u5c42\u7684\u8ba1\u7b97\u4e2d\u662f\u5426\u4f7f\u7528\u6279\u5904\u7406\u65b9\u6cd5":50,"\u6307\u793a\u5f53\u6307\u5b9a\u8f6e\u7684\u6d4b\u8bd5\u6a21\u578b\u4e0d\u5b58\u5728\u65f6":50,"\u6307\u793a\u662f\u5426\u4f7f\u7528\u591a\u7ebf\u7a0b\u6765\u8ba1\u7b97\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc":50,"\u6307\u793a\u662f\u5426\u5f00\u542f\u53c2\u6570\u670d\u52a1\u5668":50,"\u6307\u793a\u662f\u5426\u663e\u793a\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u65e5\u5fd7\u7ec6\u8282":50,"\u6307\u793a\u662f\u5426\u68c0\u67e5\u6240\u6709\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u7a00\u758f\u53c2\u6570\u7684\u5206\u5e03\u662f\u5747\u5300\u7684":50,"\u6307\u793a\u6d4b\u8bd5\u4efb\u52a1":67,"\u6307\u793a\u6d4b\u8bd5\u4efb\u52a1\u7684\u6807\u8bb0":67,"\u6309\u542f\u53d1\u5f0f\u635f\u5931\u7684\u5927\u5c0f\u9012\u589e\u6392\u5e8f":50,"\u6309\u7167\u4e0b\u9762\u6b65\u9aa4\u5373\u53ef":56,"\u6309\u94ae":43,"\u633a":39,"\u633a\u597d":39,"\u6355\u83b7":64,"\u635f\u5931\u51fd\u6570":53,"\u635f\u5931\u51fd\u6570\u5373\u4e3a\u7f51\u7edc\u7684\u4f18\u5316\u76ee\u6807":53,"\u635f\u5931\u51fd\u6570\u548c\u8bc4\u4f30\u5668":53,"\u6362":39,"\u6392\u6210\u4e00\u5217\u7684\u591a\u4e2a\u5143\u7d20":38,"\u63a5\u4e0b\u6765":[64,68],"\u63a5\u4e0b\u6765\u53ef\u4ee5\u8003\u8651\u4e0b\u65f6\u95f4\u7ebf\u7684\u5206\u6790":47,"\u63a5\u4e0b\u6765\u5c31\u53ef\u4ee5\u4f7f\u7528":47,"\u63a5\u4e0b\u6765\u6211\u4eec\u521b\u5efa\u4e00\u4e2a\u539f\u59cb":43,"\u63a5\u4e0b\u6765\u6211\u4eec\u53d6\u6d88\u5bf9":43,"\u63a5\u4e0b\u6765\u6211\u4eec\u5c06\u5c55\u793a\u5982\u4f55\u7528paddlepaddle\u8bad\u7ec3\u4e00\u4e2a\u6587\u672c\u5206\u7c7b\u6a21\u578b":64,"\u63a5\u4e0b\u6765\u7b49\u5f85":43,"\u63a5\u53d7":67,"\u63a5\u53d7\u7684\u4e1c\u897f":67,"\u63a5\u53d7\u8005":67,"\u63a5\u53e3":[27,28],"\u63a5\u53e3\u540d\u79f0":48,"\u63a5\u53e3\u5c42\u505a\u8fc7\u591a\u5c01\u88c5":28,"\u63a5\u53e3\u63d0\u53d6\u7684\u7ed3\u679c\u662f\u4e00\u81f4\u7684":62,"\u63a5\u53e3\u662f":26,"\u63a5\u53e3\u6709\u4e00\u4e2a":31,"\u63a5\u53e3\u6765\u52a0\u8f7d\u6570\u636e":64,"\u63a5\u53e3\u6765\u52a0\u8f7d\u8be5\u6587\u4ef6":62,"\u63a5\u53e3\u6765\u6253\u5f00\u8be5\u6587\u4ef6":62,"\u63a5\u53e3\u8bbe\u7f6e\u795e\u7ecf\u7f51\u7edc\u6240\u4f7f\u7528\u7684\u8bad\u7ec3\u53c2\u6570\u548c":53,"\u63a5\u7740\u6211\u4eec\u5c31\u80fd\u591f\u6253\u5f00\u6d4f\u89c8\u5668\u5728":34,"\u63a7\u5236":50,"\u63a7\u5236\u5982\u4f55\u6539\u53d8\u6a21\u578b\u53c2\u6570":32,"\u63a7\u5236\u662f\u5426\u751f\u6210\u5355\u5143\u6d4b\u8bd5":34,"\u63a7\u5236\u751f\u6210\u7684\u751f\u4ea7\u955c\u50cf\u662f\u5426\u652f\u6301avx":34,"\u63a7\u5236\u751f\u6210\u7684\u751f\u4ea7\u955c\u50cf\u662f\u5426\u652f\u6301gpu":34,"\u63a8\u5bfc\u8be5\u5c42\u524d\u5411\u548c\u540e\u5411\u4f20\u9012\u7684\u65b9\u7a0b":44,"\u63a8\u8350":39,"\u63a8\u8350\u4f7f\u7528":3,"\u63a8\u8350\u6e05\u7406\u6574\u4e2a\u7f16\u8bd1\u76ee\u5f55":33,"\u63a8\u8350\u76f4\u63a5\u5b58\u653e\u5230\u8bad\u7ec3\u76ee\u5f55":2,"\u63a8\u8350\u7cfb\u7edf":48,"\u63a8\u9001\u5230\u8fdc\u7a0b\u4ed3\u5e93":43,"\u63a8\u9500\u5458":65,"\u63cf\u8ff0":33,"\u63cf\u8ff0\u7f51\u7edc\u7ed3\u6784\u548c\u4f18\u5316\u7b97\u6cd5":64,"\u63cf\u8ff0\u95ee\u9898":43,"\u63cf\u8ff0kubernetes\u4e0a\u8fd0\u884c\u7684\u4f5c\u4e1a":54,"\u63d0\u4ea4\u65b9\u5f0f\u53ef\u53c2\u89c1":45,"\u63d0\u4f9b":48,"\u63d0\u4f9b\u4e86\u4e00\u4e2a\u542f\u52a8\u811a\u672c":56,"\u63d0\u4f9b\u4e86\u547d\u4ee4\u6837\u4f8b\u6765\u8fd0\u884c":48,"\u63d0\u4f9b\u4e86\u81ea\u52a8\u5316\u811a\u672c\u6765\u542f\u52a8\u4e0d\u540c\u8282\u70b9\u4e2d\u7684\u6240\u6709":48,"\u63d0\u4f9b\u51e0\u4e4e\u6240\u6709\u8bad\u7ec3\u7684\u5185\u90e8\u8f93\u51fa\u65e5\u5fd7":48,"\u63d0\u4f9b\u6269\u5c55\u7684\u957f\u5ea6\u4fe1\u606f":38,"\u63d0\u4f9b\u6700\u65b0":34,"\u63d0\u4f9b\u7ed9\u7528\u6237\u672c\u5730\u8f6c\u6362\u7684\u5e93":26,"\u63d0\u4f9b\u7ed9paddle\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e":26,"\u63d0\u4f9b\u8bad\u7ec3\u8fc7\u7a0b\u7684":48,"\u63d0\u51fa\u7684\u4ee3\u7801\u9700\u6c42":60,"\u63d0\u793a":31,"\u64cd\u4f5c":[39,53],"\u64cd\u6301\u5bb6\u52a1\u8005":65,"\u652f\u6301\u4e3b\u6d41x86\u5904\u7406\u5668\u5e73\u53f0":36,"\u652f\u6301\u5355\u673a\u6a21\u5f0f\u548c\u591a\u673a\u6a21\u5f0f":53,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":[40,41],"\u652f\u6301nvidia":36,"\u652f\u6301rbd":54,"\u653e\u5728\u8fd9\u4e2a\u76ee\u5f55\u91cc\u7684\u6587\u4ef6\u5176\u5b9e\u662f\u4fdd\u5b58\u5230\u4e86mfs\u4e0a":56,"\u653e\u5fc3":39,"\u6545\u800c\u662f\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":39,"\u6548\u679c\u603b\u7ed3":64,"\u6559\u7a0b\u6587\u6863\u5230":48,"\u6559\u80b2\u5de5\u4f5c\u8005":65,"\u6570":[41,67],"\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":41,"\u6570\u636e":69,"\u6570\u636e\u4e0b\u8f7d\u4e4b\u540e":61,"\u6570\u636e\u4e2d0":31,"\u6570\u636e\u5217\u8868":62,"\u6570\u636e\u5c06\u4fdd\u5b58\u5728":60,"\u6570\u636e\u5c42":[32,66],"\u6570\u636e\u5e94\u8be5\u5728\u542f\u52a8\u96c6\u7fa4\u4f5c\u4e1a\u4e4b\u524d\u51c6\u5907\u597d":48,"\u6570\u636e\u63d0\u4f9b\u5668":49,"\u6570\u636e\u63d0\u4f9b\u811a\u672c\u4ec5\u4ec5\u662f\u8bfb\u53d6meta":66,"\u6570\u636e\u63d0\u4f9b\u811a\u672c\u7684\u7ec6\u8282\u6587\u6863\u53ef\u4ee5\u53c2\u8003":66,"\u6570\u636e\u7684\u6574\u6570id":42,"\u6570\u636e\u76ee\u5f55\u4e2d\u7684\u6240\u6709\u6587\u4ef6\u88ab":48,"\u6570\u636e\u7c7b\u578b":5,"\u6570\u636e\u7f13\u5b58\u7684\u7b56\u7565":3,"\u6570\u636e\u8bbf\u95ee":1,"\u6570\u636e\u8bfb\u53d6\u5747\u4ea4\u7531\u5176\u4ed6\u8bed\u8a00\u5b8c\u6210":27,"\u6570\u636e\u8bfb\u53d6\u7a0b\u5e8f\u5f80\u5f80\u5b9a\u4e49\u5728\u4e00\u4e2a\u5355\u72ecpython\u811a\u672c\u6587\u4ef6\u91cc":53,"\u6570\u636e\u8f93\u5165":41,"\u6570\u636e\u8f93\u5165\u683c\u5f0f":3,"\u6570\u636e\u96c6":65,"\u6570\u636e\u96c6\u63cf\u8ff0":66,"\u6570\u636e\u96c6\u6587\u4ef6\u5939\u540d\u79f0":69,"\u6570\u636e\u9884\u5904\u7406\u4efb\u52a1":26,"\u6570\u636e\u9884\u5904\u7406\u5b8c\u6210\u4e4b\u540e":64,"\u6570\u636e\u9884\u6d4b":67,"\u6570\u6910\u5b9a\u4e49":68,"\u6570\u6910\u8bf4\u660e\u6587\u6863":68,"\u6570\u6910\u96c6\u548c":68,"\u6570\u76ee":52,"\u6574\u4f53":39,"\u6574\u4f53\u6570\u636e\u548c\u539f\u59cb\u6570\u636e\u5b8c\u5168\u4e00\u6837":39,"\u6574\u4f53\u7684\u7ed3\u6784\u56fe\u5982\u4e0b":56,"\u6574\u6570":44,"\u6574\u6570\u6807\u7b7e":3,"\u6574\u6d01":39,"\u6587\u4e66\u5de5\u4f5c":65,"\u6587\u4ef6":[27,43,55,67],"\u6587\u4ef6\u4e2d":[56,62,65,67],"\u6587\u4ef6\u4e2d\u6307\u5b9a\u6a21\u578b\u8def\u5f84\u548c\u8f93\u51fa\u7684\u76ee\u5f55":62,"\u6587\u4ef6\u4e2d\u6307\u5b9a\u8981\u63d0\u53d6\u7279\u5f81\u7684\u7f51\u7edc\u5c42\u7684\u540d\u5b57":62,"\u6587\u4ef6\u4e2d\u7684":62,"\u6587\u4ef6\u4e2d\u7684\u6bcf\u884c\u90fd\u5fc5\u987b\u662f\u4e00\u4e2a\u53e5\u5b50":69,"\u6587\u4ef6\u4e3a":[31,69],"\u6587\u4ef6\u4e4b\u5916":43,"\u6587\u4ef6\u4e5f\u53ef\u4ee5\u7528\u4e8e\u5bf9\u6837\u672c\u8fdb\u884c\u9884\u6d4b":62,"\u6587\u4ef6\u5185\u5bb9\u4e3a":27,"\u6587\u4ef6\u5206\u5272\u4e3a\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6587\u4ef6":66,"\u6587\u4ef6\u540d\u4e3a\u6b64uuid":25,"\u6587\u4ef6\u540d\u79f0\u4e3a":66,"\u6587\u4ef6\u59390":56,"\u6587\u4ef6\u5939\u4e2d\u7684\u6bcf\u4e2a\u6587\u4ef6\u7684\u6bcf\u4e00\u884c\u5305\u542b\u4e24\u90e8\u5206":69,"\u6587\u4ef6\u5f00\u5934":53,"\u6587\u4ef6\u7684\u5206\u9694\u7b26\u4e3a":66,"\u6587\u4ef6\u7684\u6539\u53d8":43,"\u6587\u4ef6\u7684\u683c\u5f0f\u53ef\u4ee5":66,"\u6587\u4ef6\u7a0d\u6709\u5dee\u522b":61,"\u6587\u4ef6\u7d22\u5f15":48,"\u6587\u4ef6\u7ed9\u51fa\u4e86\u5b8c\u6574\u4f8b\u5b50":64,"\u6587\u4ef6\u9700\u8981\u9884\u5148\u88ab\u8f6c\u6362\u6210paddlepaddle\u96c6\u7fa4\u5185\u90e8\u7684\u5b58\u50a8\u683c\u5f0f":26,"\u6587\u4ef6model":52,"\u6587\u5b57\u7684\u4ea4\u4e92\u5f0f\u6587\u6863":34,"\u6587\u672c\u4e2d\u7684\u5355\u8bcd\u7528\u7a7a\u683c\u5206\u9694":64,"\u6587\u672c\u4fe1\u606f\u5c31\u662f\u4e00\u4e2a\u5e8f\u5217\u6570\u636e":3,"\u6587\u672c\u5206\u7c7b\u95ee\u9898":64,"\u6587\u672c\u5377\u79ef\u5206\u53ef\u4e3a\u4e09\u4e2a\u6b65\u9aa4":64,"\u6587\u672c\u5377\u79ef\u91c7\u6837\u5c42":66,"\u6587\u672c\u6295\u5f71\u5c42":66,"\u6587\u672c\u683c\u5f0f\u7684\u5b9e\u4f8b\u6587\u4ef6":68,"\u6587\u6863":[31,53],"\u6587\u6863\u7f16\u5199\u7b49\u5de5\u4f5c":34,"\u6587\u6863\u81ea\u52a8\u5206\u7c7b\u548c\u95ee\u7b54":67,"\u6587\u6863\u90fd\u662f\u901a\u8fc7":45,"\u6587\u7ae0":56,"\u65b0":39,"\u65b0\u5efa\u4e00\u4e2a\u6743\u91cd":44,"\u65b0\u624b\u5165\u95e8":59,"\u65b0\u624b\u5165\u95e8\u7ae0\u8282":30,"\u65b9\u4fbf":39,"\u65b9\u4fbf\u4eca\u540e\u7684\u5d4c\u5165\u5f0f\u79fb\u690d\u5de5\u4f5c":33,"\u65b9\u4fbf\u5f00\u53d1\u8005\u76f4\u63a5\u767b\u5f55\u5230\u955c\u50cf\u4e2d\u8fdb\u884c\u5f00\u53d1":34,"\u65b9\u4fbf\u6d4b\u8bd5\u4eba\u5458\u6d4b\u8bd5paddle\u7684\u884c\u4e3a":30,"\u65b9\u5f0f1":31,"\u65b9\u5f0f2":31,"\u65b9\u6848\u6765\u6807\u8bb0\u6bcf\u4e2a\u53c2\u6570":67,"\u65b9\u6cd5\u4e00":52,"\u65b9\u6cd5\u4e09":52,"\u65b9\u6cd5\u4e8c":52,"\u65c1\u8fb9":39,"\u65c5\u6e38\u7f51\u7ad9":68,"\u65e0":39,"\u65e0\u4e1a\u4eba\u58eb":65,"\u65e0\u5ef6\u8fdf":50,"\u65e0\u6cd5\u505a\u5230\u5bf9\u4e8e\u5404\u79cd\u8bed\u8a00\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u7684\u9002\u914d":27,"\u65e0\u8bba\u5728\u672c\u5730\u8fd8\u662f\u5728\u4e91\u7aef":26,"\u65e0\u8bba\u662f\u5728\u672c\u5730\u6216\u662f\u4e91\u7aef\u8f6c\u6362":26,"\u65e5\u5fd7\u5c06\u4fdd\u5b58\u5728":68,"\u65e8\u5728\u5efa\u7acb\u4e00\u4e2a\u53ef\u4ee5\u88ab\u534f\u540c\u8c03\u81f3\u6700\u4f18\u7ffb\u8bd1\u6548\u679c\u7684\u5355\u795e\u7ecf\u5143\u7f51\u7edc":69,"\u65e9\u9910":39,"\u65f6":[25,31,38,42,44,50,56],"\u65f6\u5019":39,"\u65f6\u52a0\u4e0a":68,"\u65f6\u5e8f\u6a21\u578b\u5747\u4f7f\u7528\u8be5\u811a\u672c":64,"\u65f6\u5e8f\u6a21\u578b\u662f\u6307\u6570\u636e\u7684\u67d0\u4e00\u7ef4\u5ea6\u662f\u4e00\u4e2a\u5e8f\u5217\u5f62\u5f0f":3,"\u65f6\u76ee\u6807\u8bed\u8a00\u7684\u6587\u4ef6":69,"\u65f6\u88ab\u8bad\u7ec3\u7684":44,"\u65f6\u8bbe\u5907id\u53f7\u7684\u5206\u914d":52,"\u65f6\u95f4":39,"\u65f6\u95f4\u6233":65,"\u65f6\u95f4\u6233\u8868\u793a\u4e3a\u4ece1970":65,"\u65f6\u95f4\u6b65\u7684\u6982\u5ff5":39,"\u6620\u5c04\u5230\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":44,"\u662f":[33,39,53],"\u662f\u4e00\u4e2a\u51681\u7684\u5411\u91cf":44,"\u662f\u4e00\u4e2a\u5185\u7f6e\u7684\u5b9a\u65f6\u5668\u5c01\u88c5":47,"\u662f\u4e00\u4e2a\u52a8\u6001\u7a0b\u5e8f\u5206\u6790\u7684\u672f\u8bed":47,"\u662f\u4e00\u4e2a\u5305\u88c5\u6570\u636e\u7684":67,"\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u662f\u4e00\u4e2a\u53cc\u5c42\u7684\u5e8f\u5217":38,"\u662f\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3\u7684\u4ee3\u7801\u751f\u6210\u5668":27,"\u662f\u4e00\u4e2a\u5c01\u88c5\u5bf9\u8c61":47,"\u662f\u4e00\u4e2a\u5f88\u6709\u7528\u7684\u53c2\u6570":52,"\u662f\u4e00\u4e2a\u7b26\u5408\u9ad8\u65af\u5206\u5e03\u7684\u968f\u673a\u53d8\u91cf":32,"\u662f\u4e00\u4e2a\u7c7b\u578b\u7684\u6807\u5fd7":28,"\u662f\u4e00\u4e2a\u7ec4\u5408\u5c42":53,"\u662f\u4e00\u4e2a\u7edf\u8ba1\u5b66\u7684\u673a\u5668\u7ffb\u8bd1\u7cfb\u7edf":69,"\u662f\u4e00\u4e2a\u914d\u7f6e\u6587\u4ef6\u7684\u4f8b\u5b50":68,"\u662f\u4e00\u4e2a\u975e\u7ebf\u6027\u7684":44,"\u662f\u4e00\u4e2apython\u7684":3,"\u662f\u4e00\u4e2aunbound":41,"\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":3,"\u662f\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":41,"\u662f\u4e00\u7ec4":54,"\u662f\u4e0d\u5305\u62ec\u6e90\u7801\u7684":55,"\u662f\u4e0d\u5e38\u89c1\u7684\u505a\u6cd5":27,"\u662f\u4e0d\u662f\u5f88\u7b80\u5355\u5462":3,"\u662f\u4e0d\u662f\u8981\u5bf9\u6570\u636e\u505ashuffl":3,"\u662f\u4e3b\u8981\u7684\u53ef\u6267\u884cpython\u811a\u672c":67,"\u662f\u4ec0\u4e48\u4e5f\u6ca1\u5173\u7cfb":3,"\u662f\u4f17\u591a\u8bef\u5dee\u4ee3\u4ef7\u51fd\u6570\u5c42\u7684\u4e00\u79cd":32,"\u662f\u4f7f\u5f97\u8981\u5171\u4eab\u7684\u53c2\u6570\u4f7f\u7528\u540c\u6837\u7684":31,"\u662f\u504f\u5dee":42,"\u662f\u5176\u5927\u5c0f":32,"\u662f\u51e0\u4e4e\u4e0d\u5360\u5185\u5b58\u7684":3,"\u662f\u539f\u59cb\u6cd5\u8bed\u6587\u4ef6":69,"\u662f\u5404\u4e2a\u5b9e\u73b0\u4e2d\u5171\u4eab\u7684\u5934\u6587\u4ef6":28,"\u662f\u5411\u91cf":44,"\u662f\u5426\u4ee5\u9006\u5e8f\u5904\u7406\u8f93\u5165\u5e8f\u5217":41,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":33,"\u662f\u5426\u4f7f\u7528\u65e7\u7684remoteparameterupdat":50,"\u662f\u5426\u4f7f\u7528\u6743\u91cd":44,"\u662f\u5426\u4f7f\u7528gpu":66,"\u662f\u5426\u4f7f\u7528gpu\u8bad\u7ec3":69,"\u662f\u5426\u5141\u8bb8\u6682\u5b58\u7565\u5fae\u591a\u4f59pool_size\u7684\u6570\u636e":3,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":33,"\u662f\u5426\u5c06\u5168\u5c40\u79cd\u5b50\u5e94\u7528\u4e8e\u672c\u5730\u7ebf\u7a0b\u7684\u968f\u673a\u6570":50,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":33,"\u662f\u5426\u5f00\u542f\u8ba1\u65f6\u529f\u80fd":33,"\u662f\u5426\u5f00\u542frdma":33,"\u662f\u5426\u6253\u5370\u7248\u672c\u4fe1\u606f":50,"\u662f\u5426\u652f\u6301gpu":33,"\u662f\u5426\u663e\u793a":50,"\u662f\u5426\u7a00\u758f":44,"\u662f\u5426\u7f16\u8bd1\u4e2d\u82f1\u6587\u6587\u6863":33,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236\u6587\u4ef6":33,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":33,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":33,"\u662f\u5426\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":33,"\u662f\u5426\u9700\u8981\u7b49\u5f85\u8be5\u8f6e\u6a21\u578b\u53c2\u6570":50,"\u662f\u56e0\u4e3ac99\u652f\u6301":27,"\u662f\u56e0\u4e3apaddle\u7684\u7f51\u7edc\u901a\u4fe1\u4e2d":53,"\u662f\u56e0\u4e3apaddlepaddle\u914d\u7f6e\u6587\u4ef6\u4e0ec":53,"\u662f\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u52a0\u8f7d\u5b57\u5178\u5e76\u5b9a\u4e49\u6570\u636e\u63d0\u4f9b\u7a0b\u5e8f\u6a21\u5757\u548c\u7f51\u7edc\u67b6\u6784\u7684\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":67,"\u662f\u5728paddlepaddle\u4e2d\u6784\u9020\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u6700\u91cd\u8981\u7684\u6982\u5ff5":42,"\u662f\u57fa\u7840\u7684\u8ba1\u7b97\u5355\u5143":32,"\u662f\u5b58\u6709\u4e00\u7cfb\u5217\u53d8\u6362\u77e9\u9635\u7684\u6743\u91cd":44,"\u662f\u5b58\u6709\u504f\u7f6e\u5411\u91cf\u7684\u6743\u91cd":44,"\u662f\u5e8f\u5217":66,"\u662f\u5f85\u6269\u5c55\u7684\u6570\u636e":38,"\u662f\u6211\u4eec":43,"\u662f\u6307":28,"\u662f\u6307\u4e00\u4e2a\u6570\u636e\u5217\u8868\u6587\u4ef6":53,"\u662f\u6307\u4e00\u7cfb\u5217\u7684\u7279\u5f81\u6570\u636e":39,"\u662f\u6307recurrent_group\u7684\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":39,"\u662f\u6570\u636e\u8f93\u5165":42,"\u662f\u6574\u4e2a\u7684\u8bcd\u5d4c\u5165":66,"\u662f\u6709\u610f\u4e49\u7684":39,"\u662f\u6784\u5efa\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u6700\u91cd\u8981\u7684\u5de5\u5177":42,"\u662f\u6a21\u578b\u53c2\u6570\u4f18\u5316\u7684\u76ee\u6807\u51fd\u6570":32,"\u662f\u6d45\u5c42\u8bed\u4e49\u89e3\u6790\u7684\u4e00\u79cd\u5f62\u5f0f":67,"\u662f\u6e90\u8bed\u8a00\u7684\u6587\u4ef6":69,"\u662f\u7528\u6237\u4f7f\u7528c":28,"\u662f\u76ee\u6807\u82f1\u8bed\u6587\u4ef6":69,"\u662f\u77e9\u9635":44,"\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u5165\u53e3":32,"\u662f\u7f51\u7edc\u548c\u6570\u636e\u914d\u7f6e\u6587\u4ef6":61,"\u662f\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u6807\u8bc6\u7b26":44,"\u662f\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26":44,"\u662f\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":44,"\u662f\u7f51\u7edc\u5c42\u8f93\u51fa\u7684\u5927\u5c0f":44,"\u662f\u8be5\u5c42\u7684\u6807\u8bc6\u7b26":44,"\u662f\u8be5\u5c42\u7684\u7c7b\u540d":44,"\u662f\u8be5\u7f51\u7edc\u5c42\u7684":44,"\u662f\u8f93\u5165":42,"\u662f\u901a\u7528\u7269\u4f53\u5206\u7c7b\u9886\u57df\u4e00\u4e2a\u4f17\u6240\u5468\u77e5\u7684\u6570\u636e\u5e93":62,"\u662f\u9700\u8981\u4e86\u89e3\u54ea\u4e9b\u6b65\u9aa4\u62d6\u6162\u4e86\u6574\u4f53":47,"\u662fc":28,"\u662fdecoder\u7684\u6570\u636e\u8f93\u5165":41,"\u662fgoogle\u5f00\u6e90\u7684\u5bb9\u5668\u96c6\u7fa4\u7ba1\u7406\u7cfb\u7edf":54,"\u662fnvidia\u6027\u80fd\u5206\u6790\u5de5\u5177":47,"\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":41,"\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":64,"\u662fpod\u5185\u7684\u5bb9\u5668\u90fd\u53ef\u4ee5\u8bbf\u95ee\u7684\u5171\u4eab\u76ee\u5f55":54,"\u662fpython\u5c01\u88c5\u7684\u7c7b\u540d":44,"\u662frnn\u72b6\u6001":42,"\u663e":64,"\u663e\u5f0f\u6307\u5b9a\u8fd4\u56de\u7684\u662f\u4e00\u4e2a28":3,"\u665a":39,"\u666e\u901a\u7528\u6237\u8bf7\u8d70\u5b89\u88c5\u6d41\u7a0b":35,"\u6682\u4e0d\u8003\u8651\u5728\u5185":31,"\u6682\u65f6\u4e0d\u8003\u8651\u591a\u4e2aparamet":25,"\u66b4\u9732\u8fd9\u4e2a\u6982\u5ff5\u5fc5\u8981\u51fd\u6570":28,"\u66f4\u591a\u5173\u4e8edocker\u7684\u5b89\u88c5\u4e0e\u4f7f\u7528":31,"\u66f4\u591a\u5185\u5bb9\u53ef\u67e5\u770b\u53c2\u8003\u6587\u732e":68,"\u66f4\u591a\u7684\u7ec6\u8282\u53ef\u4ee5\u5728\u6587\u732e\u4e2d\u627e\u5230":68,"\u66f4\u597d\u5730\u5b8c\u6210\u4e00\u4e9b\u590d\u6742\u7684\u8bed\u8a00\u7406\u89e3\u4efb\u52a1":41,"\u66f4\u5feb":42,"\u66f4\u65b0":31,"\u66f4\u65b0\u6a21\u5f0f":31,"\u66f4\u65b9\u4fbf\u7684\u8bbe\u7f6e\u65b9\u5f0f":31,"\u66f4\u8be6\u7ec6\u6570\u636e\u683c\u5f0f\u548c\u7528\u4f8b\u8bf7\u53c2\u8003":64,"\u66f4\u8be6\u7ec6\u7684\u4f7f\u7528":53,"\u66f4\u8be6\u7ec6\u7684\u7f51\u7edc\u914d\u7f6e\u8fde\u63a5\u8bf7\u53c2\u8003":64,"\u66f4\u8be6\u7ec6\u7684\u8bf4\u660e":64,"\u66f4\u8fdb\u4e00\u6b65":41,"\u66f4\u9ad8":42,"\u66ff\u6211\u4eec\u5b8c\u6210\u4e86\u539f\u59cb\u8f93\u5165\u6570\u636e\u7684\u62c6\u5206":41,"\u6700":39,"\u6700\u4e0d\u540c\u7684\u7279\u8272\u662f\u5b83\u5e76\u6ca1\u6709\u5c06\u8f93\u5165\u8bed\u53e5\u7f16\u7801\u4e3a\u4e00\u4e2a\u5355\u72ec\u7684\u5b9a\u957f\u5411\u91cf":69,"\u6700\u4e3b\u8981\u7684\u5de5\u4f5c\u5c31\u662f\u89e3\u6790\u51fa":56,"\u6700\u4f73\u63a8\u8350":3,"\u6700\u540e":[3,43,44,48,64,68],"\u6700\u540e\u4e00\u4e2a":38,"\u6700\u540e\u4e00\u90e8\u5206\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u914d\u7f6e":32,"\u6700\u540e\u5220\u9664":30,"\u6700\u540e\u6211\u4eec\u4f7f\u7528\u94fe\u5f0f\u6cd5\u5219\u8ba1\u7b97":44,"\u6700\u597d\u7684\u6a21\u578b\u662f":68,"\u6700\u5c11\u663e\u793a\u591a\u5c11\u4e2a\u8282\u70b9":50,"\u6700\u5e38\u89c1\u7684\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662fexcept":27,"\u6700\u65b0\u7684\u4ee3\u7801":43,"\u6700\u65b0log":68,"\u6700\u7ec8":44,"\u6700\u7ec8\u5b9e\u73b0\u4e00\u4e2a\u5c42\u6b21\u5316\u7684\u590d\u6742rnn":41,"\u6700\u7ec8\u751f\u6210\u540d\u4e3apaddl":34,"\u6700\u7ec8\u7684\u8f93\u51fa\u7ed3\u679c":41,"\u6700\u7ec8\u8d8b\u4e8e\u63a5\u8fd1":32,"\u6708\u6e56":39,"\u6709":39,"\u6709\u4e00\u4e2a\u57fa\u672c\u7684\u8ba4\u8bc6":54,"\u6709\u4e24\u4e2a\u7279\u6b8a\u6807\u8bc6":69,"\u6709\u4e86meta\u914d\u7f6e\u6587\u4ef6\u4e4b\u540e":66,"\u6709\u4e9b\u5c42\u53ef\u80fd\u9700\u8981\u9ad8\u7cbe\u5ea6\u6765\u4fdd\u8bc1\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u6b63\u786e\u6267\u884c":44,"\u6709\u4e9b\u5c42\u6216\u8005\u6fc0\u6d3b\u9700\u8981\u505a\u5f52\u4e00\u5316\u4ee5\u4fdd\u8bc1\u5b83\u4eec\u7684\u8f93\u51fa\u7684\u548c\u662f\u4e00\u4e2a\u5e38\u6570":44,"\u6709\u4e9b\u7279\u5f81\u7684\u53d6\u503c\u8fbe\u5230\u6570\u767e\u4e07":31,"\u6709\u4e9b\u7535\u5f71id\u53ef\u80fd\u4e0e\u5b9e\u9645\u7535\u5f71\u4e0d\u76f8\u7b26\u5408":65,"\u6709\u5173":39,"\u6709\u5173\u5982\u4f55\u7f16\u5199\u6570\u636e\u63d0\u4f9b\u7a0b\u5e8f\u7684\u66f4\u591a\u7ec6\u8282\u63cf\u8ff0":42,"\u6709\u5173kubernetes\u76f8\u5173\u6982\u5ff5\u4ee5\u53ca\u5982\u4f55\u642d\u5efa\u548c\u914d\u7f6ekubernetes\u96c6\u7fa4":56,"\u6709\u52a9\u4e8e\u7406\u89e3\u7528\u6237\u5bf9\u4e0d\u540c\u516c\u53f8":68,"\u6709\u52a9\u4e8e\u8bca\u65ad\u5206\u5e03\u5f0f\u9519\u8bef":48,"\u6709\u56db\u4e2a\u8bad\u7ec3\u8fdb\u7a0b":53,"\u6709\u65f6\u79f0\u4e3a":68,"\u6709\u6807\u51c6\u7684":27,"\u6709\u7684\u65f6\u5019":27,"\u6709\u7684\u65f6\u5019\u7b80\u7b80\u5355\u5355\u7684\u6539\u53d8\u5c31\u80fd\u5728\u6027\u80fd\u4e0a\u4ea7\u751f\u660e\u663e\u7684\u4f18\u5316\u6548\u679c":47,"\u670d\u52a1":39,"\u670d\u52a1\u5458":39,"\u672a\u5305\u542b\u5728\u5b57\u5178\u4e2d\u7684\u5355\u8bcd":69,"\u672a\u6807\u8bb0\u7684\u8bc4\u4ef7\u6837\u672c":68,"\u672a\u77e5\u8bcd":60,"\u672c\u4f8b\u4e2d\u4e3a0":60,"\u672c\u4f8b\u4e2d\u4e3a32":60,"\u672c\u4f8b\u4e2d\u4e3a4":60,"\u672c\u4f8b\u4e2d\u4f7f\u7528for\u5faa\u73af\u8fdb\u884c\u591a\u6b21\u8c03\u7528":3,"\u672c\u4f8b\u4e2d\u7684\u539f\u59cb\u6570\u636e\u4e00\u5171\u670910\u4e2a\u6837\u672c":39,"\u672c\u4f8b\u4e2d\u7684\u8f93\u5165\u7279\u5f81\u662f\u8bcdid\u7684\u5e8f\u5217":3,"\u672c\u4f8b\u6839\u636e\u7f51\u7edc\u914d\u7f6e\u4e2d":3,"\u672c\u4f8b\u6bcf\u884c\u4fdd\u5b58\u4e00\u6761\u6837\u672c":64,"\u672c\u4f8b\u7531\u6613\u5230\u96be\u5c55\u793a4\u79cd\u4e0d\u540c\u7684\u6587\u672c\u5206\u7c7b\u7f51\u7edc\u914d\u7f6e":64,"\u672c\u4f8b\u7684":3,"\u672c\u4f8b\u7684\u6240\u6709\u5b57\u7b26\u90fd\u5c06\u8f6c\u6362\u4e3a\u8fde\u7eed\u6574\u6570\u8868\u793a\u7684id\u4f20\u7ed9\u6a21\u578b":64,"\u672c\u4f8b\u91c7\u7528\u82f1\u6587\u60c5\u611f\u5206\u7c7b\u7684\u6570\u636e":3,"\u672c\u4f8b\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":64,"\u672c\u5217\u8868\u8bf4\u660epaddle\u53d1\u7248\u4e4b\u524d\u9700\u8981\u6d4b\u8bd5\u7684\u529f\u80fd\u70b9":30,"\u672c\u5730\u6d4b\u8bd5":49,"\u672c\u5730\u8bad\u7ec3":49,"\u672c\u5730\u8bad\u7ec3\u7684\u5b9e\u9a8c":52,"\u672c\u5b9e\u4f8b\u4e2d":60,"\u672c\u5c0f\u8282\u6211\u4eec\u5c06\u4ecb\u7ecd\u6a21\u578b\u7f51\u7edc\u7ed3\u6784":64,"\u672c\u5c42\u5c3a\u5bf8":62,"\u672c\u5c42\u6709\u56db\u4e2a\u53c2\u6570":62,"\u672c\u6559\u7a0b\u4e2d\u6211\u4eec\u7ed9\u51fa\u4e86\u4e09\u4e2aresnet\u6a21\u578b":62,"\u672c\u6559\u7a0b\u5c06\u4ecb\u7ecd\u4f7f\u7528\u6df1\u5ea6\u53cc\u5411\u957f\u77ed\u671f\u8bb0\u5fc6":67,"\u672c\u6559\u7a0b\u5c06\u6307\u5bfc\u4f60\u5982\u4f55\u5728":42,"\u672c\u6559\u7a0b\u5c06\u6307\u5bfc\u60a8\u5b8c\u6210\u957f\u671f\u77ed\u671f\u8bb0\u5fc6":68,"\u672c\u6559\u7a0b\u63d0\u4f9b\u4e86\u4e00\u4e2a\u7528\u4e8eimagenet\u4e0a\u7684\u5377\u79ef\u5206\u7c7b\u7f51\u7edc\u6a21\u578b":62,"\u672c\u6587\u4e2d\u6240\u6709\u7684\u4f8b\u5b50":39,"\u672c\u6587\u4e2d\u7531\u4e8e\u8f93\u5165\u6570\u636e\u662f\u968f\u673a\u751f\u6210\u7684\u4e0d\u9700\u8981\u8bfb\u8f93\u5165\u6587\u4ef6":32,"\u672c\u6587\u4e2d\u793a\u4f8b\u6240\u4f7f\u7528\u7684\u5355\u5143\u6d4b\u8bd5\u6587\u4ef6\u662f":39,"\u672c\u6587\u4ee5paddlepaddle\u7684\u53cc\u5c42rnn\u5355\u5143\u6d4b\u8bd5\u4e3a\u793a\u4f8b":39,"\u672c\u6587\u53ea\u4f7f\u7528\u4e86\u9ed8\u8ba4\u547d\u540d\u7a7a\u95f4":54,"\u672c\u6587\u5c06\u4ecb\u7ecd\u5728kubernetes\u5bb9\u5668\u7ba1\u7406\u5e73\u53f0\u4e0a\u5feb\u901f\u6784\u5efapaddlepaddle\u5bb9\u5668\u96c6\u7fa4":56,"\u672c\u6587\u6863\u4ecb\u7ecd\u5982\u4f55\u5728paddlepaddle\u5e73\u53f0\u4e0a":60,"\u672c\u6587\u6863\u5185\u4e0d\u91cd\u590d\u4ecb\u7ecd":54,"\u672c\u6587\u6863\u63cf\u8ff0paddl":28,"\u672c\u6587\u9996\u5148\u4ecb\u7ecdtrainer\u8fdb\u7a0b\u4e2d\u7684\u4e00\u4e9b\u4f7f\u7528\u6982\u5ff5":53,"\u672c\u6765":39,"\u672c\u6b21\u8bad\u7ec3\u6587\u4ef6\u6240\u5728\u76ee\u5f55":56,"\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6\u53ef\u4ee5\u5199\u6210":56,"\u672c\u6b21\u8bad\u7ec3\u8981\u6c42\u67093\u4e2apaddlepaddle\u8282\u70b9":56,"\u672c\u6b21\u8bd5\u9a8c":64,"\u672c\u793a\u4f8b\u4e2d\u4f7f\u7528\u7684\u539f\u59cb\u6570\u636e\u5982\u4e0b":39,"\u672c\u793a\u4f8b\u610f\u56fe\u4f7f\u7528\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":39,"\u672c\u793a\u4f8b\u7684\u9884\u6d4b\u7ed3\u679c":68,"\u672c\u7bc7\u6559\u7a0b\u5728paddlepaddle\u4e2d\u91cd\u73b0\u4e86\u8fd9\u4e00\u826f\u597d\u7684\u8bad\u7ec3\u7ed3\u679c":69,"\u672c\u7bc7\u6559\u7a0b\u5c06\u4f1a\u6307\u5bfc\u4f60\u901a\u8fc7\u8bad\u7ec3\u4e00\u4e2a":69,"\u672c\u8d28\u4e0a\u4e0e\u8bad\u7ec3\u6a21\u578b\u4e00\u6837":69,"\u673a\u5668\u7684\u8bbe\u5907":52,"\u673a\u5668\u7ffb\u8bd1":[30,63],"\u6743\u91cd\u66f4\u65b0\u7684\u68af\u5ea6":50,"\u6761\u4ef6\u4e0b":54,"\u6765":39,"\u6765\u4ee3\u66ff":43,"\u6765\u4fdd\u8bc1\u8bad\u7ec3\u8fc7\u7a0b\u53ef\u4ee5\u4ece\u4e2d\u95f4\u72b6\u6001\u91cd\u65b0\u542f\u52a8":25,"\u6765\u505a\u68af\u5ea6\u68c0\u67e5":44,"\u6765\u505ableu\u8bc4\u4f30":69,"\u6765\u505c\u6b62\u8bad\u7ec3":66,"\u6765\u5206\u6790\u6267\u884c\u6587\u4ef6":47,"\u6765\u5206\u79bb\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6587\u4ef6":66,"\u6765\u5206\u9694\u6bcf\u4e00\u884c":66,"\u6765\u521d\u59cb\u5316\u53c2\u6570":31,"\u6765\u5b89\u88c5":48,"\u6765\u5b9a\u4e49\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":42,"\u6765\u5bf9\u6bd4\u5206\u6790\u4e24\u8005\u8bed\u4e49\u76f8\u540c\u7684\u539f\u56e0":39,"\u6765\u5e2e\u52a9\u4f60\u7406\u89e3paddlepaddle\u7684\u5185\u90e8\u8fd0\u884c\u673a\u5236":64,"\u6765\u5f00\u542f\u672c\u5730\u7684\u8bad\u7ec3":68,"\u6765\u5f97\u5230\u67d0\u4e2a\u7279\u5b9a\u53c2\u6570\u7684\u68af\u5ea6\u77e9\u9635":44,"\u6765\u6307\u5b9a\u7f51\u7edc\u5c42\u7684\u6570\u76ee":62,"\u6765\u63a5\u53d7\u4e0d\u4f7f\u7528\u7684\u51fd\u6570\u4ee5\u4fdd\u8bc1\u517c\u5bb9\u6027":3,"\u6765\u6ce8\u518c\u8be5\u5c42":44,"\u6765\u6df7\u5408\u4f7f\u7528gpu\u548ccpu\u8ba1\u7b97\u7f51\u7edc\u5c42\u7684\u53c2\u6570":52,"\u6765\u751f\u6210\u5e8f\u5217":69,"\u6765\u7684\u79d2\u6570":65,"\u6765\u786e\u4fdd\u628a":27,"\u6765\u786e\u5b9a\u5bf9\u5e94\u5173\u7cfb":3,"\u6765\u81ea\u5b9a\u4e49\u4f20\u6570\u636e\u7684\u8fc7\u7a0b":2,"\u6765\u83b7\u5f97\u8f93\u51fa\u7684\u68af\u5ea6":44,"\u6765\u8868\u793a":42,"\u6765\u8868\u793a\u53c2\u6570\u4f4d\u7f6e":67,"\u6765\u8868\u793a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":42,"\u6765\u8868\u793apaddle\u5185\u90e8\u7c7b":27,"\u6765\u8ba1\u7b97\u68af\u5ea6":44,"\u6765\u8bb2\u89e3\u5982\u4f55\u4f7f\u7528\u53cc\u5c42rnn":39,"\u6765\u8bbe\u7f6e":31,"\u6765\u8bf4\u660epydataprovider2\u7684\u7b80\u5355\u4f7f\u7528\u573a\u666f":3,"\u6765\u8c03\u6574":43,"\u6765\u8fd0\u884c":48,"\u6765\u8fd0\u884c\u6027\u80fd\u5206\u6790\u548c\u8c03\u4f18":47,"\u6765\u8fd0\u884c\u955c\u50cf":34,"\u6765\u8fdb\u884c\u8ba8\u8bba":28,"\u6765\u9884\u6d4b\u8fd9\u4e2a\u4e2d\u95f4\u7684\u8bcd":31,"\u676f\u5b50":39,"\u6784\u6210\u4e86\u8f93\u51fa\u53cc\u5c42\u5e8f\u5217\u7684\u7b2ci\u4e2a":38,"\u6784\u9020":56,"\u6784\u9020paddl":5,"\u67b6\u6784\u5bf9\u591a\u4e2a\u8282\u70b9\u7684":53,"\u67b6\u6784\u6765\u8bad\u7ec3\u60c5\u611f\u5206\u6790\u6a21\u578b":68,"\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u4e00\u4e2a\u8f93\u5165\u4e3a\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7f51\u7edc\u4e2d\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u8f93\u51fa":39,"\u67d0\u4e9b\u53c2\u6570\u53ea\u53ef\u7528\u4e8e\u7279\u5b9a\u7684\u5c42\u4e2d":49,"\u67e5\u770b":[43,64],"\u67e5\u770b\u5b89\u88c5\u540e\u7684paddl":36,"\u67e5\u770b\u5f53\u524d\u72b6\u6001":43,"\u67e5\u770b\u5f53\u524d\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":43,"\u67e5\u770b\u6587\u4ef6\u5177\u4f53\u88ab\u4fee\u6539\u7684\u5185\u5bb9":43,"\u67e5\u770bjob\u7684\u8be6\u7ec6\u60c5\u51b5":55,"\u6807\u51c6\u5dee\u4e3a":31,"\u6807\u51c6\u8868\u793apaddle\u7248\u672c\u53f7":30,"\u6807\u51c6lstm\u4ee5\u6b63\u5411\u5904\u7406\u8be5\u5e8f\u5217":67,"\u6807\u793a\u56fe\u7247\u662f\u5f69\u8272\u56fe\u6216\u7070\u5ea6\u56fe":61,"\u6807\u793a\u662f\u5426\u4e3a\u5f69\u8272\u56fe\u7247":61,"\u6807\u7b7e0\u8868\u793a\u8d1f\u9762\u7684\u8bc4\u8bba":68,"\u6807\u7b7e1\u8868\u793a\u6b63\u9762\u7684\u8bc4\u8bba":68,"\u6807\u7b7e\u4e0b\u627e\u5230\u6700\u65b0\u7684paddle\u955c\u50cf\u7248\u672c":34,"\u6807\u7b7e\u6587\u4ef6":67,"\u6807\u7b7e\u65b9\u6848\u6765\u81ea":67,"\u6807\u8bb0":53,"\u6807\u8bb0\u7f51\u7edc\u8f93\u51fa\u7684\u51fd\u6570\u4e3a":53,"\u6807\u8bc6\u662f\u5426\u4e3a\u8fde\u7eed\u7684batch\u8ba1\u7b97":50,"\u6839\u636e\u4f60\u7684\u4efb\u52a1":52,"\u6839\u636e\u524d\u6587\u7684\u63cf\u8ff0":56,"\u6839\u636e\u5728\u6a21\u578b\u914d\u7f6e\u6587\u4ef6\u4e2d\u4f7f\u7528\u7684\u540d\u4e3a":48,"\u6839\u636e\u6570\u636e\u91cf\u89c4\u6a21":65,"\u6839\u636e\u7528\u6237\u6307\u5b9a\u7684\u5b57\u5178":60,"\u6839\u636e\u7d22\u5f15\u77e9\u9635\u548c\u5b57\u5178\u6253\u5370\u6587\u672c":42,"\u6839\u636e\u7f51\u7edc\u914d\u7f6e\u4e2d\u7684":50,"\u6839\u636e\u8fd9\u4e9b\u53c2\u6570\u7684\u4f7f\u7528\u573a\u5408":49,"\u6839\u636e\u9ed8\u8ba4\u503c\u9012\u589e":50,"\u6839\u636e\u9ed8\u8ba4\u7aef\u53e3\u53f7\u9012\u589e":50,"\u6839\u636ecpu":34,"\u6839\u636ejob\u5bf9\u5e94\u7684pod\u4fe1\u606f":55,"\u683c\u5f0f":50,"\u683c\u5f0f\u5982\u4e0b":64,"\u683c\u5f0f\u8bf4\u660e":60,"\u68af\u5ea6\u4f1a\u5c31\u5730":44,"\u68af\u5ea6\u53c2\u6570\u7684\u5206\u5757\u6570\u76ee":50,"\u68af\u5ea6\u5c31\u53ef\u4ee5\u901a\u8fc7\u8fd9\u4e2a\u65b9\u7a0b\u8ba1\u7b97\u5f97\u5230":44,"\u68af\u5ea6\u670d\u52a1\u5668\u7684\u6570\u91cf":50,"\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5\u901a\u8fc7\u6709\u9650\u5dee\u5206\u6cd5\u6765\u9a8c\u8bc1\u4e00\u4e2a\u5c42\u7684\u68af\u5ea6":44,"\u68af\u5ea6\u68c0\u67e5\u7684\u8f93\u5165\u6570\u636e\u7684\u6279\u6b21\u5927\u5c0f":44,"\u68c0\u67e5\u70b9\u4fdd\u5b58\u7a0b\u5e8f\u6d41\u7a0b":25,"\u68d2":64,"\u697c\u5c42":39,"\u6a21\u5757":61,"\u6a21\u5757\u4e2d\u7684":3,"\u6a21\u5757\u63a5\u7ba1\u4e86shuffl":53,"\u6a21\u5757\u901a\u4fe1\u7684\u6700\u57fa\u7840\u534f\u8bae\u662fprotobuf":53,"\u6a21\u578b":67,"\u6a21\u578b\u4e00\u76f4\u4e0d\u6536\u655b":31,"\u6a21\u578b\u4fdd\u5b58\u5728\u76ee\u5f55":68,"\u6a21\u578b\u5171\u5305\u542b1":60,"\u6a21\u578b\u5217\u8868\u6587\u4ef6":67,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\u901a\u8fc7\u5b9a\u671f\u5411\u78c1\u76d8\u4e0a\u4fdd\u5b58\u4e00\u4efd\u5b58\u50a8\u5728paramet":25,"\u6a21\u578b\u53ca\u53c2\u6570\u4f1a\u88ab\u4fdd\u5b58\u5728\u8def\u5f84":61,"\u6a21\u578b\u5b58\u50a8\u8def\u5f84":64,"\u6a21\u578b\u5c06\u4fdd\u5b58\u5728\u76ee\u5f55":67,"\u6a21\u578b\u5c31\u8bad\u7ec3\u6210\u529f\u4e86":69,"\u6a21\u578b\u6570\u636e\u68c0\u67e5\u70b9\u7684\u5b9e\u73b0":25,"\u6a21\u578b\u6587\u4ef6\u5c06\u88ab\u5199\u5165\u8282\u70b9":48,"\u6a21\u578b\u6765\u5c06\u6cd5\u8bed\u7ffb\u8bd1\u6210\u82f1\u8bed":69,"\u6a21\u578b\u6765\u6307\u5bfc\u4f60\u5b8c\u6210\u8fd9\u4e9b\u6b65\u9aa4":42,"\u6a21\u578b\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":42,"\u6a21\u578b\u7684\u4ee3\u7801\u53ef\u4ee5\u5728":42,"\u6a21\u578b\u7684\u7ed3\u6784\u548c\u8bad\u7ec3\u8fc7\u7a0b":60,"\u6a21\u578b\u7684\u7f16\u7801\u5668\u90e8\u5206\u5982\u4e0b\u6240\u793a":42,"\u6a21\u578b\u88ab\u4fdd\u5b58\u5728":66,"\u6a21\u578b\u8bad\u7ec3\u4f1a\u770b\u5230\u7c7b\u4f3c\u4e0a\u9762\u8fd9\u6837\u7684\u65e5\u5fd7\u4fe1\u606f":64,"\u6a21\u578b\u8bad\u7ec3\u548c\u6700\u540e\u7684\u7ed3\u679c\u9a8c\u8bc1":32,"\u6a21\u578b\u8def\u5f84":[62,67],"\u6a21\u578b\u8f93\u51fa\u8def\u5f84":67,"\u6a21\u578b\u914d\u7f6e":[1,53],"\u6a21\u578b\u914d\u7f6e\u89e3\u6790":27,"\u6a21\u578b\u91c7\u7528":60,"\u6a21\u578b\u9884\u6d4b":5,"\u6b21":39,"\u6b22\u8fce\u901a\u8fc7":43,"\u6b63\u5219\u65b9\u6cd5\u7b49":53,"\u6b63\u6837\u672c":64,"\u6b63\u786e\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":31,"\u6b63\u8d1f\u5bf9\u9a8c\u8bc1":49,"\u6b63\u9762\u7684\u8bc4\u8bba\u7684\u5f97\u5927\u4e8e\u7b49\u4e8e7":68,"\u6b63\u9762\u8bc4\u4ef7\u6837\u672c":68,"\u6b64\u5904":60,"\u6b64\u5904\u90fd\u4e3a2":39,"\u6b64\u6559\u7a0b\u5c06\u5411\u60a8\u5206\u6b65\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528\u5185\u7f6e\u7684\u5b9a\u65f6\u5de5\u5177":47,"\u6b64\u6570\u636e\u96c6\u5305\u542b\u7535\u5f71\u8bc4\u8bba\u53ca\u5176\u76f8\u5173\u8054\u7684\u7c7b\u522b\u6807\u7b7e":68,"\u6b64\u65f6\u60a8\u53ef\u4ee5\u8fd0\u884c\u8fd9\u4e2a\u547d\u4ee4\u5728\u5f00\u53d1\u673a\u4e0a\u8fdb\u884c\u6d4b\u8bd5\u8fd0\u884c":34,"\u6b64\u65f6master\u5c06\u8d1f\u8d23\u542f\u52a8\u4e00\u4e2a\u65b0\u7684train":25,"\u6bb5\u843d\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u53cc\u5c42\u7684\u5e8f\u5217":41,"\u6bcf100\u4e2abatch\u6253\u5370\u4e00\u6b21\u7edf\u8ba1\u4fe1\u606f":68,"\u6bcf100\u4e2abatch\u663e\u793a\u53c2\u6570\u7edf\u8ba1":67,"\u6bcf20\u4e2abatch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":68,"\u6bcf20\u4e2abatch\u8f93\u51fa\u65e5\u5fd7":67,"\u6bcf\u4e00\u4e2a":30,"\u6bcf\u4e00\u4e2a\u4efb\u52a1\u6d41\u7a0b\u90fd\u53ef\u4ee5\u88ab\u5212\u5206\u4e3a\u5982\u4e0b\u4e94\u4e2a\u6b65\u9aa4":64,"\u6bcf\u4e00\u4e2a\u6587\u4ef6\u662f\u6570\u636e\u96c6\u7684\u4e00\u4e2ashard":26,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65":39,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u95f4\u7684\u795e\u7ecf\u7f51\u7edc\u5177\u6709\u4e00\u5b9a\u7684\u76f8\u5173\u6027":39,"\u6bcf\u4e00\u4e2a\u6d4b\u8bd5\u5468\u671f\u6d4b\u8bd5\u4e00\u6b21\u6240\u6709\u6570\u636e":66,"\u6bcf\u4e00\u4e2a\u8282\u70b9\u90fd\u6709\u76f8\u540c\u7684\u65e5\u5fd7\u7ed3\u6784":48,"\u6bcf\u4e00\u4e2akey\u7531":66,"\u6bcf\u4e00\u7ec4\u5185\u7684\u6240\u6709\u53e5\u5b50\u548clabel":39,"\u6bcf\u4e00\u884c\u8868\u793a\u4e00\u4e2a\u5b9e\u4f8b":68,"\u6bcf\u4e2a":[42,48,67],"\u6bcf\u4e2a\u503c\u7684\u7c7b\u578b\u53ef\u4ee5\u662f\u6574\u5f62":26,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u6bcf\u4e2a\u5355\u5c42rnn":41,"\u6bcf\u4e2a\u5355\u8bcd\u7684\u9884\u6d4b\u9519\u8bef\u7387":69,"\u6bcf\u4e2a\u53e5\u5b50\u53c8\u662f\u5355\u8bcd\u7684\u6570\u7ec4":39,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u5f00\u59cb\u6807\u8bb0\u5f00\u5934":42,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u7ed3\u675f\u6807\u8bb0\u7ed3\u5c3e":42,"\u6bcf\u4e2a\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u4e00\u81f4":39,"\u6bcf\u4e2a\u5b50\u6587\u4ef6\u5939\u4e0b\u5b58\u50a8\u76f8\u5e94\u5206\u7c7b\u7684\u56fe\u7247":61,"\u6bcf\u4e2a\u5b57\u5178\u5305\u542b\u603b\u517130000\u4e2a\u5355\u8bcd":69,"\u6bcf\u4e2a\u5b57\u5178\u90fd\u6709dictsize\u4e2a\u5355\u8bcd":69,"\u6bcf\u4e2a\u5c42\u5728\u5176":44,"\u6bcf\u4e2a\u5c42\u90fd\u6709\u4e00\u4e2a\u6216\u591a\u4e2ainput":64,"\u6bcf\u4e2a\u6279\u6b21\u6570\u636e":50,"\u6bcf\u4e2a\u6574\u6570\u5217\u8868\u88ab\u89c6\u4e3a\u4e00\u4e2a\u6574\u6570\u5e8f\u5217":42,"\u6bcf\u4e2a\u6587\u4ef6\u5939\u90fd\u5305\u542b\u6cd5\u8bed\u5230\u82f1\u8bed\u7684\u5e73\u884c\u8bed\u6599\u5e93":69,"\u6bcf\u4e2a\u6587\u4ef6\u662f\u4e00\u4e2a\u7535\u5f71\u8bc4\u8bba":68,"\u6bcf\u4e2a\u6587\u672c\u6587\u4ef6\u5305\u542b\u4e00\u4e2a\u6216\u8005\u591a\u4e2a\u5b9e\u4f8b":68,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":41,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u90fd\u7528\u4e86\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u7ed3\u679c":39,"\u6bcf\u4e2a\u6743\u91cd\u5bf9\u5e94\u4e00\u4e2a\u8f93\u5165":44,"\u6bcf\u4e2a\u6837\u672c\u7531\u4e24\u90e8\u5206\u7ec4\u6210":39,"\u6bcf\u4e2a\u6837\u672c\u95f4\u7528\u7a7a\u884c\u5206\u5f00":39,"\u6bcf\u4e2a\u6d4b\u8bd5\u5468\u671f\u6d4b\u8bd5":66,"\u6bcf\u4e2a\u7279\u5f81\u7684meta\u914d\u7f6e":66,"\u6bcf\u4e2a\u72b6\u6001":41,"\u6bcf\u4e2a\u7c7b\u522b\u4e2d\u968f\u673a\u62bd\u53d6\u4e8610\u5f20\u56fe\u7247":61,"\u6bcf\u4e2a\u7c7b\u5305\u542b6000\u5f20":61,"\u6bcf\u4e2a\u7ebf\u7a0b":50,"\u6bcf\u4e2a\u7ebf\u7a0b\u5206\u914d\u5230128\u4e2a\u6837\u672c\u7528\u4e8e\u8bad\u7ec3":50,"\u6bcf\u4e2a\u8282\u70b9\u6709\u4e24\u4e2a6\u6838cpu":69,"\u6bcf\u4e2a\u8bad\u7ec3\u8282\u70b9\u5fc5\u987b\u6307\u5b9a\u4e00\u4e2a\u552f\u4e00\u7684id\u53f7":50,"\u6bcf\u4e2a\u8bb0\u5fc6\u5355\u5143\u5305\u542b\u56db\u4e2a\u4e3b\u8981\u7684\u5143\u7d20":68,"\u6bcf\u4e2a\u8bc4\u8bba\u7684\u7f51\u5740":68,"\u6bcf\u4e2a\u8f93\u5165\u90fd\u662f\u4e00\u4e2a":44,"\u6bcf\u4e2a\u8f93\u51fa\u8282\u70b9\u90fd\u8fde\u63a5\u5230\u6240\u6709\u7684\u8f93\u5165\u8282\u70b9\u4e0a":44,"\u6bcf\u4e2a\u91cc\u9762\u90fd\u5305\u542b202mb\u7684\u5168\u90e8\u7684\u6a21\u578b\u53c2\u6570":69,"\u6bcf\u4e2adata":26,"\u6bcf\u4e2alayer\u8fd4\u56de\u7684\u90fd\u662f\u4e00\u4e2a":53,"\u6bcf\u4e2aparamet":25,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":64,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":64,"\u6bcf\u4e2apod\u5305\u542b\u4e00\u4e2apaddlepaddle\u5bb9\u5668":56,"\u6bcf\u4e2ashard\u5206\u522b\u5b58\u50a8\u5728\u5176\u4e2d\u4e00\u53f0paramet":25,"\u6bcf\u4f4d\u7528\u6237\u81f3\u5c11\u670920\u6761\u8bc4\u5206":65,"\u6bcf\u5c42\u4e0a\u53ea\u80fd\u4fdd\u5b58\u56fa\u5b9a\u6570\u76ee\u4e2a\u6700\u597d\u7684\u72b6\u6001":50,"\u6bcf\u5c42\u4f7f\u7528\u7684gpu\u53f7\u4f9d\u8d56\u4e8e\u53c2\u6570train":52,"\u6bcf\u5f53\u6a21\u578b\u5728\u7ffb\u8bd1\u8fc7\u7a0b\u4e2d\u751f\u6210\u4e86\u4e00\u4e2a\u5355\u8bcd":69,"\u6bcf\u5f53\u7cfb\u7edf\u9700\u8981\u65b0\u7684\u6570\u636e\u8bad\u7ec3\u65f6":53,"\u6bcf\u6279\u6b21":50,"\u6bcf\u6b21":34,"\u6bcf\u6b21\u63d0\u4ea4\u4ee3\u7801":43,"\u6bcf\u6b21\u6d4b\u8bd5\u90fd\u6d4b\u8bd5\u6240\u6709\u6570\u636e":68,"\u6bcf\u6b21\u751f\u62101\u4e2a\u5e8f\u5217":69,"\u6bcf\u6b21\u8bfb\u53d6\u4e00\u6761\u6570\u636e\u540e":64,"\u6bcf\u6b21\u8f93\u51fa\u4e00\u4e2adata":26,"\u6bcf\u6b21\u90fd\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":3,"\u6bcf\u884c\u5b58\u50a8\u4e00\u4e2a\u8bcd":60,"\u6bcf\u884c\u5b58\u50a8\u7684\u662f\u4e00\u4e2a\u6837\u672c\u7684\u7279\u5f81":62,"\u6bcf\u884c\u6253\u537032\u4e2a\u53c2\u6570\u4ee5":60,"\u6bcf\u884c\u8868\u793a\u4e00\u4e2a\u6279\u6b21\u4e2d\u7684\u5355\u4e2a\u8f93\u5165":44,"\u6bcf\u884c\u90fd\u662f\u4e00\u4e2a\u6cd5\u8bed\u6216\u8005\u82f1\u8bed\u7684\u53e5\u5b50":69,"\u6bcf\u8f6e\u4f1a\u5c06\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u8bad\u7ec3\u6837\u672c\u4f7f\u7528\u4e00\u6b21":50,"\u6bcf\u8f6e\u7ed3\u675f\u65f6\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":50,"\u6bcf\u8f6e\u90fd\u4f1a\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c":50,"\u6bcf\u8fd0\u884c\u591a\u5c11\u4e2a\u6279\u6b21\u6267\u884c\u4e00\u6b21\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u68c0\u67e5":50,"\u6bcf\u969410\u5206\u949f":25,"\u6bcf\u9694\u591a\u5c11batch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":64,"\u6bcfdot":50,"\u6bcflog":50,"\u6bcfsave":50,"\u6bcftest":50,"\u6bd4\u5982":[26,31,34,43,64],"\u6bd4\u5982\u4e00\u53e5\u8bdd\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5355\u8bcd":39,"\u6bd4\u5982\u5982\u679c\u8981build\u4e00\u4e2a\u4e0d\u4f9d\u8d56gpu":43,"\u6bd4\u5982\u6bcf\u969410\u5206\u949f\u6700\u65b0\u7684\u5feb\u7167":25,"\u6bd4\u5982\u6d41\u5f0f\u6570\u636e\u5904\u7406":26,"\u6bd4\u5982\u8bbe\u7f6e\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u53c2\u6570\u521d\u59cb\u5316\u65b9\u5f0f\u548cbias\u521d\u59cb\u5316\u65b9\u5f0f":31,"\u6bd4\u5982\u901a\u8fc78080\u7aef\u53e3":54,"\u6bd4\u5982ide\u914d\u7f6e\u91cc":43,"\u6bd4\u5982imagenet\u8fd9\u4e2a\u6570\u636e\u96c6\u53ef\u80fd\u88ab\u5206\u62101000\u4e2ashard":26,"\u6bd4\u5982mnist":26,"\u6bd4\u8f83\u5bb9\u6613\u5728\u5927\u6a21\u578b\u4e0b\u6ea2\u51fa":53,"\u6c34\u6e29":39,"\u6c49\u5ead":39,"\u6c60\u5316\u5c42":61,"\u6ca1":39,"\u6ca1\u6709\u4f5c\u7528":3,"\u6ca1\u6709\u4f7f\u7528avx\u6307\u4ee4\u96c6":36,"\u6ca1\u6709\u5b9e\u9645\u610f\u4e49":60,"\u6ca1\u6709\u6d4b\u8bd5\u6570\u636e":3,"\u6ca1\u6709\u8fdb\u884c\u6b63\u786e\u6027\u7684\u68c0\u67e5":65,"\u6ca1\u6709\u8fdb\u884c\u7ed3\u6784\u7684\u5fae\u8c03":66,"\u6cd5\u8bed":69,"\u6ce8":25,"\u6ce8\u610f":[3,33,34,42,44,45,56,61],"\u6ce8\u610f\u4e0a\u8ff0\u547d\u4ee4\u4e2d":56,"\u6ce8\u610f\u5230\u6211\u4eec\u5df2\u7ecf\u5047\u8bbe\u673a\u5668\u4e0a\u67094\u4e2agpu":52,"\u6ce8\u610f\u5e94\u8be5\u786e\u4fdd\u9ed8\u8ba4\u6a21\u578b\u8def\u5f84":68,"\u6ce8\u610f\u9884\u6d4b\u6570\u636e\u901a\u5e38\u4e0d\u5305\u542blabel":5,"\u6ce8\u610fnode":56,"\u6ce8\u91ca\u6389":68,"\u6cf3\u6c60":39,"\u6d41":39,"\u6d41\u7a0b\u6765\u63d0\u4ea4\u4ee3\u7801":43,"\u6d44":39,"\u6d4b\u8bd5":43,"\u6d4b\u8bd5\u6570\u636e":48,"\u6d4b\u8bd5\u6570\u636e\u4e5f\u5305\u542b":48,"\u6d4b\u8bd5\u6570\u636e\u548c\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":48,"\u6d4b\u8bd5\u6570\u636e\u548c\u751f\u6210\u6570\u636e":69,"\u6d4b\u8bd5\u6570\u636e\u653e\u7f6e\u5728\u5de5\u4f5c\u7a7a\u95f4\u4e2d\u4e0d\u540c\u76ee\u5f55\u7684\u8981\u6c42":48,"\u6d4b\u8bd5\u6570\u636e\u7684\u6240\u6709\u76f8\u5bf9\u6216\u7edd\u5bf9\u6587\u4ef6\u8def\u5f84":48,"\u6d4b\u8bd5\u6570\u6910\u96c6":68,"\u6d4b\u8bd5\u65f6\u6307\u5b9a\u7684\u5b58\u50a8\u6a21\u578b\u5217\u8868\u7684\u6587\u4ef6":50,"\u6d4b\u8bd5\u65f6\u9ed8\u8ba4\u4e0dshuffl":3,"\u6d4b\u8bd5\u662f":43,"\u6d4b\u8bd5\u6837\u672c":48,"\u6d4b\u8bd5\u6a21\u578b\u662f\u6307\u4f7f\u7528\u8bad\u7ec3\u51fa\u7684\u6a21\u578b\u8bc4\u4f30\u5df2\u6807\u8bb0\u7684\u9a8c\u8bc1\u96c6":68,"\u6d4b\u8bd5\u7684\u6a21\u578b\u5305\u62ec\u4ece\u7b2cm\u8f6e\u5230\u7b2cn":52,"\u6d4b\u8bd5\u811a\u672c\u662f":67,"\u6d4b\u8bd5\u96c6\u548c\u8bad\u7ec3\u96c6\u76ee\u5f55\u5305\u542b\u4e0b\u9762\u7684\u6587\u4ef6":68,"\u6d4b\u8bd5docker\u955c\u50cf":30,"\u6d4b\u8bd5model_list":49,"\u6d4b\u8bd5save_dir":49,"\u6d4f\u89c8\u4ee3\u7801":34,"\u6d6a\u6f2b\u7247":65,"\u6d6e\u70b9\u578b\u6570\u636e":26,"\u6d6e\u70b9\u5f02\u5e38\u901a\u5e38\u7684\u539f\u56e0\u662f\u6d6e\u70b9\u6570\u6ea2\u51fa":31,"\u6d6e\u70b9\u6570\u5360\u7528\u7684\u5b57\u8282\u6570":60,"\u6d6e\u70b9\u7a00\u758f\u6570\u636e":44,"\u6dd8\u5b9d\u7b49":68,"\u6df1\u5ea6\u53cc\u5411lstm\u5c42\u63d0\u53d6softmax\u5c42\u7684\u7279\u5f81":67,"\u6df1\u5ea6\u5b66\u4e60\u5165\u95e8\u8bfe\u7a0b":37,"\u6df7\u5408":67,"\u6df7\u5408\u5f53\u524d\u8bcd\u5411\u91cf\u548cattention\u52a0\u6743\u7f16\u7801\u5411\u91cf":42,"\u6dfb\u52a0\u542f\u52a8\u811a\u672c":56,"\u6e05\u7406\u6389\u8001\u65e7\u7684paddlepaddle\u5b89\u88c5\u5305":31,"\u6e29\u99a8":39,"\u6e90":69,"\u6e90\u4ee3\u7801":[34,64],"\u6e90\u4ee3\u7801\u683c\u5f0f":43,"\u6e90\u5b57\u5178":69,"\u6e90\u5e8f\u5217":42,"\u6e90\u7801":34,"\u6e90\u7801\u4e0edemo":55,"\u6e90\u7801\u7f16\u8bd1\u6587\u6863":45,"\u6e90\u8bed\u8a00\u5230\u76ee\u6807\u8bed\u8a00\u7684\u5e73\u884c\u8bed\u6599\u5e93\u6587\u4ef6":69,"\u6e90\u8bed\u8a00\u548c\u76ee\u6807\u8bed\u8a00\u5171\u4eab\u76f8\u540c\u7684\u7f16\u7801\u5b57\u5178":60,"\u6e90\u8bed\u8a00\u548c\u76ee\u6807\u8bed\u8a00\u90fd\u662f\u76f8\u540c\u7684\u8bed\u8a00":60,"\u6e90\u8bed\u8a00\u77ed\u8bed\u548c\u76ee\u6807\u8bed\u8a00\u77ed\u8bed\u7684\u5b57\u5178\u5c06\u88ab\u5408\u5e76":60,"\u6ee4\u6ce2\u5668\u6838\u5728\u5782\u76f4\u65b9\u5411\u4e0a\u7684\u5c3a\u5bf8":62,"\u6ee4\u6ce2\u5668\u6838\u5728\u6c34\u5e73\u65b9\u5411\u4e0a\u7684\u5c3a\u5bf8":62,"\u6f14\u793a\u4e2d\u4f7f\u7528\u7684":67,"\u6f14\u793a\u91c7\u7528":67,"\u6fc0\u6d3b":44,"\u6fc0\u6d3b\u51fd\u6570":53,"\u6fc0\u6d3b\u51fd\u6570\u4e3asoftmax":53,"\u6fc0\u6d3b\u51fd\u6570\u7c7b\u578b":64,"\u6fc0\u6d3b\u65b9\u7a0b":44,"\u6fc0\u6d3b\u7684\u7c7b\u578b":44,"\u6fc0\u6d3b\u7c7b\u578b\u7b49":53,"\u7075\u6d3b\u6027\u548c\u53ef\u6269\u5c55\u6027":0,"\u70ed\u60c5":39,"\u7136\u540e":[47,48,60,66],"\u7136\u540e\u4ea4\u7ed9\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u51fd\u6570":32,"\u7136\u540e\u4ea4\u7ed9step\u51fd\u6570":41,"\u7136\u540e\u4ecb\u7ecdpserver\u8fdb\u7a0b\u4e2d\u6982\u5ff5":53,"\u7136\u540e\u4f60\u53ea\u9700\u8981\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4":69,"\u7136\u540e\u4f7f\u7528\u4e0b\u9762\u7684\u811a\u672c":68,"\u7136\u540e\u518d\u505a\u4e00\u6b21\u6587\u672c\u5377\u79ef\u7f51\u7edc\u64cd\u4f5c":66,"\u7136\u540e\u5229\u7528\u89c2\u6d4b\u6570\u636e\u8c03\u6574":32,"\u7136\u540e\u52a0":53,"\u7136\u540e\u5355\u51fb":43,"\u7136\u540e\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":56,"\u7136\u540e\u53ef\u4ee5\u8f6c\u6362\u4e3a\u56fe\u7247":62,"\u7136\u540e\u5728":69,"\u7136\u540e\u5728\u4e0b\u4e00\u4e2a\u65f6\u95f4\u6b65\u8f93\u5165\u7ed9\u53e6\u4e00\u4e2a\u795e\u7ecf\u5143":39,"\u7136\u540e\u5728\u6d4f\u89c8\u5668\u4e2d\u8f93\u5165\u4ee5\u4e0b\u7f51\u5740":34,"\u7136\u540e\u5728\u89e3\u7801\u88ab\u7ffb\u8bd1\u7684\u8bed\u53e5\u65f6":69,"\u7136\u540e\u5728dataprovider\u91cc\u9762\u6839\u636e\u8be5\u5730\u5740\u52a0\u8f7d\u5b57\u5178":31,"\u7136\u540e\u5728etcd\u7684":25,"\u7136\u540e\u5b9a\u4e49":42,"\u7136\u540e\u5c06\u6784\u5efa\u6210\u529f\u7684\u955c\u50cf\u4e0a\u4f20\u5230\u955c\u50cf\u4ed3\u5e93":56,"\u7136\u540e\u5f97\u5230\u5e73\u5747\u91c7\u6837\u7684\u7ed3\u679c":66,"\u7136\u540e\u6211\u4eec\u5229\u7528\u591a\u8f93\u5165\u7684":66,"\u7136\u540e\u6211\u4eec\u53d1\u73b0pass":69,"\u7136\u540e\u6211\u4eec\u5806\u53e0\u4e00\u5bf9\u5bf9\u7684":67,"\u7136\u540e\u6211\u4eec\u6c42\u8fd9\u4e24\u4e2a\u7279\u5f81\u7684\u4f59\u5f26\u76f8\u4f3c\u5ea6":66,"\u7136\u540e\u6240\u6709\u7528":43,"\u7136\u540e\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4":62,"\u7136\u540e\u628a\u8fd9\u4e2a\u5305\u542b\u4e86\u8bad\u7ec3\u6570\u636e\u7684container\u4fdd\u5b58\u4e3a\u4e00\u4e2a\u65b0\u7684\u955c\u50cf":55,"\u7136\u540e\u63d0\u4ea4\u65b0\u6dfb\u52a0\u7684":43,"\u7136\u540e\u63d0\u53d6\u9690\u85cflstm\u5c42\u7684\u6240\u6709\u65f6\u95f4\u6b65\u957f\u7684\u6700\u5927\u8bcd\u5411\u91cf\u4f5c\u4e3a\u6574\u4e2a\u5e8f\u5217\u7684\u8868\u793a":68,"\u7136\u540e\u662f\u5bf9\u5e94\u7684\u82f1\u8bed\u5e8f\u5217":69,"\u7136\u540e\u6dfb\u52a0\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42":66,"\u7136\u540e\u70b9\u51fb":43,"\u7136\u540e\u7528\u5bc6\u7801":34,"\u7136\u540e\u7528pickle\u547d\u4ee4\u5c06\u7279\u5f81":66,"\u7136\u540e\u7533\u660e\u4e00\u4e2a\u5b58\u50a8\u5377":56,"\u7136\u540e\u89c2\u5bdf\u5230\u8f93\u51fa\u7684\u53d8\u5316\u4e3a":44,"\u7136\u540e\u89e3\u538b":69,"\u7136\u540e\u89e3\u7801\u5668\u901a\u8fc7\u8fd9\u4e2a\u5411\u91cf\u751f\u6210\u4e00\u4e2a\u76ee\u6807\u8bed\u53e5":69,"\u7136\u540e\u8f93\u51fa\u9884\u6d4b\u5206\u6570":66,"\u7136\u540e\u8fd4\u56de\u7ed9paddlepaddle\u8fdb\u7a0b":3,"\u7136\u540e\u8fdb\u884c\u968f\u673a\u6253\u4e71":66,"\u7136\u540e\u901a\u8fc7\u51fd\u6570":56,"\u7136\u540e\u901a\u8fc7\u81ea\u8eab\u7684ip\u5730\u5740\u5728":56,"\u7136\u800c":[42,50],"\u7136\u800c\u6709\u4e9b\u8bc4\u8bba\u4e0a\u4e0b\u6587\u975e\u5e38\u957f":68,"\u7248\u672c":36,"\u7248\u672c\u5206\u652f":30,"\u7248\u672c\u53f7":30,"\u7248\u672c\u53f7rc":30,"\u7248\u672c\u5728":43,"\u7248\u672cfork\u51fa\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f":30,"\u7279\u522b\u611f\u8c22paddlepaddle\u7684":0,"\u7279\u522b\u662f\u5728lstm\u7b49rnn\u4e2d":31,"\u7279\u522b\u662f\u5f53\u76f8\u540c\u7684\u8bcd\u5728\u53e5\u5b50\u4e2d\u51fa\u73b0\u591a\u4e8e\u4e00\u6b21\u65f6":67,"\u7279\u5f81":66,"\u7279\u5f81\u56fe\u5747\u503c":62,"\u7279\u5f81\u56fe\u65b9\u5dee":62,"\u7279\u5f81\u5c06\u4f1a\u5b58\u5230":62,"\u7279\u5f81\u6587\u4ef6":67,"\u7279\u5f81\u7684\u7c7b\u578b\u548c\u7ef4\u5ea6":66,"\u72af\u7f6a\u7247":65,"\u73af\u5883\u53d8\u91cf":56,"\u73af\u5883\u53d8\u91cf\u6765\u6307\u5b9a\u7279\u5b9a\u7684gpu":31,"\u73b0\u5728\u6211\u4eec\u53ef\u4ee5\u5f00\u59cbpaddle\u8bad\u7ec3\u4e86":66,"\u73b0\u9636\u6bb5paddle\u6709\u4e00\u4e2a\u95ee\u9898\u662f":27,"\u751a\u81f3\u4e0d\u540c\u7ade\u4e89\u5bf9\u624b\u4ea7\u54c1\u7684\u504f\u597d":68,"\u751a\u81f3\u53ef\u4ee5\u76f4\u63a5\u914d\u7f6e\u4e00\u4e2a\u5b8c\u6574\u7684lstm":53,"\u751a\u81f3\u80fd\u89e3\u91ca\u4e3a\u4ec0\u4e48\u67d0\u4e2a\u64cd\u4f5c\u82b1\u4e86\u5f88\u957f\u65f6\u95f4":47,"\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u8bad\u7ec3\u6570\u636e\u96c6\u901a\u5e38\u4f53\u79ef\u5f88\u5927":26,"\u751f\u4ea7\u73af\u5883\u7684\u65e5\u5fd7\u6570\u636e\u4f1a\u901a\u8fc7\u5b9e\u65f6\u6d41\u7684\u65b9\u5f0f":26,"\u751f\u4ea7\u955c\u50cf":[34,43],"\u751f\u4ea7\u955c\u50cf\u7684\u751f\u6210\u5206\u4e3a\u4e24\u6b65":34,"\u751f\u6210":56,"\u751f\u6210\u5404\u79cd\u8bed\u8a00\u7684\u7ed1\u5b9a\u4ee3\u7801":27,"\u751f\u6210\u540e\u7684\u6587\u6863\u5206\u522b\u5b58\u50a8\u5728\u7f16\u8bd1\u76ee\u5f55\u7684":45,"\u751f\u6210\u5e8f\u5217\u7684\u6700\u5927\u957f\u5ea6":42,"\u751f\u6210\u5f53\u524d\u5c42\u7684\u6240\u6709\u540e\u7ee7\u72b6\u6001":69,"\u751f\u6210\u6570\u636e\u51fd\u6570\u63a5\u53e3":53,"\u751f\u6210\u6570\u636e\u7684\u76ee\u5f55":69,"\u751f\u6210\u6587\u6863":27,"\u751f\u6210\u751f\u4ea7\u955c\u50cf\u53ea\u9700\u8981\u8fd0\u884c":34,"\u751f\u6210\u7684":26,"\u751f\u6210\u7684\u6570\u636e\u5c06\u4f1a\u5b58\u50a8\u5728\u8fd9\u4e2avolume\u4e0b":56,"\u751f\u6210\u7684\u6570\u636e\u7f13\u5b58\u5728\u5185\u5b58\u91cc":31,"\u751f\u6210\u7684\u7ed3\u679c\u6587\u4ef6":69,"\u751f\u6210\u7684html\u7248\u672c\u7684c":34,"\u751f\u6210\u7684meta\u914d\u7f6e\u6587\u4ef6\u5982\u4e0b\u6240\u793a":66,"\u751f\u6210\u7ed3\u679c\u6587\u4ef6\u7684\u8def\u5f84":42,"\u751f\u6210\u7ed9\u5b9asstable\u6587\u4ef6\u5bf9\u5e94\u7684data":26,"\u751f\u6210\u7f51\u7edc\u5c42\u914d\u7f6e":44,"\u751f\u6210\u81ea\u5df1\u76ee\u5f55\u4e0b\u7684\u4ed3\u5e93":43,"\u751f\u6210\u8bad\u7ec3\u9700\u8981\u7684\u6837\u672c":66,"\u751f\u6210\u8fd0\u884c\u7a0b\u5e8f":34,"\u751f\u6210api\u6587\u6863":27,"\u751f\u6210docker\u955c\u50cf\u7684\u65b9\u5f0f\u6709\u4e24\u4e2a":34,"\u7528":[65,66,67],"\u75280\u548c1\u8868\u793a":3,"\u7528\u4e86\u4e24\u4e2a\u6708\u4e4b\u540e\u8fd9\u4e2a\u663e\u793a\u5668\u5c4f\u5e55\u788e\u4e86":64,"\u7528\u4e8e":48,"\u7528\u4e8e\u4e5f\u53ef\u4ee5\u4e0a\u4f20label":26,"\u7528\u4e8e\u5207\u5206\u5355\u5355\u8bcd\u548c\u6807\u70b9\u7b26\u53f7":68,"\u7528\u4e8e\u521d\u59cb\u5316\u53c2\u6570\u548c\u8bbe\u7f6e":44,"\u7528\u4e8e\u5c06\u4e0b\u4e00\u884c\u7684\u6570\u636e\u8f93\u5165\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovider2":3,"\u7528\u4e8e\u5c06\u53c2\u6570\u4f20\u9012\u7ed9\u7f51\u7edc\u914d\u7f6e":52,"\u7528\u4e8e\u5c06\u8bcdid\u8f6c\u6362\u4e3a\u8bcd\u7684\u5b57\u5178\u6587\u4ef6":42,"\u7528\u4e8e\u6307\u5b9a\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":50,"\u7528\u4e8e\u653e\u7f6e":48,"\u7528\u4e8e\u6784\u6210\u65b0\u7684\u8bcd\u8868":60,"\u7528\u4e8e\u6807\u8bc6\u751f\u6210\u7684\u6587\u4ef6\u4e2d\u7684\u76f8\u5e94\u8f93\u51fa":42,"\u7528\u4e8e\u7a00\u758f\u8bad\u7ec3\u4e2d":50,"\u7528\u4e8e\u7edf\u8ba1\u8bcd\u9891\u7684bow\u6a21\u578b\u7279\u5f81":68,"\u7528\u4e8e\u81ea\u5b9a\u4e49\u6bcf\u6761\u6570\u636e\u7684batch":3,"\u7528\u4e8e\u8ba1\u7b97\u7f16\u7801\u5411\u91cf\u7684\u52a0\u6743\u548c":42,"\u7528\u4e8e\u8bbe\u7f6e\u8bad\u7ec3\u7b97\u6cd5":61,"\u7528\u4e8e\u8bfb\u53d6\u8bad\u7ec3":48,"\u7528\u4e8e\u96c6\u7fa4\u901a\u4fe1\u901a\u9053\u7684\u7aef\u53e3\u6570":48,"\u7528\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7f16\u7801":42,"\u7528\u547d\u4ee4":53,"\u7528\u591a\u5bf9\u6548\u679c\u5b8c\u5168\u76f8\u540c\u7684":39,"\u7528\u6237":48,"\u7528\u62371\u7684\u7279\u5f81":66,"\u7528\u6237\u4e5f\u53ef\u4ee5\u5728c":2,"\u7528\u6237\u53ea\u9700\u5b9a\u4e49rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":41,"\u7528\u6237\u53ea\u9700\u6267\u884c":67,"\u7528\u6237\u53ea\u9700\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u5c31\u53ef\u4ee5\u4e0b\u8f7d\u5e76\u5904\u7406\u539f\u59cb\u6570\u636e":67,"\u7528\u6237\u53ef\u4ee5\u4e0a\u4f20\u81ea\u5df1\u7684\u6570\u636e\u96c6":26,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528\u5f00\u53d1\u955c\u50cf\u4ee3\u66ff\u914d\u7f6e\u672c\u5730\u73af\u5883":34,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528ssh\u767b\u5f55\u5230\u8fd9\u53f0\u670d\u52a1\u5668\u4e0a\u5e76\u6267\u884c":34,"\u7528\u6237\u53ef\u4ee5\u5206\u522b\u67e5\u770b\u6700\u65b0\u7684":45,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003":53,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003sphinx\u6559\u7a0b\u8fdb\u884c\u4e66\u5199":45,"\u7528\u6237\u53ef\u4ee5\u5728\u8f93\u51fa\u7684\u6587\u672c\u6a21\u578b\u4e2d\u770b\u5230":60,"\u7528\u6237\u53ef\u4ee5\u5b89\u5168\u7684\u91ca\u653e\u67d0\u4e2ac":28,"\u7528\u6237\u53ef\u4ee5\u6839\u636e\u8bad\u7ec3\u65e5\u5fd7":64,"\u7528\u6237\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u8fd9\u4e2a\u52a8\u6001\u5e93\u6765\u5f15\u5165paddl":28,"\u7528\u6237\u53ef\u4ee5\u7f16\u5199\u7a0b\u5e8f\u5b8c\u6210\u8f6c\u6362":26,"\u7528\u6237\u53ef\u4ee5\u81ea\u5b9a\u4e49beam":50,"\u7528\u6237\u53ef\u4ee5\u8bbe\u7f6e":52,"\u7528\u6237\u53ef\u4ee5\u901a\u8fc7\u7b80\u5355\u4f7f\u7528python\u63a5\u53e3":2,"\u7528\u6237\u53ef\u4ee5\u901a\u8fc7\u7f51\u9875\u6d4f\u89c8\u6587\u6863":34,"\u7528\u6237\u53ef\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e\u5b83\u4eec":33,"\u7528\u6237\u53ef\u5728cmake\u7684\u547d\u4ee4\u884c\u4e2d":33,"\u7528\u6237\u5728\u4f7f\u7528paddlepaddl":31,"\u7528\u6237\u5728\u9996\u6b21\u7f16\u8bd1\u5b89\u88c5paddlepaddle\u65f6":45,"\u7528\u6237\u5b9a\u4e49\u7684\u53c2\u6570":3,"\u7528\u6237\u5c06\u914d\u7f6e\u4e0e\u8bad\u7ec3\u6570\u636e\u5207\u5206\u597d\u653e\u5728\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u9884\u5148\u5206\u914d\u597d\u7684\u76ee\u5f55\u4e2d":56,"\u7528\u6237\u5e94\u8be5\u63d0\u4f9b\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":67,"\u7528\u6237\u5f3a\u5236\u6307\u5b9a\u7279\u5b9a\u7684python\u7248\u672c":31,"\u7528\u6237\u6307\u5b9a\u65b0\u7684\u5b57\u5178\u7684\u8def\u5f84":60,"\u7528\u6237\u6587\u4ef6\u4e2d\u6709\u56db\u79cd\u7c7b\u578b\u7684\u5b57\u6bb5":66,"\u7528\u6237\u7279\u5f81":66,"\u7528\u6237\u8fd8\u53ef\u4ee5\u6839\u636e\u6982\u7387\u5206\u5e03\u77e9\u9635\u5b9e\u73b0\u67f1\u641c\u7d22\u6216\u7ef4\u7279\u6bd4\u89e3\u7801":67,"\u7528\u6237\u901a\u8fc7c":28,"\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":52,"\u7528\u6237\u9700\u8981\u6307\u5b9a\u672c\u673a\u4e0apython\u7684\u8def\u5f84":31,"\u7528\u6237\u9700\u8981\u9996\u5148\u786e\u8ba4py_paddle\u5305\u5df2\u7ecf\u5b89\u88c5":45,"\u7528\u6237\u9884\u6d4b\u7684\u547d\u4ee4\u884c\u754c\u9762\u5982\u4e0b":66,"\u7528\u6237id":65,"\u7528\u6237id\u8303\u56f4\u4ece1\u52306040":65,"\u7528\u6765\u4ece\u53c2\u6570\u670d\u52a1\u5668\u9884\u53d6\u53c2\u6570\u77e9\u9635\u76f8\u5e94\u7684\u884c":44,"\u7528\u6765\u4f30\u8ba1\u7ebf\u6027\u51fd\u6570\u7684\u53c2\u6570w":32,"\u7528\u6765\u5177\u4f53\u63cf\u8ff0":66,"\u7528\u6765\u5177\u4f53\u8bf4\u660e\u6570\u636e\u96c6\u7684\u5b57\u6bb5\u548c\u6587\u4ef6\u683c\u5f0f":66,"\u7528\u6765\u8ba1\u7b97\u6a21\u578b\u7684\u8bef\u5dee":32,"\u7528\u8fd9\u4e2a\u955c\u50cf\u521b\u5efa\u7684\u5bb9\u5668\u9700\u8981\u6709\u4ee5\u4e0b\u4e24\u4e2a\u529f\u80fd":56,"\u7531":41,"\u7531\u4e8e\u4e0d\u540c\u7684paddle\u7684\u7248\u672c\u53ef\u80fd\u9700\u8981\u4e0d\u540c\u7684\u4f9d\u8d56\u548c\u5de5\u5177":34,"\u7531\u4e8e\u5b83\u5185\u90e8\u5305\u542b\u4e86\u6bcf\u7ec4\u6570\u636e\u4e2d\u7684\u6240\u6709\u53e5\u5b50":39,"\u7531\u4e8e\u5bb9\u5668\u4e4b\u95f4\u5171\u4eabnet":54,"\u7531\u4e8e\u5bf9parameters\u7684\u66f4\u65b0\u9700\u8981\u83b7\u53d6parameters\u5185\u5b58\u7684":25,"\u7531\u4e8e\u5df2\u7ecf\u77e5\u9053\u4e86\u771f\u5b9e\u7b54\u6848":32,"\u7531\u4e8e\u610f\u5916\u7684\u526f\u672c\u8bb0\u5f55\u548c\u6d4b\u8bd5\u8bb0\u5f55":65,"\u7531\u4e8e\u6211\u4eec\u60f3\u8981\u7684\u53d8\u6362\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":39,"\u7531\u4e8e\u6211\u4eec\u652f\u6301\u8bad\u7ec3\u6570\u636e\u6709\u4e0d\u540c\u7684\u6279\u6b21\u5927\u5c0f":44,"\u7531\u4e8e\u6570\u636e\u8bb8\u53ef\u7684\u539f\u56e0":67,"\u7531\u4e8e\u6807\u51c6\u7684\u7ffb\u8bd1\u7ed3\u679c\u5df2\u7ecf\u4e0b\u8f7d\u5230\u8fd9\u91cc":69,"\u7531\u4e8e\u6bcf\u4e2a\u5377\u79ef\u5c42\u540e\u9762\u8fde\u63a5\u7684\u662fbatch":62,"\u7531\u4e8e\u8fd9\u4e2a\u5730\u5740\u4f1a\u88abdataprovider\u4f7f\u7528":2,"\u7531\u4e8e\u8fd9\u6837\u505a\u53ef\u4ee5\u907f\u514d\u5f88\u591a\u6b7b\u9501\u95ee\u9898":3,"\u7531\u4e8e\u96c6\u7fa4\u4e2d\u540c\u65f6\u5b58\u5728\u4e24\u53f0\u673a\u5668\u6545\u969c\u7684\u6982\u7387\u6781\u4f4e":25,"\u7531\u4e8e\u987a\u5e8f\u8c03\u7528\u8fd9\u4e9bgenerator\u4e0d\u4f1a\u51fa\u73b0\u4e0a\u8ff0\u95ee\u9898":3,"\u7531\u4e8ec":27,"\u7531\u4e8epaddlepaddle\u5df2\u7ecf\u5b9e\u73b0\u4e86\u4e30\u5bcc\u7684\u7f51\u7edc\u5c42":32,"\u7531\u4e8estep":41,"\u7531\u4e8etest_data\u5305\u542b\u4e24\u6761\u9884\u6d4b\u6570\u636e":5,"\u7531\u8bcd\u8bed\u6784\u6210\u7684\u53e5\u5b50":38,"\u7531grouplen":65,"\u7535\u5f711\u7684\u7279\u5f81":66,"\u7535\u5f71\u4fe1\u606f\u4ee5\u53ca\u7535\u5f71\u8bc4\u5206":65,"\u7535\u5f71\u540d\u5b57\u6bb5":66,"\u7535\u5f71\u540d\u79f0":65,"\u7535\u5f71\u548c\u7528\u6237":66,"\u7535\u5f71\u548c\u7528\u6237\u6709\u8bb8\u591a\u7684\u7279\u5f81":66,"\u7535\u5f71\u5927\u90e8\u5206\u662f\u624b\u5de5\u8f93\u5165\u6570\u636e":65,"\u7535\u5f71\u7279\u5f81":66,"\u7535\u5f71\u7c7b\u578b":65,"\u7535\u5f71\u7c7b\u578b\u5982\u7b26\u5408\u591a\u79cd\u7528\u7ba1\u9053\u7b26\u53f7":65,"\u7535\u5f71id":65,"\u7535\u5f71id\u8303\u56f4\u4ece1\u52303952":65,"\u7535\u8111":39,"\u767e\u4e07\u6570\u636e\u96c6":65,"\u7684":[39,43,48,55,56,64,68],"\u768410\u7ef4\u6574\u6570\u6807\u7b7e":3,"\u7684\u4e00\u4e2a\u7b80\u5355\u8c03\u7528\u5982\u4e0b":41,"\u7684\u4e00\u4e2a\u7ebf\u6027\u51fd\u6570":32,"\u7684\u4e00\u79cd":69,"\u7684\u4e3a0":50,"\u7684\u4e3b\u8981\u90e8\u5206":67,"\u7684\u4efb\u4e00\u4e00\u79cd":31,"\u7684\u4efb\u52a1":69,"\u7684\u4f5c\u7528":53,"\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":38,"\u7684\u504f\u7f6e\u5411\u91cf":44,"\u7684\u5185\u5b58":31,"\u7684\u5185\u5bb9\u5982\u4e0b\u6240\u793a":69,"\u7684\u5185\u5bb9\u6765\u5b9a\u5236imag":56,"\u7684\u5185\u6838block\u4f7f\u7528\u60c5\u51b5":47,"\u7684\u51fd\u6570":53,"\u7684\u5206\u7c7b\u4efb\u52a1\u4e2d\u8d62\u5f97\u4e86\u7b2c\u4e00\u540d":62,"\u7684\u522b\u540d":[6,7,13,14,15],"\u7684\u5355\u8bcd\u7ea7\u522b\u7684\u5377\u79ef\u795e\u7ecf\u7f51\u7edc":66,"\u7684\u53cd\u5411\u4f20\u64ad\u5c06\u4f1a\u6253\u5370\u65e5\u5fd7\u4fe1\u606f":50,"\u7684\u53d8\u6362\u77e9\u9635":44,"\u7684\u53e5\u5b50\u7684\u60c5\u611f":68,"\u7684\u540d\u5b57":3,"\u7684\u540d\u79f0\u76f8\u540c":42,"\u7684\u540e\u7f00":65,"\u7684\u5411\u91cf":44,"\u7684\u542f\u52a8\u53c2\u6570":56,"\u7684\u542f\u52a8\u53c2\u6570\u5e76\u6267\u884c\u8fdb\u7a0b":56,"\u7684\u547d\u540d\u98ce\u683c\u5e76\u4e0d\u80fd\u9002\u5e94\u5176\u4ed6\u7b2c\u4e09\u65b9\u8bed\u8a00":27,"\u7684\u5730\u5740":54,"\u7684\u5730\u65b9":43,"\u7684\u5747\u5300\u5206\u5e03":31,"\u7684\u5934\u6587\u4ef6":27,"\u7684\u5dee\u8ddd\u4e0d\u65ad\u51cf\u5c0f":32,"\u7684\u5e73\u5747\u503c":38,"\u7684\u5e8f\u5217\u5f62\u72b6\u4e00\u81f4":39,"\u7684\u5f00\u53d1\u5de5\u4f5c\u90fd\u5e94\u8be5\u5728\u4e00\u4e2a\u65b0\u7684\u5206\u652f\u4e0a\u5b8c\u6210":43,"\u7684\u603b":48,"\u7684\u63a5\u53e3\u6837\u5f0f":27,"\u7684\u63cf\u8ff0\u8bf4\u660e\u4e2d":43,"\u7684\u6570\u636e":53,"\u7684\u6570\u636e\u6d41\u56fe":26,"\u7684\u6570\u636e\u8bfb\u53d6\u811a\u672c\u548c\u7c7b\u4f3c\u4e8e":64,"\u7684\u6570\u76ee\u4e00\u81f4":38,"\u7684\u6587\u4ef6\u4e5f\u5e26\u5230\u65b0\u5206\u652f\u4e0a":43,"\u7684\u65b9\u6cd5\u5df2\u88ab\u8bc1\u660e\u662f\u4e00\u79cd\u5f3a\u5927\u7684\u6a21\u578b":69,"\u7684\u65b9\u7a0b":44,"\u7684\u65f6\u5019\u5982\u679c\u62a5\u4e00\u4e9b\u4f9d\u8d56\u672a\u627e\u5230\u7684\u9519\u8bef\u662f\u6b63\u5e38\u7684":36,"\u7684\u65f6\u95f4\u6b65\u4fe1\u606f\u6210\u6b63\u6bd4":31,"\u7684\u66f4\u8be6\u7ec6\u51c6\u786e\u7684\u5b9a\u4e49":39,"\u7684\u6700\u5c0f\u503c":50,"\u7684\u6700\u65b0\u4ee3\u7801\u5e76\u66f4\u65b0\u5f53\u524d\u5206\u652f":43,"\u7684\u67b6\u6784\u7684\u793a\u4f8b":42,"\u7684\u6837\u5f0f":43,"\u7684\u6838\u5fc3\u662f\u8bbe\u8ba1step\u51fd\u6570\u7684\u8ba1\u7b97\u903b\u8f91":41,"\u7684\u6bb5\u843d\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":41,"\u7684\u6d4b\u8bd5\u6570\u636e\u96c6":67,"\u7684\u6e90\u7801\u4ee5\u53ca\u751f\u6210\u6587\u6863\u9700\u8981\u591a\u79cd\u5f00\u53d1\u5de5\u5177":43,"\u7684\u6e90\u7801\u91cc\u4f7f\u7528\u4e86":27,"\u7684\u7248\u672c\u53f7":60,"\u7684\u7279\u5f81":62,"\u7684\u72b6\u6001":41,"\u7684\u7528\u6237\u53c2\u8003":48,"\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":66,"\u7684\u76f8\u5173\u6587\u6863\u8fdb\u884c\u914d\u7f6e":53,"\u7684\u771f\u5b9e\u5173\u7cfb\u4e3a":32,"\u7684\u77e9\u9635":44,"\u7684\u795e\u7ecf\u7f51\u7edc\u673a\u5668\u7ffb\u8bd1":69,"\u7684\u7a20\u5bc6\u5411\u91cf\u4f5c\u4e3a\u8f93\u5165":44,"\u7684\u7aef\u5230\u7aef\u7cfb\u7edf\u6765\u89e3\u51b3srl\u4efb\u52a1":67,"\u7684\u7b2ci\u4e2a\u503c":44,"\u7684\u7b2cj\u4e2a\u503c":44,"\u7684\u7d22\u5f15\u6587\u4ef6\u5f15\u7528\u8bad\u7ec3":48,"\u7684\u7ed3\u6784\u5982\u4e0b":66,"\u7684\u7ef4\u5ea6":60,"\u7684\u884c\u6570\u5e94\u8be5\u4e00\u81f4":69,"\u7684\u89c4\u8303":27,"\u7684\u8bad\u7ec3\u6a21\u578b\u811a\u672c":64,"\u7684\u8bdd":31,"\u7684\u8def\u5f84\u4e2d":68,"\u7684\u8f93\u5165":41,"\u7684\u8f93\u51fa":47,"\u7684\u8f93\u51fa\u4fe1\u606f\u5165\u624b\u662f\u4e2a\u4e0d\u9519\u7684\u9009\u62e9":47,"\u7684\u8f93\u51fa\u51fd\u6570\u8fd4\u56de\u7684\u662f\u4e0b\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa\u8bcd\u7684":42,"\u7684\u8f93\u51fa\u683c\u5f0f":39,"\u7684\u8f93\u51fa\u88ab\u7528\u4f5c":42,"\u7684\u8fd4\u56de\u503c\u4e00\u81f4":65,"\u7684\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":43,"\u7684\u90e8\u5206":48,"\u7684\u914d\u7f6e":[53,60],"\u7684\u9875\u9762\u5220\u9664\u8fdc\u7a0b\u4ed3\u5e93\u7684\u5206\u652f":43,"\u7684docker\u955c\u50cf":34,"\u76ee\u524d":[41,67],"\u76ee\u524d\u4f7f\u7528":43,"\u76ee\u524d\u53ea\u8003\u8651\u52a8\u6001\u6269\u5bb9trainer\u6570\u91cf":25,"\u76ee\u524d\u5d4c\u5165python\u89e3\u91ca\u5668":27,"\u76ee\u524d\u5df2\u88ab\u767e\u5ea6\u5185\u90e8\u591a\u4e2a\u4ea7\u54c1\u7ebf\u5e7f\u6cdb\u4f7f\u7528":0,"\u76ee\u524d\u652f\u6301\u4e24\u79cd":38,"\u76ee\u524d\u652f\u6301fail":50,"\u76ee\u524d\u8be5\u53c2\u6570\u4ec5\u7528\u4e8eaucvalidationlayer\u548cpnpairvalidationlayer\u5c42":50,"\u76ee\u524d\u8fd8\u672a\u652f\u6301":41,"\u76ee\u524dpaddle\u7684\u8fdb\u7a0b\u6a21\u578b\u662fc":27,"\u76ee\u524dpaddlepaddle\u7684develop\u5206\u652f\u7684\u6587\u6863\u662f\u81ea\u52a8\u89e6\u53d1\u66f4\u65b0\u7684":45,"\u76ee\u5f55":[48,55,56,68],"\u76ee\u5f55\u4e0b":[28,44,64,69],"\u76ee\u5f55\u4e0b\u627e\u5230":64,"\u76ee\u5f55\u4e0b\u7684demo\u8bad\u7ec3\u51fa\u6765":5,"\u76ee\u5f55\u4e0b\u7684python\u5305":31,"\u76ee\u5f55\u4e2d":[48,66],"\u76ee\u5f55\u4e2d\u7684":[47,48],"\u76ee\u5f55\u4e2dpaddl":56,"\u76ee\u5f55\u4f1a\u51fa\u73b0\u5982\u4e0b\u51e0\u4e2a\u65b0\u7684\u6587\u4ef6":67,"\u76ee\u5f55\u5c31\u6210\u4e3a\u4e86\u5171\u4eab\u5b58\u50a8":56,"\u76ee\u5f55\u7ed3\u6784\u5982\u4e0b":69,"\u76ee\u5f55\u91cc\u63d0\u4f9b\u4e86\u8be5\u6570\u636e\u7684\u4e0b\u8f7d\u811a\u672c\u548c\u9884\u5904\u7406\u811a\u672c":64,"\u76ee\u6807":69,"\u76ee\u6807\u51fd\u6570\u662f\u6807\u7b7e\u7684\u4ea4\u53c9\u71b5":67,"\u76ee\u6807\u5411\u91cf":42,"\u76ee\u6807\u5b57\u5178":69,"\u76f4\u5230\u8bad\u7ec3\u6536\u655b\u4e3a\u6b62":31,"\u76f4\u5230\u903c\u8fd1\u771f\u5b9e\u89e3":32,"\u76f4\u63a5\u4f7f\u7528c\u8bed\u8a00\u7684":27,"\u76f4\u63a5\u5220\u9664\u8fd9\u4e2a\u53c2\u6570\u5373\u53ef":28,"\u76f4\u63a5\u5bfc\u51fa\u5230c\u7684\u63a5\u53e3\u6bd4\u8f83\u56f0\u96be":27,"\u76f4\u63a5\u8fd0\u884c":34,"\u76f4\u63a5\u8fd4\u56de\u8ba1\u7b97\u7ed3\u679c":5,"\u76f4\u63a5\u8fdb\u5165\u8bad\u7ec3\u6a21\u578b\u7ae0\u8282":64,"\u76f8\u5173\u6982\u5ff5\u662f":3,"\u76f8\u5173\u7684\u9e1f\u7c7b\u6570\u636e\u96c6\u53ef\u4ee5\u4ece\u5982\u4e0b\u5730\u5740\u4e0b\u8f7d":61,"\u76f8\u5173\u8bba\u6587":67,"\u76f8\u53cd":69,"\u76f8\u540c\u540d\u5b57\u7684\u53c2\u6570":31,"\u76f8\u5bf9":39,"\u76f8\u5bf9\u4e8epaddlepaddle\u7a0b\u5e8f\u8fd0\u884c\u65f6\u7684\u8def\u5f84":2,"\u76f8\u5bf9mnist\u800c\u8a00":3,"\u76f8\u5e94\u7684\u6570\u636e\u8bfb\u53d6\u811a\u672c\u548c\u8bad\u7ec3\u6a21\u578b\u811a\u672c":64,"\u76f8\u5e94\u7684\u6570\u636e\u8fed\u4ee3\u5668\u5982\u4e0b":67,"\u76f8\u5e94\u7684\u6807\u8bb0\u53e5\u5b50\u662f":67,"\u76f8\u5f53":39,"\u77e9\u9635":49,"\u7814\u7a76\u4eba\u5458\u5206\u6790\u4e86\u51e0\u4e2a\u5173\u4e8e\u6d88\u8d39\u8005\u4fe1\u5fc3\u548c\u653f\u6cbb\u89c2\u70b9\u7684\u8c03\u67e5":68,"\u7814\u7a76\u751f":65,"\u786e\u4fdd\u7f16\u8bd1\u5668\u9009\u9879":43,"\u793a":64,"\u793a\u4f8b":[31,62],"\u793a\u4f8b3\u5bf9\u4e8e\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u6570\u636e\u5b8c\u5168\u76f8\u540c":39,"\u793a\u4f8b3\u7684\u914d\u7f6e\u4f7f\u7528\u4e86\u5355\u5c42rnn\u548c\u53cc\u5c42rnn":39,"\u793a\u4f8b3\u7684\u914d\u7f6e\u5206\u522b\u4e3a":39,"\u793e\u533a\u53c2\u4e0e\u56f0\u96be":27,"\u793e\u533a\u8d21\u732e\u4ee3\u7801\u5b66\u4e60\u6210\u672c\u9ad8":27,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u7684\u53c2\u6570":25,"\u795e\u7ecf\u7f51\u7edc\u5728\u8bad\u7ec3\u7684\u65f6\u5019":31,"\u795e\u7ecf\u7f51\u7edc\u673a\u5668\u7ffb\u8bd1":69,"\u795e\u7ecf\u7f51\u7edc\u7684\u67d0\u4e00\u5c42":53,"\u795e\u7ecf\u7f51\u7edc\u7684\u7f51\u7edc\u7ed3\u6784\u4e2d\u5177\u6709\u6709\u5411\u73af\u7ed3\u6784":39,"\u795e\u7ecf\u7f51\u7edc\u7684\u8bad\u7ec3\u672c\u8eab\u662f\u4e00\u4e2a\u975e\u5e38\u6d88\u8017\u5185\u5b58\u548c\u663e\u5b58\u7684\u5de5\u4f5c":31,"\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":32,"\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e\u4e3b\u8981\u5305\u62ec\u7f51\u7edc\u8fde\u63a5":53,"\u79bb":39,"\u79bb\u7ebf\u6279\u5904\u7406":26,"\u79d1\u5b66\u5bb6":65,"\u79d1\u5e7b\u7247":65,"\u79f0\u4e3a":[42,43,53],"\u79f0\u4e3a\u5f00\u53d1\u955c\u50cf":34,"\u79f0\u4e3a\u751f\u4ea7\u955c\u50cf":34,"\u79f0\u4e3a\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6":53,"\u79f0\u4e4b\u4e3a\u53cc\u5c42\u5e8f\u5217\u7684\u4e00\u4e2a\u5b50\u5e8f\u5217":38,"\u79f0\u4e4b\u4e3a\u96c6\u675f\u5927\u5c0f":50,"\u7a00\u758f\u6570\u636e\u7684\u683c\u5f0f":44,"\u7a00\u758f\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":56,"\u7a00\u758f\u768401\u5411\u91cf":3,"\u7a00\u758f\u7684\u5411\u91cf":3,"\u7a00\u758f\u77e9\u9635\u7684\u4e58\u79ef\u5e94\u7528\u4e8e\u524d\u5411\u4f20\u64ad\u8fc7\u7a0b":52,"\u7a0b\u5e8f\u4ece\u6b64\u76ee\u5f55\u62f7\u8d1d\u6587\u4ef6\u5230\u5bb9\u5668\u5185\u8fdb\u884c\u8bad\u7ec3":56,"\u7a0b\u5e8f\u505c\u6b62":50,"\u7a0b\u5e8f\u5458":65,"\u7a0b\u5e8f\u76f4\u63a5\u9000\u51fa":50,"\u7a0d\u505a\u8be6\u7ec6\u8bf4\u660e":53,"\u7a20\u5bc6\u5411\u91cf":44,"\u7a20\u5bc6\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":56,"\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":3,"\u7a97\u6237":39,"\u7acb\u523b\u9000\u51fa":31,"\u7aef\u53e3":48,"\u7aef\u53e3\u6570":48,"\u7aef\u53e3\u9644\u52a0\u5230\u4e3b\u673a\u540d\u4e0a":48,"\u7aef\u81ea\u5b9a\u4e49\u4e00\u4e2a":2,"\u7aef\u8bfb\u53d6\u6570\u636e":31,"\u7b2c":39,"\u7b2c\u4e00\u4e2a\u53c2\u6570\u662fsettings\u5bf9\u8c61":3,"\u7b2c\u4e00\u4e2a\u65b9\u6cd5\u7684\u597d\u5904\u662f\u7b80\u5355\u5feb\u6377":34,"\u7b2c\u4e00\u4e2a\u6837\u672c\u540c\u65f6encode\u4e24\u6761\u6570\u636e\u6210\u4e24\u4e2a\u5411\u91cf":39,"\u7b2c\u4e00\u4e2apass\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":3,"\u7b2c\u4e00\u4e2atag\u4e3a":30,"\u7b2c\u4e00\u5929":39,"\u7b2c\u4e00\u6b65\u662f\u8fd0\u884c":34,"\u7b2c\u4e00\u884c\u4ece":68,"\u7b2c\u4e00\u884c\u5b58\u7684\u662f\u56fe\u50cf":62,"\u7b2c\u4e00\u884c\u662f":60,"\u7b2c\u4e00\u884c\u7684":69,"\u7b2c\u4e00\u90e8\u5206\u5b9a\u4e49\u4e86\u6570\u636e\u8f93\u5165":32,"\u7b2c\u4e00\u90e8\u5206\u662f\u56fe\u7247\u7684\u6807\u7b7e":3,"\u7b2c\u4e09":68,"\u7b2c\u4e09\u5217\u662f\u751f\u6210\u7684\u82f1\u8bed\u5e8f\u5217":69,"\u7b2c\u4e09\u6b65":62,"\u7b2c\u4e09\u6b65\u5b8c\u6210\u540e":30,"\u7b2c\u4e8c":68,"\u7b2c\u4e8c\u4e2a\u4e3a":30,"\u7b2c\u4e8c\u4e2a\u65b9\u6cd5\u7684\u597d\u5904\u662fdockerfile\u53ef\u4ee5\u628a\u6574\u4e2a\u751f\u6210\u6d41\u7a0b\u63cf\u8ff0\u5f88\u6e05\u695a":34,"\u7b2c\u4e8c\u5217\u662f\u96c6\u675f\u641c\u7d22\u7684\u5f97\u5206":69,"\u7b2c\u4e8c\u6b65":[60,62],"\u7b2c\u4e8c\u6b65\u662f\u8fd0\u884c":34,"\u7b2c\u4e8c\u884c\u5b58\u7684\u662f\u56fe\u50cf":62,"\u7b2c\u4e8c\u90e8\u5206\u4e3b\u8981\u662f\u9009\u62e9\u5b66\u4e60\u7b97\u6cd5":32,"\u7b2c\u4e8c\u90e8\u5206\u662f28":3,"\u7b2ci\u884c\u7b2cj\u5217\u7684\u6570\u503c":44,"\u7b49":28,"\u7b49\u5168\u90e8\u9759\u6001\u5e93\u4e2d\u7684\u76ee\u6807\u6587\u4ef6\u5168\u90e8\u6253\u5305\u540e\u4ea7\u751f\u7684\u6587\u4ef6":28,"\u7b49\u5176\u4ed6":69,"\u7b49\u53c2\u6570":56,"\u7b49\u591a\u79cd\u516c\u6709\u4e91\u73af\u5883":54,"\u7b49\u5f85\u8fd9\u4e2a\u7a0b\u5e8f\u6267\u884c\u6210\u529f\u5e76\u8fd4\u56de0\u5219\u6210\u529f\u9000\u51fa":54,"\u7b49\u6587\u4ef6":28,"\u7b49\u7b49":[64,69],"\u7b49\u90fd\u5c5e\u4e8e\u4e00\u4e2a\u547d\u540d\u7a7a\u95f4":54,"\u7b80\u4ecb":46,"\u7b80\u5355\u6765\u8bf4":47,"\u7b80\u5355\u7684\u5168\u8fde\u63a5\u7f51\u7edc":31,"\u7b80\u5355\u7684\u57fa\u4e8e\u5b57\u6bcd\u7684\u8bcd\u5d4c\u5165":66,"\u7b80\u5355\u7684\u6027\u80fd\u5206\u6790":47,"\u7b80\u5355\u7684\u6574\u4e2a\u8bcd\u5d4c\u5165":66,"\u7b80\u5355\u7684pydataprovider2\u6837\u4f8b\u5c31\u8bf4\u660e\u5b8c\u6bd5\u4e86":3,"\u7b80\u5355\u7684yaml\u6587\u4ef6\u5982\u4e0b":55,"\u7b80\u76f4":39,"\u7b97\u6cd5":[31,32,42,68],"\u7b97\u6cd5\u4e2d\u7684beam\u5927\u5c0f":42,"\u7b97\u6cd5\u914d\u7f6e":68,"\u7ba1\u7406\u4eba\u5458":65,"\u7ba1\u7406\u5458":65,"\u7c7b\u4f3c":[28,38],"\u7c7b\u4f3c\u5730":67,"\u7c7b\u4f5c\u4e3a\u53c2\u6570\u7684\u62bd\u8c61":44,"\u7c7b\u522b\u4e2a\u6570":61,"\u7c7b\u522b\u4e2d\u7684\u53c2\u6570\u53ef\u7528\u4e8e\u6240\u6709\u573a\u5408":49,"\u7c7b\u522bid":64,"\u7c7b\u522bid\u548c\u6587\u672c\u4fe1\u606f\u7528":64,"\u7c7b\u540d\u548cc":27,"\u7c7b\u578b":[27,50,66],"\u7c7b\u578b\u53ef\u4ee5\u662fpaddlepaddle\u652f\u6301\u7684\u4efb\u610f\u8f93\u5165\u6570\u636e\u7c7b\u578b":38,"\u7c7b\u578b\u662fsparse_binary_vector":3,"\u7c7b\u578b\u662fsparse_float_vector":3,"\u7c7b\u578b\u6765\u8bbe\u7f6e":3,"\u7c7b\u578b\u7684":39,"\u7c7b\u7684\u6784\u9020\u51fd\u6570\u548c\u6790\u6784\u51fd\u6570":44,"\u7c7b\u9700\u8981\u5b9e\u73b0\u521d\u59cb\u5316":44,"\u7cfb\u7edf\u7f16\u8bd1wheel\u5305\u7684\u65f6\u5019":31,"\u7cfb\u7edfc":53,"\u7d2f\u52a0\u6c42\u548c":53,"\u7ea2\u697c\u68a6":60,"\u7eaa\u5f55\u7247":65,"\u7eafcpu\u955c\u50cf\u4ee5\u53cagpu\u955c\u50cf\u90fd\u4f1a\u7528\u5230avx\u6307\u4ee4\u96c6":34,"\u7ebf\u6027\u56de\u5f52\u7684\u8f93\u5165\u662f\u4e00\u6279\u70b9":32,"\u7ebf\u6027\u56de\u5f52\u7684\u8f93\u51fa\u662f\u4ece\u8fd9\u6279\u70b9\u4f30\u8ba1\u51fa\u6765\u7684\u53c2\u6570":32,"\u7ebf\u6027\u8ba1\u7b97\u7f51\u7edc\u5c42":32,"\u7ebf\u7a0bid\u53f7":52,"\u7ec4\u6210":53,"\u7ec6\u8282\u63cf\u8ff0":51,"\u7ecf\u5e38\u4f1a\u6d88\u8017\u657010gb\u7684\u5185\u5b58\u548c\u6570gb\u7684\u663e\u5b58":31,"\u7ed3\u5408":54,"\u7ed3\u675f\u6807\u8bb0":42,"\u7ed3\u6784\u5982\u4e0b":68,"\u7ed3\u6784\u5982\u4e0b\u56fe":60,"\u7ed3\u679c\u4fdd\u5b58\u5728":67,"\u7ed3\u679c\u53d1\u73b0\u5b83\u4eec\u4e0e\u540c\u65f6\u671f\u7684twitter\u6d88\u606f\u4e2d\u7684\u60c5\u7eea\u8bcd\u9891\u7387\u76f8\u5173":68,"\u7ed3\u8bba":27,"\u7ed9":39,"\u7ed9\u51fa\u56fe\u7247\u5c3a\u5bf8":61,"\u7ed9\u51fa\u8f93\u5165\u6570\u636e\u6240\u5728\u8def\u5f84":61,"\u7ed9\u5b9a\u52a8\u8bcd":67,"\u7ed9\u5b9a\u7684\u6587\u672c\u53ef\u4ee5\u662f\u4e00\u4e2a\u6587\u6863":68,"\u7ed9\u5b9aencoder\u8f93\u51fa\u548c\u5f53\u524d\u8bcd":41,"\u7ee7\u7eed\u6df1\u5165\u4e86\u89e3":53,"\u7ee7\u7eed\u8bad\u7ec3\u6216\u9884\u6d4b":3,"\u7ef4\u57fa\u767e\u79d1\u4e2d\u6587\u9875\u9762":39,"\u7ef4\u57fa\u767e\u79d1\u9875\u9762":39,"\u7ef4\u5ea6\u4e3aword_dim":64,"\u7ef4\u5ea6\u662f\u7c7b\u522b\u4e2a\u6570":64,"\u7ef4\u5ea6\u662f\u8bcd\u5178\u5927\u5c0f":64,"\u7ef4\u62a4":54,"\u7ef4\u7a7a\u95f4":42,"\u7ef4\u7a7a\u95f4\u5b8c\u6210":42,"\u7f13\u5b58\u6c60\u7684\u51cf\u5c0f":31,"\u7f13\u5b58\u8bad\u7ec3\u6570\u636e\u5230\u5185\u5b58":3,"\u7f16\u5199\u597d\u6570\u636e\u63d0\u4f9b\u811a\u672c\u540e":66,"\u7f16\u5199\u5b8cyaml\u6587\u4ef6\u540e":56,"\u7f16\u5199\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6":56,"\u7f16\u53f7":66,"\u7f16\u53f7\u4ece0\u5f00\u59cb":31,"\u7f16\u53f7\u5b57\u6bb5":66,"\u7f16\u7801\u5411\u91cf":42,"\u7f16\u7801\u5668\u8f93\u51fa":42,"\u7f16\u7801\u6e90\u5e8f\u5217":42,"\u7f16\u89e3\u7801\u6a21\u578b\u5c06\u4e00\u4e2a\u6e90\u8bed\u53e5\u7f16\u7801\u4e3a\u4e00\u4e2a\u5b9a\u957f\u7684\u5411\u91cf":69,"\u7f16\u8bd1":[34,43],"\u7f16\u8bd1\u5668\u6ca1\u6709":27,"\u7f16\u8bd1\u578b\u8bed\u8a00":27,"\u7f16\u8bd1\u5b8c\u6210\u4e4b\u540e":45,"\u7f16\u8bd1\u5b8c\u6210\u540e":45,"\u7f16\u8bd1\u6210\u52a8\u6001\u5e93":50,"\u7f16\u8bd1\u6d41\u7a0b\u4e3b\u8981\u63a8\u8350\u9ad8\u7ea7\u7528\u6237\u67e5\u770b":35,"\u7f16\u8bd1\u751f\u6210":45,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684docker\u53d1\u884c\u955c\u50cf":30,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684ubuntu":30,"\u7f16\u8bd1\u9009\u9879":33,"\u7f16\u8bd1c":28,"\u7f16\u8bd1master\u5206\u652f\u7684docker\u53d1\u884c\u955c\u50cf":30,"\u7f16\u8bd1ubuntu\u7684deb\u5305":30,"\u7f16\u8f91":54,"\u7f29\u653e\u53c2\u6570":62,"\u7f51\u7edc":[67,68],"\u7f51\u7edc\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":54,"\u7f51\u7edc\u540d\u79f0":64,"\u7f51\u7edc\u5c42\u53ef\u4ee5\u6709\u591a\u4e2a\u8f93\u5165":44,"\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u4e3a":44,"\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":44,"\u7f51\u7edc\u5c42\u7684\u7ec6\u8282\u53ef\u4ee5\u901a\u8fc7\u4e0b\u9762\u8fd9\u4e9b\u4ee3\u7801\u7247\u6bb5\u6765\u6307\u5b9a":44,"\u7f51\u7edc\u5c42\u7684\u8f93\u51fa\u662f\u7ecf\u8fc7\u6fc0\u6d3b\u51fd\u6570\u4e4b\u540e\u7684\u503c":50,"\u7f51\u7edc\u5c42\u914d\u7f6e\u5305\u542b\u4ee5\u4e0b\u51e0\u9879":44,"\u7f51\u7edc\u63a5\u53e3\u5361":48,"\u7f51\u7edc\u6a21\u5757":62,"\u7f51\u7edc\u6a21\u578b\u5c06\u8f93\u51fa\u6807\u7b7e\u7684\u6982\u7387\u5206\u5e03":67,"\u7f51\u7edc\u7684\u8bad\u7ec3\u8fc7\u7a0b":68,"\u7f51\u7edc\u7684\u8f93\u51fa\u4e3a\u795e\u7ecf\u7f51\u7edc\u7684\u4f18\u5316\u76ee\u6807":53,"\u7f51\u7edc\u7684\u8f93\u51fa\u4e5f\u53ef\u901a\u8fc7":53,"\u7f51\u7edc\u7ed3\u6784\u5982\u4e0b\u56fe\u6240\u793a":66,"\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e\u4e09\u90e8\u5206":53,"\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e\u8fd9\u4e09\u90e8\u5206\u8be5\u6982\u5ff5":53,"\u7f51\u7edc\u8fde\u63a5":53,"\u7f51\u7edc\u901a\u4fe1":44,"\u7f51\u7edc\u914d\u7f6e":[48,64,68],"\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":[62,67],"\u800c":[32,42,55,66],"\u800c\u4e0d\u4f7f\u7528imdb\u6570\u6910\u96c6\u4e2d\u7684imdb":68,"\u800c\u4e0d\u5fc5\u5728\u610fpaddl":28,"\u800c\u4e0d\u652f\u6301pypy\u89e3\u91ca\u5668":27,"\u800c\u4e0d\u662f\u4f7f\u7528\u540c\u6b65":48,"\u800c\u4e0d\u662f\u65b0\u6570\u636e\u9a71\u52a8\u7cfb\u7edf":53,"\u800c\u4e0d\u662f\u6e90\u7801\u76ee\u5f55\u91cc":31,"\u800c\u4e0d\u662f\u7279\u5f81\u7684\u96c6\u5408":39,"\u800c\u4e0d\u662f\u7ec4\u5408\u4e0a\u4e0b\u6587\u7ea7\u522b\u4fe1\u606f":68,"\u800c\u4e0d\u66b4\u9732\u6982\u5ff5\u7684\u5b9e\u73b0":28,"\u800c\u4e0d\u7528\u5173\u5fc3\u6570\u636e\u5982\u4f55\u4f20\u8f93":3,"\u800c\u4e14":69,"\u800c\u4e4b\u524d\u7684\u53c2\u6570\u5c06\u4f1a\u88ab\u5220\u9664":50,"\u800c\u4ece\u5e94\u7528\u7684\u89d2\u5ea6":47,"\u800c\u4f18\u5316\u6027\u80fd\u7684\u9996\u8981\u4efb\u52a1":47,"\u800c\u5176\u4ed6\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":52,"\u800c\u51fa\u73b0\u9636\u6bb5\u6027\u7684\u8fd0\u884c\u505c\u6ede":25,"\u800c\u53cc\u5c42rnn\u662f\u53ef\u4ee5\u5904\u7406\u8fd9\u79cd\u8f93\u5165\u6570\u636e\u7684\u7f51\u7edc\u7ed3\u6784":39,"\u800c\u53f3\u56fe\u7684\u74f6\u9888\u8fde\u63a5\u6a21\u5757\u7528\u4e8e50\u5c42":62,"\u800c\u5728cpp\u91cc\u9762\u5b9e\u73b0\u8fd9\u4e2ac\u7684\u63a5\u53e3":27,"\u800c\u591a\u8bed\u8a00\u63a5\u53e3\u9700\u8981\u76f4\u63a5\u8bfb\u53d6\u751f\u6210\u7684\u4e8c\u8fdb\u5236":27,"\u800c\u5927\u591a\u6570\u65b9\u6cd5\u53ea\u662f\u5229\u7528n":68,"\u800c\u5bf9\u4e8e\u53cc\u5c42\u5e8f\u5217":39,"\u800c\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5185\u5c42\u7279\u5f81\u6570\u636e\u800c\u8a00":39,"\u800c\u5bf9\u4e8egolang":27,"\u800c\u5bf9\u4e8egolang\u9519\u8bef\u5904\u7406\u5e94\u8be5\u4f7f\u7528\u8fd4\u56de\u503c":27,"\u800c\u5c06\u8fd9\u4e2a\u6bb5\u843d\u7684\u6bcf\u4e00\u53e5\u8bdd\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u5411\u91cf":39,"\u800c\u5f53\u524d\u5df2\u7ecf\u67095":47,"\u800c\u662f\u76f4\u63a5\u4ece\u5185\u5b58\u7684\u7f13\u5b58\u91cc\u8bfb\u53d6\u6570\u636e":31,"\u800c\u662f\u76f4\u63a5\u4fee\u6539paddl":28,"\u800c\u66f4\u6df1\u5165\u7684\u5206\u6790":47,"\u800c\u6709\u4e9b\u53c2\u6570\u9700\u8981\u5728\u96c6\u7fa4\u591a\u673a\u8bad\u7ec3\u4e2d\u4f7f\u7528\u7b49":49,"\u800c\u6ca1\u6709\u77ed\u65f6\u8bb0\u5fc6\u7684\u635f\u5931":68,"\u800c\u6e90\u5e8f\u5217\u7684\u7f16\u7801\u5411\u91cf\u53ef\u4ee5\u88ab\u65e0\u8fb9\u754c\u7684memory\u8bbf\u95ee":42,"\u800c\u7a00\u758f\u66f4\u65b0\u5728\u53cd\u5411\u4f20\u64ad\u4e4b\u540e\u7684\u6743\u91cd\u66f4\u65b0\u65f6\u8fdb\u884c":52,"\u800c\u7cfb\u7edf\u4e2d\u7684":31,"\u800c\u8fd9\u4e00\u53e5\u8bdd\u5c31\u53ef\u4ee5\u8868\u793a\u6210\u8fd9\u4e9b\u4f4d\u7f6e\u7684\u6570\u7ec4":39,"\u800c\u8fd9\u4e2acontext\u53ef\u80fd\u4f1a\u975e\u5e38\u5927":3,"\u800c\u8fd9\u6bcf\u4e00\u4e2a\u6570\u7ec4\u5143\u7d20":39,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":33,"\u800cpaddlepaddle\u5219\u4f1a\u5e2e\u7528\u6237\u505a\u4ee5\u4e0b\u5de5\u4f5c":3,"\u800crnn\u662f\u6700\u6d41\u884c\u7684\u9009\u62e9":41,"\u800cswig\u53ea\u80fd\u7b80\u5355\u7684\u66b4\u9732c":27,"\u800cweight":61,"\u804c\u4e1a":65,"\u804c\u4e1a\u4ece\u4e0b\u9762\u6240\u5217\u4e2d\u9009\u62e9":65,"\u80fd\u591f\u5904\u7406\u53cc\u5c42\u5e8f\u5217":41,"\u80fd\u591f\u5bf9\u53cc\u5411\u5e8f\u5217\u8fdb\u884c\u5904\u7406\u7684\u6709":41,"\u80fd\u591f\u627e\u5230\u8fd9\u91cc\u4f7f\u7528\u7684\u6240\u6709\u6570\u636e":64,"\u80fd\u591f\u8bb0\u5f55\u4e0a\u4e00\u4e2asubseq":41,"\u80fd\u83b7\u53d6":48,"\u811a\u672c":[48,61,66],"\u811a\u672c\u4fdd\u5b58\u5728":61,"\u811a\u672c\u5f00\u59cb\u65f6":56,"\u811a\u672c\u63d0\u4f9b\u4e86\u4e00\u4e2a\u9884\u6d4b\u63a5\u53e3":68,"\u811a\u672c\u65f6\u9700\u8981\u52a0\u4e0a":68,"\u811a\u672c\u8fd0\u884c\u5b8c\u6210\u540e":61,"\u81ea\u52a8\u5173\u95ed\u5bf9\u5e94\u7684":43,"\u81ea\u52a8\u5730\u5c06\u8fd9\u4e9b\u9009\u9879\u5e94\u7528\u5230":48,"\u81ea\u52a8\u5b8c\u6210\u8fd9\u4e00\u8fc7\u7a0b":41,"\u81ea\u52a8\u6302\u8f7d\u5206\u5e03\u5f0f\u5b58\u50a8\u76ee\u5f55":25,"\u81ea\u52a8\u751f\u6210":45,"\u81ea\u52a8\u83b7\u53d6\u4e0a\u4e00\u4e2a\u751f\u6210\u7684\u8bcd":42,"\u81ea\u5e95\u5411\u4e0a\u6cd5":64,"\u81ea\u7136\u8bed\u8a00\u5904\u7406\u7b49":52,"\u81ea\u7531\u804c\u4e1a\u8005":65,"\u81f3\u4e8e\u4e3a\u4ec0\u4e48\u9700\u8981c":28,"\u81f3\u6b64":[3,39,43],"\u8212\u9002":39,"\u826f\u597d\u7684\u6587\u6863":27,"\u827a\u672f\u5bb6":65,"\u8282\u70b9\u4e2d\u7684":48,"\u82e5":44,"\u82e5\u5e72\u4e2a\u53e5\u5b50\u6784\u6210\u4e00\u4e2a\u6bb5\u843d":38,"\u82e5\u6709\u4e0d\u4e00\u81f4\u4e4b\u5904":47,"\u82e5\u6709\u5fc5\u8981":44,"\u82e5\u8f93\u51fa\u662f\u5355\u5c42\u5e8f\u5217":38,"\u82e5\u8f93\u51fa\u662f\u53cc\u5c42\u5e8f\u5217":38,"\u82f1\u6587\u6587\u6863":45,"\u82f1\u6587\u6587\u6863\u76ee\u5f55":45,"\u82f1\u8bed":69,"\u8303\u56f4":52,"\u83b7\u53d6":43,"\u83b7\u53d6\u5229\u7528":64,"\u83b7\u53d6\u5b57\u5178\u7ef4\u5ea6":68,"\u83b7\u53d6\u6700\u65b0\u7684\u68c0\u67e5\u70b9\u7684\u6587\u4ef6uuid":25,"\u83b7\u53d6\u8be5\u6761\u6837\u672c\u7c7b\u522bid":64,"\u83b7\u53d6\u901a\u8fc7":68,"\u83b7\u53d6trainer":56,"\u83b7\u5f97\u53c2\u6570\u5c3a\u5bf8":44,"\u867d\u7136":32,"\u867d\u7136\u4e0d\u9f13\u52b1\u8fd9\u6837":28,"\u867d\u7136\u6bcf\u4e2agenerator\u5728\u6ca1\u6709\u8c03\u7528\u7684\u65f6\u5019":3,"\u867d\u7136\u8fd9\u4e9b\u6587\u4ef6\u5e76\u975e\u90fd\u9700\u8981\u96c6\u7fa4\u8bad\u7ec3":48,"\u867d\u7136paddle\u770b\u8d77\u6765\u5305\u542b\u4e86\u4f17\u591a\u53c2\u6570":49,"\u884c":60,"\u884c\u4f18\u5148\u6b21\u5e8f\u5b58\u50a8":62,"\u884c\u5185\u4f7f\u7528":3,"\u884c\u653f\u5de5\u4f5c":65,"\u884c\u955c\u50cf\u5305\u62ec\u7eafcpu\u7248\u672c\u548cgpu\u7248\u672c\u4ee5\u53ca\u5176\u5bf9\u5e94\u7684\u975eavx\u7248\u672c":34,"\u8868\u660e\u4e86\u8fd9\u4e9b\u884c\u7684\u6807\u53f7":44,"\u8868\u660e\u8fd9\u4e2a\u5c42\u7684\u4e00\u4e2a\u5b9e\u4f8b\u662f\u5426\u9700\u8981\u504f\u7f6e":44,"\u8868\u793a\u4e00\u4e2akubernetes\u96c6\u7fa4\u4e2d\u7684\u4e00\u4e2a\u5de5\u4f5c\u8282\u70b9":54,"\u8868\u793a\u4e3adeviceid":52,"\u8868\u793a\u5728\u96c6\u7fa4\u4f5c\u4e1a":48,"\u8868\u793a\u5973\u6027":65,"\u8868\u793a\u5c06\u5916\u5c42\u7684outer_mem\u4f5c\u4e3a\u5185\u5c42memory\u7684\u521d\u59cb\u72b6\u6001":39,"\u8868\u793a\u5f53\u524d\u96c6\u7fa4\u4f5c\u4e1a\u7684\u8282\u70b9":48,"\u8868\u793a\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":48,"\u8868\u793a\u751f\u6210\u6570\u636e\u7684\u5e8f\u5217id":69,"\u8868\u793a\u7528\u4e8e\u8bad\u7ec3\u6216\u9884\u6d4b":3,"\u8868\u793a\u7537\u6027":65,"\u8868\u793a\u7684\u6bcf\u4e2a\u5355\u8bcd":64,"\u8868\u793a\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u7684\u5206\u7c7b\u9519\u8bef":68,"\u8868\u793a\u8bad\u7ec3\u4e86xx\u4e2a\u6837\u672c":68,"\u8868\u793a\u8bad\u7ec3\u4e86xx\u4e2abatch":68,"\u8868\u793a\u8bfb\u8005\u6240\u4f7f\u7528\u7684docker\u955c\u50cf\u4ed3\u5e93\u5730\u5740":56,"\u8868\u793a\u8fc7\u4e8620\u4e2abatch":64,"\u8868\u793a\u8fc7\u4e862560\u4e2a\u6837\u672c":64,"\u8868\u793a\u8fd9\u4e2ajob\u7684\u540d\u5b57":56,"\u88ab":43,"\u88ab\u6269\u5c55\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"\u88ab\u653e\u5728":44,"\u88ab\u79f0\u4e3a":42,"\u88ab\u79f0\u4e3a\u6570\u636e\u63d0\u4f9b\u5668":53,"\u897f\u90e8\u7247":65,"\u8981\u4e0b\u8f7d\u548c\u89e3\u538b\u6570\u636e\u96c6":66,"\u8981\u4e0b\u8f7d\u89e3\u538b\u8fd9\u4e2a\u6a21\u578b":69,"\u8981\u4f7f\u7528\u547d\u4ee4\u884c\u5206\u6790\u5de5\u5177":47,"\u8981\u5728\u5df2\u6709\u7684kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884cpaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":56,"\u8981\u5728\u6240\u6709\u8282\u70b9\u4e0a\u5b58\u5728":48,"\u8981\u5bf9\u4e00\u4e2a\u56fe\u7247\u7684\u8fdb\u884c\u5206\u7c7b\u9884\u6d4b":61,"\u8981\u5c06\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6\u8f6c\u5316\u4e3ameta\u914d\u7f6e\u6587\u4ef6":66,"\u8981\u6c42\u5355\u5c42\u5e8f\u5217\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee":38,"\u8981\u751f\u6210\u7684\u76ee\u6807\u5e8f\u5217":41,"\u8981\u8c03\u7528":44,"\u89e3\u51b3\u529e\u6cd5\u662f":31,"\u89e3\u51b3\u65b9\u6848\u662f":31,"\u89e3\u538b":69,"\u89e3\u6790\u5668\u80fd\u901a\u8fc7\u6587\u4ef6\u7684\u6269\u5c55\u540d\u81ea\u52a8\u8bc6\u522b\u6587\u4ef6\u7684\u683c\u5f0f":66,"\u89e3\u6790\u6570\u636e\u96c6\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5b57\u6bb5":66,"\u89e3\u6790\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":5,"\u89e3\u6790\u73af\u5883\u53d8\u91cf\u5f97\u5230":56,"\u89e3\u6790\u8bad\u7ec3\u6a21\u578b\u65f6\u7528\u7684\u914d\u7f6e\u6587\u4ef6":5,"\u89e3\u7801\u5668\u4f7f\u7528":42,"\u89e3\u7801\u5668\u57fa\u4e8e\u7f16\u7801\u6e90\u5e8f\u5217\u548c\u6700\u540e\u751f\u6210\u7684\u76ee\u6807\u8bcd\u9884\u6d4b\u4e0b\u4e00\u76ee\u6807\u8bcd":42,"\u89e3\u7801\u5668\u662f\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":42,"\u89e3\u7801\u5668\u6839\u636e\u4e0a\u4e0b\u6587\u5411\u91cf\u9884\u6d4b\u51fa\u4e00\u4e2a\u76ee\u6807\u5355\u8bcd":69,"\u89e3\u91ca":64,"\u89e3\u91ca\u578b\u8bed\u8a00\u53ea\u80fd\u8c03\u7528\u52a8\u6001\u5e93":27,"\u89e3\u91ca\u6027\u8bed\u8a00\u5b9e\u9645\u8fd0\u884c\u7684\u4e8c\u8fdb\u5236\u662f\u89e3\u91ca\u5668\u672c\u8eab":27,"\u8ba1\u7b97":42,"\u8ba1\u7b97\u504f\u7f6e\u7684\u68af\u5ea6":44,"\u8ba1\u7b97\u5355\u5143\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u6838\u5fc3":32,"\u8ba1\u7b97\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":42,"\u8ba1\u7b97\u53d8\u6362\u77e9\u9635\u7684\u5927\u5c0f\u548c\u683c\u5f0f":44,"\u8ba1\u7b97\u5f53\u524d\u5c42\u6743\u91cd\u7684\u68af\u5ea6":44,"\u8ba1\u7b97\u6bcf\u4e2a\u8bcd\u7684\u8bcd\u5411\u91cf":42,"\u8ba1\u7b97\u6fc0\u6d3b\u51fd\u6570\u7684\u68af\u5ea6":44,"\u8ba1\u7b97\u7684\u7ec6\u8282\u5c06\u5728\u4e0b\u9762\u7684\u5c0f\u8282\u7ed9\u51fa":44,"\u8ba1\u7b97\u8bef\u5dee\u51fd\u6570":32,"\u8ba1\u7b97\u8f6c\u6362\u77e9\u9635\u548c\u8f93\u5165\u7684\u68af\u5ea6":44,"\u8ba1\u7b97\u8f93\u5165\u548c\u53c2\u6570\u7684\u68af\u5ea6":44,"\u8ba1\u7b97\u8f93\u5165\u5c42\u7684\u504f\u5dee":44,"\u8ba1\u7b97\u8f93\u51fa":44,"\u8ba1\u7b97\u8fd9\u4e2a\u6587\u4ef6\u7684md5":25,"\u8ba9\u6a21\u578b\u80fd\u591f\u5f97\u5230\u8bad\u7ec3\u66f4\u65b0":64,"\u8ba9\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fdb\u884c\u8bad\u7ec3\u6216\u9884\u6d4b":2,"\u8ba9\u8fd9\u4e2a\u793a\u4f8b\u53d8\u5f97\u66f4\u597d":66,"\u8ba9paddle\u6838\u5fc3\u4e2d":28,"\u8bad\u7ec3":[49,68],"\u8bad\u7ec3\u4e0e\u5e94\u7528":1,"\u8bad\u7ec3\u4efb\u52a1\u7684\u8fd0\u884c\u53ef\u80fd\u4f1a\u5360\u6ee1trainer\u548cparamet":25,"\u8bad\u7ec3\u4f5c\u4e1a":48,"\u8bad\u7ec3\u53ca\u6d4b\u8bd5\u8bef\u5dee\u66f2\u7ebf\u56fe\u4f1a\u88ab":61,"\u8bad\u7ec3\u53ef\u4ee5\u8bbe\u7f6e\u4e3atrue":67,"\u8bad\u7ec3\u540e":67,"\u8bad\u7ec3\u548c\u7eaf\u4f7f\u7528":30,"\u8bad\u7ec3\u5931\u8d25\u65f6\u53ef\u4ee5\u68c0\u67e5\u9519\u8bef\u65e5\u5fd7":48,"\u8bad\u7ec3\u597d\u4e00\u4e2a\u6df1\u5c42\u795e\u7ecf\u7f51\u7edc\u901a\u5e38\u8981\u8017\u8d39\u975e\u5e38\u957f\u7684\u65f6\u95f4":47,"\u8bad\u7ec3\u5b8c\u6210\u540e":32,"\u8bad\u7ec3\u6570\u636e\u548c\u6d4b\u8bd5\u6570\u636e\u7684\u76ee\u5f55":69,"\u8bad\u7ec3\u6570\u636e\u662f":3,"\u8bad\u7ec3\u6570\u636e\u6709\u95ee\u9898":31,"\u8bad\u7ec3\u6570\u636e\u7684\u683c\u5f0f\u5f80\u5f80\u5404\u4e0d\u76f8\u540c":53,"\u8bad\u7ec3\u6570\u6910\u96c6":68,"\u8bad\u7ec3\u65f6":56,"\u8bad\u7ec3\u65f6\u6240\u9700\u8bbe\u7f6e\u7684\u4e3b\u8981\u53c2\u6570\u5982\u4e0b":64,"\u8bad\u7ec3\u65f6\u9ed8\u8ba4shuffl":3,"\u8bad\u7ec3\u6a21\u578b\u4e4b\u524d":68,"\u8bad\u7ec3\u6a21\u578b\u540e":42,"\u8bad\u7ec3\u6a21\u578b\u6b63\u786e\u6027":30,"\u8bad\u7ec3\u7684\u635f\u5931\u51fd\u6570\u9ed8\u8ba4\u6bcf\u969410\u4e2abatch\u6253\u5370\u4e00\u6b21":69,"\u8bad\u7ec3\u7684\u811a\u672c\u662f":67,"\u8bad\u7ec3\u7b97\u6cd5":53,"\u8bad\u7ec3\u7b97\u6cd5\u901a\u5e38\u5b9a\u4e49\u5728\u53e6\u4e00\u5355\u72ecpython\u6587\u4ef6\u4e2d":53,"\u8bad\u7ec3\u7ed3\u675f\u540e\u67e5\u770b\u8f93\u51fa\u7ed3\u679c":56,"\u8bad\u7ec3\u811a\u672c":64,"\u8bad\u7ec3\u811a\u672c\u7b49\u7b49":64,"\u8bad\u7ec3\u81f3\u591a":66,"\u8bad\u7ec3\u8282\u70b9\u6570\u91cf":56,"\u8bad\u7ec3\u8bed\u8a00\u6a21\u578b\u8ddd\u79bb":31,"\u8bad\u7ec3\u8f6e\u6b21":64,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u53c2\u6570\u6216\u8005\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u7684\u68af\u5ea6\u5c3a\u5ea6\u8fc7\u5927":31,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6d4b\u8bd5test_period":49,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4e3a\u672c\u5730\u6a21\u5f0f":50,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4f7f\u7528gpu":50,"\u8bad\u7ec3\u8fdb\u7a0b":53,"\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u8bbe\u5907\u5c5e\u6027\u5c06\u4f1a\u65e0\u6548":50,"\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6\u4e3b\u8981\u5305\u62ec\u6570\u636e\u6e90":53,"\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6\u7684\u6570\u636e\u6e90\u914d\u7f6e\u4e2d\u6307\u5b9adataprovider\u6587\u4ef6\u540d\u5b57":53,"\u8bad\u7ec3\u9636\u6bb5":53,"\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\u6587\u4ef6\u5217\u8868":68,"\u8bad\u7ec3\u96c6\u5df2\u7ecf\u505a\u4e86\u968f\u673a\u6253\u4e71\u6392\u5e8f\u800c\u6d4b\u8bd5\u96c6\u6ca1\u6709":68,"\u8bad\u7ec3\u96c6\u5df2\u7ecf\u968f\u673a\u6253\u4e71":68,"\u8bad\u7ec3\u96c6\u5e73\u5747\u503c":61,"\u8bad\u7ec3dot_period":49,"\u8bb0\u5f55\u4e0b\u6240\u6709\u5931\u8d25\u7684\u4f8b\u5b50":30,"\u8bb0\u5fc6\u6a21\u5757":42,"\u8bba\u6587":62,"\u8bbe\u4e3a\u5df2\u90e8\u7f72\u7684\u5de5\u4f5c\u7a7a\u95f4\u76ee\u5f55":48,"\u8bbe\u4e3a\u672c\u5730":48,"\u8bbe\u7f6e":28,"\u8bbe\u7f6e\u4e3a":44,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528\u672c\u5730\u8bad\u7ec3\u6216\u8005\u4f7f\u7528\u96c6\u7fa4\u4e0a\u7684\u4e00\u4e2a\u8282\u70b9":50,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528gpu\u6a21\u5f0f":50,"\u8bbe\u7f6e\u4efb\u52a1\u7684\u6a21\u5f0f\u4e3a\u6d4b\u8bd5":69,"\u8bbe\u7f6e\u4fdd\u5b58\u6a21\u578b\u7684\u8f93\u51fa\u8def\u5f84":69,"\u8bbe\u7f6e\u5168\u5c40\u5b66\u4e60\u7387":68,"\u8bbe\u7f6e\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":3,"\u8bbe\u7f6e\u5185\u5b58\u4e2d\u6700\u5c0f\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":3,"\u8bbe\u7f6e\u53c2\u6570\u7684\u540d\u5b57":31,"\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570":[31,46,66],"\u8bbe\u7f6e\u5b57\u5178\u6587\u4ef6":68,"\u8bbe\u7f6e\u5de5\u4f5c\u6a21\u5f0f\u4e3a\u8bad\u7ec3":68,"\u8bbe\u7f6e\u5e73\u5747sgd\u7a97\u53e3":68,"\u8bbe\u7f6e\u6210":31,"\u8bbe\u7f6e\u6210\u4e00\u4e2a\u5c0f\u4e00\u4e9b\u7684\u503c":31,"\u8bbe\u7f6e\u6570\u636e\u904d\u5386\u6b21\u6570":67,"\u8bbe\u7f6e\u6807\u7b7e\u7c7b\u522b\u5b57\u5178":68,"\u8bbe\u7f6e\u6a21\u578b\u8def\u5f84":68,"\u8bbe\u7f6e\u7684\u547d\u4ee4\u884c\u53c2\u6570":68,"\u8bbe\u7f6e\u795e\u7ecf\u7f51\u7edc\u7684\u914d\u7f6e\u6587\u4ef6":69,"\u8bbe\u7f6e\u7c7b\u522b\u6570":68,"\u8bbe\u7f6e\u7ebf\u7a0b\u6570":[67,68],"\u8bbe\u7f6e\u7f51\u7edc\u914d\u7f6e":68,"\u8bbe\u7f6e\u8f93\u51fa\u7684\u5c3a\u5bf8":44,"\u8bbe\u7f6e\u8f93\u51fa\u8def\u5f84\u4ee5\u4fdd\u5b58\u8bad\u7ec3\u5b8c\u6210\u7684\u6a21\u578b":68,"\u8bbe\u7f6e\u8fd9\u4e2apydataprovider2\u8fd4\u56de\u4ec0\u4e48\u6837\u7684\u6570\u636e":3,"\u8bbe\u7f6e\u9ed8\u8ba4\u8bbe\u5907\u53f7\u4e3a0":52,"\u8bbe\u7f6ebatch":68,"\u8bbe\u7f6ecpu\u7ebf\u7a0b\u6570\u6216\u8005gpu\u8bbe\u5907\u6570":69,"\u8bbe\u7f6egpu":50,"\u8bbe\u7f6epass":68,"\u8bbe\u7f6epasses\u7684\u6570\u91cf":69,"\u8bbf\u95eekubernetes\u7684\u63a5\u53e3\u6765\u67e5\u8be2\u6b64job\u5bf9\u5e94\u7684\u6240\u6709pod\u4fe1\u606f":56,"\u8bc4\u4ef7\u9884\u6d4b\u7684\u6548\u679c":32,"\u8bc4\u4f30\u5668":53,"\u8bc4\u4f30\u5668\u53ef\u4ee5\u8bc4\u4ef7\u6a21\u578b\u7ed3\u679c":53,"\u8bc4\u4f30\u8be5\u4ea7\u54c1\u7684\u8d28\u91cf":64,"\u8bc4\u5206":[65,66],"\u8bc4\u5206\u6587\u4ef6\u7684\u6bcf\u4e00\u884c\u4ec5\u4ec5\u63d0\u4f9b\u7535\u5f71\u6216\u7528\u6237\u7684\u7f16\u53f7\u6765\u4ee3\u8868\u76f8\u5e94\u7684\u7535\u5f71\u6216\u7528\u6237":66,"\u8bc4\u5206\u88ab\u8c03\u6574\u4e3a5\u661f\u7684\u89c4\u6a21":65,"\u8bc6\u522b\u6570\u5b57":30,"\u8bcd\u5411\u91cf":[30,60],"\u8bcd\u5411\u91cf\u6a21\u578b":63,"\u8bcd\u5411\u91cf\u6a21\u578b\u540d\u79f0":60,"\u8bcd\u672c\u8eab\u548c\u8bcd\u9891":60,"\u8bcd\u9891\u6700\u9ad8\u7684":69,"\u8bd5\u7740\u8ba9\u8f93\u51fa\u7684\u5206\u6790\u6570\u636e\u548c\u7406\u8bba\u503c\u5bf9\u5e94":47,"\u8be5":[48,67],"\u8be5\u51fd\u6570\u5177\u6709\u4e24\u4e2a\u53c2\u6570":3,"\u8be5\u51fd\u6570\u5728\u521d\u59cb\u5316\u7684\u65f6\u5019\u4f1a\u88ab\u8c03\u7528":3,"\u8be5\u51fd\u6570\u7684\u529f\u80fd\u662f":3,"\u8be5\u53c2\u6570\u5728\u7f51\u7edc\u914d\u7f6e\u7684output":50,"\u8be5\u53c2\u6570\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":50,"\u8be5\u53c2\u6570\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":50,"\u8be5\u53c2\u6570\u5fc5\u987b\u80fd\u88abflag":50,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u65e5\u5fd7\u622a\u65ad\u4fe1\u606f":50,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u9519\u8bef\u622a\u65ad\u65e5\u5fd7":50,"\u8be5\u53c2\u6570\u7528\u4e8e\u6307\u5b9a\u52a8\u6001\u5e93\u8def\u5f84":50,"\u8be5\u53c2\u6570\u7684\u610f\u601d\u662f\u8bad\u7ec3num":50,"\u8be5\u53c2\u6570\u9ed8\u8ba4\u4e3anull":50,"\u8be5\u5bf9\u8c61\u5177\u6709\u4ee5\u4e0b\u4e24\u4e2a\u5c5e\u6027":3,"\u8be5\u5c42\u4ec5\u9700\u8981\u8fd9\u4e9b\u975e\u96f6\u6837\u672c\u4f4d\u7f6e\u6240\u5bf9\u5e94\u7684\u53d8\u6362\u77e9\u9635\u7684\u90a3\u4e9b\u884c":44,"\u8be5\u5c42\u795e\u7ecf\u5143\u4e2a\u6570":64,"\u8be5\u622a\u65ad\u4f1a\u5f71\u54cd":50,"\u8be5\u6279\u6b21\u7684\u8f93\u5165\u4e2d\u4ec5\u6709\u4e00\u4e2a\u5b50\u96c6\u662f\u975e\u96f6\u7684":44,"\u8be5\u63a5\u53e3\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bfb\u53d6\u6570\u636e":3,"\u8be5\u63a5\u53e3\u53ef\u7528\u4e8e\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":33,"\u8be5\u6570\u636e\u53ca\u6709\u5f88\u591a\u4e0d\u540c\u7684\u7248\u672c":65,"\u8be5\u6570\u636e\u96c6":60,"\u8be5\u6570\u636e\u96c6\u4e8e2003\u5e742\u6708\u53d1\u5e03":65,"\u8be5\u6570\u636e\u96c6\u5305\u542b\u4e00\u4e9b\u7528\u6237\u4fe1\u606f":65,"\u8be5\u6570\u76ee\u662f\u63d0\u524d\u5b9a\u4e49\u597d\u7684":50,"\u8be5\u6587\u4ef6\u53ef\u4ee5\u4ece\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6\u751f\u6210":66,"\u8be5\u6587\u4ef6\u662f\u7531cpickle\u4ea7\u751f\u7684":62,"\u8be5\u6587\u4ef6\u662fpython\u7684pickle\u5bf9\u8c61":66,"\u8be5\u6587\u4ef6\u8d1f\u8d23\u4ea7\u751f\u56fe\u7247\u6570\u636e\u5e76\u4f20\u9012\u7ed9paddle\u7cfb\u7edf":61,"\u8be5\u6a21\u578b\u4f9d\u7136\u4f7f\u7528\u903b\u8f91\u56de\u5f52\u5206\u7c7b\u7f51\u7edc\u7684\u6846\u67b6":64,"\u8be5\u6a21\u578b\u5728\u957f\u8bed\u53e5\u7ffb\u8bd1\u7684\u573a\u666f\u4e0b\u6548\u679c\u63d0\u5347\u66f4\u52a0\u660e\u663e":69,"\u8be5\u6a21\u578b\u7684\u7f51\u7edc\u914d\u7f6e\u5982\u4e0b":32,"\u8be5\u6a21\u578b\u7684\u8bf4\u660e\u5982\u4e0b\u56fe\u6240\u793a":42,"\u8be5\u6a21\u578b\u7f51\u7edc\u53ea\u662f\u7528\u4e8e\u8fdb\u884cdemo\u5c55\u793apaddle\u5982\u4f55\u5de5\u4f5c":66,"\u8be5\u793a\u4f8b\u5c06\u5c55\u793apaddle\u5982\u4f55\u8fdb\u884c\u8bcd\u5411\u91cf\u5d4c\u5165":66,"\u8be5\u793a\u4f8b\u7684\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":66,"\u8be5\u7b97\u6cd5\u6bcf\u6279\u91cf":32,"\u8be5\u7c7b\u7684\u5b9e\u73b0\u7ec6\u8282\u5728":44,"\u8be5\u811a\u672c\u4ec5\u4ec5\u662f\u5f00\u59cb\u4e00\u4e2apaddle\u8bad\u7ec3\u8fc7\u7a0b":66,"\u8be5\u811a\u672c\u4f1a\u751f\u6210\u4e00\u4e2adot\u6587\u4ef6":62,"\u8be5\u811a\u672c\u5c06\u8f93\u51fa\u9884\u6d4b\u5206\u7c7b\u7684\u6807\u7b7e":61,"\u8be5\u8bed\u53e5\u4f1a\u4e3a\u6bcf\u4e2a\u5c42\u521d\u59cb\u5316\u5176\u6240\u9700\u8981\u7684\u53d8\u91cf\u548c\u8fde\u63a5":44,"\u8be5layer\u5c06\u591a\u4e2a\u8f93\u5165":53,"\u8be5python\u4ee3\u7801\u53ef\u4ee5\u751f\u6210protobuf\u5305":53,"\u8be6\u7ec6\u4ecb\u7ecd\u53ef\u4ee5\u53c2\u8003":39,"\u8be6\u7ec6\u4fe1\u606f\u8bf7\u68c0\u67e5":48,"\u8be6\u7ec6\u5185\u5bb9\u8bf7\u53c2\u89c1":64,"\u8be6\u7ec6\u53ef\u4ee5\u53c2\u8003":53,"\u8be6\u7ec6\u5730\u5c55\u793a\u4e86\u6574\u4e2a\u7279\u5f81\u63d0\u53d6\u7684\u8fc7\u7a0b":62,"\u8be6\u7ec6\u6587\u6863\u53c2\u8003":31,"\u8be6\u7ec6\u7684\u53c2\u6570\u89e3\u91ca":64,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":33,"\u8be6\u7ec6\u89c1":38,"\u8bed\u4e49\u89d2\u8272\u6807\u6ce8":[63,67],"\u8bed\u610f\u89d2\u8272\u6807\u6ce8":30,"\u8bed\u8a00\u6a21\u578b":60,"\u8bf4\u660e":[25,33,56],"\u8bf4\u660e\u6bcf\u4e2a\u7279\u5f81\u6587\u4ef6\u5177\u4f53\u5b57\u6bb5\u662f":66,"\u8bf4\u660e\u8fd9\u4e2a\u5c42\u7684\u8f93\u5165":44,"\u8bf7\u4e0d\u8981\u6df7\u6dc6":53,"\u8bf7\u518d\u6b21\u786e\u8ba4py_paddle\u5305\u5df2\u7ecf\u5b89\u88c5":45,"\u8bf7\u53c2\u7167\u7f51\u7edc\u914d\u7f6e\u7684\u6587\u6863\u4e86\u89e3\u66f4\u8be6\u7ec6\u7684\u4fe1\u606f":52,"\u8bf7\u53c2\u8003":[3,28,31,39,42,44,45,53,64],"\u8bf7\u53c2\u8003\u5982\u4e0b\u8868\u683c":64,"\u8bf7\u53c2\u8003\u9875\u9762":66,"\u8bf7\u53c2\u8003layer\u6587\u6863":61,"\u8bf7\u53c2\u89c1":43,"\u8bf7\u53c2\u9605":42,"\u8bf7\u53c2\u9605\u60c5\u611f\u5206\u6790\u7684\u6f14\u793a\u4ee5\u4e86\u89e3\u6709\u5173\u957f\u671f\u77ed\u671f\u8bb0\u5fc6\u5355\u5143\u7684\u66f4\u591a\u4fe1\u606f":67,"\u8bf7\u5b89\u88c5cuda":36,"\u8bf7\u5c06with_doc\u9009\u9879\u5173\u95ed":45,"\u8bf7\u628a\u6587\u6863\u91cc\u547d\u4ee4\u4e2d\u7684paddlepaddl":34,"\u8bf7\u6307\u5b9a\u8be5\u76ee\u5f55":50,"\u8bf7\u67e5\u770b":60,"\u8bf7\u6ce8\u610f":[34,42,55,60],"\u8bf7\u770b\u4e0b\u9762\u7684\u4f8b\u5b50":52,"\u8bf7\u786e\u4fdd":43,"\u8bf7\u8bb0\u4f4f":48,"\u8bf7\u8fd0\u884c":34,"\u8bf7\u9009\u62e9\u6b63\u786e\u7684\u7248\u672c":31,"\u8bf8\u5982\u56fe\u50cf\u5206\u7c7b":52,"\u8bfb\u53d612\u4e2a\u91c7\u6837\u6570\u636e\u8fdb\u884c\u968f\u673a\u68af\u5ea6\u8ba1\u7b97\u6765\u66f4\u65b0\u66f4\u65b0":32,"\u8bfb\u53d6\u6570\u636e":3,"\u8bfb\u53d6\u6bcf\u4e00\u884c":3,"\u8bfb\u53d6volume\u4e2d\u7684\u6570\u636e\u8fdb\u884c\u8fd9\u6b21\u5206\u5e03\u5f0f\u8bad\u7ec3":56,"\u8bfb\u8005\u53ef\u4ee5\u67e5\u770b":56,"\u8bfb\u8005\u9700\u8981\u66ff\u6362\u6210\u81ea\u5df1\u4f7f\u7528\u7684\u4ed3\u5e93\u5730\u5740":56,"\u8c03\u7528":[44,61],"\u8c03\u7528\u4e00\u6b21":3,"\u8c03\u7528\u4e0a\u9762\u7684process\u51fd\u6570\u83b7\u5f97\u89c2\u6d4b\u6570\u636e":32,"\u8c03\u7528\u7684pydataprovider2\u662f":3,"\u8c03\u7528\u7b2c\u4e8c\u6b21\u7684\u65f6\u5019":3,"\u8c03\u7528\u8be5\u51fd\u6570\u540e":44,"\u8c03\u7528\u8fd9\u4e2apydataprovider2\u7684\u65b9\u6cd5":3,"\u8c13\u8bcd\u4e0a\u4e0b\u6587":67,"\u8d1f\u6837\u672c":64,"\u8d1f\u9762\u7684\u8bc4\u8bba\u7684\u5f97\u5206\u5c0f\u4e8e\u7b49\u4e8e4":68,"\u8d1f\u9762\u8bc4\u4ef7\u6837\u672c":68,"\u8d21\u732e\u6587\u6863":45,"\u8d44\u6e90\u5bf9\u8c61\u7684\u540d\u5b57\u662f\u552f\u4e00\u7684":54,"\u8d77":39,"\u8def\u5f84\u4e0b":[32,62],"\u8df3\u8f6c\u5230":43,"\u8f6c\u4e3ajpeg\u6587\u4ef6\u5e76\u5b58\u5165\u7279\u5b9a\u7684\u76ee\u5f55":61,"\u8f6c\u6362\u751f\u6210\u7684\u6587\u4ef6\u540d\u4f1a\u662f\u4ee5\u4e0b\u683c\u5f0f":26,"\u8f6c\u6362\u8fc7\u6765\u7684":62,"\u8f6e":66,"\u8f83":39,"\u8f93\u5165":[38,42],"\u8f93\u5165\u5168\u662f\u5176\u4ed6layer\u7684\u8f93\u51fa":53,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u5355\u5c42\u5e8f\u5217":41,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u53cc\u5c42\u5e8f\u5217":41,"\u8f93\u5165\u56fe\u7247\u7684\u9ad8\u5ea6\u53ca\u5bbd\u5ea6":61,"\u8f93\u5165\u5c42\u5c3a\u5bf8":62,"\u8f93\u5165\u6570\u636e\u4e3a\u4e00\u4e2a\u5b8c\u6574\u7684\u65f6\u95f4\u5e8f\u5217":39,"\u8f93\u5165\u6570\u636e\u4e3a\u5728\u5355\u5c42rnn\u6570\u636e\u91cc\u9762":39,"\u8f93\u5165\u6570\u636e\u6574\u4f53\u4e0a\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":39,"\u8f93\u5165\u6570\u636e\u7684\u5b57\u5178\u7ef4\u6570\u662f1\u767e\u4e07":52,"\u8f93\u5165\u6570\u6910\u6240\u5728\u76ee\u5f55":68,"\u8f93\u5165\u6587\u672c":60,"\u8f93\u5165\u6587\u672c\u4e2d\u6ca1\u6709\u5934\u90e8":60,"\u8f93\u5165\u662f\u5426\u662f\u8f6c\u7f6e\u7684":44,"\u8f93\u5165\u662f\u7531\u4e00\u4e2alist\u4e2d\u7684\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u7ec4\u6210\u7684":44,"\u8f93\u5165\u7279\u5f81\u56fe\u7684\u901a\u9053\u6570\u76ee":62,"\u8f93\u5165\u7684":60,"\u8f93\u5165\u7684\u539f\u59cb\u6570\u636e\u96c6\u8def\u5f84":69,"\u8f93\u5165\u7684\u540d\u5b57":44,"\u8f93\u5165\u7684\u5927\u5c0f":44,"\u8f93\u5165\u7684\u6587\u672c\u683c\u5f0f\u5982\u4e0b":60,"\u8f93\u5165\u7684\u6587\u672c\u8bcd\u5411\u91cf\u6a21\u578b\u540d\u79f0":60,"\u8f93\u5165\u7684\u7c7b\u578b":44,"\u8f93\u5165\u95e8":68,"\u8f93\u5165\u9884\u6d4b\u6837\u672c":68,"\u8f93\u5165n\u4e2a\u5355\u8bcd":64,"\u8f93\u51fa":[38,42],"\u8f93\u51fa\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":41,"\u8f93\u51fa\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":41,"\u8f93\u51fa\u4e3an\u4e2aword_dim\u7ef4\u5ea6\u5411\u91cf":64,"\u8f93\u51fa\u51fd\u6570":42,"\u8f93\u51fa\u5e8f\u5217\u7684\u7c7b\u578b":38,"\u8f93\u51fa\u5e8f\u5217\u7684\u8bcd\u8bed\u6570\u548c\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":41,"\u8f93\u51fa\u5e94\u8be5\u7c7b\u4f3c\u5982\u4e0b":66,"\u8f93\u51fa\u6587\u4ef6\u7684\u683c\u5f0f\u8bf4\u660e":60,"\u8f93\u51fa\u65e5\u5fd7\u4fdd\u5b58\u5728\u8def\u5f84":68,"\u8f93\u51fa\u65e5\u5fd7\u8bf4\u660e\u5982\u4e0b":68,"\u8f93\u51fa\u67092\u5217":60,"\u8f93\u51fa\u7279\u5f81\u56fe\u7684\u901a\u9053\u6570\u76ee":62,"\u8f93\u51fa\u7684\u4e8c\u8fdb\u5236\u8bcd\u5411\u91cf\u6a21\u578b\u540d\u79f0":60,"\u8f93\u51fa\u7684\u6587\u672c\u6a21\u578b\u540d\u79f0":60,"\u8f93\u51fa\u7684\u68af\u5ea6":50,"\u8f93\u51fa\u76ee\u5f55":62,"\u8f93\u51fa\u7ed3\u679c\u53ef\u80fd\u4f1a\u968f\u7740\u5bb9\u5668\u7684\u6d88\u8017\u800c\u88ab\u5220\u9664":55,"\u8fbe\u5230\u5bb9\u707e\u7684\u76ee\u7684":25,"\u8fc7\u4e86\u4e00\u4e2a\u5f88\u7b80\u5355\u7684recurrent_group":39,"\u8fc7\u5b8c\u6240\u6709\u8bad\u7ec3\u6570\u636e\u5373\u4e3a\u4e00\u4e2apass":31,"\u8fd0":34,"\u8fd0\u884c":36,"\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u5373\u53ef":66,"\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4\u6765\u4e0b\u8f7d\u8fd9\u4e2a\u811a\u672c":69,"\u8fd0\u884c\u4ee5\u4e0b\u6307\u4ee4":34,"\u8fd0\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4\u4e0b\u8f7d\u548c\u83b7\u53d6\u6211\u4eec\u7684\u5b57\u5178\u548c\u9884\u8bad\u7ec3\u6a21\u578b":60,"\u8fd0\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4\u4e0b\u8f7d\u6570\u636e\u96c6":60,"\u8fd0\u884c\u4ee5\u4e0b\u8bad\u7ec3\u547d\u4ee4":32,"\u8fd0\u884c\u4ee5\u53ca\u53d1\u5e03\u60a8\u7684ai\u7a0b\u5e8f":34,"\u8fd0\u884c\u5206\u5e03\u5f0f\u4f5c\u4e1a":48,"\u8fd0\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":[31,46,64],"\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5":34,"\u8fd0\u884c\u5931\u8d25":52,"\u8fd0\u884c\u5b8c\u4ee5\u4e0a\u547d\u4ee4":60,"\u8fd0\u884c\u5b8c\u6210\u540e":48,"\u8fd0\u884c\u5b8c\u811a\u672c":68,"\u8fd0\u884c\u6210\u529f\u4ee5\u540e":60,"\u8fd0\u884c\u6210\u529f\u540e\u76ee\u5f55":68,"\u8fd0\u884c\u65e5\u5fd7":48,"\u8fd0\u884c\u7684\u4e00\u4e9b\u53c2\u6570\u901a\u8fc7\u8fd9\u79cd\u65b9\u5f0f\u4f20\u9012\u5230\u5bb9\u5668\u5185":56,"\u8fd1":39,"\u8fd1\u671f\u63d0\u51fa\u7684nmt\u6a21\u578b\u901a\u5e38\u90fd\u5c5e\u4e8e\u7f16\u89e3\u7801\u6a21\u578b":69,"\u8fd4\u56de":[8,9,10,11,16,17,20,22],"\u8fd4\u56de0":3,"\u8fd4\u56de8\u4e2a\u7279\u5f81list\u548c1\u4e2a\u6807\u7b7elist":67,"\u8fd4\u56de\u4e00\u6761\u5b8c\u6574\u7684\u6837\u672c":3,"\u8fd4\u56de\u6570\u636e\u7684\u6bcf\u4e00\u6761\u6837\u672c\u7ed9":66,"\u8fd4\u56de\u65f6":3,"\u8fd4\u56de\u7684\u662f":3,"\u8fd4\u56de\u7684\u987a\u5e8f\u9700\u8981\u548cinput_types\u4e2d\u5b9a\u4e49\u7684\u987a\u5e8f\u4e00\u81f4":3,"\u8fd4\u56de\u7b2c\u4e8c\u6b65":30,"\u8fd4\u56de\u7b2ci\u4e2a\u8f93\u5165\u77e9\u9635":44,"\u8fd4\u56de\u7c7b\u578b":[8,9,10,11,16,17,20,22],"\u8fd8\u4f1a":39,"\u8fd8\u4f1a\u8f93\u51fa\u4e00\u4e2a":43,"\u8fd8\u662f":39,"\u8fd8\u6709":39,"\u8fd8\u80fd\u5904\u7406\u5176\u4ed6\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e":68,"\u8fd8\u91c7\u7528\u4e86\u4e24\u4e2a\u5176\u4ed6\u7279\u5f81":67,"\u8fd8\u9700\u8981\u8bbe\u7f6e\u4e0b\u9762\u4e24\u4e2a\u53c2\u6570":53,"\u8fd8\u9700\u8981\u8fdb\u884c\u9884\u5904\u7406":61,"\u8fd9":[31,39,64],"\u8fd9\u4e00\u5757\u7684\u8017\u65f6\u6bd4\u4f8b\u771f\u7684\u592a\u9ad8":47,"\u8fd9\u4e00\u5c42\u8fdb\u884c\u5c01\u88c5":28,"\u8fd9\u4e00\u6982\u5ff5\u4e0d\u518d\u7410\u788e":28,"\u8fd9\u4e00\u8fc7\u7a0b\u5bf9\u7528\u6237\u662f\u5b8c\u5168\u900f\u660e\u7684":41,"\u8fd9\u4e09\u4e2a\u5206\u652f":30,"\u8fd9\u4e09\u4e2a\u6b65\u9aa4\u53ef\u914d\u7f6e\u4e3a":64,"\u8fd9\u4e0e\u672c\u5730\u8bad\u7ec3\u76f8\u540c":48,"\u8fd9\u4e24\u4e2a\u6587\u4ef6\u5939\u4e0b\u5404\u81ea\u670910\u4e2a\u5b50\u6587\u4ef6\u5939":61,"\u8fd9\u4e24\u4e2a\u6807\u51c6":67,"\u8fd9\u4e24\u4e2a\u9700\u8981\u4e0e":53,"\u8fd9\u4e2a":[39,54],"\u8fd9\u4e2a\u4efb\u52a1\u7684\u914d\u7f6e\u4e3a":31,"\u8fd9\u4e2a\u4efb\u52a1\u7684dataprovider\u4e3a":31,"\u8fd9\u4e2a\u51fd\u6570\u7684":42,"\u8fd9\u4e2a\u51fd\u6570\u8fdb\u884c\u53d8\u6362":39,"\u8fd9\u4e2a\u51fd\u6570\u9700\u8981\u8bbe\u7f6e":42,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u7684\u8fde\u63a5\u53c2\u6570\u4e0epaddle\u7684\u5176\u4ed6\u4e8c\u8fdb\u5236":28,"\u8fd9\u4e2a\u5305\u91cc\u9762\u5305\u542b\u4e86\u6a21\u578b\u914d\u7f6e\u9700\u8981\u7684\u5404\u4e2a\u6a21\u5757":53,"\u8fd9\u4e2a\u53c2\u6570\u4e5f\u4e0d\u4f1a\u4e00\u5e76\u5220\u9664":28,"\u8fd9\u4e2a\u5411\u91cf\u4e0e\u6e90\u4e2d\u641c\u7d22\u51fa\u7684\u4f4d\u7f6e\u548c\u6240\u6709\u4e4b\u524d\u751f\u6210\u7684\u76ee\u6807\u5355\u8bcd\u6709\u5173":69,"\u8fd9\u4e2a\u5730\u5740\u5219\u4e3a\u5b83\u7684\u7edd\u5bf9\u8def\u5f84\u6216\u76f8\u5bf9\u8def\u5f84":2,"\u8fd9\u4e2a\u5730\u5740\u6765\u8868\u793a\u6b64\u6b65\u9aa4\u6240\u6784\u5efa\u51fa\u7684\u955c\u50cf":56,"\u8fd9\u4e2a\u57fa\u7c7b":44,"\u8fd9\u4e2a\u5934\u6587\u4ef6\u4e0d\u5047\u8bbe\u5176\u4ed6\u6587\u4ef6\u7684\u5f15\u7528\u987a\u5e8f":28,"\u8fd9\u4e2a\u5b57\u5178\u662f\u6574\u6570\u6807\u7b7e\u548c\u5b57\u7b26\u4e32\u6807\u7b7e\u7684\u4e00\u4e2a\u5bf9\u5e94":68,"\u8fd9\u4e2a\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":41,"\u8fd9\u4e2a\u63a5\u53e3\u9700\u8981\u505a\u5230":27,"\u8fd9\u4e2a\u6570\u636e\u4e5f\u88ab\u5355\u5c42rnn\u7f51\u7edc\u76f4\u63a5\u4f7f\u7528":39,"\u8fd9\u4e2a\u6570\u636e\u5217\u8868\u6587\u4ef6\u4e2d\u5305\u542b\u7684\u662f\u6bcf\u4e00\u4e2a\u8bad\u7ec3\u6216\u8005\u6d4b\u8bd5\u6587\u4ef6\u7684\u8def\u5f84":53,"\u8fd9\u4e2a\u6570\u91cf\u79f0\u4e3abeam":69,"\u8fd9\u4e2a\u6587\u4ef6\u5177\u6709\u72ec\u7279\u7684\u8bed\u6cd5":27,"\u8fd9\u4e2a\u663e\u793a\u5668\u5f88\u68d2":64,"\u8fd9\u4e2a\u6a21\u578b\u5bf9\u4e8e\u7f16\u89e3\u7801\u6a21\u578b\u6765\u8bf4":69,"\u8fd9\u4e2a\u76ee\u5f55\u4e2d\u9664\u4e86":28,"\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u5355\u5143\u5c31\u53ebmemori":39,"\u8fd9\u4e2a\u7a0b\u5e8f\u662f\u60a8\u5728\u5f00\u53d1\u673a\u4e0a\u4f7f\u7528\u5f00\u53d1\u955c\u50cf\u5b8c\u6210\u5f00\u53d1":34,"\u8fd9\u4e2a\u7c7b\u7684\u53c2\u6570\u5305\u62ec":44,"\u8fd9\u4e2a\u7c7b\u9700\u8981\u7ee7\u627f":44,"\u8fd9\u4e2a\u7cfb\u7edf\u5c06srl\u4efb\u52a1\u89c6\u4e3a\u5e8f\u5217\u6807\u6ce8\u95ee\u9898":67,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u4e2d\u7684\u53e6\u4e00\u4e2a\u9879\u76ee\u662f":28,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u5305\u542b\u4e24\u4e2a\u9879\u76ee":28,"\u8fd9\u4e2a\u8282\u70b9\u53ef\u4ee5\u662f\u7269\u7406\u673a\u6216\u8005\u865a\u62df\u673a":54,"\u8fd9\u4e2a\u8868\u683c":54,"\u8fd9\u4e2a\u8fc7\u7a0b\u5bf9\u7528\u6237\u4e5f\u662f\u900f\u660e\u7684":41,"\u8fd9\u4e2a\u8fc7\u7a0b\u5c31\u662f\u6a21\u578b\u7684\u8bad\u7ec3\u8fc7\u7a0b":32,"\u8fd9\u4e2a\u8fc7\u7a0b\u9664\u4e86\u7f16\u8bd1paddlepaddle\u4e3a":43,"\u8fd9\u4e2a\u914d\u7f6e\u4e0e":60,"\u8fd9\u4e2a\u914d\u7f6e\u6587\u4ef6":54,"\u8fd9\u4e2a\u914d\u7f6e\u6587\u4ef6\u7f51\u7edc\u7531":53,"\u8fd9\u4e2a\u914d\u7f6e\u662f\u5426\u7528\u6765\u751f\u6210":69,"\u8fd9\u4e2a\u955c\u50cf\u5305\u542b\u4e86paddle\u76f8\u5173\u7684\u5f00\u53d1\u5de5\u5177\u4ee5\u53ca\u7f16\u8bd1\u548c\u8fd0\u884c\u73af\u5883":34,"\u8fd9\u4e2a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u751f\u6210\u4e00\u7cfb\u5217\u6743\u91cd":42,"\u8fd9\u4e2a\u95ee\u9898\u662fpydataprovider\u8bfb\u6570\u636e\u65f6\u5019\u7684\u903b\u8f91\u95ee\u9898":3,"\u8fd9\u4e2a\u9759\u6001\u5e93\u5305\u542b\u4e86paddle\u7684\u5168\u90e8\u7b26\u53f7":28,"\u8fd9\u4e2adataprovider\u8f83\u590d\u6742":3,"\u8fd9\u4e2ainstance\u53ef\u4ee5\u662f\u5355\u4e2a\u503c":26,"\u8fd9\u4e2ajob\u624d\u7b97\u6210\u529f\u7ed3\u675f":56,"\u8fd9\u4e2alayer\u7684\u8f93\u51fa\u4f1a\u4f5c\u4e3a\u6574\u4e2a":41,"\u8fd9\u4e5f\u4f1a\u6781\u5927\u51cf\u5c11\u6570\u636e\u8bfb\u5165\u7684\u8017\u65f6":31,"\u8fd9\u4e9b":48,"\u8fd9\u4e9b\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1\u901a\u5e38\u4f1a\u628a\u6570\u636e\u5207\u5272\u6210\u591a\u4e2a\u5206\u7247\u5206\u5e03\u5f0f\u7684\u5b58\u50a8\u5728\u591a\u4e2a\u8282\u70b9\u4e4b\u4e0a":26,"\u8fd9\u4e9b\u53c2\u6570\u7684\u5177\u4f53\u63cf\u8ff0":56,"\u8fd9\u4e9b\u53c2\u6570\u7684\u7b80\u77ed\u4ecb\u7ecd\u5982\u4e0b":66,"\u8fd9\u4e9b\u540d\u5b57\u5fc5\u987b\u8981\u5199\u5bf9":44,"\u8fd9\u4e9b\u6570\u636e\u4f1a\u88ab\u7528\u6765\u66f4\u65b0\u53c2\u6570":31,"\u8fd9\u4e9b\u6570\u636e\u4f7f\u7528\u7684\u5185\u5b58\u4e3b\u8981\u548c\u4e24\u4e2a\u53c2\u6570\u6709\u5173\u7cfb":31,"\u8fd9\u4e9b\u6587\u4ef6\u5c06\u4f1a\u88ab\u4fdd\u5b58\u5728":62,"\u8fd9\u4e9b\u6a21\u578b\u90fd\u662f\u7531\u539f\u4f5c\u8005\u63d0\u4f9b\u7684\u6a21\u578b":62,"\u8fd9\u4e9b\u7279\u5f81\u503c\u4e0e\u4e0a\u8ff0\u4f7f\u7528c":62,"\u8fd9\u4e9b\u7279\u5f81\u548c\u6807\u7b7e\u5b58\u50a8\u5728":67,"\u8fd9\u4e9b\u7279\u5f81\u6570\u636e\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u6709\u610f\u4e49\u7684":39,"\u8fd9\u4efd\u6559\u7a0b\u5c55\u793a\u4e86\u5982\u4f55\u5728paddlepaddle\u4e2d\u5b9e\u73b0\u4e00\u4e2a\u81ea\u5b9a\u4e49\u7684\u7f51\u7edc\u5c42":44,"\u8fd9\u4efd\u7b80\u77ed\u7684\u4ecb\u7ecd\u5c06\u5411\u4f60\u5c55\u793a\u5982\u4f55\u5229\u7528paddlepaddle\u6765\u89e3\u51b3\u4e00\u4e2a\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u95ee\u9898":32,"\u8fd9\u4f1a\u63d0\u793a\u5f53\u524d\u76ee\u5f55\u7684\u4e00\u4e9b\u53d8\u5316":43,"\u8fd9\u4f1a\u81ea\u52a8\u8fdb\u884c\u7f51\u7edc\u914d\u7f6e\u4e2d\u58f0\u660e\u7684\u6fc0\u6d3b\u64cd\u4f5c":44,"\u8fd9\u4f7f\u5f97nmt\u6a21\u578b\u5f97\u4ee5\u89e3\u653e\u51fa\u6765":69,"\u8fd9\u4fbf\u662f\u4e00\u79cd\u53cc\u5c42rnn\u7684\u8f93\u5165\u6570\u636e":39,"\u8fd9\u51e0\u4e2a\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":33,"\u8fd9\u53ef\u4ee5\u8ba9\u5176\u4ed6\u4eba\u77e5\u9053\u8fd9\u6b21\u63d0\u4ea4\u505a\u4e86\u54ea\u4e9b\u6539\u53d8":43,"\u8fd9\u53ef\u4ee5\u901a\u8fc7":43,"\u8fd9\u548c\u5355\u5c42rnn\u7684\u914d\u7f6e\u662f\u7b49\u4ef7\u7684":39,"\u8fd9\u56db\u4e2a\u7b80\u5355\u7684\u7279\u5f81\u662f\u6211\u4eec\u7684srl\u7cfb\u7edf\u6240\u9700\u8981\u7684":67,"\u8fd9\u56db\u6761\u6570\u636e\u540c\u65f6\u5904\u7406\u7684\u53e5\u5b50\u6570\u91cf\u4e3a":39,"\u8fd9\u5728\u5f88\u5927\u7a0b\u5ea6\u4e0a\u4f18\u4e8e\u5148\u524d\u7684\u6700\u5148\u8fdb\u7684\u7cfb\u7edf":67,"\u8fd9\u5728\u6784\u9020\u975e\u5e38\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u662f\u6709\u7528\u7684":42,"\u8fd9\u5bf9\u4e8e\u901a\u5e38\u7684java\u7684\u5f00\u53d1\u8005\u6765\u8bf4":27,"\u8fd9\u5c06\u82b1\u8d39\u6570\u5206\u949f\u7684\u65f6\u95f4":69,"\u8fd9\u5c31\u662f":53,"\u8fd9\u5df2\u7ecf\u5728":68,"\u8fd9\u610f\u5473\u7740":42,"\u8fd9\u610f\u5473\u7740\u6a21\u578b\u5728\u8bad\u7ec3\u6570\u636e\u4e0a\u4e0d\u65ad\u7684\u6539\u8fdb":32,"\u8fd9\u610f\u5473\u7740\u9664\u4e86\u6307\u5b9adevic":52,"\u8fd9\u65f6\u8fdb\u884c\u77e9\u9635\u4e58\u6cd5\u8fd0\u7b97\u5c31\u53ef\u80fd\u5bfc\u81f4\u6d6e\u70b9\u6570\u6ea2\u51fa":31,"\u8fd9\u662f\u4e00\u4e2a\u57fa\u4e8e\u7edf\u8ba1\u7684\u673a\u5668\u7ffb\u8bd1\u7cfb\u7edf":68,"\u8fd9\u662f\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":38,"\u8fd9\u662f\u56e0\u4e3a":27,"\u8fd9\u662f\u56e0\u4e3a\u5b83\u53d1\u6398\u51fa\u4e86\u56fe\u7247\u7684\u4e24\u7c7b\u91cd\u8981\u4fe1\u606f":61,"\u8fd9\u662f\u666e\u901a\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u4ee3\u7801":39,"\u8fd9\u662f\u76ee\u524dcmake\u5bfb\u627epython\u7684\u903b\u8f91\u5b58\u5728\u7f3a\u9677":31,"\u8fd9\u662f\u96c6\u675f\u641c\u7d22\u7684\u7ed3\u679c":69,"\u8fd9\u6765\u81ea\u4e8epaddlepaddle\u7684\u5185\u5b58\u4e2d":69,"\u8fd9\u6837":[28,32,48],"\u8fd9\u6837\u4fdd\u8bc1":30,"\u8fd9\u6837\u505a\u53ef\u4ee5\u6781\u5927\u7684\u51cf\u5c11\u5185\u5b58\u5360\u7528":31,"\u8fd9\u6837\u5355\u4e2a\u5b50\u7ebf\u7a0b\u7684\u957f\u5ea6\u5c31\u4e0d\u4f1a\u6ea2\u51fa\u4e86":53,"\u8fd9\u6837\u53ef\u4ee5\u51cf\u5c0fgpu\u5185\u5b58":52,"\u8fd9\u6837\u5bb9\u5668\u7684":56,"\u8fd9\u6837\u5c31\u4f1a\u751f\u6210\u4e24\u4e2a\u6587\u4ef6":66,"\u8fd9\u6837\u5c31\u53ef\u4ee5\u5728\u4e91\u7aef\u6267\u884c\u591a\u79cd\u6570\u636e\u7c7b\u8ba1\u7b97\u4efb\u52a1":26,"\u8fd9\u6837\u5f00\u53d1\u4eba\u5458\u53ef\u4ee5\u4ee5\u4e00\u81f4\u7684\u65b9\u5f0f\u5728\u4e0d\u540c\u7684\u5e73\u53f0\u4e0a\u5de5\u4f5c":34,"\u8fd9\u6837\u7684\u88c5\u9970\u5668":44,"\u8fd9\u6837\u7684\u8bdd":55,"\u8fd9\u6837\u7684\u8bdd\u6bcf\u4f4d\u7528\u6237\u5728\u6d4b\u8bd5\u6587\u4ef6\u4e2d\u5c06\u4e0e\u8bad\u7ec3\u6587\u4ef6\u542b\u6709\u540c\u6837\u7684\u4fe1\u606f":66,"\u8fd9\u6b63\u662f\u5b83\u4eec\u901f\u5ea6\u5feb\u7684\u539f\u56e0":47,"\u8fd9\u6bb5\u7b80\u77ed\u7684\u914d\u7f6e\u5c55\u793a\u4e86paddlepaddle\u7684\u57fa\u672c\u7528\u6cd5":32,"\u8fd9\u7528\u4e8e\u5728\u591a\u7ebf\u7a0b\u548c\u591a\u673a\u4e0a\u66f4\u65b0\u53c2\u6570":44,"\u8fd9\u79cd\u521d\u59cb\u5316\u65b9\u5f0f\u5728\u4e00\u822c\u60c5\u51b5\u4e0b\u4e0d\u4f1a\u4ea7\u751f\u5f88\u5dee\u7684\u7ed3\u679c":31,"\u8fd9\u79cd\u60c5\u51b5\u4e0b\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":44,"\u8fd9\u79cd\u65b9\u5f0f\u5fc5\u987b\u4f7f\u7528paddle\u5b58\u50a8\u7684\u6a21\u578b\u8def\u5f84\u683c\u5f0f":52,"\u8fd9\u79cd\u751f\u6210\u6280\u672f\u53ea\u7528\u4e8e\u7c7b\u4f3c\u89e3\u7801\u5668\u7684\u751f\u6210\u8fc7\u7a0b":42,"\u8fd9\u79cd\u7c7b\u578b\u7684\u8f93\u5165\u5fc5\u987b\u901a\u8fc7":41,"\u8fd9\u79cd\u96c6\u7fa4\u8282\u70b9\u7ba1\u7406\u65b9\u5f0f\u4f1a\u5728\u5c06\u6765\u4f7f\u7528":56,"\u8fd9\u7bc7\u6587\u6863":43,"\u8fd9\u7bc7\u6587\u7ae0":69,"\u8fd9\u7ec4\u8bed\u4e49\u76f8\u540c\u7684\u793a\u4f8b\u914d\u7f6e\u5982\u4e0b":39,"\u8fd9\u901a\u8fc7\u83b7\u5f97\u53cd\u5411\u5faa\u73af\u7f51\u7edc\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":42,"\u8fd9\u90fd\u9700\u8981\u8fd9\u4e2a\u63a5\u53e3\u6309\u7167\u7ea6\u5b9a\u4fd7\u6210\u7684\u89c4\u5219\u6765\u6ce8\u91ca\u5b8c\u5907":27,"\u8fd9\u91cc":[31,34,42,43,53,54,56,62,67],"\u8fd9\u91cc\u4e5f\u53ef\u53eb\u5206\u7c7b\u5c42":53,"\u8fd9\u91cc\u4ecb\u7ecd\u5982\u751f\u6210\u9020\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf":34,"\u8fd9\u91cc\u4ee5":64,"\u8fd9\u91cc\u4f7f\u7528\u4e00\u4e2a\u57fa\u4e8emomentum\u7684\u968f\u673a\u68af\u5ea6\u4e0b\u964d":32,"\u8fd9\u91cc\u4f7f\u7528\u4e86\u4e09\u79cd\u7f51\u7edc\u5355\u5143":32,"\u8fd9\u91cc\u4f7f\u7528\u4e86paddlepaddle\u7684python\u63a5\u53e3\u6765\u52a0\u8f7d\u6570\u6910":68,"\u8fd9\u91cc\u4f7f\u7528\u4e86paddlepaddle\u9884\u5b9a\u4e49\u597d\u7684rnn\u5904\u7406\u51fd\u6570":39,"\u8fd9\u91cc\u4f7f\u7528\u7b80\u5355\u7684":31,"\u8fd9\u91cc\u5229\u7528\u5b83\u5efa\u6a21\u53d8\u91cf\u4e4b\u95f4\u7684\u7ebf\u6027\u5173\u7cfb":32,"\u8fd9\u91cc\u53ea\u52a0\u8f7d":69,"\u8fd9\u91cc\u53ea\u7b80\u5355\u4ecb\u7ecd\u4e86\u5355\u673a\u8bad\u7ec3":64,"\u8fd9\u91cc\u5bf9":53,"\u8fd9\u91cc\u5c55\u793a\u5982\u4f55\u4f7f\u7528\u89c2\u6d4b\u6570\u636e\u6765\u62df\u5408\u8fd9\u4e00\u7ebf\u6027\u5173\u7cfb":32,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528":66,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7684\u662f\u4e00\u4e2a\u5c0f\u7684vgg\u7f51\u7edc":61,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7684\u662fgpu\u6a21\u5f0f\u8fdb\u884c\u8bad\u7ec3":61,"\u8fd9\u91cc\u6211\u4eec\u5728movielens\u6570\u636e\u96c6\u63cf\u8ff0\u4e00\u79cd":66,"\u8fd9\u91cc\u6211\u4eec\u5c55\u793a\u4e00\u4efd\u7b80\u5316\u8fc7\u7684\u4ee3\u7801":44,"\u8fd9\u91cc\u6211\u4eec\u7528paddl":34,"\u8fd9\u91cc\u6211\u4eec\u901a\u8fc7\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2ajob\u6765\u4e0b\u8f7d\u5e76\u5207\u5272\u6570\u636e":56,"\u8fd9\u91cc\u6307\u5b9a\u8bcd\u5178":64,"\u8fd9\u91cc\u6570\u636e\u5c42\u6709\u4e24\u4e2a":32,"\u8fd9\u91cc\u662f\u4e00\u4e2a\u4f8b\u5b50":69,"\u8fd9\u91cc\u6709\u4e00\u4e9b\u4e0d\u540c\u7684\u53c2\u6570\u9700\u8981\u6307\u5b9a":69,"\u8fd9\u91cc\u68c0\u9a8c\u8fd0\u884c\u65f6\u95f4\u6a21\u578b\u7684\u6536\u655b":48,"\u8fd9\u91cc\u6bcf\u4e2a5\u4e2abatch\u6253\u5370\u4e00\u4e2a\u70b9":69,"\u8fd9\u91cc\u6bcf\u9694100\u4e2abatch\u663e\u793a\u4e00\u6b21\u53c2\u6570\u7edf\u8ba1\u4fe1\u606f":69,"\u8fd9\u91cc\u6bcf\u969410\u4e2abatch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":69,"\u8fd9\u91cc\u7684\u5217\u51fa\u7684\u548c":61,"\u8fd9\u91cc\u76f4\u63a5\u901a\u8fc7\u9884\u6d4b\u811a\u672c":64,"\u8fd9\u91cc\u7b80\u5355\u4ecb\u7ecdlayer":53,"\u8fd9\u91cc\u7ed9\u51fa\u96c6\u4e2d\u5e38\u89c1\u7684\u90e8\u7f72\u65b9\u6cd5":54,"\u8fd9\u91cc\u8bbe\u7f6e\u4e3a\u4f7f\u7528cpu":69,"\u8fd9\u91cc\u8bbe\u7f6e\u4e3afals":69,"\u8fd9\u91cc\u8bbe\u7f6e\u4e3atrue":69,"\u8fd9\u91cc\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":64,"\u8fd9\u91cc\u9700\u8981\u7528\u6237\u989d\u5916\u6ce8\u610f":25,"\u8fdb\u4e3b\u4ed3\u5e93\u540e":43,"\u8fdb\u5165":68,"\u8fdb\u5165\u5bb9\u5668":55,"\u8fdb\u5165\u5f00\u53d1\u955c\u50cf\u5e76\u5f00\u59cb\u5de5\u4f5c":34,"\u8fdb\u7a0b":53,"\u8fdb\u7a0b\u4e2d\u53ef\u4ee5\u542f\u52a8\u591a\u4e2a\u5b50\u7ebf\u7a0b\u53bb\u63a5\u53d7":53,"\u8fdb\u7a0b\u4e4b\u540e":53,"\u8fdb\u7a0b\u5171\u7ed1\u5b9a\u591a\u5c11\u4e2a\u7aef\u53e3\u7528\u6765\u505a\u7a20\u5bc6\u66f4\u65b0":53,"\u8fdb\u7a0b\u542f\u52a8\u7684\u5fc5\u8981\u53c2\u6570":56,"\u8fdb\u7a0b\u7684":48,"\u8fdb\u7a0b\u7684\u542f\u52a8\u53c2\u6570":56,"\u8fdb\u7a0b\u7684\u8fd0\u884c\u73af\u5883":56,"\u8fdb\u7a0b\u7aef\u53e3\u662f":53,"\u8fdb\u7a0b\u9700\u8981\u7684":56,"\u8fdb\u800c\u8fdb\u884c\u4ee3\u7801\u8bc4\u5ba1":30,"\u8fdb\u884c\u4e86":39,"\u8fdb\u884c\u4f7f\u7528":61,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6848":56,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6cd5":56,"\u8fdb\u884c\u540c\u6b65":53,"\u8fdb\u884c\u5f00\u53d1":43,"\u8fdb\u884c\u62c6\u89e3":39,"\u8fdb\u884c\u6fc0\u6d3b\u64cd\u4f5c":44,"\u8fdb\u884c\u8bfb\u5165\u548c\u9884\u5904\u7406\u4ece\u800c\u5f97\u5230\u771f\u5b9e\u8f93\u5165":32,"\u8fdb\u884c\u9884\u6d4b":64,"\u8fdb\u9636\u6307\u5357":59,"\u8fde\u63a5":41,"\u8fde\u63a5\u4e09\u4e2alstm\u9690\u85cf\u5c42":68,"\u9000\u4f11\u4eba\u5458":65,"\u9000\u51fa\u5bb9\u5668":55,"\u9002\u4e2d":39,"\u9002\u5408\u81ea\u5df1\u5b9e\u9a8c":34,"\u9009":39,"\u9009\u62e9":39,"\u9009\u62e9\u6d4b\u8bd5\u7ed3\u679c\u6700\u597d\u7684\u6a21\u578b\u6765\u9884\u6d4b":64,"\u9009\u62e9\u76ee\u6807\u5206\u652f":43,"\u9009\u62e9\u8def\u5f84\u6765\u52a8\u6001\u52a0\u8f7dnvidia":50,"\u9009\u62e9\u8fc7\u540e\u7684":69,"\u9009\u62e9\u9002\u5408\u60a8\u7684\u573a\u666f\u7684\u5408\u9002\u65b9\u6848":54,"\u9009\u62e9cephfs\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u670d\u52a1":26,"\u9009\u81ea\u4e0b\u5217\u7c7b\u578b":65,"\u9009\u9879":[33,60],"\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":49,"\u901a\u5e38":[28,48,68],"\u901a\u5e38\u4f1a\u4f7f\u7528\u73af\u5883\u53d8\u91cf\u914d\u7f6ejob\u7684\u914d\u7f6e\u4fe1\u606f":56,"\u901a\u5e38\u4f7f\u7528\u7a00\u758f\u8bad\u7ec3\u6765\u52a0\u901f\u8ba1\u7b97\u8fc7\u7a0b":52,"\u901a\u5e38\u505a\u6cd5\u662f\u4ece\u4e00\u4e2a\u6bd4\u8f83\u5927\u7684learning_rate\u5f00\u59cb\u8bd5":31,"\u901a\u5e38\u540d\u5b57\u662f":43,"\u901a\u5e38\u5728\u9ad8\u7ea7\u60c5\u51b5\u4e0b":53,"\u901a\u5e38\u60c5\u51b5\u4e0b":47,"\u901a\u5e38\u6211\u4eec\u4f1a\u5b89\u88c5ceph\u7b49\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u6765\u5b58\u50a8\u8bad\u7ec3\u6570\u636e":55,"\u901a\u5e38\u6307\u5c06\u4e00\u4e2a\u6574\u4f53\u62c6\u5206\u6210\u591a\u4efd\u7684\u5176\u4e2d\u7684\u4e00\u4efd":25,"\u901a\u5e38\u662f\u4e00\u4e2apython\u51fd\u6570":53,"\u901a\u5e38\u6bcf\u4e2a\u914d\u7f6e\u6587\u4ef6\u90fd\u4f1a\u5305\u62ec":53,"\u901a\u5e38\u6bcf\u4e2ajob\u5305\u62ec\u4e00\u4e2a\u6216\u8005\u591a\u4e2apod":54,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u4f7f\u7528":42,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u5c06\u914d\u7f6e\u5b58\u4e8e":44,"\u901a\u5e38\u8981\u6c42\u65f6\u95f4\u6b65\u4e4b\u95f4\u5177\u6709\u4e00\u4e9b\u4f9d\u8d56\u6027":39,"\u901a\u5e38\u90fd\u4f1a\u4f7f\u7528\u4e0b\u9762\u8fd9\u4e9b\u547d\u4ee4\u884c\u53c2\u6570":52,"\u901a\u7528":49,"\u901a\u77e5":39,"\u901a\u77e5\u7cfb\u7edf\u4e00\u8f6e\u6570\u636e\u8bfb\u53d6\u7ed3\u675f":53,"\u901a\u8fc7":[31,39,43,44,48,53,64],"\u901a\u8fc7\u4e24\u4e2a\u5d4c\u5957\u7684":41,"\u901a\u8fc7\u4ea4\u66ff\u4f7f\u7528\u5377\u79ef\u548c\u6c60\u5316\u5904\u7406":61,"\u901a\u8fc7\u4f7f\u7528":33,"\u901a\u8fc7\u51fd\u6570":56,"\u901a\u8fc7\u5377\u79ef\u64cd\u4f5c\u4ece\u56fe\u7247\u6216\u7279\u5f81\u56fe\u4e2d\u63d0\u53d6\u7279\u5f81":61,"\u901a\u8fc7\u547d\u4ee4\u884c\u53c2\u6570":31,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u7684\u8f93\u51fa":41,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa":41,"\u901a\u8fc7\u6240\u6709\u5355\u5143\u6d4b\u8bd5":43,"\u901a\u8fc7\u6240\u6709\u8bad\u7ec3\u96c6\u4e00\u6b21\u79f0\u4e3a\u4e00\u904d":68,"\u901a\u8fc7\u67e5\u770b\u4e70\u5bb6\u5bf9\u67d0\u4e2a\u4ea7\u54c1\u7684\u8bc4\u4ef7\u53cd\u9988":64,"\u901a\u8fc7\u6a21\u578b\u63a8\u65adapi\u7684\u5b9e\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u6837\u4f8b":28,"\u901a\u8fc7\u7f16\u8bd1\u4f1a\u751f\u6210py_paddle\u8f6f\u4ef6\u5305":5,"\u901a\u8fc7\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u6765\u6307\u5b9a":44,"\u901a\u8fc7\u8c03\u7528":5,"\u901a\u8fc7\u914d\u7f6e\u7c7b\u4f3c\u4e8e":64,"\u901a\u8fc7data":41,"\u901a\u8fc7volum":54,"\u903b\u8f91\u56de\u5f52":64,"\u9053\u6b49":39,"\u9069":39,"\u9075\u5faa\u4ee5\u4e0b\u6d41\u7a0b":30,"\u9075\u5faa\u5982\u4e0b\u7684\u683c\u5f0f":65,"\u9075\u5faa\u6587\u7ae0":60,"\u90a3\u4e48":[28,41,44],"\u90a3\u4e480\u5c42\u5e8f\u5217\u5373\u4e3a\u4e00\u4e2a\u8bcd\u8bed":41,"\u90a3\u4e48\u53ef\u4ee5\u8ba4\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":31,"\u90a3\u4e48\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4e0d\u4f1a\u6267\u884c\u6d4b\u8bd5\u64cd\u4f5c":2,"\u90a3\u4e48\u5982\u4f55\u5224\u65ad\u8bad\u7ec3\u4e0d\u6536\u655b\u5462":31,"\u90a3\u4e48\u5e38\u6570\u8f93\u51fa\u6240\u80fd\u8fbe\u5230\u7684\u6700\u5c0fcost\u662f":31,"\u90a3\u4e48\u5f53check\u51fa\u6570\u636e\u4e0d\u5408\u6cd5\u65f6":3,"\u90a3\u4e48\u6211\u4eec\u53ef\u4ee5\u5224\u65ad\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":31,"\u90a3\u4e48\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u6536\u96c6\u5e02\u573a\u4e0a\u623f\u5b50\u7684\u5927\u5c0f\u548c\u4ef7\u683c":32,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528":42,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u65b9\u6cd5":42,"\u90a3\u4e48\u6536\u655b\u53ef\u80fd\u5f88\u6162":31,"\u90a3\u4e48\u6700\u597d\u5c06\u6570\u636e\u6587\u4ef6\u5728\u6bcf\u6b21\u8bfb\u53d6\u4e4b\u524d\u505a\u4e00\u6b21shuffl":31,"\u90a3\u4e48\u7528\u6237\u9700\u8981\u5728\u672c\u5730\u7f16\u8bd1\u5b89\u88c5paddlepaddl":45,"\u90a3\u4e48\u8bad\u7ec3\u6709\u53ef\u80fd\u4e0d\u6536\u655b":31,"\u90a3\u4e48\u8be5\u4f18\u5316\u7b97\u6cd5\u81f3\u5c11\u9700\u8981":31,"\u90a3\u4e48fc1\u548cfc2\u5c42\u5c06\u4f1a\u4f7f\u7528\u7b2c1\u4e2agpu\u6765\u8ba1\u7b97":52,"\u90a3\u4e48paddlepaddle\u4f1a\u6839\u636elayer\u7684\u58f0\u660e\u987a\u5e8f":3,"\u90a3\u4e5f\u5c31\u4e0d\u9700\u8981\u6025\u7740\u4f18\u5316\u6027\u80fd\u5566":47,"\u90a3\u4f30\u8ba1\u8fd9\u91cc\u7684\u6f5c\u529b\u5c31\u6ca1\u5565\u597d\u6316\u7684\u4e86":47,"\u90a3\u51cf\u5c11\u5b66\u4e60\u738710\u500d\u7ee7\u7eed\u8bd5\u9a8c":31,"\u90a3\u6211\u4f1a\u671f\u671b\u5206\u6790\u5de5\u5177\u7edf\u8ba1\u5230\u901f\u5ea6\u662f100gb":47,"\u90a3\u7a0b\u5e8f\u5206\u6790\u5de5\u5177\u662f\u5fc5\u4e0d\u53ef\u5c11\u7684\u5229\u5668":47,"\u90ae\u7f16":65,"\u90e8\u7f72\u548c\u914d\u7f6e\u6bd4\u8f83\u7b80\u5355":54,"\u90fd":39,"\u90fd\u4f1a\u4ea7\u751f\u5f53\u524d\u5c42\u72b6\u6001\u7684\u6240\u6709\u7ee7\u627f\u7ed3\u679c":50,"\u90fd\u4f7f\u7528\u968f\u673a\u503c\u521d\u59cb\u5316":32,"\u90fd\u53ea\u662f\u4ecb\u7ecd\u53cc\u5c42rnn\u7684api\u63a5\u53e3":39,"\u90fd\u662f\u4e94\u4f4d\u7684\u6570\u5b57":26,"\u90fd\u662f\u5bf9layer1\u5143\u7d20\u7684\u62f7\u8d1d":38,"\u90fd\u662f\u5c06\u6bcf\u4e00\u53e5\u5206\u597d\u8bcd\u540e\u7684\u53e5\u5b50":39,"\u90fd\u662fabi\u8c03\u7528\u6807\u51c6\u7684":27,"\u90fd\u7528":43,"\u90fd\u9700\u8981\u5199\u63d0\u4ea4\u8bf4\u660e":43,"\u90fd\u9700\u8981\u8c03\u7528\u4e00\u6b21":44,"\u914d\u5408\u4f7f\u7528":53,"\u914d\u7f6e":68,"\u914d\u7f6e\u4e86\u7f51\u7edc":66,"\u914d\u7f6e\u51fa\u975e\u5e38\u590d\u6742\u7684\u7f51\u7edc":53,"\u914d\u7f6e\u521b\u5efa\u5b8c\u6bd5\u540e":61,"\u914d\u7f6e\u5982\u4e0b":60,"\u914d\u7f6e\u6253\u5f00":47,"\u914d\u7f6e\u6570\u636e\u6e90":53,"\u914d\u7f6e\u6587\u4ef6":64,"\u914d\u7f6e\u6587\u4ef6\u63a5\u53e3\u662ffc_layer":44,"\u914d\u7f6e\u6a21\u578b\u6587\u4ef6":60,"\u914d\u7f6e\u7b49\u6587\u4ef6\u7684\u76ee\u5f55\u89c6\u4e3a":48,"\u914d\u7f6e\u7b80\u5355\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u4f8b\u5b50":42,"\u914d\u7f6e\u7f51\u7edc\u5c42\u7684\u8f93\u5165":44,"\u914d\u7f6eapi":38,"\u9152\u5e97":39,"\u91c7\u6837\u5c42":66,"\u91c7\u7528":67,"\u91c7\u7528\u53e6\u4e00\u79cd\u65b9\u6cd5\u6765\u5806\u53e0lstm\u5c42":67,"\u91c7\u7528\u5747\u5300\u5206\u5e03\u6216\u8005\u9ad8\u65af\u5206\u5e03\u521d\u59cb\u5316":50,"\u91c7\u7528multi":31,"\u91ca\u653e\u5bf9paramters\u5185\u5b58\u7684\u9501\u5b9a":25,"\u91cc\u4ecb\u7ecd\u4e86\u7528paddle\u6e90\u7801\u4e2d\u7684\u811a\u672c\u4e0b\u8f7d\u8bad\u7ec3\u6570\u636e\u7684\u8fc7\u7a0b":55,"\u91cc\u4f1a\u7ee7\u7eed\u5b89\u88c5":36,"\u91cc\u6240\u6709\u7684\u7b26\u53f7\u90fd\u5199\u5165\u81ea\u5df1\u7684\u7a0b\u5e8f\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cc":27,"\u91cc\u6307\u5b9a\u56fe\u50cf\u6570\u636e\u5217\u8868":62,"\u91cc\u7684\u65e5\u5fd7":48,"\u91cc\u901a\u8fc7train_list\u548ctest_list\u6307\u5b9a\u662f\u8bad\u7ec3\u6587\u4ef6\u5217\u8868\u548c\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":53,"\u91cc\u9762\u6db5\u76d6\u4e86":34,"\u91cc\u9762\u6db5\u76d6\u4e86paddlepaddle\u8fd0\u884c\u6240\u9700\u7684\u6240\u6709\u73af\u5883":34,"\u91cd\u547d\u540d\u6210":27,"\u91cd\u65b0\u7f16\u8bd1paddlepaddl":47,"\u9488\u5bf9\u4efb\u52a1\u8fd0\u884c\u5b8c\u6210\u540e\u5bb9\u5668\u81ea\u52a8\u9000\u51fa\u7684\u573a\u666f":55,"\u9488\u5bf9\u5185\u5b58\u548c\u663e\u5b58":31,"\u9488\u5bf9\u6587\u672c":66,"\u94fe\u63a5\u4f55\u79cdblas\u5e93\u7b49":33,"\u94fe\u63a5\u5230\u81ea\u5df1\u7684\u7a0b\u5e8f\u91cc":27,"\u94fe\u63a5\u5f85\u8865\u5145":64,"\u9500\u552e":65,"\u9519\u8bef\u5904\u7406":27,"\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662f\u8fd4\u56de\u503c":27,"\u9519\u8bef\u5904\u7406\u7684\u65b9\u5f0f\u4e5f\u4e0d\u5c3d\u76f8\u540c":27,"\u9519\u8bef\u7387":64,"\u9519\u8bef\u7684define_py_data_sources2\u7c7b\u4f3c":31,"\u955c\u50cf\u91cc\u6709":55,"\u957f\u5ea6":31,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570":42,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u8f93\u51famemori":42,"\u95ee\u9898":32,"\u95f4\u9694":64,"\u9650\u5236\u5957\u63a5\u5b57\u53d1\u9001\u7f13\u51b2\u533a\u7684\u5927\u5c0f":50,"\u9650\u5236\u5957\u63a5\u5b57\u63a5\u6536\u7f13\u51b2\u533a\u7684\u5927\u5c0f":50,"\u9664\u4e86":3,"\u9664\u4e86boot_lay":39,"\u9664\u53bbdata\u5c42":64,"\u9664\u6784\u9020\u67d0\u79cd\u7c7b\u578b\u7684\u51fd\u6570":28,"\u9664\u8bcd\u5411\u91cf\u6a21\u578b\u5916\u7684\u53c2\u6570\u5c06\u4f7f\u7528\u6b63\u6001\u5206\u5e03\u968f\u673a\u521d\u59cb\u5316":60,"\u9664\u96f6\u7b49\u95ee\u9898":31,"\u968f\u540e\u53ef\u4ee5\u7528\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf\u5f00build":43,"\u968f\u673a\u521d\u59cb\u4e0d\u5b58\u5728\u7684\u53c2\u6570":67,"\u968f\u673a\u6570\u7684\u79cd\u5b50":50,"\u968f\u673a\u6570seed":49,"\u968f\u7740\u8f6e\u6570\u589e\u52a0\u8bef\u5dee\u4ee3\u4ef7\u51fd\u6570\u7684\u8f93\u51fa\u5728\u4e0d\u65ad\u7684\u51cf\u5c0f":32,"\u9694\u5f00":62,"\u96c6":65,"\u96c6\u675f\u641c\u7d22\u4e2d\u7684\u6269\u5c55\u5e7f\u5ea6":69,"\u96c6\u675f\u641c\u7d22\u4f7f\u7528\u5e7f\u5ea6\u4f18\u5148\u641c\u7d22\u6765\u6784\u5efa\u641c\u7d22\u6811":69,"\u96c6\u675f\u641c\u7d22\u4f7f\u7528\u5e7f\u5ea6\u4f18\u5148\u641c\u7d22\u7684\u65b9\u5f0f\u6784\u5efa\u67e5\u627e\u6811":50,"\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673a\u4f7f\u7528cpu\u7684paddle\u8bad\u7ec3\u4f5c\u4e1a":55,"\u96c6\u7fa4\u4f5c\u4e1a\u4e2d\u6240\u6709\u8fdb\u7a0b\u7684\u73af\u5883\u8bbe\u7f6e":48,"\u96c6\u7fa4\u4f5c\u4e1a\u5c06\u4f1a\u5728\u51e0\u79d2\u540e\u542f\u52a8":48,"\u96c6\u7fa4\u5de5\u4f5c":48,"\u96c6\u7fa4\u6d4b\u8bd5":49,"\u96c6\u7fa4\u8bad\u7ec3":49,"\u96c6\u7fa4\u8fdb\u7a0b":48,"\u96c6\u7fa4\u901a\u4fe1\u4fe1\u9053\u7684\u7aef\u53e3\u6570":48,"\u96c6\u7fa4\u901a\u4fe1\u901a\u9053\u7684":48,"\u96c6\u7fa4\u901a\u4fe1\u901a\u9053\u7684\u7aef\u53e3\u53f7":48,"\u9700\u5728nvvp\u754c\u9762\u4e2d\u9009\u4e0a\u624d\u80fd\u5f00\u542f":47,"\u9700\u8981\u4e0e":53,"\u9700\u8981\u4f7f\u7528\u5176\u5236\u5b9a\u7684\u65b9\u5f0f\u6302\u8f7d\u540e\u5e76\u5bfc\u5165\u6570\u636e":56,"\u9700\u8981\u4fdd\u5b58\u4e00\u4e9b\u9884\u7f6e\u7684\u516c\u5f00\u6570\u636e\u96c6":26,"\u9700\u8981\u4fdd\u6301\u5f53\u524d\u5206\u652f\u76ee\u5f55":43,"\u9700\u8981\u5148\u6302\u8f7d\u5230\u670d\u52a1\u5668node\u4e0a\u518d\u901a\u8fc7kubernet":54,"\u9700\u8981\u540c\u6b65\u539f\u4ed3\u5e93":43,"\u9700\u8981\u542f\u52a8":53,"\u9700\u8981\u542f\u52a8\u7684\u8282\u70b9\u4e2a\u6570\u4ee5\u53ca":56,"\u9700\u8981\u5728":48,"\u9700\u8981\u5728\u521b\u5efa\u5bb9\u5668\u524d\u6302\u8f7d\u5377\u4ee5\u4fbf\u6211\u4eec\u4fdd\u5b58\u8bad\u7ec3\u7ed3\u679c":55,"\u9700\u8981\u5728\u7cfb\u7edf\u91cc\u5148\u5b89\u88c5\u597ddocker\u5de5\u5177\u5305":45,"\u9700\u8981\u5728cmake\u7684\u65f6\u5019":28,"\u9700\u8981\u5b89\u88c5graphviz\u6765\u8f6c\u6362dot\u6587\u4ef6\u4e3a\u56fe\u7247":62,"\u9700\u8981\u5bf9":54,"\u9700\u8981\u5c06\u5176parameter\u8bbe\u7f6e\u6210":31,"\u9700\u8981\u5c06\u6807\u8bb0\u7684\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6837\u672c\u79fb\u52a8\u5230\u53e6\u4e00\u4e2a\u8def\u5f84":68,"\u9700\u8981\u5c06bugfix\u7684\u5206\u652f\u540c\u65f6merge\u5230":30,"\u9700\u8981\u5f15\u7528":28,"\u9700\u8981\u6307\u5b9a\u4e0e\u67d0\u4e00\u4e2a\u8f93\u5165\u7684\u5e8f\u5217\u4fe1\u606f\u662f\u4e00\u81f4\u7684":39,"\u9700\u8981\u660e\u786e\u6307\u5b9a":50,"\u9700\u8981\u6709\u4e00\u4e2a\u5916\u90e8\u7684\u5b58\u50a8\u670d\u52a1\u6765\u4fdd\u5b58\u8bad\u7ec3\u6240\u9700\u6570\u636e\u548c\u8bad\u7ec3\u8f93\u51fa":54,"\u9700\u8981\u6709\u7a33\u5b9a\u7684\u5bfc\u51fa\u7b26\u53f7":27,"\u9700\u8981\u6839\u636e\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u6765\u7ed1\u5b9a\u4e00\u4e2a":56,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[30,50,53,66],"\u9700\u8981\u6ce8\u610f\u7684\u662f\u68af\u5ea6\u68c0\u67e5\u4ec5\u4ec5\u9a8c\u8bc1\u4e86\u68af\u5ea6\u7684\u8ba1\u7b97":44,"\u9700\u8981\u6ce8\u610f\u7684\u662fpaddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5b50\u5e8f\u5217\u6570\u76ee\u4e00\u6837\u7684\u591a\u8f93\u5165\u53cc\u5c42rnn":39,"\u9700\u8981\u88ab\u66b4\u9732\u5230\u5176\u4ed6\u8bed\u8a00":28,"\u9700\u8981\u9075\u5faa\u4ee5\u4e0b\u7ea6\u5b9a":41,"\u9700\u8981import\u8fd9\u4e9b\u51fd\u6570":53,"\u9700\u8981python\u63a5\u53e3\u91cc\u5904\u7406shuffl":53,"\u975e\u5e38\u6570":44,"\u975e\u96f6\u6570\u5b57\u7684\u4e2a\u6570":44,"\u97f3\u4e50\u5267":65,"\u987a\u5e8f":39,"\u9884\u5904\u7406\u6570\u636e\u4e00\u822c\u7684\u547d\u4ee4\u4e3a":66,"\u9884\u5904\u7406\u811a\u672c":68,"\u9884\u5b9a\u4e49\u7f51\u7edc":68,"\u9884\u5b9a\u4e49\u7f51\u7edc\u5982\u56fe3\u6240\u793a":68,"\u9884\u63d0\u4ea4\u94a9\u5b50":43,"\u9884\u6d4b\u540e":67,"\u9884\u6d4b\u63a5\u53e3\u811a\u672c":68,"\u9884\u6d4b\u6982\u7387\u53d6\u5e73\u5747":62,"\u9884\u6d4b\u7a0b\u5e8f\u5c06\u8bfb\u53d6\u7528\u6237\u7684\u8f93\u5165":66,"\u9884\u6d4b\u7ed3\u679c\u4ee5\u6587\u672c\u7684\u5f62\u5f0f\u4fdd\u5b58\u5728":64,"\u9884\u6d4b\u811a\u672c\u662f":67,"\u9884\u6d4b\u9636\u6bb5":53,"\u9884\u6d4bid":64,"\u9884\u6d4bimdb\u7684\u672a\u6807\u8bb0\u8bc4\u8bba\u7684\u4e00\u4e2a\u5b9e\u4f8b\u5982\u4e0b":68,"\u9884\u8bad\u7ec3\u6a21\u578b\u4f7f\u7528\u7684\u5b57\u5178\u7684\u8def\u5f84":60,"\u9884\u8bad\u7ec3\u8bcd\u5411\u91cf\u5b57\u5178\u6a21\u578b\u7684\u8def\u5f84":60,"\u989c\u8272\u901a\u9053\u987a\u5e8f\u4e3a":62,"\u989d\u5916\u7684\u53c2\u6570":64,"\u9996\u5148":[3,32,39,42,44,60,62,64,67,68],"\u9996\u5148\u4e0b\u8f7dcifar":61,"\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":33,"\u9996\u5148\u5b89\u88c5\u5e76\u5728\u5f53\u524d\u76ee\u5f55\u8fd0\u884c\u5b83":43,"\u9996\u5148\u5b89\u88c5paddlepaddl":68,"\u9996\u5148\u5bf9\u8f93\u5165\u505a\u4e00\u4e2a\u5c0f\u7684\u6270\u52a8":44,"\u9996\u5148\u6211\u4eec\u9700\u8981\u63a8\u5bfc\u8be5\u7f51\u7edc\u5c42\u7684":44,"\u9996\u5148\u662f\u6cd5\u8bed\u5e8f\u5217":69,"\u9996\u5148\u901a\u8fc7":43,"\u9a71\u52a8":45,"\u9ad8\u4e2d\u6bd5\u4e1a\u751f":65,"\u9ad8\u4eae\u90e8\u5206":39,"\u9ad8\u53ef\u7528":54,"\u9ad8\u5ea6\u652f\u6301\u7075\u6d3b\u548c\u9ad8\u6548\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":42,"\u9ad8\u6548\u6027":0,"\u9ad8\u65af\u5206\u5e03":31,"\u9ed1\u8272\u7535\u5f71":65,"\u9ed8\u8ba4":[3,50,69],"\u9ed8\u8ba4\u4e00\u4e2apass\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":64,"\u9ed8\u8ba4\u4e0d\u663e\u793a":50,"\u9ed8\u8ba4\u4e0d\u8bbe\u7f6e":41,"\u9ed8\u8ba4\u4e3a0":[50,52],"\u9ed8\u8ba4\u4e3a1":[3,52],"\u9ed8\u8ba4\u4e3a100":52,"\u9ed8\u8ba4\u4e3a4096mb":50,"\u9ed8\u8ba4\u4e3a\u4e0d\u4f7f\u7528":66,"\u9ed8\u8ba4\u4e3a\u7b2c\u4e00\u4e2a\u8f93\u5165":41,"\u9ed8\u8ba4\u4e3anull":50,"\u9ed8\u8ba4\u4f7f\u7528\u591a\u7c7b\u4ea4\u53c9\u71b5\u635f\u5931\u51fd\u6570\u548c\u5206\u7c7b\u9519\u8bef\u7387\u7edf\u8ba1\u8bc4\u4f30\u5668":53,"\u9ed8\u8ba4\u4f7f\u7528concurrentremoteparameterupdat":50,"\u9ed8\u8ba4\u503c":[33,38,52],"\u9ed8\u8ba4\u521d\u59cb\u72b6\u4e3a0":41,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b":[31,48,68],"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u4f7f\u7528\u6b64\u7f51\u7edc":68,"\u9ed8\u8ba4\u6307\u5b9a\u7b2c\u4e00\u4e2a\u8f93\u5165":39,"\u9ed8\u8ba4\u662f0":53,"\u9ed8\u8ba4\u662f1":53,"\u9ed8\u8ba4\u7528\u6765\u5207\u5206\u5355\u8bb0\u548c\u6807\u70b9\u7b26\u53f7":68,"\u9ed8\u8ba4\u7684":55,"\u9ed8\u8ba4\u8bbe\u7f6e\u4e3a\u771f":52,"\u9ed8\u8ba4\u914d\u7f6e\u5982\u4e0b":48,"adamax\u7b49":64,"amazon\u7535\u5b50\u4ea7\u54c1\u8bc4\u8bba\u6570\u636e":64,"api\u4e2d\u4f7f\u7528":27,"api\u5bf9\u6bd4\u4ecb\u7ecd":40,"api\u5bfc\u51fa\u7684\u52a8\u6001\u5e93":28,"api\u5bfc\u51fa\u7684\u9759\u6001\u5e93":28,"api\u63a5\u53d7\u7684\u7c7b\u578b\u5168\u662f":28,"api\u63a5\u53e3":54,"api\u63a5\u53e3\u7684\u53c2\u6570\u8f6c\u53d1\u7ed9":28,"api\u6587\u6863\u751f\u6210\u8fc7\u7a0b\u4f9d\u8d56\u4e8epy_paddl":45,"api\u65f6":28,"api\u65f6\u6240\u552f\u4e00\u9700\u8981\u5f15\u5165\u7684\u5934\u6587\u4ef6":28,"api\u662f\u591a\u8bed\u8a00api\u7684\u57fa\u7840\u90e8\u5206":28,"api\u66b4\u9732\u7684\u7c7b\u578b":28,"api\u751f\u6210\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u4f1a\u88ab\u5b89\u88c5\u5230":28,"api\u7684\u5b9e\u4f8b":28,"api\u7684\u5b9e\u73b0\u7ec6\u8282":28,"api\u7684\u63a5\u53e3":28,"api\u7684\u65f6\u5019\u63a8\u8350paddle\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":28,"api\u7684\u7f16\u8bd1\u9009\u9879\u9ed8\u8ba4\u5173\u95ed":28,"api\u76ee\u5f55\u7ed3\u6784\u5982\u4e0a\u56fe\u8868\u6240\u793a":28,"api\u83b7\u5f97\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u5b9e\u4f8b":28,"async_sgd\u8fdb\u884c\u8bad\u7ec3\u65f6":31,"atlas\u7684\u8def\u5f84":33,"avx\u7684\u955c\u50cf":34,"awselasticblockstore\u7b49":54,"batch\u4e2d\u5305\u542b":31,"batches\u4e2a\u6279\u6b21\u4fdd\u5b58\u4e00\u6b21\u53c2\u6570":50,"batches\u6b21":50,"bin\u548c\u8bc4\u5206\u6587\u4ef6":66,"blas\u7684\u8def\u5f84":33,"block\u6784\u6210\u4e00\u4e2amodel":25,"book\u4e00\u5b9a\u662f\u60a8\u6700\u597d\u7684\u9009\u62e9":34,"book\u4e2d\u6240\u6709\u7ae0\u8282\u529f\u80fd\u7684\u6b63\u786e\u6027":30,"book\u662f\u4e3a\u7528\u6237\u548c\u5f00\u53d1\u8005\u5236\u4f5c\u7684\u4e00\u4e2a\u4ea4\u4e92\u5f0f\u7684jupyt":34,"book\u7684docker\u955c\u50cf":34,"bool\u578b\u53c2\u6570":3,"boolean":[10,16,27],"bugfix\u5206\u652f\u4e5f\u662f\u5728\u5f00\u53d1\u8005\u81ea\u5df1\u7684fork\u7248\u672c\u5e93\u7ef4\u62a4":30,"bugfix\u5206\u652f\u9700\u8981\u5206\u522b\u7ed9\u4e3b\u7248\u672c\u5e93\u7684":30,"build\u6307\u4ee4\u6309\u7167dockerfile\u751f\u6210\u955c\u50cf":34,"build\u8fd9\u4e2a\u547d\u4ee4\u7684":34,"byte":31,"c99\u662f\u76ee\u524dc\u6700\u5e7f\u6cdb\u7684\u4f7f\u7528\u6807\u51c6":27,"c\u6709\u6807\u51c6\u7684abi":27,"c\u8bed\u8a00\u662f\u6709\u5bfc\u51fa\u7b26\u53f7\u7684\u6807\u51c6\u7684":27,"caoying\u7684pul":69,"case":[10,16,28,29,47],"class":[7,10,12,14,15,16,17,18,19,20,22,23,27,31,44,68],"cmake\u4e2d\u5c06":47,"cmake\u627e\u5230\u7684python\u5e93\u548cpython\u89e3\u91ca\u5668\u7248\u672c\u53ef\u80fd\u6709\u4e0d\u4e00\u81f4\u73b0\u8c61":31,"cmake\u7f16\u8bd1\u65f6":33,"cmake\u914d\u7f6e\u4e2d\u5c06":47,"conf\u4f5c\u4e3a\u914d\u7f6e":69,"const":44,"container\u4e2d":55,"core\u4e2d\u7684\u6a21\u578b\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u53c2\u6570":28,"core\u4e2d\u8fd9\u4e00\u7c7b\u578b\u63a5\u53e3\u7684\u667a\u80fd\u6307\u9488":28,"core\u662f\u5426\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u5b9e\u4f8b":28,"core\u6982\u5ff5":28,"cost\u8fd8\u5927\u4e8e\u8fd9\u4e2a\u6570":31,"count\u4e2agpu\u4e0a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97\u67d0\u4e00\u5c42":52,"count\u548cgpu":52,"cpu\u7248\u672c":36,"cuda\u5e73\u53f0":36,"cuda\u5e93":50,"cudnn\u5e93":[33,50],"dat\u4e2d":66,"data\u5230\u5206\u5e03\u5f0f\u5b58\u50a8\u8865\u5145\u8bad\u7ec3\u6570\u636e":26,"data\u76ee\u5f55\u4e2d\u5b58\u653e\u5207\u5206\u597d\u7684\u6570\u636e":56,"dataprovider\u5171\u8fd4\u56de\u4e24\u4e2a\u6570\u636e":39,"dataprovider\u5171\u8fd4\u56de\u4e24\u7ec4\u6570\u636e":39,"dataprovider\u662f\u88ab\u7cfb\u7edf\u8c03\u7528":53,"dataprovider\u662fpaddlepaddle\u7cfb\u7edf\u7684\u6570\u636e\u63d0\u4f9b\u5668":53,"dataprovider\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":2,"dataprovider\u7684\u4ecb\u7ecd":[4,64],"dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":31,"dataprovider\u8fd4\u56de\u7a7a\u6570\u636e":53,"dataprovider\u91cc\u5b9a\u4e49\u6570\u636e\u8bfb\u53d6\u51fd\u6570":53,"deb\u5305":30,"deb\u5305\u7f16\u8bd1\u95ee\u9898":30,"deb\u5b89\u88c5\u5305":36,"decay\u5219\u4e3a0":61,"decoder\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4f1a\u5f15\u7528\u5168\u90e8\u7ed3\u679c":41,"decoder\u63a5\u53d7\u4e24\u4e2a\u8f93\u5165":41,"decoder\u6bcf\u6b21\u9884\u6d4b\u4ea7\u751f\u4e0b\u4e00\u4e2a\u6700\u53ef\u80fd\u7684\u8bcd\u8bed":41,"decoer\u67b6\u6784":41,"default":[7,9,10,11,12,15,16,17,19,20,22,23,52,55,56,68],"demo\u9884\u6d4b\u8f93\u51fa\u5982\u4e0b":5,"dictionary\u662f\u4ece\u7f51\u7edc\u914d\u7f6e\u4e2d\u4f20\u5165\u7684dict\u5bf9\u8c61":3,"dictionary\u7531\u89e3\u6790\u81ea\u52a8\u751f\u6210":66,"dir\u4e2d\u670916\u4e2a\u5b50\u76ee\u5f55":69,"docker\u5b58\u5728\u95ee\u9898":34,"docker\u5b89\u88c5\u8bf7\u53c2\u8003":45,"docker\u7684\u5b98\u7f51":45,"docker\u955c\u50cf\u6765\u670d\u52a1html\u4ee3\u7801":34,"dockerfile\u4f4d\u4e8epaddlepaddl":34,"dockers\u8bbe\u7f6e":34,"dropout\u7684\u6bd4\u4f8b":44,"elec\u6d4b\u8bd5\u96c6":64,"embedding\u6a21\u578b\u9700\u8981\u7a0d\u5fae\u6539\u53d8\u63d0\u4f9b\u6570\u636e\u7684python\u811a\u672c":64,"encode\u6210\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":39,"encoder\u548cdecoder\u53ef\u4ee5\u662f\u80fd\u591f\u5904\u7406\u5e8f\u5217\u7684\u4efb\u610f\u795e\u7ecf\u7f51\u7edc\u5355\u5143":41,"encoder\u8f93\u51fa":41,"entropy\u4f5c\u4e3acost":31,"evaluator\u7684\u503c\u4f4e\u4e8e0":69,"export":[31,34,36,61],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":3,"false\u7684\u60c5\u51b5":3,"fc1\u548cfc2\u5c42\u5728gpu\u4e0a\u8ba1\u7b97":52,"fc3\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":52,"final":[11,17,66],"float":[3,7,9,10,12,15,16,18,20,32,47,62,66],"float\u7b49":52,"function":[8,10,11,12,16,17,18,20,23,24,29,42,68],"gen\u6587\u4ef6\u5939\u4e2d\u7684\u6587\u4ef6\u5217\u8868":69,"generator\u4fbf\u4f1a\u5b58\u4e0b\u5f53\u524d\u7684\u4e0a\u4e0b\u6587":3,"generator\u81f3\u5c11\u9700\u8981\u8c03\u7528\u4e24\u6b21\u624d\u4f1a\u77e5\u9053\u662f\u5426\u505c\u6b62":3,"git\u6d41\u5206\u652f\u6a21\u578b":43,"github\u9996\u9875":43,"golang\u53ef\u4ee5\u4f7f\u7528":27,"golang\u7684":27,"gpu\u4e8c\u8fdb\u5236\u6587\u4ef6":33,"gpu\u5219\u8fd8\u9700\u8981\u9ad8\u5e76\u884c\u6027":47,"gpu\u53cc\u7f13\u5b58":3,"gpu\u548c\u975eavx\u533a\u5206\u4e86\u5982\u4e0b4\u4e2a\u955c\u50cf":34,"gpu\u548ccpu\u901a\u4fe1":53,"gpu\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18":46,"gpu\u6838\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u6307\u5b9a":50,"gpu\u7248\u672c":36,"gpu\u7248\u672c\u5e76\u60f3\u4f7f\u7528":67,"gpu\u7684docker\u955c\u50cf\u7684\u65f6\u5019":31,"gram\u7ea7\u522b\u7684\u77e5\u8bc6":68,"group\u6559\u7a0b":40,"gru\u6216lstm":42,"gru\u6a21\u578b":64,"gru\u6a21\u578b\u914d\u7f6e":64,"h\u5e76\u4e0d\u56f0\u96be":27,"hot\u7a20\u5bc6\u5411\u91cf":66,"html\u5373\u53ef\u8bbf\u95ee\u672c\u5730\u6587\u6863":45,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":3,"id\u4e3a0\u7684\u6982\u7387":64,"id\u4e3a1\u7684\u6982\u7387":64,"id\u6307\u5b9a\u4f7f\u7528\u54ea\u4e2agpu\u6838":50,"id\u6307\u5b9a\u7684gpu":52,"id\u65e0\u6548":50,"image\u91cc":55,"imagenet\u6570\u636e\u96c6\u7b49":26,"images\u6570\u636e\u96c6\u4e0a\u4f20\u5230\u4e91\u7aef\u7684":26,"imdb\u6570\u636e\u96c6\u5305\u542b25":68,"imdb\u6709\u4e24\u4e2a\u6807\u7b7e":68,"imdb\u7684\u6570\u6910\u96c6":68,"import":[3,5,9,10,16,23,31,32,45,53,60,61,62,66,68,69],"include\u4e0b\u9700\u8981\u5305\u542bcbla":33,"include\u4e0b\u9700\u8981\u5305\u542bmkl":33,"init_hook\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":3,"instance\u4e0e\u751f\u6210\u6570\u636e\u96c6\u65f6":26,"instance\u5305\u6db5\u4e24\u4e2a\u503c":26,"instance\u662f\u4e00\u6a21\u4e00\u6837\u7684":26,"int":[3,7,9,10,11,12,15,16,17,20,27,28,29,39,44,52,64,66,67],"interface\u6587\u4ef6\u7684\u5199\u6cd5\u975e\u5e38":27,"issue\u7f16\u53f7":43,"job\u542f\u52a8\u540e\u4f1a\u521b\u5efa\u8fd9\u4e9bpod\u5e76\u5f00\u59cb\u6267\u884c\u4e00\u4e2a\u7a0b\u5e8f":54,"job\u5b8c\u6210\u8f6c\u6362":26,"job\u6216\u8005\u5e94\u7528\u7a0b\u5e8f\u5728\u5bb9\u5668\u4e2d\u8fd0\u884c\u65f6\u751f\u6210\u7684\u6570\u636e\u4f1a\u5728\u5bb9\u5668\u9500\u6bc1\u65f6\u6d88\u5931":54,"job\u662f\u672c\u6b21\u8bad\u7ec3\u5bf9\u5e94\u7684job":56,"job\u7684\u540d\u5b57":56,"kubernetes\u4e3a\u8fd9\u6b21\u8bad\u7ec3\u521b\u5efa\u4e863\u4e2apod\u5e76\u4e14\u8c03\u5ea6\u5230\u4e863\u4e2anode\u4e0a\u8fd0\u884c":56,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":46,"kubernetes\u5355\u673a\u8bad\u7ec3":46,"kubernetes\u53ef\u4ee5\u5728\u7269\u7406\u673a\u6216\u865a\u62df\u673a\u4e0a\u8fd0\u884c":54,"kubernetes\u53ef\u4ee5\u901a\u8fc7yaml\u6587\u4ef6\u6765\u521b\u5efa\u76f8\u5173\u5bf9\u8c61":56,"kubernetes\u5c31\u4f1a\u521b\u5efa3\u4e2apod\u4f5c\u4e3apaddlepaddle\u8282\u70b9\u7136\u540e\u62c9\u53d6\u955c\u50cf":56,"kubernetes\u63d0\u4f9b\u4e86\u591a\u79cd\u96c6\u7fa4\u90e8\u7f72\u7684\u65b9\u6848":54,"kubernetes\u652f\u6301\u591a\u79cdvolum":54,"kubernetes\u6709job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u652f\u6301":55,"kubernetes\u96c6\u7fa4\u5c31\u662f\u7531node\u8282\u70b9\u4e0emaster\u8282\u70b9\u7ec4\u6210\u7684":54,"label\u662f\u539f\u59cb\u6570\u636e\u4e2d\u5bf9\u4e8e\u6bcf\u4e00\u53e5\u8bdd\u7684\u5206\u7c7b\u6807\u7b7e":39,"labels\u662f\u6bcf\u7ec4\u5185\u6bcf\u4e2a\u53e5\u5b50\u7684\u6807\u7b7e":39,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":38,"layer\u62ff\u5230\u7684\u7528\u6237\u8f93\u5165":41,"layer\u7c7b\u53ef\u4ee5\u81ea\u52a8\u8ba1\u7b97\u4e0a\u9762\u7684\u5bfc\u6570":44,"layer\u91cc\u9762\u53ef\u4ee5\u5b9a\u4e49\u53c2\u6570\u5c5e\u6027":53,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u548catlas\u4e24\u4e2a\u5e93":33,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u5e93":33,"lib\u4e0b\u9700\u8981\u5305\u542bopenblas\u5e93":33,"lib\u76ee\u5f55\u4e0b\u9700\u8981\u5305\u542bmkl_cor":33,"list\u4e2d\u7684\u6bcf\u4e00\u884c\u90fd\u4f20\u9012\u7ed9process\u51fd\u6570":3,"list\u4f5c\u4e3a\u68c0\u67e5\u5217\u8868":30,"list\u5199\u5165\u90a3\u4e2a\u6587\u672c\u6587\u4ef6\u7684\u5730\u5740":3,"list\u548ctest":2,"list\u5982\u4e0b\u6240\u793a":52,"list\u5b58\u653e\u5728\u672c\u5730":2,"list\u6216\u8005test":53,"list\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b\u5217\u8868":52,"long":[10,11,16,17,20],"lstm\u67b6\u6784\u7684\u6700\u5927\u4f18\u70b9\u662f\u5b83\u53ef\u4ee5\u5728\u957f\u65f6\u95f4\u95f4\u9694\u5185\u8bb0\u5fc6\u4fe1\u606f":68,"lstm\u6a21\u578b":64,"lstm\u6a21\u578b\u914d\u7f6e":64,"lstm\u7f51\u7edc\u7c7b\u4f3c\u4e8e\u5177\u6709\u9690\u85cf\u5c42\u7684\u6807\u51c6\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":68,"memory\u4e0d\u80fd\u72ec\u7acb\u5b58\u5728":41,"memory\u4e5f\u53ef\u4ee5\u5177\u6709":42,"memory\u4e5f\u53ef\u4ee5\u662f\u5e8f\u5217":42,"memory\u53ea\u80fd\u5728":41,"memory\u53ef\u4ee5\u7f13\u5b58\u4e0a\u4e00\u4e2a\u65f6\u523b\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u8f93\u51fa":39,"memory\u6307\u5411\u4e00\u4e2alay":41,"memory\u662f\u5728\u5355\u6b65\u51fd\u6570\u4e2d\u5faa\u73af\u4f7f\u7528\u7684\u72b6\u6001":42,"memory\u662fpaddlepaddle\u5b9e\u73b0rnn\u65f6\u5019\u4f7f\u7528\u7684\u4e00\u4e2a\u6982\u5ff5":39,"memory\u7684":42,"memory\u7684\u521d\u59cb\u72b6\u6001":41,"memory\u7684\u65f6\u95f4\u5e8f\u5217\u957f\u5ea6\u4e00\u81f4\u7684\u60c5\u51b5":39,"memory\u7684\u66f4\u591a\u8ba8\u8bba\u8bf7\u53c2\u8003\u8bba\u6587":41,"memory\u7684\u8f93\u51fa\u5b9a\u4e49\u5728":42,"memory\u7684i":41,"memory\u9ed8\u8ba4\u521d\u59cb\u5316\u4e3a0":41,"mkl\u7684\u8def\u5f84":33,"mkl_sequential\u548cmkl_intel_lp64\u4e09\u4e2a\u5e93":33,"mnist\u662f\u4e00\u4e2a\u5305\u542b\u670970":3,"mode\u548cattent":69,"mode\u7684python\u51fd\u6570":69,"model\u505a\u5206\u652f\u7ba1\u7406":30,"model\u53ef\u4ee5\u901a\u8fc7":5,"model\u6765\u5b9e\u73b0\u624b\u5199\u8bc6\u522b\u7684\u9884\u6d4b\u4ee3\u7801":5,"movielens\u6570\u636e\u96c6":66,"name\u662f\u4f53\u88c1":66,"name\u662f\u5e74\u9f84":66,"name\u662f\u6027\u522b":66,"name\u662f\u7535\u5f71\u540d":66,"name\u662f\u804c\u4e1a":66,"name\u7ec4\u5408\u53ef\u4ee5\u627e\u5230\u672c\u6b21\u8bad\u7ec3\u9700\u8981\u7684\u6587\u4ef6\u8def\u5f84":56,"ndarray\u7c7b\u578b\u7684\u503c\u548c\u6574\u578b\u7684\u503c":26,"new":[10,16,20,24,29,43,44],"nfs\u7684\u90e8\u7f72\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":54,"nmt\u6a21\u578b\u53d7\u5236\u4e8e\u6e90\u8bed\u53e5\u7684\u7f16\u7801":69,"noavx\u7248\u672c":36,"normalization\u5c42":62,"normalization\u5c42\u7684\u53c2\u6570":62,"note\u7684\u4e66\u5199":30,"notebook\u662f\u4e00\u4e2a\u5f00\u6e90\u7684web\u7a0b\u5e8f":34,"null":[10,16,44,50,66],"num_gradient_servers\u53c2\u6570":56,"openblas\u7684\u8def\u5f84":33,"operator\u7684\u6982\u5ff5":53,"out\u4e0b\u5305\u542b":61,"out\u7684\u6587\u4ef6\u5939":61,"outer_mem\u662f\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":39,"output\u6587\u4ef6\u5939\u5b58\u653e\u8bad\u7ec3\u7ed3\u679c\u4e0e\u65e5\u5fd7":56,"packages\u91cc\u9762":31,"packages\u91cc\u9762\u7684python\u5305":31,"paddepaddle\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":33,"paddle\u4e00\u4e2a\u52a8\u6001\u5e93\u53ef\u4ee5\u5728\u4efb\u4f55linux\u7cfb\u7edf\u4e0a\u8fd0\u884c":27,"paddle\u4e2d\u7684\u4e00\u6761pass\u8868\u793a\u8bad\u7ec3\u6570\u636e\u96c6\u4e2d\u6240\u6709\u7684\u6837\u672c\u4e00\u6b21":69,"paddle\u4e2d\u7ecf\u5e38\u4f1a\u5c06\u65f6\u95f4\u5e8f\u5217\u6210\u4e3a":39,"paddle\u4e8c\u8fdb\u5236\u5728\u8fd0\u884c\u65f6\u6355\u83b7\u4e86\u6d6e\u70b9\u6570\u5f02\u5e38":31,"paddle\u4f7f\u7528git":30,"paddle\u5185\u5d4c\u7684python\u89e3\u91ca\u5668\u548c\u5916\u90e8\u4f7f\u7528\u7684python\u5982\u679c\u7248\u672c\u4e0d\u540c":27,"paddle\u5185\u90e8\u7684\u7c7b\u4e3ac":27,"paddle\u5f00\u53d1\u8fc7\u7a0b\u4f7f\u7528":30,"paddle\u66ff\u6362\u6210":34,"paddle\u6bcf\u6b21\u53d1\u65b0\u7684\u7248\u672c":30,"paddle\u6bcf\u6b21\u53d1\u7248\u672c\u9996\u5148\u8981\u4fdd\u8bc1paddl":30,"paddle\u7684\u4e3b\u7248\u672c\u5e93\u9075\u5faa":30,"paddle\u7684\u5404\u7248\u672c\u955c\u50cf\u53ef\u4ee5\u53c2\u8003":55,"paddle\u7684\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0\u5305\u62ec\u4e00\u4e0b\u51e0\u4e2a\u65b9\u9762":27,"paddle\u7684\u7c7b\u578b\u5168\u90e8\u9000\u5316\u6210":28,"paddle\u7684\u94fe\u63a5\u65b9\u5f0f\u6bd4\u8f83\u590d\u6742":27,"paddle\u7684c":28,"paddle\u7684dock":55,"paddle\u7684docker\u5f00\u53d1\u955c\u50cf\u5e26\u6709\u4e00\u4e2a\u901a\u8fc7":34,"paddle\u8bad\u7ec3\u4efb\u52a1":26,"paddle\u8def\u5f84\u4e0b":28,"paddle\u955c\u50cf":55,"paddle\u9700\u8981\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3":27,"paddle\u9700\u8981\u66b4\u9732\u7684api\u5f88\u591a":28,"paddle\u9759\u6001\u5e93\u94fe\u63a5\u590d\u6742":27,"paddle_\u7c7b\u578b\u540d":28,"paddle_\u7c7b\u578b\u540d_\u51fd\u6570\u540d":28,"paddlepaddle\u4e2d":[38,41],"paddlepaddle\u4e2d\u7684\u4e00\u4e2apass\u610f\u5473\u7740\u5bf9\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u6837\u672c\u8fdb\u884c\u4e00\u6b21\u8bad\u7ec3":68,"paddlepaddle\u4e2d\u7684\u8bb8\u591alayer\u5e76\u4e0d\u5728\u610f\u8f93\u5165\u662f\u5426\u662f\u65f6\u95f4\u5e8f\u5217":39,"paddlepaddle\u4f1a\u5728\u8c03\u7528\u8bfb\u53d6\u6570\u636e\u7684python\u811a\u672c\u4e4b\u524d":64,"paddlepaddle\u4f7f\u7528\u5747\u503c0":31,"paddlepaddle\u4f7f\u7528avx":31,"paddlepaddle\u4f7f\u7528swig\u5bf9\u5e38\u7528\u7684\u9884\u6d4b\u63a5\u53e3\u8fdb\u884c\u4e86\u5c01\u88c5":5,"paddlepaddle\u4fdd\u7559\u6dfb\u52a0\u53c2\u6570\u7684\u6743\u529b":3,"paddlepaddle\u5148\u4ece\u4e00\u4e2a\u6587\u4ef6\u5217\u8868\u91cc\u83b7\u5f97\u6570\u636e\u6587\u4ef6\u5730\u5740":32,"paddlepaddle\u5305\u62ec\u5f88\u591a\u635f\u5931\u51fd\u6570\u548c\u8bc4\u4f30\u8d77":53,"paddlepaddle\u53d1\u5e03\u65b0\u7248\u672c\u7684\u65f6\u5019\u90fd\u4f1a\u53d1\u5e03\u5bf9\u5e94\u7248\u672c\u7684\u751f\u4ea7\u955c\u50cf\u4ee5\u53ca\u5f00\u53d1\u955c\u50cf":34,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528cudnn":33,"paddlepaddle\u53ef\u4ee5\u6267\u884c\u7528\u6237\u7684python\u811a\u672c\u7a0b\u5e8f\u6765\u8bfb\u53d6\u5404\u79cd\u683c\u5f0f\u7684\u6570\u636e\u6587\u4ef6":64,"paddlepaddle\u53ef\u4ee5\u6bd4\u8f83\u7b80\u5355\u7684\u5224\u65ad\u54ea\u4e9b\u8f93\u51fa\u662f\u5e94\u8be5\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684":39,"paddlepaddle\u53ef\u4ee5\u901a\u8fc7\u8be5\u673a\u5236\u5224\u65ad\u662f\u5426\u5df2\u7ecf\u6536\u96c6\u9f50\u6240\u6709\u7684\u68af\u5ea6":44,"paddlepaddle\u5728\u5b9e\u73b0rnn\u7684\u65f6\u5019":39,"paddlepaddle\u591a\u673a\u91c7\u7528\u7ecf\u5178\u7684":53,"paddlepaddle\u5b58\u7684\u662f\u6709\u503c\u4f4d\u7f6e\u7684\u7d22\u5f15":3,"paddlepaddle\u5b9a\u4e49\u7684\u53c2\u6570":3,"paddlepaddle\u5c06\u4ee5\u8bbe\u7f6e\u53c2\u6570\u7684\u65b9\u5f0f\u6765\u8bbe\u7f6e":64,"paddlepaddle\u5c06\u5728\u89c2\u6d4b\u6570\u636e\u96c6\u4e0a\u8fed\u4ee3\u8bad\u7ec330\u8f6e":32,"paddlepaddle\u5c06\u6bcf\u4e2a\u6a21\u578b\u53c2\u6570\u4f5c\u4e3a\u4e00\u4e2anumpy\u6570\u7ec4\u5355\u72ec\u5b58\u4e3a\u4e00\u4e2a\u6587\u4ef6":32,"paddlepaddle\u5c06train":3,"paddlepaddle\u5f00\u53d1\u955c\u50cf\u5c31\u88ab\u6784\u5efa\u5b8c\u6bd5\u4e86":34,"paddlepaddle\u63d0\u4f9b\u4e13\u7528\u7684":26,"paddlepaddle\u63d0\u4f9b\u4e86\u57fa\u4e8e":53,"paddlepaddle\u63d0\u4f9b\u4e86\u5f88\u591a\u4f18\u79c0\u7684\u5b66\u4e60\u7b97\u6cd5":32,"paddlepaddle\u63d0\u4f9b\u4e86ubuntu":36,"paddlepaddle\u63d0\u4f9b\u6570\u4e2a\u9884\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6765\u8fdb\u884c\u5b89\u88c5":35,"paddlepaddle\u652f\u6301\u4ee5\u4e0b\u4efb\u610f\u4e00\u79cdblas\u5e93":33,"paddlepaddle\u652f\u6301\u5927\u91cf\u7684\u8ba1\u7b97\u5355\u5143\u548c\u4efb\u610f\u6df1\u5ea6\u7684\u7f51\u7edc\u8fde\u63a5":32,"paddlepaddle\u652f\u6301\u975e\u5e38\u591a\u7684\u4f18\u5316\u7b97\u6cd5":31,"paddlepaddle\u652f\u6301sparse\u7684\u8bad\u7ec3":31,"paddlepaddle\u6587\u6863\u4f7f\u7528":45,"paddlepaddle\u6587\u6863\u9700\u8981\u51c6\u5907\u7684\u73af\u5883\u76f8\u5bf9\u8f83\u590d\u6742":45,"paddlepaddle\u662f\u4e00\u4e2a\u6700\u65e9\u7531\u767e\u5ea6\u79d1\u5b66\u5bb6\u548c\u5de5\u7a0b\u5e08\u5171\u540c\u7814\u53d1\u7684\u5e76\u884c\u5206\u5e03\u5f0f\u6df1\u5ea6\u5b66\u4e60\u5e73\u53f0":0,"paddlepaddle\u662f\u4e00\u4e2a\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6":53,"paddlepaddle\u662f\u6e90\u4e8e\u767e\u5ea6\u7684\u4e00\u4e2a\u6df1\u5ea6\u5b66\u4e60\u5e73\u53f0":32,"paddlepaddle\u6bcf\u6b21\u53d1\u5e03\u65b0\u7248\u672c\u90fd\u4f1a\u53d1\u5e03\u5bf9\u5e94\u7684\u5f00\u53d1\u955c\u50cf\u4f9b\u5f00\u53d1\u8005\u76f4\u63a5\u4f7f\u7528":34,"paddlepaddle\u7684\u5185\u5b58\u5360\u7528\u4e3b\u8981\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u65b9\u9762":31,"paddlepaddle\u7684\u53c2\u6570\u4f7f\u7528\u540d\u5b57":31,"paddlepaddle\u7684\u6570\u636e\u5305\u62ec\u56db\u79cd\u4e3b\u8981\u7c7b\u578b":3,"paddlepaddle\u7684\u6587\u6863\u5305\u62ec\u82f1\u6587\u6587\u6863":45,"paddlepaddle\u7684\u6587\u6863\u6784\u5efa\u6709\u76f4\u63a5\u6784\u5efa\u548c\u57fa\u4e8edocker\u6784\u5efa\u4e24\u79cd\u65b9\u5f0f":45,"paddlepaddle\u7684\u6e90\u7801":43,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":35,"paddlepaddle\u7684bas":44,"paddlepaddle\u7684docker\u5bb9\u5668\u4f7f\u7528\u65b9\u5f0f":35,"paddlepaddle\u7684trainer\u8fdb\u7a0b\u91cc\u5185\u5d4c\u4e86python\u89e3\u91ca\u5668":53,"paddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":39,"paddlepaddle\u76ee\u524d\u552f\u4e00\u5b98\u65b9\u652f\u6301\u7684\u8fd0\u884c\u7684\u65b9\u5f0f\u662fdocker\u5bb9\u5668":34,"paddlepaddle\u76ee\u524d\u5df2\u7ecf\u5f00\u653e\u6e90\u7801":0,"paddlepaddle\u76ee\u524d\u63d0\u4f9b\u4e24\u79cd\u53c2\u6570\u521d\u59cb\u5316\u7684\u65b9\u5f0f":31,"paddlepaddle\u8c03\u7528process\u51fd\u6570\u6765\u8bfb\u53d6\u6570\u636e":64,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u68af\u5ea6\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":41,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u8bef\u5dee\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":41,"paddlepaddle\u955c\u50cf\u9700\u8981\u63d0\u4f9b":56,"paddlepaddle\u9700\u8981\u7528\u6237\u5728\u7f51\u7edc\u914d\u7f6e":2,"paddlepaddle\u9700\u8981\u7684\u6240\u6709\u7f16\u8bd1\u5de5\u5177":34,"pass\u4e2a\u6a21\u578b\u5230\u7b2c":50,"pass\u5230":69,"pass\u5c06\u4e0d\u8d77\u4f5c\u7528":50,"pass\u8f6e\u5f00\u59cb\u8bad\u7ec3":50,"pass\u8f6e\u7684\u6a21\u578b\u7528\u4e8e\u6d4b\u8bd5":50,"passes\u8f6e":50,"patch\u53f7":30,"patch\u53f7\u52a0\u4e00":30,"path\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b":52,"period\u4e2a\u6279\u6b21\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":50,"period\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":50,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u53c2\u6570\u7edf\u8ba1":50,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u7b26\u53f7":50,"period\u4e2abatch\u5904\u7406\u7684\u5f53\u524d\u635f\u5931":68,"period\u4e2abatch\u7684\u5206\u7c7b\u9519\u8bef":68,"period\u6574\u9664":50,"period\u8f6e\u4fdd\u5b58\u8bad\u7ec3\u53c2\u6570":50,"pod\u4e2d\u7684\u5bb9\u5668\u5171\u4eabnet":54,"pod\u662fkubernetes\u7684\u6700\u5c0f\u8c03\u5ea6\u5355\u5143":54,"pooling\u5bf9\u7279\u5f81\u56fe\u4e0b\u91c7\u6837":61,"process\u51fd\u6570\u4f1a\u7528yield\u8bed\u53e5\u8f93\u51fa\u8fd9\u6761\u6570\u636e":64,"prod\u7684\u751f\u4ea7\u955c\u50cf":34,"pserver\u8fdb\u7a0b\u7528\u4e8e\u534f\u8c03\u591a\u4e2atrainer\u8fdb\u7a0b\u4e4b\u95f4\u7684\u901a\u4fe1":53,"public":[20,44,55,68],"py_paddle\u91cc\u9762\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5de5\u5177\u7c7b":5,"pydataprovider2\u4f1a\u5c3d\u53ef\u80fd\u591a\u7684\u4f7f\u7528\u5185\u5b58":3,"pydataprovider2\u63d0\u4f9b\u4e86\u4e24\u79cd\u7b80\u5355\u7684cache\u7b56\u7565":3,"pydataprovider2\u662fpaddlepaddle\u4f7f\u7528python\u63d0\u4f9b\u6570\u636e\u7684\u63a8\u8350\u63a5\u53e3":3,"pydataprovider2\u7684\u4f7f\u7528":[2,4,31,42,53,64,66],"pydataprovider\u4f7f\u7528\u7684\u662f\u5f02\u6b65\u52a0\u8f7d":31,"python\u4ee3\u7801\u5c06\u968f\u673a\u4ea7\u751f2000\u4e2a\u89c2\u6d4b\u70b9":32,"python\u5305":45,"python\u53ef\u4ee5\u89e3\u9664\u6389\u5185\u90e8\u53d8\u91cf\u7684\u5f15\u7528":3,"python\u5c01\u88c5\u7684\u5b9e\u73b0\u4f7f\u5f97\u6211\u4eec\u53ef\u4ee5\u5728\u914d\u7f6e\u6587\u4ef6\u4e2d\u4f7f\u7528\u65b0\u5b9e\u73b0\u7684\u7f51\u7edc\u5c42":44,"python\u811a\u672c\u91cc\u5b9a\u4e49\u4e86\u6a21\u578b\u914d\u7f6e":53,"query\u6539\u5199":69,"rate\u4e3a0":69,"rate\u4e3a5":69,"rate\u88ab\u8bbe\u7f6e\u4e3a0":61,"reader\u7684\u4f7f\u7528\u65b9\u5f0f\u90fd\u662f\u4e00\u81f4\u7684":26,"reader\u8f93\u51fa\u7684data":26,"recommendation\u6587\u4ef6\u5939\u5185\u5b58\u653e\u8bad\u7ec3\u6587\u4ef6":56,"release\u9875\u9762":30,"repo\u7684\u6839\u76ee\u5f55":34,"research\u5b9e\u9a8c\u5ba4\u641c\u96c6\u6574\u7406":65,"resnet\u6a21\u578b":63,"return":[3,10,11,16,17,19,20,22,23,26,32,39,42,44,56,62,64,66],"rnn\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":39,"rnn\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u901a\u8fc7\u4e86\u4e00\u4e2alstm\u7f51\u7edc":39,"rnn\u603b\u662f\u5f15\u7528\u4e0a\u4e00\u65f6\u523b\u9884\u6d4b\u51fa\u7684\u8bcd\u7684\u8bcd\u5411\u91cf":41,"rnn\u6a21\u578b":64,"rnn\u76f8\u5173\u6a21\u578b":46,"s3\u4e4b\u7c7b\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u4e4b\u4e0a":26,"search\u7684\u65b9\u6cd5":50,"sentences\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684\u6570\u636e":39,"seq\u53c2\u6570\u5fc5\u987b\u4e3afals":41,"server\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":50,"server\u4e4b\u4e0a":25,"server\u4e4b\u95f4\u7684\u7f51\u7edc\u5e26\u5bbd":25,"server\u4f1a\u6682\u505c\u53c2\u6570\u66f4\u65b0\u5e76\u7b49\u5f85":25,"server\u4f1a\u83b7\u53d6parameters\u5185\u5b58\u7684":25,"server\u5185\u5b58\u4e2d\u7684\u6a21\u578b\u6570\u636e\u7684\u5b8c\u6574\u955c\u50cf":25,"server\u540c\u6b65\u7684\u4fdd\u5b58\u4e00\u4e2a\u7279\u5b9a\u65f6\u95f4\u70b9\u7684\u5168\u5c40\u68c0\u67e5\u70b9":25,"server\u5728\u96c6\u7fa4\u4e2d\u542f\u52a8\u540e":25,"server\u6545\u969c\u540e\u88abkubernetes\u91cd\u65b0\u542f\u52a8":25,"server\u6b64\u65f6\u8fd8\u9700\u8981\u901a\u8fc7\u7f51\u7edc\u8bbf\u95ee\u5206\u5e03\u5f0f\u5b58\u50a8\u4ee5\u4fdd\u5b58\u5feb\u7167":25,"server\u751f\u6210\u4e00\u4e2auuid":25,"server\u7684\u5355\u70b9\u6216\u591a\u70b9\u540c\u65f6\u6545\u969c":25,"server\u7684\u6570\u636e\u5feb\u7167":25,"server\u7684\u68c0\u67e5\u70b9\u5404\u81ea\u72ec\u7acb\u4fdd\u5b58":25,"server\u7b2c\u4e00\u6b21\u542f\u52a8\u6216\u4efb\u610f\u65f6\u95f4paramet":25,"sh\u6765\u8bad\u7ec3\u6a21\u578b":61,"sh\u6765\u8fdb\u884c\u6784\u5efa":45,"sh\u8c03\u7528\u4e86":62,"short":[10,11,16,17],"simd\u6307\u4ee4\u63d0\u9ad8cpu\u6267\u884c\u6548\u7387":31,"size\u4e3a1":69,"size\u4e3a50":69,"size\u4e3a512":50,"size\u53ef\u80fd\u4f1a\u5bf9\u8bad\u7ec3\u7ed3\u679c\u4ea7\u751f\u5f71\u54cd":31,"size\u5927\u5c0f\u4e3a128":68,"size\u662f3":69,"size\u672c\u8eab\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u8d85\u53c2\u6570":31,"size\u7684\u503c":3,"softmax\u5c42":60,"softmax\u6fc0\u6d3b\u7684\u8f93\u51fa\u7684\u548c\u603b\u662f1":44,"sparse\u8bad\u7ec3\u9700\u8981\u8bad\u7ec3\u7279\u5f81\u662f":31,"srl\u4f5c\u4e3a\u5f88\u591a\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d\u7684\u4e2d\u95f4\u6b65\u9aa4\u662f\u5f88\u6709\u7528\u7684":67,"ssh\u65b9\u5f0f\u7684\u4e00\u4e2a\u4f18\u70b9\u662f\u6211\u4eec\u53ef\u4ee5\u4ece\u591a\u4e2a\u7ec8\u7aef\u8fdb\u5165\u5bb9\u5668":34,"ssh\u8fdb\u5165\u5bb9\u5668":34,"static":[10,28],"step\u51fd\u6570\u4e2d\u7684memori":41,"step\u51fd\u6570\u5185\u90e8\u53ef\u4ee5\u81ea\u7531\u7ec4\u5408paddlepaddle\u652f\u6301\u7684\u5404\u79cdlay":41,"subseq\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":38,"super":44,"swig\u652f\u6301\u7684\u8bed\u8a00\u6216\u8005\u89e3\u91ca\u5668\u6709\u5c40\u9650":27,"swig\u66b4\u9732\u7684\u63a5\u53e3\u4fdd\u7559\u4e86c":27,"swig\u751f\u6210\u7684\u4ee3\u7801\u4e0d\u80fd\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":27,"swig\u76f4\u63a5\u8bfb\u53d6c":27,"swig\u9700\u8981\u5199\u4e00\u4e2ainterface\u6587\u4ef6":27,"swig_paddle\u4e2d\u7684\u9884\u6d4b\u63a5\u53e3\u7684\u53c2\u6570\u662f\u81ea\u5b9a\u4e49\u7684c":5,"switch":28,"t\u6307\u5b9a\u4e86\u751f\u6210\u7684\u955c\u50cf\u7684\u540d\u5b57":34,"tag\u4e3a":30,"test\u548cgen\u8fd9\u4e09\u4e2a\u6587\u4ef6\u5939\u662f\u56fa\u5b9a\u7684":69,"tests\u7684paddlepaddl":43,"tflops\u4e86":47,"trainer\u8fdb\u7a0b\u4f1a\u8c03\u7528dataprovider\u51fd\u6570\u8fd4\u56de\u6570\u636e":53,"trainer\u8fdb\u7a0b\u53ef\u4ee5\u5229\u7528\u8fd9\u4e2a\u89e3\u91ca\u5668\u6267\u884cpython\u811a\u672c":53,"true":[7,9,10,11,12,15,16,17,19,20,22,23,29,31,39,42,44,52,56,62,66,67,68,69],"true\u8868\u793a\u53cd\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":42,"try":[12,18,24,29,31,66],"type\u5b57\u6bb5\u5747\u4e0d\u5c3d\u76f8\u540c":28,"type\u662fon":66,"ubuntu\u5b89\u88c5\u5305\u7684\u529f\u80fd\u6b63\u786e\u6027":30,"ubuntu\u7684deb\u5b89\u88c5\u5305\u7b49":35,"ubuntu\u90e8\u7f72paddlepaddl":35,"update\u53c2\u6570\u65f6\u624d\u6709\u6548":50,"utf8\u7f16\u7801":60,"uts\u7b49linux":54,"v2\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2a\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":33,"vocab\u4e2d\u6bcf\u4e2a\u5207\u5206\u5355\u8bcd\u7684\u9884\u671f\u8bc4\u7ea7":68,"vocab\u505a\u4e3a\u5b57\u5178":68,"void":[27,28,44],"volume\u6302\u8f7d\u5230\u5bb9\u5668\u4e2d":54,"w0\u548c":62,"wbias\u662f\u9700\u8981\u5b66\u4e60\u7684\u53c2\u6570":62,"while":[7,9,15,20,29,56,69],"words\u5373\u4e3a\u8fd9\u4e2a\u6570\u636e\u4e2d\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":39,"words\u662f\u539f\u59cb\u6570\u636e\u4e2d\u7684\u6bcf\u4e00\u53e5\u8bdd":39,"x\u548cwindow":34,"x\u548cwindows\u4e0a\u7684\u786c\u4ef6\u8d44\u6e90":34,"yaml\u6587\u4ef6\u4e2d\u5404\u4e2a\u5b57\u6bb5\u7684\u5177\u4f53\u542b\u4e49":56,"yaml\u6587\u4ef6\u63cf\u8ff0\u4e86\u8fd9\u6b21\u8bad\u7ec3\u4f7f\u7528\u7684docker\u955c\u50cf":56,"zero\u4e09\u79cd\u64cd\u4f5c":50,AGE:55,AWS:[26,54,57,58],Abs:6,And:[9,10,12,16,18,20,29],But:[10,11,16,17,31],EOS:[10,16],For:[5,8,9,10,12,16,18,20,23,29,36,47],IDs:20,NFS:54,Not:[23,24],One:[9,10,11,17,22],QoS:55,TLS:23,That:[10,16,20,29],The:[3,7,8,9,10,11,12,14,15,16,17,18,20,22,23,24,28,29,44,56,64,66,67,69],Their:[10,16,24],Then:[10,16,66],There:[9,10,16,20,22,23,24],Use:[23,29],Used:[11,17],Using:[24,68],Will:[20,22],With:[10,11,16,17],Yes:34,___embedding_0__:56,___embedding_1__:56,__init__:44,__list_to_map__:66,__main__:[5,62],__meta__:66,__mse_cost_0__:56,__name__:[5,62],__rnn_step__:42,_link:[11,17],_proj:[10,16],_recurrent_group:42,_res2_1_branch1_bn:62,_source_language_embed:[42,60],_target_language_embed:[42,60],aaaaa:26,abc:[10,16],abl:[10,16,23],about:[10,11,16,17,67,69],abov:[3,10,16,23,24,47],abs:[11,17],acceler:25,accept:[20,23,29,67],access:[10,11,16,17,23],accord:[9,10,16],accrod:[11,17],accumul:24,accuraci:9,acl:68,aclimdb:68,across:[10,16],act:[10,11,16,17,31,32,39,42,53],act_typ:64,activ:[4,10,11,16,17,21,53,64],activi:[11,17],actual:[10,16],adadelta:[12,31,64],adagrad:[12,64],adam:[12,23,64,68],adamax:12,adamoptim:[53,60,64,68,69],adapt:[9,12,18],add:[10,11,16,17,20,43,66],add_input:44,add_test:44,add_to:[10,16],add_unittest_without_exec:44,addbia:44,added:[9,43],addit:[10,11,16,17],address:24,addrow:44,addto:10,addtolay:[10,16],adversari:29,affect:[10,16],afi:3,after:[10,16,20],again:[23,24],age:[20,56,66],agg_level:[10,16,38,39],aggregatelevel:[10,16,38,39],aircraft:69,airplan:61,aistat:[10,16],alex:[10,16,68],alexnet_pass1:52,alexnet_pass2:52,algo_hrnn_demo:39,algorithm:[10,12,16,18,60,68,69],align:[10,11,16,17,20,69],all:[3,7,9,10,12,15,16,18,23,24,28,41,56,66,67,68],alloc:[7,15],allow:[23,64],allow_only_one_model_on_one_gpu:[49,50,52],almost:[11,17],alreadi:[24,31,36],also:[9,10,11,16,17,20,23,29,47,64],alwai:[10,11,16,17,22,29,56],amazon:55,ambigu:29,amd64:54,analysi:[67,68],ani:[10,11,16,17,20,23,24,29],annot:67,annual:67,anoth:[10,16,23],anyth:[20,29,67],api:[20,22,23,30,47,53,56,59,68],api_pydataprovider2_sequential_model:8,api_trainer_config:66,apiserv:54,apivers:[54,55,56],apo:69,append:[3,22,29,39,42,56,66],append_gradient_machin:22,appleyard:47,appli:[10,11,16,17],applic:[47,55],approach:[10,16,64],apt:[36,61],arbitrari:[10,16],architectur:69,archiv:[20,27,28],arg:[3,8,9,10,11,12,16,17,20,31,32,56,61,62,64,66,67,68],arg_nam:[10,16],argpars:56,args_ext:56,argument:[3,8,10,16,20,56,64,66,67],argumentpars:56,argv:62,around:[3,10,16],arrai:[5,10,16,20,22,29,62],articl:43,arxiv:[10,11,16,17,68],ask:24,assert:5,assign:[10,16],associ:67,assum:[10,16],astyp:29,asyc:24,async:[12,24,49],async_count:[49,50],async_lagged_grad_discard_ratio:50,async_lagged_ratio_default:[49,50],async_lagged_ratio_min:[49,50],asynchron:24,atla:33,atlas_root:33,attenion:[11,17],attent:[10,11,16,17,69],attr:[7,11,15,16,17],attribut:[4,10,11,16,17,21],auc:[9,49],author:54,authorized_kei:48,auto:[27,44,47],automat:[10,16,23,69],automaticli:[10,16],automobil:61,avail:24,averag:[9,10,12,16,19,67],average_test_period:[49,50,67],averagepool:[10,16],avg:[13,38,47,64],avgcost:[9,64,66,68,69],avgpool:[10,16,64],avoid:[24,47],avx:34,await:55,awar:[23,24],azur:54,b2t:60,b363:55,b8561f5c79193550d64fa47418a9e67ebdd71546186e840f88de5026b8097465:55,back:24,backward:[10,11,14,16,17,44],backward_first:42,backwardactiv:44,bag:64,baidu:[10,16,55],balasubramanyan:68,bank:67,bardward:[11,17],bare:[54,55],barrierstatset:47,base:[6,12,16,17,19,20,23],baseactiv:[10,11,16],baseev:22,basematrix:44,basenam:9,basepool:13,basepoolingtyp:[10,11,16,17],baseregular:12,basestr:[7,8,9,10,11,15,16,17,19,22,66],bash:[34,43,45,55,56],basic:[10,16,22],batch:[9,10,11,12,16,17,18,20,22,23,24,26,32,48,55,56,61,62,64,66,68,69],batch_0:62,batch_id:22,batch_norm:[10,17],batch_norm_lay:11,batch_norm_typ:[10,16],batch_read:[26,29],batch_siz:[12,20,22,31,32,48,53,60,61,64,66,68,69],batchsiz:[10,16,44],bbbbb:26,bcd:[10,16],bcebo:20,beam:[10,16,42,69],beam_gen:[10,16,42],beam_search:[22,41,42],beam_siz:[10,16,42,49,50,52],beamsiz:69,becaus:[10,16,20,23,24,29,39],been:67,befor:[10,11,16,17,24,29,31,66],begin:[9,10,16],beginiter:[22,23],beginn:42,beginpass:[22,23],begintrain:23,being:[24,29],belong:[10,16],below:[10,16,20,24,29],benefit:[11,17],bengio:[10,16],bertolami:68,besid:[10,16,20],best:[8,10,16,66],best_model_path:67,besteffort:55,beta1:[12,18],beta2:[12,18],beta:62,better:[10,11,16,17],between:[10,12,16,18,24,28,69],bgr:62,bi_lstm:[11,17],bia:[10,11,12,16,17,18,44,53,62],bias:[10,16],bias_attr:[10,11,16,17,31,32,39,42],bias_param:31,bias_param_attr:[11,17],biases_:44,biasparameter_:44,biassiz:44,bidi:55,bidirect:[11,17,67,68],bidirectional_lstm_net:68,bigger:24,bilinear:[10,16],bilinear_interpol:[10,16],bilinearfwdbwd:47,bin:[34,36,48,54,55,56,66],binari:[9,10,16,20,64],bird:61,bitext:69,blank:[10,16],bleu:69,block:[10,16,25],block_expand:10,block_i:[10,16],block_x:[10,16],bn_attr:17,bn_bias_attr:[11,17],bn_layer_attr:11,bn_param_attr:[11,17],bollen:68,book:20,bool:[7,9,10,11,12,15,16,17,19,20,44,50,52,64,66,68],boot:[10,16,41,42],boot_bia:[10,16],boot_bias_active_typ:[10,16],boot_lay:[10,16,39,42],boot_with_const_id:[10,16],bos_id:[10,16,42],both:[7,10,11,14,15,16,17,23,24],bottom:64,bow:[26,64],branch:[10,16,23,30,43],brelu:6,brendan:68,broadcast:24,broken:43,browser:34,bryan:68,buf_siz:20,buffer:[20,29],buffered_read:29,bug:43,build:[20,34,43,56,57,58,69],build_dict:20,build_doc:45,built:47,bunk:68,c11:27,c703c041:43,c99:28,cach:[31,64,66,67],cache_pass_in_mem:[3,31,64,66,67],cachetyp:[3,31,64,66,67],calc_batch_s:[3,67],calcul:[9,10,11,12,16,17,18,24],call:[10,11,16,17,23,47,56,64],callabl:[10,16,20],callback:44,calrnn:39,caltech:61,can:[7,8,9,10,11,15,16,17,20,23,24,29,47,64],can_over_batch_s:[3,67],candid:[10,16],capi:27,capi_prvi:28,caption:69,card:48,care:[11,17,29],cat:[34,56,61,62,68],categori:[10,16,20,24,64],categorig:20,categoryfil:55,caus:24,ccb2_pc30:69,cde:[10,16],cdn:20,ceil:[10,16],ceil_mod:[10,16],cell:[10,11,16,17],ceph:[26,54],certif:[23,54],cffi:27,cfg:55,cgo:27,chain:20,challeng:24,chanc:23,chang:[10,16,20,29,43,68],channel:[10,16,47],char_bas:66,check:[3,20,31,36,43,44,50],check_align:20,check_eq:44,check_fail_continu:3,check_l:44,check_sparse_distribution_batch:[49,50],check_sparse_distribution_in_pserv:[49,50],check_sparse_distribution_ratio:[49,50],check_sparse_distribution_unbalance_degre:[49,50],checkgrad:50,checkgrad_ep:50,checkout:43,chunk:9,chunk_schem:9,chunktyp:9,cifar:61,cifar_vgg_model:61,claimnam:56,clang:[27,34,43],class1:68,class2:68,class_dim:68,classic:[10,16],classif:[10,16,64,68,69],classifi:[9,62],classification_cost:[31,39,53,61,64],classification_error_evalu:[64,68,69],classification_threshold:9,clean:[31,43],client:54,clip:[7,12,15,50,64],clock:[10,16],clone:34,close:[3,29,43],cloud:24,cluster:[23,24,48,54,56],cluster_train:48,cmake:[28,31,33,43,45,47],cmakelist:44,cmatrix:[27,28],cna:[10,16],cnn:[10,16,55,64],code:[3,5,20,23,29,34,44,55,66],coded_stream:31,codedinputstream:31,coeff:[10,16],coeffici:[10,16],collabor:24,collect:[10,16,20,22],collectbia:44,color:[61,62],colour:20,column:[9,10,16,29],com:[10,11,16,17,20,34,36,43,54,55,62],combin:[10,11,16,17,20,22,66],command:[44,52,55,56,57,58],commandlin:[47,56],comment:[11,17,39,56,64],commit:55,common_util:[48,66],commun:24,compil:36,complet:[10,11,16,17,20,22,24,55,56],complex:[11,17,29],complic:[10,16],compos:[20,23],composenotalign:20,comput:[10,11,16,17,23,24,67,68],conat:16,conat_lay:10,concat:[10,69],concat_lay:42,concaten:[11,17],concept:23,concern:23,concurr:24,condit:[10,16,55],conf:[5,10,16,31,39,48,56,60,62,69],conf_paddle_gradient_num:56,conf_paddle_n:56,conf_paddle_port:56,conf_paddle_ports_num:56,conf_paddle_ports_num_spars:56,config:[7,10,11,15,16,17,32,44,48,49,50,52,53,54,55,56,61,64,66,67,68,69],config_:50,config_arg:[49,50,52,62,64,67,68],config_bas:[16,17,22],config_fil:67,config_gener:[48,66],config_lay:44,config_pars:[5,44],configprotostr:31,configur:[8,10,16,44,60,62,69],conflict:43,confront:69,conll05st:67,conll:[20,67],connect:[11,17,55,64],connectionist:[10,16,68],connor:68,consequ:[10,11,16,17],consid:[9,10,12,16,18],consider:[11,17],consist:[10,16,20,29],construct:[23,66],construct_featur:66,consum:24,contact:24,contain:[3,8,9,10,11,16,17,19,20,22,23,55,56,64],context:[3,10,11,16,17,20,42,54],context_attr:[11,17],context_len:[10,11,16,17,64,66],context_proj_layer_nam:11,context_proj_nam:17,context_proj_param_attr:[11,17],context_project:[11,17,66],context_start:[10,11,16,17,64],continu:24,contrast:[10,16],control:[7,15,55,69],conv:[11,17],conv_act:[11,17],conv_attr:17,conv_batchnorm_drop_r:[11,17],conv_bias_attr:[11,17],conv_filter_s:[11,17],conv_layer_attr:11,conv_num_filt:[11,17],conv_op:[10,16],conv_pad:[11,17],conv_param_attr:[11,17],conv_shift:10,conv_strid:[11,17],conv_with_batchnorm:[11,17],conveni:23,convert:[3,5,20,26,29,64,66],convlay:[10,16],convolut:[10,11,16,17,53],convoper:[10,16],convtran:[10,16],convtranslay:[10,16],cool:43,coordin:24,copi:[22,23,66],core:[7,15,28],corpora:69,corpu:[20,67],correct:[9,10,16],correctli:[9,20],correspoind:23,correspond:23,corss_entropi:23,cos:[10,16],cos_sim:66,cosin:[10,16],cost:[12,18,22,23,32,53,66,68,69],cost_id:[10,16],could:[9,10,16,20,22,23,29],couldn:36,count:[24,29,47,50,52,55,66,67,68,69],counter:24,cpickl:66,cpp:[27,28,31,39,44,47,56,64,66,69],cpu:[3,7,10,15,16,30,34,36,47,52,55,56,67],cpuinfo:34,cpusparsematrix:28,crash:[24,47],creat:[7,10,15,16,20,22,23,24,43,44,55,56],create_bias_paramet:44,create_input_paramet:44,createfromconfigproto:5,creator:[20,26],crf:[10,67],crf_decod:10,critic:68,crlf:43,crop:62,crop_siz:62,cross:[10,16,31,64],cross_entropi:[16,23],cross_entropy_with_selfnorm:16,crt:54,csc:44,cslm:69,csr:44,ctc:10,ctc_layer:9,ctest:[34,43],ctrl:[48,66],ctx:67,ctx_0:67,ctx_0_slot:67,ctx_n1:67,ctx_n1_slot:67,ctx_n2:67,ctx_n2_slot:67,ctx_p1:67,ctx_p1_slot:67,ctx_p2:67,ctx_p2_slot:67,cub:61,cuda:[36,47,48,50],cuda_dir:[49,50],cuda_so:[31,34],cuda_visible_devic:31,cudaconfigurecal:47,cudadevicegetattribut:47,cudaeventcr:47,cudaeventcreatewithflag:47,cudafre:47,cudagetdevic:47,cudagetdevicecount:47,cudagetdeviceproperti:47,cudagetlasterror:47,cudahostalloc:47,cudalaunch:47,cudamalloc:47,cudamemcpi:47,cudaprofilerstart:47,cudaprofilerstop:47,cudaprofilestop:47,cudaruntimegetvers:47,cudasetdevic:47,cudasetupargu:47,cudastreamcr:47,cudastreamcreatewithflag:47,cudastreamsynchron:47,cudeviceget:47,cudevicegetattribut:47,cudevicegetcount:47,cudevicegetnam:47,cudevicetotalmem:47,cudnn:[10,16,19],cudnn_batch_norm:[10,16],cudnn_conv:[10,16],cudnn_conv_workspace_limit_in_mb:[49,50],cudnn_convt:[10,16],cudnn_dir:[49,50],cudnnv5:33,cudrivergetvers:47,cuinit:47,cumul:[10,16],curl:54,current:[3,10,12,16,24,54,64],current_word:42,currentcost:[9,64,66,68,69],currentev:[9,64,66,68,69],curv:23,custom:23,custom_batch_read:29,cutoff:20,cycl:24,cyclic:[10,16],cython:27,dalla:3,dan:67,darwin:54,dat:[20,48,66],data:[3,8,11,12,17,18,22,23,26,31,36,39,48,49,53,55,56,57,60,61,62,64,66,67,68,69],data_config:5,data_dir:[60,61,68,69],data_feed:20,data_fil:32,data_initialz:64,data_lay:[3,9,31,32,39,42,53,61,64,66,67],data_nam:20,data_provid:8,data_read:[20,29],data_reader_creator_random_imag:29,data_sourc:8,data_typ:[16,20],databas:20,datadim:[10,16],datalay:[10,16],dataprovid:[2,8,31,32,42,48,53,56,64,66,67],dataprovider_:64,dataprovider_bow:64,dataprovider_emb:64,dataproviderconvert:5,datasci:[10,16],dataset:[26,29,62,64,65,68,69],datasourc:[4,66],date:67,db_lstm:67,dcudnn_root:33,dead:24,deb:[36,43],decai:[12,18],decid:[23,29],declar:[10,11,16],decod:[10,11,16,17,41,42,69],decoder_boot:42,decoder_group_nam:42,decoder_input:42,decoder_mem:42,decoder_prev:[11,17],decoder_s:42,decoder_st:[11,17,42],deconv:[10,16],deconvolut:[10,16],decor:[3,20],deep:[10,16,47,61,62],deer:61,def:[3,5,10,16,20,23,26,29,31,32,39,42,44,56,62,64,66,67],defalut:[10,16],default_decor:56,default_devic:52,default_valu:52,defin:[3,8,9,10,11,16,17,20,23,24,29,31,64,66],define_py_data_sources2:[3,8,31,32,53,61,62,64,66],defini:69,definit:[20,24,60],degre:[10,16],del:66,delar:64,delet:[24,43],delimit:[9,66],demand:24,demo:[5,10,16,20,42,48,55,57,60,61,62,64,66,68,69],dens:[10,16,20,66],dense_vector:[3,5,16,20,32,66],dense_vector_sequ:20,depend:24,deriv:[14,23],descent:[10,12,16,24],describ:[23,55,64],descript:56,deseri:22,design:[10,16,20,27],desir:[24,55],detail:[7,10,11,12,15,16,17,18],detect:[9,43],determin:[10,16,20],dev:[31,34,43,61,66,69],develop:[30,43,69],deviat:[7,15],devic:[7,15,31,34,52],deviceid:52,devid:[10,16],dez:68,dfs:11,dict:[3,8,20,22,31,39,56,64,66,68,69],dict_dim:[31,39,68],dict_fil:[9,39,42,64,67],dict_nam:8,dict_siz:20,dictionai:64,dictionari:[3,8,9,10,16,20,22,23,31,64,69],dictrionari:64,dictsiz:69,diff:43,differ:[8,9,10,16,24],digit:[10,16],dim:[20,44,60,68],dimens:[10,14,16,19,20,31,64],dimes:[10,16],din:66,dir:[48,56,62,66,67,68,69],direct:[10,11,16,17],directli:[11,17],directori:[26,43,47,55],disabl:31,discard:[20,24,43,50],discount:[10,16],discov:24,discuss:23,disput:69,dist_train:23,distanc:9,distribut:[10,16,50,57,58],distribute_test:[49,50],disucss:23,divid:[12,18],diy_beam_search_prob_so:[49,50],dmkl_root:33,do_forward_backward:29,doc:[5,11,17,20,34,45,56],doc_cn:45,docker:[30,31,34,43,55,56,57,58],docker_build:23,docker_push:23,dockerfil:[34,43,56],dockerhub:34,document:[11,17],documentari:3,doe:[11,17,24,29],doesn:[7,10,15,16,20,23,29],dog:[61,62],don:[11,17,23,29],done:[10,11,16,17,24,47,56],dot:69,dot_period:[50,52,56,61,66,68,69],dotmuloper:[10,16],dotmulproject:[10,16],doubl:50,down:47,download:[20,24,55],download_cifar:61,doxygen:43,dpkg:36,dpython_execut:31,dpython_include_dir:31,dpython_librari:31,drop_rat:[7,15,53],dropout:[7,10,15,16],dropout_lay:10,dropout_r:[11,17],drwxr:55,dso_handl:36,dtoh:47,dtype:[5,32,62],dubai:69,due:66,dure:[3,10,16,24,64,69],dwith_c_api:28,dwith_gpu:33,dwith_profil:47,dwith_python:28,dwith_swig_pi:28,dwith_tim:47,dynam:[3,28,29],dynamic_cast:44,each:[3,9,10,16,19,20,22,24,29,64,66],each_feature_vector:14,each_meta:66,each_pixel_str:3,each_sequ:[10,16,38,39],each_time_step_output:14,each_timestep:[10,16,38],each_word:3,eaqual:[10,16],eas:[20,29],easi:29,easier:[23,29],easili:[23,29],ec2:54,echo:[31,34,66,68],edit:9,edu:[20,55,61],efg:[10,16],either:[10,16,20,22,23,64],electron:55,elem_dim:[10,16],element:[9,10,11,16,17,20,22,29],elif:[23,66],els:[10,16,23,34,39,44,62,64,66],emb1:39,emb2:39,emb:[31,39,55,64],emb_group:39,emb_sum:31,embed:[10,23,60,66,68],embedding_lay:[31,39,42,64,66],embedding_nam:[10,16,42],embedding_s:[10,16,42],empir:[10,16],emplace_back:44,empti:[9,20,24,32],enabl:[7,15,47],enable_grad_shar:[49,50],enable_parallel_vector:50,enc_proj:[11,17,42],enc_seq:[11,17],enc_vec:42,encod:[11,17,39,69],encoded_proj:[11,17,42],encoded_sequ:[11,17,42],encoded_vector:42,encoder1:39,encoder1_expand:39,encoder1_rep:39,encoder2:39,encoder2_rep:39,encoder_last:[10,16],encoder_proj:42,encoder_s:42,end:[9,10,16,29,42,43,67,68],end_pass:23,enditer:[22,23],endpass:[22,23],endtrain:23,english:[10,16,69],ensembl:[11,17],ensur:24,entir:[10,11,16,17],entri:20,entropi:[10,16,64],enumer:[10,14,31,64,66],enumerate_data_types_of_data_lay:20,env:[31,56],environ:[23,31,47,55],eol:43,eos:10,eos_id:[10,16,42],epsilon:[12,18],equal:[10,11,12,16,17,24,39],equat:[10,11,12,16,17,18],equival:[10,16,23],error:[7,9,10,12,15,16,18,23,31,50,64,66,68,69],error_clipping_threshold:[7,15],errorr:9,especi:[11,17],essenc:23,essenti:[10,16,23],estim:[10,16,23],eta:55,etc:[12,20,29,69],etcd:24,eth0:[48,53,56],eval:[9,64,66,68,69],eval_bleu:69,evalu:[4,10,16,22,47,48,53,66,68,69],evaluate_pass:68,evaluator_bas:9,even:[23,29],event:55,event_handl:[22,23],everi:[9,10,11,16,17,20,23,24],exactli:[9,10,11,16,17],exampl:[8,9,10,11,12,16,17,18,20,22,29,62,64],exc_path:31,exceed:[10,16],except:[20,66],excluded_chunk_typ:9,exconv:[10,16],exconvt:[10,16],exdb:20,exe:54,exec:34,execut:24,exist:[23,24,29,68],exit:55,exp:6,expand:[10,39],expand_a:[10,16,38,39],expand_lay:39,expand_level:[10,16,38],expandconvlay:[10,16],expandlevel:[10,16,38],expect:[10,16],expir:24,explain:[9,24],explan:[10,16],explicit:44,explicitli:23,explor:[10,16],exponenti:14,express:23,extend:66,extens:12,extent:28,extern:[27,28],extra:[10,11,15,16,17],extra_input:16,extra_lay:22,extraattr:[7,15,52,53],extraattribut:[16,17],extraattributenon:16,extract:[10,16,60,62,67],extract_fea_c:62,extract_fea_pi:62,extralayerattribut:[7,10,11,15,16],extralayeroutput:11,extrapaddl:17,extrem:[10,16],f1205:31,f120da72:55,fa0wx:55,fabric:48,facotr:[10,16],factor:[7,10,12,15,16,18],factori:27,fail:[24,31,36,50,55],failur:24,fake_imag:29,fals:[7,9,10,11,12,15,16,17,18,20,29,31,32,39,42,44,52,55,60,64,66,67,68,69],false_label:29,false_read:29,faq:59,fast:[10,16,47],faster:[10,11,16,17,24],fbd1f2bb71f4:55,fc1:[44,52],fc2:52,fc3:52,fc4:52,fc_act:[11,17],fc_attr:[11,17],fc_bias_attr:[11,17],fc_layer:[16,31,32,39,52,53,64,66],fc_layer_nam:11,fc_mat:22,fc_name:17,fc_param:31,fc_param_attr:[11,17],fclayer:44,fdata:[39,67],fea:62,fea_output:62,feat:68,featur:[3,10,14,16,20,43,62,64,66,67],feature_a:31,feature_b:31,feature_map:66,feed:[11,17,20,22,23],feeder:20,fernan:68,festiv:3,fetch:[20,24,43],few:[24,29],fewer:[10,16],fg0:[10,16],field:[10,16,22,66],figur:[23,60,69],file1:69,file2:69,file:[3,9,10,16,20,22,23,24,26,28,29,43,62,64,65,66,68,69],file_list:3,file_nam:[32,39,62,64,67],filenam:[3,26,31,66],filer:[10,16],fill:[10,16,24,64],filter:[10,16],filter_s:[10,11,16,17,53],filter_size_i:[10,16],find:[10,12,16,18,24,36],fine:[7,15],finish:[24,55],first:[10,16,20,23,24,64,66],first_seq:42,firstn:20,firstseen:55,fit:20,fix:[7,15,27,43],flag:20,flexiabl:29,flexibl:[10,11,16,17,23],flight:69,float32:[5,20,29,32,62],floor:[10,16],flow:30,fly:64,folder:[26,69],follow:[9,10,11,12,16,17,18,20,23,24,29,57,58,66],forbid:23,force_load:27,forget:[12,18,23],form:[11,12,17,18],format:[9,43,44],former:23,formula:[10,11,16,17],formular:[10,16],forward:[11,14,17,44],forwardactiv:44,forwardtest:5,found:[10,16],frame:9,framework:[23,64],free:20,french:69,frequenc:20,frequent:29,frog:61,from:[3,5,10,11,16,17,20,22,24,26,29,31,32,34,41,47,53,55,60,61,64,66,67,68,69],from_sequ:38,from_timestep:[10,16,38],fromfil:[29,32,62],fulfil:47,full:[10,16,24],full_matrix_project:[11,17,39,42,53],fulli:64,fullmatrixproject:[10,16],fully_matrix_project:[11,17],fullyconnectedlay:44,func:[3,20],further:[10,16],fusion:66,gain:[10,16],gamma:62,gan:23,gate:[10,11,16,17],gate_act:[10,11,16,17,39],gate_recurr:[10,16],gather:[10,16,66],gauss:[7,15],gcc:[27,34],gce:54,gcepersistentdisk:54,gdebi:36,gen:[10,16,69],gen_conf:69,gen_data:69,gen_result:69,gen_trans_fil:42,gender:[20,56,66],gener:[3,9,10,11,16,17,20,22,23,24,26,29,47,52,56,60,64,69],generated_word_embed:[10,16],generatedinput:[10,16,41,42],genr:[56,66],gereat:9,get:[3,10,11,16,17,20,22,36,44,55,61,64,66,67,68],get_batch_s:67,get_best_pass:68,get_config_arg:[52,64,66,68],get_data:[55,64,67],get_dict:20,get_embed:20,get_imdb:68,get_input_lay:44,get_model:62,get_movie_title_dict:20,get_output_attr:17,get_output_layer_attr:11,get_sample_from_lin:31,get_shap:22,get_word_dict:20,getbatchs:44,getenv:[23,56],gethostbynam:56,gethostnam:56,getidmap:56,getinput:44,getinputgrad:44,getinputvalu:44,getoutputgrad:44,getoutputvalu:44,getparameterptr:44,getpodlist:56,getsiz:44,gettranspos:44,getw:44,getweight:44,getwgrad:44,gildea:67,gist:[11,17],git:[30,34,43],github:[10,11,16,17,34,36,43,62],give:[3,24],given:[20,22,29,64],global:[7,12,15,23,24,47,66],global_learning_r:[7,15],globalstat:47,globalstatinfo:47,globe:3,glusterf:54,godoc:27,goe:[10,11,16,17,24],good:[10,16,29],goodfellow13:[10,16],googl:[23,31],googleapi:54,gpu:[7,10,12,15,16,19,30,31,34,36,47,52,62,67,68,69],gpu_id:[31,50,52],gpugpu_id:49,grab:24,grad:50,grad_share_block_num:[49,50],gradient:[7,9,10,12,15,16,18,22,24,50,56,64],gradient_clipping_threshold:[7,12,15,64,68],gradient_machin:[22,28],gradient_serv:53,gradientmachin:[5,22,28,56,66,69],gradientserv:53,gram:60,graph:[10,16,22,24],grave:68,greater:[10,16],grep:[34,68],groudtruth:42,ground:[9,10,16,69],group:[11,17],group_id:66,group_input:[39,42],grouplen:[20,65],gru:[10,16,64],gru_attr:17,gru_bias_attr:[11,17],gru_decod:42,gru_decoder_with_attent:42,gru_encoder_decod:[60,69],gru_layer_attr:11,gru_memori:[11,17],gru_siz:64,gru_step:[17,42],gru_step_lay:[11,42],grumemori:[11,17,42],gserver:[10,16,44],gsizex:47,guid:55,gur_group:[11,17],gzip:55,hadoop:23,handl:[23,29],handler:22,handwrit:68,harvest:64,has:[10,11,12,16,17,18,20,23,24,47,64,67],has_kei:22,hassubseq:39,have:[9,10,11,16,17,20,23,24,29],hdf:26,head:68,header:[28,32,62,66],height:[10,16,20,27,29,44],held:24,hello:23,help:[5,43],helper:[8,10,11,16,17],here:[7,10,11,15,16,17,20,23,24,29],heurist:[10,16,69],hidden1:53,hidden2:53,hidden:[10,11,16,17,31,66],hidden_a:31,hidden_b:31,hidden_dim:39,hidden_s:[11,17,66],hierach:41,hierarch:[10,16,39],high:[7,15],highest:20,highli:20,him:23,hint:5,hl_dso_load:36,hl_get_sync_flag:44,hold:[23,24],home:[26,48,55,56],hook2:39,hook:[39,66,67],horizont:[10,16],hors:61,horst:68,host:[48,55],hostnetwork:56,hostpath:[54,55,56],hot:64,hous:[3,20],how:[7,10,15,16,23,24],howardjohnson:39,howev:[11,17,29],howto:56,hpp:27,html:[20,61],htod:47,http:[10,11,16,17,20,34,36,43,54,55,61,62,65,69],huber:[10,16],huge:[10,16],huina:68,hyper:[10,16],hyperplan:20,i0601:66,i0706:69,i0719:69,i1116:56,i1117:47,ib0:48,ics:20,icwsm:68,id_input:[9,42],idea:[10,16,29],identityoffsetproject:[10,16],identityproject:[10,16],idmap:56,ids:[9,10,16,31,64,66],idx:44,ieee:68,ignor:[3,9,10,16],ijcnlp:68,illustr:24,ilsvrc:62,imag:[19,20,23,29,43,55,56,57,58,61,62,69],image_a:29,image_b:29,image_classif:61,image_fil:29,image_lay:29,image_list_provid:62,image_nam:23,image_path:29,image_provid:61,image_reader_cr:29,image_s:62,imagenet:26,imagepullpolici:56,imageri:[10,16],images_reader_cr:29,imdber:68,img:[3,10,16,53,61],img_conv:17,img_conv_lay:11,img_norm_typ:10,img_pool:17,img_pool_lay:11,img_siz:61,imgsiz:47,imgsizei:47,imgsizex:47,immutable_paramet:23,implement:[10,11,12,16,17,18,20,27,28],importerror:66,inarg:5,inc_path:31,includ:[10,11,16,17,20,23,27,28,43,47],incorrect:[10,16],increas:[24,31],incupd:44,inde:[20,29],independ:[10,16],index:[9,10,16,19,20,22,24,39,66],indexslot:[10,16],indic:[9,10,16],individu:24,industri:24,infer:[23,24,27],infiniband:48,info:[9,10,16,20,39,44,48,56],inform:[9,20],ininst:23,init:[7,15,44,56,66,67],init_hook:[39,64,66,67],init_hook_wrapp:8,init_model_path:[49,50,52,60,64,67],initi:[3,7,10,15,16,20,50,64],initial_max:[7,15,31],initial_mean:[7,10,15,16,31],initial_min:[7,15,31],initial_std:[7,10,15,16,31],initpaddl:5,inlcud:[11,17],inner:[31,39],inner_:39,inner_mem:39,inner_param_attr:[11,17],inner_rnn_output:39,inner_rnn_st:39,inner_rnn_state_:39,inner_step:39,inner_step_impl:39,input1:[10,11,16,17],input2:[10,16],input:[3,9,10,11,14,16,17,19,20,22,29,31,32,38,39,41,42,44,52,53,56,60,61,64,66,67,69],input_data:44,input_data_target:44,input_featur:14,input_fil:[32,67],input_hassub_sequence_data:44,input_id:[10,16],input_imag:[11,17,61],input_index:44,input_label:44,input_lay:[10,44],input_nam:23,input_sequence_data:44,input_sequence_label:44,input_sparse_float_value_data:44,input_sparse_non_value_data:44,input_t:44,input_typ:[31,32,39,42,64,66],inputdef:44,inputlayers_:44,inputtyp:20,insert:43,insid:[9,10,16,24,29],instal:[31,36,43,48,55,61,66],instanc:[10,12,16,24,26],instanti:24,instead:[10,16,19,29],int32:[50,53],integ:[3,9,10,16,20,27,64],integer_sequ:31,integer_valu:[3,20,31,39,64],integer_value_sequ:[3,20,39,42,64,67],integer_value_sub_sequ:39,integr:67,inter:[10,16],intercept:[10,16],interfac:[7,10,11,15,16,17,48],intergr:[10,16],intern:[10,11,16,17,20,22],internet:24,interpol:10,interpret:9,introduc:24,invalid:29,invok:[3,10,16,22,47,66],iob:9,ioe:9,ip_str:56,ipc:54,ips:56,ipt:[10,16,31,39,42],ipython:23,is_async:12,is_gener:[10,16,60,69],is_kei:66,is_layer_typ:10,is_loc:22,is_predict:[64,66,68],is_seq:[10,16,42,66],is_sequ:66,is_stat:[7,15],is_test:[62,67,68],is_train:3,isinst:5,ispodallrun:56,isspars:44,item:[10,16,20,22,29,56],iter:[3,10,11,12,16,17,18,20,22,23,24,29],its:[3,9,10,11,16,17,23,24,36,47],itself:[11,17,24],java:27,jeremi:47,jie:[67,68],jmlr:[10,16],job:[9,20,48,49,50,52,54,56,62,64,67,68,69],job_dispatch_packag:48,job_id:20,job_mod:60,job_nam:56,job_namespac:56,job_path:56,job_path_output:56,job_workspac:48,jobnam:56,jobpath:56,jobselector:56,johan:68,join:[24,39],joint:69,jointli:[11,17,69],journal:[67,68],jpg:62,json:[48,55,66],jth:[11,17],jupyt:34,just:[9,10,11,14,16,17,20],jypyt:23,k8s:56,k8s_data:56,k8s_job:23,k8s_token:23,k8s_train:56,k8s_user:23,kafka:26,kaim:[10,16],kaimingh:62,kebilinearinterpbw:47,kebilinearinterpfw:47,keep:[10,16,24],kei:[3,20,22,24,43,47,54,56,66],kernel:[10,16],key1:50,key2:50,keyword:56,kill:24,kind:[23,24,54,55,56],kingsburi:67,know:[11,17,23],kriz:[20,61],ksimonyan:[11,17],kube:54,kube_cluster_tl:23,kube_ctrl_start_job:23,kube_list_containers_in_job_and_return_current_containers_rank:23,kubeadm:54,kubectl:[54,55,56],kuberent:24,kubernet:[23,24,46,48,56,57,58],kubernetes_service_host:23,kwarg:[3,9,10,11,12,16,17,18,20,39,64,66,67],l1_rate:[7,15],l2_rate:[7,15],l2regular:[53,61,64,68],label:[3,9,10,12,16,18,20,22,29,31,32,39,53,55,61,62,64,66,67,68],label_dict:67,label_dim:[10,16,39,64],label_fil:[29,67],label_lay:[10,29],label_list:67,label_path:29,label_slot:67,labeledbow:68,labelselector:56,lag:50,lake:3,lambdacost:[10,16],lambdarank:[10,16],languag:[10,16,20,60],larg:[19,20,43,69],larger:[7,9,10,12,15,16],last:[9,10,11,16,17,38,39],last_seq:39,last_time_step_output:[10,16],lastseen:55,later:64,latest:[10,16,24,31,55,56],launcher:23,layer1:[10,11,16,17,38],layer2:[10,16,38],layer3:[10,16],layer:[4,5,7,9,11,15,17,19,20,21,22,29,38,41,42,44,53,62,64,66,67],layer_0:44,layer_attr:[10,16,42,52,53],layer_num:[52,62],layer_s:[10,16],layer_typ:[10,16],layerbas:44,layerconfig:44,layergradutil:44,layermap:44,layeroutout:[10,16],layeroutput:[9,11,16,53,66],layers_test:31,lbl:[9,61],ld_library_path:[36,48],learn:[7,9,10,11,12,15,16,17,18,20,23,29,34,47,61,62,67,68,69],learnabl:[10,16],learning_method:[12,32,53,60,61,64,66,68,69],learning_r:[7,12,15,31,32,53,60,61,64,66,68,69],leas:24,least:[9,10,16,24],lecun:20,left:[10,16],leman:69,len:[3,10,16,39,42,44,56,64,66,67],length:[10,11,16,17,20,55],less:[10,16,23],less_than:23,let02:55,let:[10,16,23],level:[7,10,15,16,41],lib64:[31,34,36,48,50],lib:[28,33,36],lib_path:31,libari:28,libcuda:[31,34],libjpeg:61,libnvidia:[31,34],libpaddl:[27,28,43],libpaddle_capi:28,libpaddle_gserv:28,libpaddle_math:28,libprotobuf:31,librari:[10,16,28,36,48,50],life:24,like:[9,10,16,20,24,29,62],limit:[10,16,20,31,47],line:[3,9,20,31,39,43,52,64,66,67],line_count:31,linear:[6,10,16],linear_comb:10,linearactiv:[10,16,32],linguist:67,link:[10,11,16,17,41,68],linux:[34,54],lipeng:60,lipton:68,list:[2,3,8,9,10,11,16,20,22,23,32,48,52,53,61,62,64,66,67,68,69],lium:69,live:24,liwicki:68,load:[10,16,23,24,32,56,62,66,67,68,69],load_data_arg:5,load_featur:62,load_feature_c:62,load_feature_pi:62,load_missing_parameter_strategi:[49,50,52,60,67],loadparamet:5,loadsave_parameters_in_pserv:[49,50],local:[7,15,24,33,36,45,48,49,50,56],localhost:[34,54],localip:56,lock:24,log:[6,31,44,48,50,55,56,61,66,67,68,69],log_barrier_abstract:[49,50],log_barrier_lowest_nod:[49,50],log_barrier_show_log:[49,50],log_clip:[49,50],log_error_clip:[49,50],log_period:[50,52,55,56,61,64,66,67,68,69],log_period_serv:[49,50],logarithm:14,logger:[3,39],logist:64,look:[3,9,64],loop:29,loss:[10,16,64],low:[10,16],lpaddle_capi_shar:28,lpaddle_capi_whol:28,lst:66,lstm:[10,16,39,42,55,64],lstm_attr:17,lstm_bias_attr:[11,17],lstm_cell_attr:[11,17],lstm_group:[11,17,39],lstm_group_input:39,lstm_input:39,lstm_last:39,lstm_layer_attr:11,lstm_nest_group:39,lstm_output:39,lstm_size:64,lstm_step:[11,17],lstmemori:[11,17,39,42],lstmemory_group:[10,16,39],ltr:[10,16],mac:[28,34],machan:[11,17],machin:[10,11,12,16,17,20,22,41,68,69],made:24,mai:[8,9,10,16,29],main:5,maintain:[10,16],major:69,make:[3,10,23,24,29,36,43,44,47,68],manag:24,mandarin:[10,16],mani:[10,11,16,17],manufactur:69,mao:68,map:[10,16,20,22,23,66],map_read:20,mapreduc:23,marcu:68,mark:67,mark_slot:67,market:68,martha:67,mask:[7,10,15,16],master:[23,30,54,68],mat:[27,28],mat_param_attr:[11,17],math:[11,17,27,44,47],matirx:[10,16],matplotlib:61,matrix:[9,10,11,16,17,20,22,27,28,44],matrixptr:44,matrixtyp:28,max:[7,10,13,15,16,20,38,47,52,66],max_id:[16,22],max_job_id:20,max_length:[10,16,42],max_movie_id:20,max_sort_s:[10,16],max_user_id:20,maxid:[9,10],maxid_lay:9,maxim:[10,16],maximum:[9,20],maxinum:19,maxout:10,maxpool:[10,16],mayb:[10,11,16,17],md5:[20,25],mean:[7,9,10,11,12,15,16,17,18,19,20,22,29,31,50,62,64,66],mean_img_s:61,mean_meta:62,mean_meta_224:62,mean_valu:62,mechan:[10,11,16,17],meet:67,mem:[10,16,39],member:23,memcpi:47,memori:[11,17,42,47,55,64],memory_nam:[10,16],memory_threshold_on_load_data:[49,50],mere:[11,17],merg:43,mergedict:[60,69],messag:[43,55],meta:[48,61,62,66],meta_config:[48,66],meta_fil:66,meta_gener:[48,66],meta_path:61,meta_to_head:66,metadata:[55,56],metal:54,metaplotlib:23,method:[3,8,10,11,12,16,18,22,66,69],mfs:56,might:[10,16],million:20,min:[7,15,47,52,66],min_pool_s:[3,31,53],mini:[10,16,20,22,24],mini_batch:29,minibatch:[10,16],minibatch_data:20,minikub:54,minim:[12,18],minimum:[10,16],minut:24,miss:67,mix:[11,17,53],mixed_attr:17,mixed_bias_attr:[11,17],mixed_lay:[11,16,39,42,53,67],mixed_layer_attr:11,mixedlayertyp:[10,16],mkl:33,mkl_root:33,ml_data:[48,66],mnist:[3,5,26,29],mnist_model:5,mnist_provid:3,mnist_random_image_batch_read:29,mnist_train:[3,29],mnist_train_batch_read:29,mnt:56,mod:67,mode:[10,16,43,56,68],model:[10,11,12,16,17,20,24,25,52,53,60,61,64,66,67,68,69],model_averag:12,model_config:5,model_list:[50,52,67,68],model_output:68,model_path:52,model_zoo:[60,62],modelaverag:12,modifi:43,modul:[3,8,11,17,20,22,31,32,53,61,62,64,66],modulo:[10,16],momentum:[7,12,15,31,64],momentumoptim:[32,61],mon:55,mono:[10,16],month:69,mood:68,moosef:54,more:[9,10,11,16,17,20,23,24,29,31,47],morin:[10,16],mose:[68,69],moses_bleu:69,mosesdecod:68,most:[10,16,20,23,29],mountpath:[55,56],move:[10,16,24],movi:[3,20,66],movie_categori:20,movie_featur:66,movie_head:66,movie_id:[56,66],movie_info:20,movie_meta:66,movie_nam:66,movie_review:20,movieinfo:20,movielen:65,moving_average_fract:[10,16],mpi:48,mse:10,mse_cost:[32,66],much:[10,16,24,29],mul:44,multi:[10,16,62,69],multi_binary_label_cross_entropi:16,multi_crop:62,multinomi:[10,16],multipl:[9,10,11,16,17,20,23],multipli:[9,10,16],must:[9,10,11,14,16,17,29,36,44],mypaddl:[55,56],name:[3,7,8,9,10,11,15,16,17,19,20,22,23,24,26,28,31,32,34,39,42,44,47,52,53,54,55,56,57,58,61,64,66,69],name_prefix:26,namespac:[27,44,54,55,56],nativ:[10,16],nchw:[10,16],ndarrai:[22,26],ndarri:22,ndcg:[10,16],ndcg_num:[10,16],necessari:[10,16,64],need:[10,11,16,17,20,23,31,47,56,64],neg:[3,9,10,16,64,67,68],neg_distribut:[10,16],nest:20,net:[10,11,16,17,68],net_conf:68,net_diagram:62,network:[4,5,7,9,10,12,15,16,18,20,21,22,23,24,29,39,48,52,56,60,61,62,66,67,68,69],network_config:52,neural:[10,11,12,16,17,18,20,22,23,24,39,41,60,66,67,68,69],neuralnetwork:[10,16],never:[20,29,55,56],next:[3,10,16,20,24],nfs:56,nfsdir:56,nginx:34,nic:[48,49,50,53,56],nine:20,nlp:[10,16],nltk:20,nmt:69,nnz:44,no_cach:3,no_sequ:[3,66],noah:68,noavx:34,node0:56,node:[10,16,54,55,56],node_0:56,node_1:56,node_2:56,nodebook:34,nodefil:48,nois:[10,16,24],non:[10,16,24],none:[3,5,7,8,9,10,11,12,15,16,17,18,19,20,22,23,32,42,62,64],norm_by_tim:[10,16],normal:[10,11,16,17,20,55,56,62],notat:[10,16],note:[7,10,11,12,15,16,17,19,22,23,29,36,68],noth:[14,22,43],novel:68,now:[10,16,24,41],np_arrai:20,ntst1213:69,ntst14:69,nullptr:[36,44],num:[10,16,48,50,56,67,68,69],num_channel:[10,11,16,17,53,61],num_chunk_typ:9,num_class:[10,11,16,17,61],num_filt:[10,11,16,17,53],num_gradient_serv:[49,50,53],num_group:[10,16],num_neg_sampl:[10,16],num_parameter_serv:23,num_pass:[22,32,49,50,52,55,56,64,66,67,68,69],num_repeat:[10,16],num_result:9,num_results_per_sampl:[10,16],num_shard:26,number:[9,10,16,20,24,26,29,69],numchunktyp:9,numdevices_:52,numlogicaldevices_:52,numofallsampl:9,numofwrongpredict:9,numpi:[20,22,26,29,32,62],numsampl:47,numtagtyp:9,nvcc:34,nvidia:[31,34],obj:[3,8,31,32,53,61,62,64,66],object:[3,7,8,9,10,11,12,15,16,17,18,20,22,23,27,47,64,66],observ:[12,18],occup:[56,66],occur:[20,22],oct:55,odd:[10,16],off:[28,33,34,36,43],offlin:[24,26],offset:[10,16,66],often:9,ograd:44,omit:[31,64],on_init:3,onc:[10,16,24],one:[3,8,9,10,11,12,14,16,17,18,19,20,23,24,29,64,67,68],one_host_dens:66,one_hot_dens:66,onli:[9,10,11,16,17,19,20,22,23,39,41],onlin:[12,18,24,26,29],open:[3,10,16,23,29,31,32,39,62,64,66,67],openbla:33,openblas_root:33,oper:[10,11,12,16,17,18,53],opinion:68,opt:[23,33,56],optim:[4,7,15,21,22,31,53],option:[9,10,16,23],order:[10,11,16,17,20,29,56],org:[10,11,16,17,20,34,65],organ:[10,16],origin:[10,16,20,43],other:[9,10,11,12,16,17,20,24,64,66],otherchunktyp:9,otherwis:[8,10,16,20,23,24,29],our:23,out:[10,16,22,23,39,41,42,53,61],out_dir:56,out_left:[10,16],out_mem:42,out_right:[10,16],out_size_i:[10,16],out_size_x:[10,16],outer:39,outer_mem:39,outer_rnn_st:39,outer_rnn_state_:39,outer_step:39,output:[7,9,10,14,15,16,17,19,20,22,23,26,29,31,32,39,42,48,52,53,55,56,60,61,62,64,66,67,68],output_:[10,16],output_dir:62,output_fil:67,output_id:[10,16],output_lay:[22,62],output_max_index:19,output_mem:[10,16,42],output_path:26,outputh:[10,16],outputw:[10,16],outsid:[3,10,11,16,17],outv:44,over:[10,11,16,17,23],overrid:24,packag:[20,31],pad:10,pad_c:[10,16],pad_h:[10,16],pad_w:[10,16],padding_attr:[10,16],padding_i:[10,16],padding_x:[10,16],paddl:[3,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,22,23,24,26,27,28,31,32,34,36,43,44,45,47,48,52,53,55,56,60,61,64,66,67,68,69],paddle_error:[27,28],paddle_matrix:[27,28],paddle_matrix_cr:28,paddle_matrix_get_shap:27,paddle_matrix_shap:27,paddle_n:[48,56],paddle_output:55,paddle_port:[48,56],paddle_ports_num:[48,56],paddle_ports_num_for_spars:48,paddle_ports_num_spars:56,paddle_process_by_paddl:56,paddle_pserver2:48,paddle_root:60,paddle_server_num:56,paddle_source_root:60,paddle_train:[28,30,48,56],paddledev:[31,34,55,56],paddlepaddl:[10,11,12,16,17,20,24,29,31,34,36,42,43,47,48,53,57,58,60,67],pair:9,palceholder_just_ignore_the_embed:60,palmer:67,paper:[10,16,69],para:60,paraconvert:60,parallel:[47,52,55,56,69],parallel_nn:[7,15,49,50],param:[7,10,15,16,66],param_attr:[10,11,16,17,31,32,42],paramattr:[7,10,15,16,31,32,42],paramet:[4,9,10,11,12,16,17,18,19,20,21,25,29,50,53,56,66,67,68,69],parameter_attribut:[10,16],parameter_block_s:[49,50],parameter_block_size_for_spars:[49,50],parameter_learning_r:[7,15],parameter_nam:[22,23],parameter_serv:23,parameterattribut:[7,10,11,15,16,17],parameterclient2:56,parametermap:44,parameters_:44,parameterset:23,parametris:[12,18],paramutil:66,paraphras:[60,69],paraphrase_data:60,paraphrase_model:60,paraphrase_modeldata:60,paraspars:44,parent:10,pars:[20,66],parse_config:5,parse_known_arg:56,parsefromstr:31,parser:56,part:[66,68],partial:[10,16],partit:[24,26],paserv:56,pass:[3,8,10,16,20,22,24,29,31,32,43,47,50,52,55,56,61,64,66,67,68,69],pass_id:22,pass_idx:29,passtyp:44,past:23,path:[9,20,22,24,29,36,48,50,54,55,56,67,68],pattern:[20,24,27,66,68],paul:67,paus:24,pave:69,pdf:[10,11,16,17],pem:23,pend:24,penn:67,per:[10,16,20,24,29],perform:[10,11,16,17,47,49],period:[24,50,67,68,69],perl:[68,69],peroid:[10,16],persistentvolum:54,persistentvolumeclaim:[54,56],person:23,pfs:26,pickl:66,picklabl:8,pid:54,piec:[10,11,16,17],pillow:61,pip:[31,43,48,61,66],pixel:[3,10,16,20,53],pixels_float:3,pixels_str:3,place:[3,24],plain:[9,10,16,22,28],plan:24,pleas:[7,10,11,12,15,16,17,18,23,24,29,31,36,56],plot:[23,61],plotcurv:61,png:61,pnpairvalid:49,pod:[54,55,56],podip:56,podlist:56,point:[24,47],polar:20,poll:68,pool3:44,pool:[4,11,17,21,53,66],pool_attr:[11,17],pool_bias_attr:[11,17],pool_layer_attr:11,pool_pad:[11,17],pool_siz:[3,10,11,16,17,53],pool_size_i:[10,16],pool_strid:[11,17],pool_typ:[10,11,16,17],pooling_lay:[11,31,64,66],pooling_typ:[10,16,31,38,64],poolingtyp:19,port:[48,49,50,53,55,56],port_num:49,ports_num:[50,53,56],ports_num_for_spars:[49,50,52,53,56],pos:[66,68],pose:24,posit:[3,9,10,16,20,64],positive_label:9,posix:54,possibl:23,potenti:47,power:10,practic:[8,10,16],pre:[10,11,16,17,20,23,60,68,69],pre_dictandmodel:60,precis:9,pred:[64,67],predetermin:[10,16],predic:[20,67],predicate_dict:67,predicate_dict_fil:67,predicate_slot:67,predict:[3,5,9,10,12,16,18,22,31,48,53,60,61,62,64,66,67,68],predict_fil:[49,50],predict_output_dir:[49,50,64],predict_sampl:5,predin:61,prefer:54,prefetch:44,prefix:[24,26],pregrad:44,premodel:60,prepar:57,preprocess:[20,48,60,61,64,66,68,69],present:[23,43],prev_batch_st:[49,50],prevent:[12,18,23,24],previou:[10,11,16,17,24],previous:[10,16],price:20,principl:23,print:[5,7,15,22,23,32],printallstatu:47,printer:9,printstatu:47,prite:9,privat:[28,43],prob:[9,22],probabilist:[10,16,60],probabl:[9,10,16,22],problem:[10,12,16,18,23],proc:34,proce:[20,24,29],proceed:[10,16,67],process2:39,process:[3,7,8,10,11,12,15,16,17,23,26,31,32,39,42,53,56,64,66,67],process_predict:64,process_test:8,process_train:8,processdata:[61,62],processor:47,prod:[34,43],produc:[11,17,20,24,29],productgraph:55,profil:47,proflier:47,prog:56,program:[20,23,26,29,47,56],progress:24,proj:[10,16],project:[10,11,16,17,28,53],promis:[10,11,16,17],prone:23,prop:67,propag:[12,18],properli:64,proposit:67,protect:44,proto:19,protobuf:31,provid:[8,10,16,20,23,31,32,39,49,53,64,66,67],prune:[10,16],ps_desir:24,pserver:[48,49,50,53,56],pserver_id:25,pserver_num_thread:[49,50],pseudo:23,psize:44,ptr:28,pull:30,purpos:[24,47],push:56,push_back:44,put:[24,64],pwd:[34,43],py_paddl:[5,20,45],pydataprovid:[31,53],pydataprovider2:[3,5,32,53,56,66],pydataproviderwrapp:8,pyramid:[10,16],pyramid_height:[10,16],python:[8,22,23,27,30,31,34,43,44,45,48,60,61,62,66,67,68,69],pythonpath:[31,61],pzo:68,queri:[10,16,69],question:[10,16,23],quick:55,quick_start:[55,56,57,64],quick_start_data:55,quickstart:55,rac:[10,16],rais:20,ramnath:68,ran:47,rand:[47,50,52,67],random:[7,10,15,16,20,26,29,32],random_imag:26,rang:[10,16,20,26,29,56,64],rank:[10,16,23,62,64],rare:3,rate:[7,9,12,15,18,20,48,56,66],ratio:50,raw:[10,16],raw_meta:66,rdma_tcp:[49,50],reach:24,read:[3,20,22,23,24,26,29,32,62,64,66],read_from_realistic_imag:23,read_from_rng:23,read_lock:25,read_mnist_imag:23,read_next_from_fil:31,read_ranking_model_data:23,reader:[22,26],reader_cr:26,reader_creator_bool:29,reader_creator_random_imag:[20,29],reader_creator_random_image_and_label:[20,29],readi:[24,55],readm:[28,43,66,68],real:29,real_process:3,realist:23,reason:[10,11,17,23,24,55],recal:9,receiv:[8,24],recognit:[10,16,62,68],recommend:[11,17,23,48,56,66],record:[66,67],recordio:23,recov:24,rectangular:[10,16],recurr:[39,40,67,68],recurrent_group:[11,17,39,41,42],recurrent_lay:11,recurrentgradientmachin:28,recurrentgroup:9,reduc:[12,18],ref:66,refer:[7,8,10,11,12,15,16,17,18,24,33],referenc:[10,16],reference_cblas_root:33,refine_unknown_arg:56,regex:66,register_gpu_profil:47,register_lay:44,register_timer_info:47,registri:55,regress:9,regular:[7,12,15,53,61,64,68],rel:[11,17],relat:[8,24,66],relationship:20,releas:[30,36,54,67],reliabl:24,relu:[6,10,16],reluactiv:10,rememb:[10,16],remot:[7,15,43,48,50,52],remov:[20,43],reorgan:[10,16],repeat:10,repres:[10,12,16,20,64],represent:64,request:[24,30,55,69],requir:[9,10,16,23,24,48,66],res5_3_branch2c_bn:62,res5_3_branch2c_conv:62,res:67,research:[10,16,20,61],reserveoutput:44,reset:[10,16,24],reshap:29,reshape_s:[10,16],residu:62,resnet_101:62,resnet_152:62,resnet_50:62,resolv:[43,55],respons:[10,16,55],rest:[10,16],restart:[24,55],restartpolici:[55,56],resu:29,result:[3,9,10,14,16,22,47,64,69],result_fil:[9,42],ret_val:66,return_seq:[11,17],reuqest:30,reus:29,reveal:23,revers:[10,11,16,17,41,42],review:[20,43,55,68],reviews_electronics_5:55,rewrit:69,rgb:[10,16],rgen:68,rho:[12,18],right:[10,16],rmsprop:[12,64],rmspropoptim:66,rnn:[10,11,16,17,41,42,49,68],rnn_bias_attr:42,rnn_layer_attr:42,rnn_out:42,rnn_state:39,rnn_state_:39,rnn_step:[10,16],rnn_use_batch:[49,50],rnnlm:20,robot:61,role:[20,23,67,68],roman:68,root:[12,18,19,34,48,55,56,60],root_dir:48,rot:[10,16],rotat:10,routin:66,routledg:68,row:[9,10,16,20],row_id:[10,16],rstrip:56,rtype:[10,16,66],run:[23,24,31,34,43,47,48,55,56,57,58,66],runinitfunct:[47,56],runtim:[3,31],s_fusion:66,s_id:66,same:[3,8,9,10,11,16,17,23,39,64],samping_id:[10,16],sampl:[3,9,20,64,66,68,69],sample_id:9,sample_num:9,santiago:68,save:[3,10,16,20,24,26,55,66,67,68,69],save_dir:[32,50,52,55,56,61,64,66,67,68,69],save_only_on:[49,50],saving_period:[49,50,56],saving_period_by_batch:[49,50,52],saw:3,scalar:[10,16],scale:[10,14,66],scalingproject:[10,16],scatter:[10,16],scheduler_factor:[7,15],scheme:[9,12],schmidhub:68,schwenk:69,scienc:68,score:[9,10,16,66],script:[20,45],seaplane_s_000978:61,search:[10,16,24,42,69],seat:69,second:[10,16,20,23,29,64,66],sed:68,see:[10,11,16,17,23,24,31,64],seed:[47,50],segment:9,sel_fc:[10,16],select:[10,16],selectiv:[10,16],selector:55,self:44,selfnorm:[10,16],semant:[20,23,30,67,68],semantic_role_label:42,semat:23,sen_len:67,send:24,sens:10,sent:[23,55],sent_id:42,sentanc:31,sentenc:[3,10,16,20,39,42,67],sentence_last_state1:39,sentence_last_state2:39,sentiment:[3,67,68],sentiment_data:68,sentiment_net:68,sentimental_provid:3,sentimental_train:3,separ:[9,64,66],seq:[10,16,39,66],seq_pool:[10,16,38],seq_text_print:9,seq_to_seq_data:[60,69],seq_typ:[20,66],seqlastin:39,seqtext_printer_evalu:42,seqtoseq:[10,16,31,42,60,69],seqtoseq_net:[10,16,42,60,69],sequel:3,sequenc:[3,9,10,11,14,16,17,19,20,31,39,41,64,66,68,69],sequence_conv_pool:64,sequence_layer_group:[10,16,39],sequence_nest_layer_group:[10,16,39],sequencegen:39,sequencesoftmax:6,sequencestartposit:[10,16],sequencetextprint:9,sequencetyp:3,sequenti:[10,16,67],seri:[11,17,39,68],serial:22,server:[23,48,50,53,54,56],serverless:24,set:[3,7,9,10,11,15,16,17,20,22,23,24,31,32,39,42,47,48,53,55,60,61,62,64,66,67,68,69],set_active_typ:44,set_default_parameter_nam:[7,15],set_drop_r:44,set_input:[10,16],set_siz:44,set_typ:44,settotalbyteslimit:31,setup:[44,64],sever:[10,16],sgd:[12,18,22,23,24,32,48,49,68],shape:[10,16,22],shard:[24,25,26],share:[10,16,28,67],shared_bia:[11,17],shared_bias:[10,16],shared_ptr:[27,28],ship:61,shold:68,shorten:[10,16],should:[9,10,12,16,20,22,23,29,41],should_be_fals:23,should_be_tru:23,should_shuffl:[3,39,67],show:[12,18,20,24,67,68,69],show_check_sparse_distribution_log:[49,50],show_layer_stat:[49,50],show_parameter_stats_period:[49,50,52,55,67,68,69],shown:[9,10,16,23],shuf:[31,66],shuffl:[20,31],side:[10,16,22],sigint:48,sigmoid:[6,10,16,17],sigmoidactiv:[10,11,39],similar:[10,16,29,66],similarli:[10,16],simpl:[9,10,11,14,16,17,20,22,56],simple_attent:42,simple_gru:64,simple_img_conv_pool:53,simple_lstm:[10,16,64],simple_rnn:[10,16,42],simpli:[10,16,23],simplifi:23,sinc:[10,16,24,29],singl:[9,11,12,17,20,24],size:[3,9,10,11,12,16,17,18,20,24,26,29,31,32,39,42,44,53,61,62,64,66,68,69],size_a:[10,16],size_b:[10,16],size_t:44,skip:[29,32,43,62],sleep:56,slide:[10,12,16,18,20,24],slope:[10,16],slot:[66,67],slot_dim:66,slot_nam:66,slottyp:66,small:20,small_messag:[49,50],small_vgg:61,smaller:[10,16,24],smith:68,smooth:[10,16],snap:55,snapshot:25,sock_recv_buf_s:[49,50],sock_send_buf_s:[49,50],socket:56,softmax:[6,10,11,16,17,23,31,42,44,64],softmax_param:31,softmax_param_attr:[11,17],softmax_selfnorm_alpha:[10,16],softmaxactiv:[31,39,42,53,64],softrelu:6,solv:23,some:[7,10,12,15,16,20,22,23],some_c_api_funct:28,some_inst:28,some_python_class:27,somecppclass:27,somedata:22,somegotyp:27,someth:[10,16],sometim:[12,18,29],soon:24,sort:[10,16,20,56,66,68],sourc:[8,10,16,20,28,29,69],source_dict_dim:42,source_language_word:42,space:9,space_seperated_tokens_from_dictionary_according_to_seq:9,space_seperated_tokens_from_dictionary_according_to_sub_seq:9,spars:[7,10,12,15,16,18,20,31,44,48,50,52,56,64],sparse_binary_vector:[3,20,31,64],sparse_binary_vector_sequ:20,sparse_float_vector:3,sparse_non_value_slot:20,sparse_upd:[7,15,31],sparse_value_slot:20,sparse_vector:[20,31],sparse_vector_sequ:20,sparseparam:44,sparseprefetchrowcpumatrix:44,spatial:[10,16],spec:[55,56],special:[10,16],specifi:[9,10,16,20,23,36,64],speech:[10,16],speed:[11,17],sphinx:[27,34,45],split:[3,10,16,39,48,64,66,67],split_count:56,spp:10,squar:[6,10,12,16,18,19],squarerootn:13,squarerootnpool:[10,16],srand:50,src:[56,69],src_backward:42,src_dict:[31,42],src_dict_path:31,src_embed:42,src_forward:42,src_id:42,src_root:5,src_word_id:42,srl:[20,67],ssd:16,ssh:[34,48],sshd:34,sstabl:[23,26],stabl:[30,54],stacked_lstm_net:68,stacked_num:68,stackexchang:[10,16],stage:43,stake:69,stale:24,standard:[7,15],stanford:[20,55],stanh:6,start:[10,16,22,24,31,50,55,56],start_paddl:56,start_pass:[49,50],start_pserv:[49,50],startpaddl:56,startup:24,stat:[47,50,67,68,69],state:[10,11,16,17,24,41,55],state_act:[10,11,16,17,39],statfulset:56,staticinput:[10,16,41,42],statist:[10,16],statset:47,statu:[9,43,47,55,56],status:55,std:[27,28,44,50],stderr:48,stdout:48,step:[10,11,12,16,17,19,24,39,41,42],stepout:39,stochast:[12,18,24],stock:68,stop:[10,16],storag:54,store:[9,10,16,20,64,66],str:[22,52,56],strategi:[19,24,50,67],street:[10,16,67],strict:29,stride:[10,16],stride_i:[10,16],stride_x:[10,16],string:[3,8,9,10,16,44,50,53],strip:[31,39,64,66,67],struct:28,structur:[20,64],stub:[10,16],stuff:43,stun:3,style:[10,16],sub:[9,10,16,20,23],sub_sequ:3,subgradi:[12,18],subnet:23,subobjectpath:55,subseq:[38,41],subsequenceinput:[10,16,39],succeed:55,success:55,successfulcr:55,sudo:[36,61],suffic:29,suggest:[10,16],sum:[9,10,12,13,16,18,25],sum_:[10,16],sum_to_one_norm:10,sumpool:[10,16,31],support:[7,9,10,12,15,16,19,20,24,29,39],sure:[36,68],swap_channel:62,swig:[27,28],swig_paddl:[5,20],symbol:[10,16,28],symlink:43,sync:24,syncflag:44,synchron:[12,18,24],syntax:29,sys:62,system:[24,31,68],t2b:60,t_i:[10,16],tab:64,tabl:[10,16],tableproject:[10,16],tag:[9,20,34],tagtyp:9,tainer_id:56,take:[3,9,10,11,16,17,23],tanh:[6,10,11,16,17,44],tanhactiv:[10,11,39,42,53],target:[10,16,20,22,69],target_dict_dim:42,target_dictionary_dim:[10,16],target_language_embed:[10,16],target_language_word:42,targetinlink:[10,16,39],task:[9,10,16,67],tbd:39,tconf:68,tcp:[50,53],tcp_rdma:53,tear:47,technic:24,tee:[55,61,66,67,68,69],tell:24,tellig:68,templat:[55,56],tempor:[10,16],tensor:10,term:[10,11,16,17,24],termin:55,tesh:67,test100:20,test10:20,test:[2,3,8,9,10,16,20,22,23,28,29,30,43,44,47,48,50,52,53,61,62,64,66,67,68,69],test_all_data_in_one_period:[55,61,66,67,68],test_compar:31,test_comparespars:31,test_comparetwonet:31,test_comparetwoopt:31,test_config_pars:31,test_data:[5,69],test_fcgrad:44,test_gpuprofil:47,test_layergrad:44,test_list:[3,8,31,32,53,61,64],test_networkcompar:31,test_part_000:68,test_pass:[49,50,52,69],test_period:[49,50,52],test_predict:31,test_pydataprovid:31,test_pydataprovider2:31,test_pydataproviderwrapp:31,test_ratio:66,test_recurrent_machine_gener:31,test_recurrentgradientmachin:[31,39],test_swig_api:31,test_train:31,test_traineronepass:31,test_wait:[49,50],testa:23,testb:23,testbilinearfwdbwd:47,testconfig:44,tester:[66,69],testfcgrad:44,testfclay:44,testlayergrad:44,testq:23,testresult:22,testutil:44,text:[3,9,11,17,20,23,60,64,68,69],text_conv:64,text_conv_pool:66,text_fil:[20,68],tflop:47,tgz:20,than:[7,9,10,11,12,15,16,17,24,31],thei:[23,24,47],them:[11,17,23,24,29,47,64,66],therein:[10,16],thi:[3,7,8,9,10,11,12,15,16,17,18,20,22,23,24,29,47,64,66,68],thing:3,think:23,third:[10,16,24],those:[24,67],thread:47,thread_local_rand_use_global_se:[49,50],threadid:52,threadloc:47,three:[9,10,12,16,24,29,62],threshold:[7,9,12,15,24,50],through:[10,16,24],throughput:47,thu:[10,16],tier:55,time:[10,11,16,17,19,20,23,24,29,39,47,50,55,56,68],timelin:[10,16],timeout:24,timer:47,timestamp:[10,16,25],timestep:[10,16],titl:[20,56,66],tmp:3,to_your_paddle_clone_path:45,todo:[9,11,17,20,24],toend:[10,16],togeth:[10,11,16,17,20,22],token:[9,10,16,23,42,68],too:20,tool:[45,56],top:[9,62],top_k:9,topolog:[20,23,24],topolopi:22,toronto:[20,61],total:[9,22,24,29,47,55,69],total_pass:29,touch:68,tourist:69,track:[10,24,43],tractabl:[10,16],tradit:[10,16],trail:20,train100:20,train10:20,train:[2,3,7,8,9,10,12,15,16,18,20,26,31,32,36,48,50,52,53,55,56,57,58,60,61,62,64,66,67,68,69],train_arg:56,train_args_dict:56,train_args_list:56,train_conf:[60,69],train_config_dir:56,train_data:69,train_list:[3,8,31,32,53,61,62,64],train_part_000:68,trainabl:[10,16],trainer:[3,5,23,25,26,32,44,50,52,53,56,66,67,68,69],trainer_config:[2,3,5,32,48,55,56,64,66,68],trainer_config_help:[3,6,7,8,9,10,11,12,13,31,32,44,53,61,66],trainer_count:[31,49,50,52,55,56,66,67,68,69],trainer_id:[50,56],trainerconfighelp:31,trainerid:56,trainerintern:[64,66,69],tran:[10,44],transact:[24,68],transform:[10,16],transform_param_attr:[11,17],translat:[10,11,16,17,60,69],transpos:[10,16],transposedfullmatrixproject:[10,16],travel:[3,11],travi:43,treat:[10,16],tree:[10,16,56],trg:69,trg_dict:42,trg_dict_path:42,trg_embed:42,trg_id:42,trg_ids_next:42,tricki:27,trn:64,truck:61,true_imag:29,true_label:29,true_read:29,truth:[9,10,16,69],tst:64,tune:[7,15,49],tupl:[8,10,11,16,20,22,29],ture:[10,16],turn:[10,16,29,41],tutori:[55,56,57,58,60,69],tweet:68,twitter:68,two:[10,11,16,17,23,29,47,64],txt:[3,44,48,54,64,66,68],type:[3,9,10,11,12,16,17,19,20,23,24,27,28,29,32,39,44,52,55,62,64,66,67],type_nam:[10,66],typedef:[27,28],typic:9,typo_freq:20,ubuntu:30,ubyt:29,uci:20,ufldl:[10,16],uid:55,uint64:27,uint64_t:27,unconstrain:68,undeterminist:47,uniform:[7,10,15,16,20,26,29],uninstal:31,uniqu:[23,24],unique_ptr:44,unit:[10,11,16,17],unittest:[28,31],univ:69,unix:48,unk:[60,69],unk_idx:[64,67],unknown:[10,16],unseg:[10,16],unsup:68,unsupbow:68,until:[24,56],untrack:43,unus:66,updat:[7,10,12,15,16,24,43,48,52],update_equ:22,updatecallback:44,upgrad:31,upload:24,upstream:43,url:[20,68],urls_neg:68,urls_po:68,urls_unsup:68,usag:[9,10,11,16,17,20,22,56,66],use:[7,8,9,10,11,12,15,16,17,19,20,22,23,24,43,47,56,62,64,66,67,68,69],use_global_stat:[10,16],use_gpu:[5,31,49,50,52,55,56,61,62,64,66,67,68,69],use_jpeg:61,use_old_updat:[49,50],use_seq:[32,66],use_seq_or_not:66,used:[3,9,10,11,12,16,17,18,19,20,22,23,24,29,47,64,66,68],useful:[10,11,16,17],usegpu:44,user:[7,9,10,11,15,16,17,20,22,23,29,54,64,66],user_featur:66,user_head:66,user_id:[56,66],user_info:20,user_meta:66,user_nam:66,userinfo:20,usernam:43,uses:24,using:[7,8,10,15,16,20,23,24,29,36,67],usr:[31,33,34,36,48,50,56],usrdict:60,usrmodel:60,usual:[10,16,20,22,47],utc:65,util:[47,56,61,66],uuid:25,v28:[10,16],valid:29,valu:[3,5,7,9,10,12,15,16,18,19,20,22,24,44,52,56,62,64,67],value1:50,value2:50,value_rang:20,vanilla:42,variabl:[10,16,20,23,55],varianc:[10,16],vector:[10,11,16,17,20,23,64,66],verb:20,veri:[10,16,19,61],version:[10,11,16,17,30,34,36,47,49,50],versu:23,vertic:[10,16],vgg:[11,17,61],vgg_16_cifar:61,via:[24,29,36,43],view:[10,16],vision:61,visipedia:61,visual:[10,16],vocab:68,volum:[34,54,55,56],volumemount:[55,56],vutbr:20,wai:[10,11,16,17,23,69],wait:[12,18,24,56],wall:67,want:[3,10,11,16,17,23,29],warn:[10,16,31,56],warp:[10,16],watch:24,wbia:62,wei:[67,68],weight:[9,10,11,12,16,17,18,44,62],weight_act:[11,17],weightlist:44,weights_:44,weights_t:44,wether:[10,16],what:[7,10,11,12,15,16,17,18,43,64],when:[3,7,9,10,12,15,16,20,22,24,47],where:[10,11,12,16,17,18,23,24],whether:[9,10,11,16,17,29,68],which:[9,10,11,12,16,17,18,20,23,24,26,29,64,66],whole:[3,9,20,27,28],whole_cont:66,whose:[10,16,20,24],why:[11,17,28,47],width:[9,10,16,20,27,29,44,69],wiki:[10,16],wikipedia:[10,16,20],wilder:3,window:[10,16,19,20,54],wise:[10,16],with_avx:[33,34,36,43],with_doc:33,with_dock:45,with_doubl:[33,36,44],with_dso:33,with_gpu:[33,34,36,43],with_predict_sdk:36,with_profil:47,with_python:[33,36],with_rdma:[33,36],with_style_check:[33,43],with_swig_pi:33,with_test:[33,34,43],with_tim:[33,36,47],within:[10,16],without:[9,10,16,24,29],wmt14:69,wmt14_data:69,wmt14_model:69,wmt_shrinked_data:20,woboq:34,won:39,wonder:3,word2vec:31,word:[3,9,10,16,20,31,39,41,64,67],word_dict:[39,64,67],word_dim:[39,64],word_id:[3,31],word_idx:20,word_slot:67,word_vector:64,word_vector_dim:[10,16,42,60],words_freq_sort:20,work:[20,23,24,29,34,39,43,55,56],workspac:48,would:[22,24,29,67],wrapper:[11,17,47],write:[20,23,24,29,67],write_lock:25,writelin:32,writer:23,wrong:29,wsj:67,wuyi:54,www:[10,16,20,61,69],xarg:[31,34,44],xgbe0:50,xgbe1:50,xiaojun:68,xrang:[29,32,44],xxbow:68,xxx:[23,62,69],xxxx:25,y_i:[10,16],yaml:[55,56],yann:20,yapf:43,yeild:22,yield:[3,20,23,26,29,31,32,39,42,64,66,67],you:[3,7,10,11,12,15,16,17,36,62,68],your:[10,16,23,31],your_param_nam:31,your_repo:56,your_source_root:28,yuyang18:[11,17,20],zachari:68,zeng:68,zero:[7,10,12,15,16,18,20,24,50],zhou:[67,68],zip:[20,56,65],zoo:60},titles:["\u5173\u4e8ePaddlePaddle","API","DataProvider\u7684\u4ecb\u7ecd","PyDataProvider2\u7684\u4f7f\u7528","API\u4e2d\u6587\u624b\u518c","\u57fa\u4e8ePython\u7684\u9884\u6d4b","Activations","Parameter Attributes","DataSources","Evaluators","Layers","Networks","Optimizers","Poolings","Activation","Parameter Attribute","Layers","Networks","Optimizer","Pooling","Data Reader Interface and DataSets","Model Configuration","Training and Inference","PaddlePaddle Design Doc","Design Doc: Distributed Training","\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\uff08Checkpointing\uff09","\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Python Data Reader Design Doc","Paddle\u53d1\u884c\u89c4\u8303","FAQ","\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u4efb\u52a1","PaddlePaddle\u7684\u7f16\u8bd1\u9009\u9879","PaddlePaddle\u7684Docker\u5bb9\u5668\u4f7f\u7528\u65b9\u5f0f","\u5b89\u88c5\u4e0e\u7f16\u8bd1","Ubuntu\u90e8\u7f72PaddlePaddle","\u65b0\u624b\u5165\u95e8","\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684Layer","\u5355\u53cc\u5c42RNN API\u5bf9\u6bd4\u4ecb\u7ecd","RNN\u76f8\u5173\u6a21\u578b","Recurrent Group\u6559\u7a0b","RNN\u914d\u7f6e","\u5982\u4f55\u8d21\u732e\u4ee3\u7801","\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42","\u5982\u4f55\u8d21\u732e/\u4fee\u6539\u6587\u6863","\u8fdb\u9636\u6307\u5357","GPU\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18","\u8fd0\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3","\u53c2\u6570\u6982\u8ff0","\u7ec6\u8282\u63cf\u8ff0","\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570","\u4f7f\u7528\u6848\u4f8b","\u57fa\u672c\u4f7f\u7528\u6982\u5ff5","Kubernetes \u7b80\u4ecb","Kubernetes\u5355\u673a\u8bad\u7ec3","Kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3","&lt;no title&gt;","&lt;no title&gt;","PaddlePaddle \u6587\u6863","\u4e2d\u6587\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u4f7f\u7528","\u56fe\u50cf\u5206\u7c7b\u6559\u7a0b","Model Zoo - ImageNet","\u5b8c\u6574\u6559\u7a0b","\u5feb\u901f\u5165\u95e8\u6559\u7a0b","MovieLens\u6570\u636e\u96c6","MovieLens\u6570\u636e\u96c6\u8bc4\u5206\u56de\u5f52\u6a21\u578b","\u8bed\u4e49\u89d2\u8272\u6807\u6ce8\u6559\u7a0b","\u60c5\u611f\u5206\u6790\u6559\u7a0b","\u6587\u672c\u751f\u6210\u6559\u7a0b"],titleterms:{"\u4e00\u4e9b\u7ec6\u8282\u7684\u8865\u5145":56,"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":26,"\u4e0b\u8f7d\u4e0e\u89e3\u538b\u7f29":69,"\u4e0b\u8f7d\u548c\u6570\u636e\u62bd\u53d6":60,"\u4e0b\u8f7d\u5e76\u89e3\u538b\u6570\u636e\u96c6":66,"\u4e0b\u8f7d\u6570\u636e":55,"\u4e0d\u4f7f\u7528":27,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":27,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":27,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":27,"\u4e2d\u6587\u5b57\u5178":60,"\u4e2d\u6587\u77ed\u8bed\u6539\u5199\u7684\u4f8b\u5b50":60,"\u4e2d\u6587\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u4f7f\u7528":60,"\u4e2d\u6587\u8bcd\u5411\u91cf\u7684\u9884\u8bad\u7ec3\u6a21\u578b":60,"\u4e3a\u4ec0\u4e48\u9700\u8981\u6027\u80fd\u5206\u6790":47,"\u4ec0\u4e48\u662f\u6027\u80fd\u5206\u6790":47,"\u4ec5\u4ec5\u4f7f\u7528void":27,"\u4ecb\u7ecd":[60,62],"\u4ece\u5feb\u7167\u6062\u590d":25,"\u4ee3\u7801\u8981\u6c42":43,"\u4efb\u52a1\u7b80\u4ecb":32,"\u4f18\u5316\u7b97\u6cd5":64,"\u4f18\u5316\u7b97\u6cd5\u914d\u7f6e":53,"\u4f7f\u7528":[43,55],"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":27,"\u4f7f\u7528\u6848\u4f8b":52,"\u4f7f\u7528\u6982\u8ff0":64,"\u4f7f\u7528\u6a21\u578b\u521d\u59cb\u5316\u7f51\u7edc":52,"\u4f7f\u7528\u73af\u5883\u53d8\u91cf":56,"\u4f7f\u7528\u7528\u6237\u6307\u5b9a\u7684\u8bcd\u5411\u91cf\u5b57\u5178":60,"\u4f7f\u7528\u8bf4\u660e":46,"\u4f7f\u7528\u8f6c\u6362\u5e93":26,"\u4f7f\u7528docker\u6784\u5efapaddlepaddle\u7684\u6587\u6863":45,"\u4f7f\u7528paddlepaddle\u751f\u6210\u6a21\u578b":69,"\u4f7f\u7528paddlepaddle\u8bad\u7ec3\u6a21\u578b":69,"\u4fdd\u6301\u672c\u5730\u4ed3\u5e93\u6700\u65b0":43,"\u4fee\u6539\u542f\u52a8\u811a\u672c":55,"\u4fee\u6539\u6587\u6863":45,"\u514b\u9686":43,"\u5173\u4e8epaddlepaddl":0,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":28,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":28,"\u5185\u5b58\u4e0d\u591f\u7528\u7684\u60c5\u51b5":3,"\u5185\u7f6e\u5b9a\u65f6\u5668":47,"\u5199\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":44,"\u51c6\u5907\u5de5\u4f5c\u7a7a\u95f4":48,"\u51c6\u5907\u5e8f\u5217\u6570\u636e":42,"\u51c6\u5907\u6570\u636e":[32,66],"\u51c6\u5907\u8bad\u7ec3\u6570\u636e":56,"\u51c6\u5907\u96c6\u7fa4\u4f5c\u4e1a\u914d\u7f6e":48,"\u51cf\u5c11\u6570\u636e\u8f7d\u5165\u7684\u8017\u65f6":31,"\u51cf\u5c11dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":31,"\u5206\u5272\u8bad\u7ec3":66,"\u5206\u5e03\u5f0f\u8bad\u7ec3":53,"\u5206\u652f\u89c4\u8303":30,"\u521b\u5efa\u672c\u5730\u5206\u652f":43,"\u521b\u5efajob":56,"\u521b\u5efapaddl":55,"\u5220\u9664\u672c\u5730\u5206\u652f":43,"\u5220\u9664\u8fdc\u7a0b\u5206\u652f":43,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":31,"\u5230\u8fdc\u7a0b\u4ed3\u5e93":43,"\u5236\u4f5c\u955c\u50cf":56,"\u5236\u4f5cdocker\u955c\u50cf":55,"\u524d\u63d0\u6761\u4ef6":48,"\u52a0\u901f\u6267\u884c":25,"\u52a0\u901f\u8bad\u7ec3\u901f\u5ea6":31,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":27,"\u52a8\u6001\u6269\u5bb9":25,"\u5355\u5143\u6d4b\u8bd5":50,"\u5355\u53cc\u5c42rnn":39,"\u5377\u79ef\u6a21\u578b":64,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc":61,"\u539f\u56e0":27,"\u539f\u56e0\u5217\u8868":27,"\u53c2\u6570\u4fe1\u606f":62,"\u53c2\u6570\u5185\u5b58":31,"\u53c2\u6570\u670d\u52a1\u5668\u548c\u5206\u5e03\u5f0f\u901a\u4fe1":50,"\u53c2\u6570\u6982\u8ff0":49,"\u53c2\u6570\u8bfb\u53d6":62,"\u53c2\u8003":3,"\u53c2\u8003\u6587\u6863":68,"\u53c2\u8003\u8d44\u6599":47,"\u53cc\u5411lstm":68,"\u53cc\u5c42rnn":39,"\u53cc\u5c42rnn\u4ecb\u7ecd":41,"\u53cc\u5c42rnn\u7684\u4f7f\u7528":41,"\u53ef\u80fd\u7684\u5185\u5b58\u6cc4\u9732\u95ee\u9898":3,"\u53ef\u80fd\u9047\u5230\u7684\u95ee\u9898":36,"\u53ef\u9009\u529f\u80fd":60,"\u5411\u7cfb\u7edf\u4f20\u9001\u6570\u636e":64,"\u5411\u91cf":50,"\u542f\u52a8\u4efb\u52a1":56,"\u542f\u52a8\u96c6\u7fa4\u4f5c\u4e1a":48,"\u547d\u4ee4\u884c\u53c2\u6570":64,"\u548c":38,"\u56fe\u50cf\u5206\u7c7b\u6559\u7a0b":61,"\u5728\u4e0d\u540c\u8bbe\u5907\u4e0a\u6307\u5b9a\u5c42":52,"\u5728paddlepaddle\u5e73\u53f0\u8bad\u7ec3\u6a21\u578b":60,"\u57fa\u4e8epython\u7684\u9884\u6d4b":5,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":53,"\u57fa\u672c\u539f\u7406":41,"\u57fa\u672c\u8981\u6c42":27,"\u5982\u4f55\u4e66\u5199paddlepaddle\u7684\u6587\u6863":45,"\u5982\u4f55\u5171\u4eab\u53c2\u6570":31,"\u5982\u4f55\u51cf\u5c11\u5185\u5b58\u5360\u7528":31,"\u5982\u4f55\u521d\u59cb\u5316\u53c2\u6570":31,"\u5982\u4f55\u52a0\u901fpaddlepaddle\u7684\u8bad\u7ec3\u901f\u5ea6":31,"\u5982\u4f55\u6307\u5b9agpu\u8bbe\u5907":31,"\u5982\u4f55\u66f4\u65b0www":45,"\u5982\u4f55\u6784\u5efapaddlepaddle\u7684\u6587\u6863":45,"\u5982\u4f55\u8d21\u732e":45,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":43,"\u5982\u4f55\u8fdb\u884c\u6027\u80fd\u5206\u6790":47,"\u5982\u4f55\u9009\u62e9sgd\u7b97\u6cd5\u7684\u5b66\u4e60\u7387":31,"\u5b50\u5e8f\u5217\u95f4\u65e0memori":39,"\u5b50\u5e8f\u5217\u95f4\u6709memori":39,"\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6":66,"\u5b89\u88c5":[36,64],"\u5b89\u88c5\u4e0e\u7f16\u8bd1":35,"\u5b89\u88c5\u6d41\u7a0b":35,"\u5b89\u88c5kubectl":54,"\u5b8c\u6574\u6559\u7a0b":63,"\u5b9e\u73b0":27,"\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42":44,"\u5b9e\u73b0\u65b9\u5f0f":28,"\u5b9e\u73b0c":44,"\u5b9e\u73b0python\u5c01\u88c5":44,"\u5bfc\u51fac":27,"\u5c06\u547d\u4ee4\u53c2\u6570\u4f20\u7ed9\u7f51\u7edc\u914d\u7f6e":52,"\u5c0f\u7ed3":3,"\u5de5\u5177":47,"\u5e38\u7528\u6a21\u578b":63,"\u5e76\u5b8c\u6210":43,"\u5efa\u7acb":43,"\u5f00\u53d1\u6807\u51c6":46,"\u5f00\u59cb\u5f00\u53d1":43,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":50,"\u5f15\u7528":67,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":25,"\u5feb\u901f\u5165\u95e8\u6559\u7a0b":64,"\u6027\u80fd\u4f18\u5316":46,"\u6027\u80fd\u5206\u6790\u5c0f\u6280\u5de7":47,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u4ecb\u7ecd":47,"\u6027\u80fd\u8c03\u4f18":50,"\u603b\u4f53\u6548\u679c\u603b\u7ed3":64,"\u60c5\u611f\u5206\u6790\u6559\u7a0b":68,"\u6216\u8005\u662f":31,"\u627e\u5230\u7684pythonlibs\u548cpythoninterp\u7248\u672c\u4e0d\u4e00\u81f4":31,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":27,"\u63a5\u53e3":62,"\u63a8\u5bfc\u65b9\u7a0b":44,"\u63a8\u6d4b\u6267\u884c":25,"\u63d0\u4ea4":43,"\u63d0\u4ea4\u955c\u50cf":55,"\u63d0\u53d6\u7535\u5f71\u6216\u7528\u6237\u7684\u7279\u5f81\u5e76\u751f\u6210python\u5bf9\u8c61":66,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":38,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":26,"\u6570\u636e\u51c6\u5907":[61,66,69],"\u6570\u636e\u63cf\u8ff0":67,"\u6570\u636e\u63d0\u4f9b":67,"\u6570\u636e\u63d0\u4f9b\u5668":53,"\u6570\u636e\u63d0\u4f9b\u811a\u672c":66,"\u6570\u636e\u652f\u6301":50,"\u6570\u636e\u683c\u5f0f\u51c6\u5907":64,"\u6570\u636e\u6e90\u914d\u7f6e":53,"\u6570\u636e\u7684\u51c6\u5907\u548c\u9884\u5904\u7406":60,"\u6570\u636e\u96c6\u7279\u5f81":65,"\u6570\u636e\u9884\u5904\u7406":69,"\u6570\u6910\u4ecb\u7ecd":68,"\u6570\u6910\u51c6\u5907":68,"\u6574\u4f53\u65b9\u6848":56,"\u6587\u4ef6\u9884\u5904\u7406":26,"\u6587\u672c\u751f\u6210":69,"\u6587\u672c\u751f\u6210\u6559\u7a0b":69,"\u6587\u6863":[34,59],"\u65b0\u624b\u5165\u95e8":37,"\u65f6\u5e8f\u6a21\u578b":64,"\u65f6\u5e8f\u6a21\u578b\u7684\u4f7f\u7528\u573a\u666f":3,"\u65f6\u95f4\u5e8f\u5217":39,"\u65f6\u95f4\u6b65":39,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":28,"\u672c\u5730\u6d4b\u8bd5":52,"\u672c\u5730\u8bad\u7ec3":52,"\u672f\u8bed":25,"\u6784\u5efa\u548c\u6d4b\u8bd5":43,"\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":55,"\u67e5\u770b\u8f93\u51fa":56,"\u6837\u4f8b\u6570\u636e":3,"\u6848\u4f8b\u4e00":52,"\u6848\u4f8b\u4e8c":52,"\u68c0\u67e5\u6a21\u578b\u8f93\u51fa":48,"\u68c0\u67e5\u96c6\u7fa4\u8bad\u7ec3\u7ed3\u679c":48,"\u6982\u8ff0":[38,41],"\u6a21\u578b":62,"\u6a21\u578b\u4e0b\u8f7d":62,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":25,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":28,"\u6a21\u578b\u68c0\u9a8c":32,"\u6a21\u578b\u7f51\u7edc\u7ed3\u6784":64,"\u6a21\u578b\u8bad\u7ec3":[61,69],"\u6a21\u578b\u8bc4\u4f30\u548c\u9884\u6d4b":66,"\u6a21\u578b\u914d\u7f6e":[39,46],"\u6a21\u578b\u914d\u7f6e\u7684\u6a21\u578b\u914d\u7f6e":39,"\u6ce8\u610f\u4e8b\u9879":3,"\u6d41\u7a0b\u4ecb\u7ecd":26,"\u6d4b\u8bd5":[50,67],"\u6d4b\u8bd5\u6587\u4ef6":66,"\u6d4b\u8bd5\u6a21\u578b":68,"\u7279\u5f81":67,"\u7279\u5f81\u63d0\u53d6":62,"\u751f\u6210\u5e8f\u5217":42,"\u751f\u6210\u6a21\u578b\u7684\u547d\u4ee4\u4e0e\u7ed3\u679c":69,"\u751f\u6210\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":41,"\u7528\u6237\u6587\u4ef6\u63cf\u8ff0":65,"\u7528\u6237\u81ea\u5b9a\u4e49\u6570\u636e\u96c6":69,"\u7528\u6237\u81ea\u5b9a\u4e49\u6570\u6910\u9884\u5904\u7406":68,"\u7535\u5f71\u6587\u4ef6\u63cf\u8ff0":65,"\u76ee\u5f55\u7ed3\u6784":28,"\u76f4\u63a5\u6784\u5efapaddlepaddle\u7684\u6587\u6863":45,"\u76f8\u5173\u6982\u5ff5":41,"\u77e9\u9635":50,"\u793a\u4f8b1":39,"\u793a\u4f8b2":39,"\u793a\u4f8b3":39,"\u793a\u4f8b4":39,"\u793a\u4f8b\u7a0b\u5e8f":26,"\u795e\u7ecf\u5143\u6fc0\u6d3b\u5185\u5b58":31,"\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e":66,"\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":67,"\u7a00\u758f\u8bad\u7ec3":52,"\u7b26\u53f7":27,"\u7b80\u4ecb":[54,69],"\u7b80\u5355\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":42,"\u7c7b":[27,44],"\u7cfb\u7edf\u6846\u56fe":53,"\u7ec3\u4e60":61,"\u7ec6\u8282\u63a2\u7a76":61,"\u7ec6\u8282\u63cf\u8ff0":50,"\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":48,"\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u4efb\u52a1":32,"\u7f16\u5199yaml\u6587\u4ef6":55,"\u7f16\u8bd1\u6d41\u7a0b":35,"\u7f16\u8bd1\u9009\u9879":28,"\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":33,"\u7f29\u5bb9":25,"\u7f51\u7edc\u53ef\u89c6\u5316":62,"\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e":53,"\u7f51\u7edc\u914d\u7f6e\u4e2d\u7684\u8c03\u7528":3,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":27,"\u80cc\u666f":27,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":50,"\u81f4\u8c22":0,"\u89c2\u6d4b\u8bcd\u5411\u91cf":60,"\u8bad\u7ec3":[50,66,67],"\u8bad\u7ec3\u5668\u914d\u7f6e\u6587\u4ef6":66,"\u8bad\u7ec3\u56e0\u6b64\u9000\u51fa\u600e\u4e48\u529e":31,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8":26,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":26,"\u8bad\u7ec3\u6a21\u578b":[32,64,68],"\u8bad\u7ec3\u6a21\u578b\u7684\u547d\u4ee4\u4e0e\u7ed3\u679c":69,"\u8bad\u7ec3\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":41,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0":31,"\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6":53,"\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570":51,"\u8bc4\u5206\u6587\u4ef6\u63cf\u8ff0":65,"\u8bcd\u5411\u91cf\u6a21\u578b":64,"\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u4fee\u6b63":60,"\u8bcd\u6c47\u8868":39,"\u8be6\u7ec6\u6559\u7a0b":47,"\u8bed\u4e49\u89d2\u8272\u6807\u6ce8\u6559\u7a0b":67,"\u8bfb\u53d6\u53cc\u5c42\u5e8f\u5217\u6570\u636e":39,"\u8f6c\u6362\u5e93":26,"\u8f93\u5165":41,"\u8f93\u5165\u4e0d\u7b49\u957f":39,"\u8f93\u5165\u793a\u4f8b":41,"\u8f93\u51fa":41,"\u8f93\u51fa\u65e5\u5fd7":64,"\u8fd0\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":48,"\u8fd0\u884c\u5bb9\u5668":55,"\u8fd0\u884cdocker":31,"\u8fd0\u884cpaddlepaddl":34,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":27,"\u8fdb\u884c\u8bad\u7ec3":[26,55],"\u8fdb\u9636\u6307\u5357":46,"\u9009\u62e9\u5b58\u50a8\u65b9\u6848":54,"\u901a\u7528":50,"\u901a\u8fc7docker\u5bb9\u5668\u5f00\u53d1paddlepaddl":34,"\u903b\u8f91\u56de\u5f52\u6a21\u578b":64,"\u9047\u5230":31,"\u90e8\u7f72kubernetes\u96c6\u7fa4":54,"\u914d\u7f6e\u4e2d\u7684\u6570\u636e\u52a0\u8f7d\u5b9a\u4e49":64,"\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":42,"\u914d\u7f6ekubectl":54,"\u914d\u7f6ekubectl\u8bbf\u95ee\u4f60\u7684kubernetes\u96c6\u7fa4":54,"\u94a9\u5b50":43,"\u9644\u5f55":64,"\u968f\u673a\u6570":50,"\u96c6\u7fa4\u8bad\u7ec3":52,"\u975e\u6cd5\u6307\u4ee4":31,"\u9884\u5904\u7406":61,"\u9884\u5904\u7406\u547d\u4ee4\u548c\u7ed3\u679c":69,"\u9884\u5904\u7406\u5de5\u4f5c\u6d41\u7a0b":69,"\u9884\u6d4b":[61,62,64,67,68],"\u9884\u6d4b\u6d41\u7a0b":5,"\u9884\u6d4bdemo":5,"\u9884\u8bad\u7ec3\u7684\u6a21\u578b":69,"api\u4e2d\u6587\u624b\u518c":4,"api\u5bf9\u6bd4\u4ecb\u7ecd":39,"beam_search\u7684\u751f\u6210":39,"blas\u8def\u5f84\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":33,"bleu\u8bc4\u4f30":69,"book\u4e2d\u6240\u6709\u7ae0\u8282":30,"bool\u578b\u7684\u7f16\u8bd1\u9009\u9879":33,"cmake\u6e90\u7801\u7f16\u8bd1":31,"cudnn\u7684\u7f16\u8bd1\u9009\u9879":33,"dataprovider\u7684\u4ecb\u7ecd":2,"dataprovider\u7684\u4f7f\u7528":3,"float":31,"gpu\u548ccpu\u6df7\u5408\u4f7f\u7528":52,"gpu\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18":47,"gpu\u955c\u50cf\u51fa\u73b0":31,"group\u6559\u7a0b":41,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":56,"kubernetes\u5355\u673a\u8bad\u7ec3":55,"meta\u6587\u4ef6":66,"meta\u914d\u7f6e\u6587\u4ef6":66,"mnist\u7684\u4f7f\u7528\u573a\u666f":3,"movielens\u6570\u636e\u96c6":65,"movielens\u6570\u636e\u96c6\u8bc4\u5206\u56de\u5f52\u6a21\u578b":66,"org\u6587\u6863":45,"paddle\u52a8\u6001\u5e93\u4e2d":27,"paddle\u53d1\u884c\u89c4\u8303":30,"paddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":30,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":27,"paddlepaddle\u53d1\u5e03\u7684docker\u955c\u50cf\u4f7f\u7528\u8bf4\u660e":34,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":33,"paddlepaddle\u7684docker\u5bb9\u5668\u4f7f\u7528\u65b9\u5f0f":34,"pod\u95f4\u901a\u4fe1":56,"pydataprovider2\u7684\u4f7f\u7528":3,"python\u63a5\u53e3":62,"python\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86":31,"python\u811a\u672c\u8bfb\u53d6\u6570\u636e":64,"return":29,"rnn\u76f8\u5173\u6a21\u578b":40,"rnn\u914d\u7f6e":42,"so\u627e\u4e0d\u5230":36,"ubuntu\u90e8\u7f72paddlepaddl":36,Abs:14,absactiv:6,activ:[6,14],adadelta:18,adadeltaoptim:12,adagrad:18,adagradoptim:12,adam:18,adamax:18,adamaxoptim:12,adamoptim:12,addto:16,addto_lay:10,aggreg:[10,16],algorithm:24,api:[1,4,28],applic:4,argument:29,async:50,attent:42,attribut:[7,15],auc_evalu:9,avg:19,avgpool:13,base:[9,10],baseactiv:6,basepool:19,basepoolingtyp:13,basesgdoptim:12,batch:29,batch_norm:16,batch_norm_lay:10,batch_siz:29,beam_search:[10,16],becaus:31,between:23,bidirectional_lstm:[11,17],big:31,bilinear_interp:16,bilinear_interp_lay:10,bla:33,block_expand:16,block_expand_lay:10,book:34,brelu:14,breluactiv:6,cach:3,capi:28,capi_priv:28,check:[10,16],checkpoint:[24,25],chunk_evalu:9,cifar:20,classif:9,classification_error_evalu:9,classification_error_printer_evalu:9,clone:43,column_sum_evalu:9,commit:43,compos:29,concat:16,concat_lay:10,config:4,configur:21,conll05:20,connect:[10,16],content:[3,28,31,38,47,53],context_project:[10,16],conv:[10,16],conv_oper:[10,16],conv_project:[10,16],conv_shift:16,conv_shift_lay:10,cos_sim:[10,16],cost:[10,16],cp27mu:31,creat:29,creator:29,crf:16,crf_decod:16,crf_decoding_lay:10,crf_layer:10,cross_channel_norm:16,cross_entropi:10,cross_entropy_cost:16,cross_entropy_with_selfnorm:10,cross_entropy_with_selfnorm_cost:16,ctc:16,ctc_error_evalu:9,ctc_layer:10,cuda:[31,33],cudnn:33,cudnnavg:19,cudnnmax:19,custom:29,dat:65,data:[10,16,20,24,29],data_lay:10,datafeed:20,dataprovid:[4,50],dataset:[20,24],datasourc:8,datatyp:20,decayedadagrad:18,decayedadagradoptim:12,decor:29,design:[23,24,29],dictionari:29,dispatch:24,distribut:[23,24],doc:[23,24,29],dotmul_oper:[10,16],dotmul_project:[10,16],driver:31,dropout_lay:[11,17],dylib:28,dynam:24,embed:16,embedding_lay:10,entri:29,eos:16,eos_lay:10,evalu:9,event:[22,23],exampl:[23,28],except:31,exp:14,expactiv:6,expand:[16,38],expand_lay:10,faq:31,fault:24,fc_layer:10,first_seq:[10,16,38],fork:43,format:24,from:23,full_matrix_project:[10,16],fulli:[10,16],gate:42,get_output:16,get_output_lay:10,gpu:50,gradient_printer_evalu:9,group:[10,16],gru:[11,17,50],gru_group:[11,17],gru_step:16,gru_step_lay:10,gru_unit:[11,17],grumemori:[10,16],handler:[23,27],how:29,hsigmoid:[10,16],huber_cost:[10,16],ident:14,identity_project:[10,16],identityactiv:6,illeg:31,imag:[10,11,16,17],imagenet:62,imdb:[20,68],img_cmrnorm:16,img_cmrnorm_lay:10,img_conv:16,img_conv_bn_pool:[11,17],img_conv_group:[11,17],img_conv_lay:10,img_pool:16,img_pool_lay:10,imikolov:20,implement:29,infer:22,ingredi:23,init_hook:3,input_typ:3,instruct:31,insuffici:31,interfac:[20,24,29],interpol:16,interpolation_lay:10,isn:29,issu:43,job:[24,55],join:[10,16],kubernet:[54,55],lambda_cost:[10,16],last_seq:[10,16,38],layer:[10,16,23],layeroutput:10,layertyp:10,libcudart:36,libcudnn:36,libpaddle_capi_shar:28,libpaddle_capi_whol:28,linear:14,linear_comb:16,linear_comb_lay:10,linearactiv:6,linux_x86_64:31,list:[25,29],log:14,logactiv:6,lstm:[11,17,50,67,68],lstm_step:16,lstm_step_lay:10,lstmemori:[10,16],lstmemory_group:[11,17],lstmemory_unit:[11,17],map:29,master:24,math:[10,16],max:19,maxframe_printer_evalu:9,maxid:16,maxid_lay:10,maxid_printer_evalu:9,maxout:16,maxout_lay:10,maxpool:13,memori:[10,16,39,41],messag:31,mini:29,minibatch:20,misc:[11,17],mix:[10,16],mixed_lay:10,mnist:20,model:[4,21,23,42,62],momentum:18,momentumoptim:12,movi:65,movielen:20,mse_cost:[10,16],multi_binary_label_cross_entropi:10,multi_binary_label_cross_entropy_cost:16,multipl:29,nce:16,nce_lay:10,need:29,network:[11,17,42],neural:42,nlp:[11,17,50],norm:[10,16],nvprof:47,nvvp:47,object:24,onli:29,optim:[12,18,24],output:11,pad:16,pad_lay:10,paddl:[29,30],paddlepaddl:[23,45,59],parallel_nn:52,paramet:[7,15,22,23,24],perform:50,platform:31,pnpair_evalu:9,point:31,pool:[10,13,16,19,38],pooling_lay:10,power:16,power_lay:10,pre:43,precision_recall_evalu:9,prefetch:29,print:9,process:24,protocol:31,provid:[3,29],pull:43,push:43,python:29,queue:24,rank:9,rank_cost:[10,16],rate:65,reader:[20,23,29],recoveri:24,recurr:[10,11,16,17,41,42],recurrent_group:[10,16],recurrent_lay:10,refer:3,reject:31,relu:14,reluactiv:6,repeat:16,repeat_lay:10,request:43,reshap:[10,16],resnet:62,rmsprop:18,rmspropoptim:12,rnn:[39,50],rotat:16,rotate_lay:10,sampl:[10,16],sampling_id:16,sampling_id_lay:10,scale:[16,24],scaling_lay:10,scaling_project:[10,16],selective_fc:16,selective_fc_lay:10,sentiment:20,seq_concat:16,seq_concat_lay:10,seq_reshap:16,seq_reshape_lay:10,seqtext_printer_evalu:9,sequenc:42,sequence_conv_pool:[11,17],sequencesoftmax:14,sequencesoftmaxactiv:6,server:24,set:12,sgd:50,share:23,shuffl:29,sigmoid:14,sigmoidactiv:6,simple_attent:[11,17],simple_gru:[11,17],simple_img_conv_pool:[11,17],simple_lstm:[11,17],singl:29,slice:[10,16],slope_intercept:16,slope_intercept_lay:10,smooth_l1_cost:[10,16],softmax:14,softmaxactiv:6,softrelu:14,softreluactiv:6,spp:16,spp_layer:10,squar:14,squareactiv:6,squarerootn:19,squarerootnpool:13,stack:68,stanh:14,stanhactiv:6,start:23,store:24,suffici:29,sum:19,sum_cost:[10,16],sum_evalu:9,sum_to_one_norm:16,sum_to_one_norm_lay:10,summar:23,sumpool:13,support:31,tabl:28,table_project:[10,16],take:29,tanh:14,tanhactiv:6,task:24,tensor:16,tensor_lay:10,text_conv_pool:[11,17],thi:31,todo:[25,26],toler:24,too:31,train:[22,23,24,29],trainer:[22,24],tran:16,trans_full_matrix_project:[10,16],trans_lay:10,tune:50,uci_h:20,updat:23,usag:29,use:29,user:[24,65],util:9,value_printer_evalu:9,version:31,vgg_16_network:[11,17],warp_ctc:16,warp_ctc_lay:10,wheel:31,whl:31,why:29,wmt14:20,zoo:62}})
\ No newline at end of file
+Search.setIndex({docnames:["about/index_cn","api/index_cn","api/v1/data_provider/dataprovider_cn","api/v1/data_provider/pydataprovider2_cn","api/v1/index_cn","api/v1/predict/swig_py_paddle_cn","api/v2/config/activation","api/v2/config/attr","api/v2/config/evaluators","api/v2/config/layer","api/v2/config/networks","api/v2/config/optimizer","api/v2/config/pooling","api/v2/data","api/v2/model_configs","api/v2/run_logic","design/api","design/cluster_train/README","design/cluster_train/checkpointing","design/cluster_train/data_dispatch","design/multi_language_interface/00.why_plain_c","design/multi_language_interface/01.inference_implementation","design/reader/README","design/releasing_process","faq/index_cn","getstarted/basic_usage/index_cn","getstarted/build_and_install/cmake/build_from_source_cn","getstarted/build_and_install/docker_install_cn","getstarted/build_and_install/index_cn","getstarted/build_and_install/ubuntu_install_cn","getstarted/index_cn","howto/deep_model/rnn/hierarchical_layer_cn","howto/deep_model/rnn/hrnn_rnn_api_compare_cn","howto/deep_model/rnn/index_cn","howto/deep_model/rnn/recurrent_group_cn","howto/deep_model/rnn/rnn_config_cn","howto/dev/contribute_to_paddle_cn","howto/dev/new_layer_cn","howto/dev/write_docs_cn","howto/index_cn","howto/optimization/gpu_profiling_cn","howto/usage/cluster/cluster_train_cn","howto/usage/cmd_parameter/arguments_cn","howto/usage/cmd_parameter/detail_introduction_cn","howto/usage/cmd_parameter/index_cn","howto/usage/cmd_parameter/use_case_cn","howto/usage/concepts/use_concepts_cn","howto/usage/k8s/k8s_basis_cn","howto/usage/k8s/k8s_cn","howto/usage/k8s/k8s_distributed_cn","howto/usage/k8s/src/k8s_data/README","howto/usage/k8s/src/k8s_train/README","index_cn","tutorials/embedding_model/index_cn","tutorials/image_classification/index_cn","tutorials/imagenet_model/resnet_model_cn","tutorials/index_cn","tutorials/quick_start/index_cn","tutorials/rec/ml_dataset_cn","tutorials/rec/ml_regression_cn","tutorials/semantic_role_labeling/index_cn","tutorials/sentiment_analysis/index_cn","tutorials/text_generation/index_cn"],envversion:50,filenames:["about/index_cn.md","api/index_cn.rst","api/v1/data_provider/dataprovider_cn.rst","api/v1/data_provider/pydataprovider2_cn.rst","api/v1/index_cn.rst","api/v1/predict/swig_py_paddle_cn.rst","api/v2/config/activation.rst","api/v2/config/attr.rst","api/v2/config/evaluators.rst","api/v2/config/layer.rst","api/v2/config/networks.rst","api/v2/config/optimizer.rst","api/v2/config/pooling.rst","api/v2/data.rst","api/v2/model_configs.rst","api/v2/run_logic.rst","design/api.md","design/cluster_train/README.md","design/cluster_train/checkpointing.md","design/cluster_train/data_dispatch.md","design/multi_language_interface/00.why_plain_c.md","design/multi_language_interface/01.inference_implementation.md","design/reader/README.md","design/releasing_process.md","faq/index_cn.rst","getstarted/basic_usage/index_cn.rst","getstarted/build_and_install/cmake/build_from_source_cn.rst","getstarted/build_and_install/docker_install_cn.rst","getstarted/build_and_install/index_cn.rst","getstarted/build_and_install/ubuntu_install_cn.rst","getstarted/index_cn.rst","howto/deep_model/rnn/hierarchical_layer_cn.rst","howto/deep_model/rnn/hrnn_rnn_api_compare_cn.rst","howto/deep_model/rnn/index_cn.rst","howto/deep_model/rnn/recurrent_group_cn.md","howto/deep_model/rnn/rnn_config_cn.rst","howto/dev/contribute_to_paddle_cn.md","howto/dev/new_layer_cn.rst","howto/dev/write_docs_cn.rst","howto/index_cn.rst","howto/optimization/gpu_profiling_cn.rst","howto/usage/cluster/cluster_train_cn.md","howto/usage/cmd_parameter/arguments_cn.md","howto/usage/cmd_parameter/detail_introduction_cn.md","howto/usage/cmd_parameter/index_cn.rst","howto/usage/cmd_parameter/use_case_cn.md","howto/usage/concepts/use_concepts_cn.rst","howto/usage/k8s/k8s_basis_cn.md","howto/usage/k8s/k8s_cn.md","howto/usage/k8s/k8s_distributed_cn.md","howto/usage/k8s/src/k8s_data/README.md","howto/usage/k8s/src/k8s_train/README.md","index_cn.rst","tutorials/embedding_model/index_cn.md","tutorials/image_classification/index_cn.md","tutorials/imagenet_model/resnet_model_cn.md","tutorials/index_cn.md","tutorials/quick_start/index_cn.rst","tutorials/rec/ml_dataset_cn.md","tutorials/rec/ml_regression_cn.rst","tutorials/semantic_role_labeling/index_cn.md","tutorials/sentiment_analysis/index_cn.md","tutorials/text_generation/index_cn.md"],objects:{},objnames:{},objtypes:{},terms:{"00012\u7684\u6a21\u578b\u6709\u7740\u6700\u9ad8\u7684bleu\u503c27":62,"0005\u4e58\u4ee5batch":54,"000\u4e2a\u5df2\u6807\u6ce8\u8fc7\u7684\u9ad8\u6781\u6027\u7535\u5f71\u8bc4\u8bba\u7528\u4e8e\u8bad\u7ec3":61,"000\u4e2a\u7528\u4e8e\u6d4b\u8bd5":61,"000\u4e2atxt\u6587\u4ef6":61,"000\u4f4d\u7528\u6237\u5bf94":58,"000\u5e45\u56fe\u50cf\u4e0a\u6d4b\u8bd5\u4e86\u6a21\u578b\u7684\u5206\u7c7b\u9519\u8bef\u7387":55,"000\u5f20\u7070\u5ea6\u56fe\u7247\u7684\u6570\u5b57\u5206\u7c7b\u6570\u636e\u96c6":3,"000\u6761\u8bc4\u4ef7":58,"000\u90e8\u7535\u5f71\u76841":58,"00186201e":5,"00m":40,"02595v1":9,"03m":40,"0424m":40,"0473v3":10,"05d":54,"0630u":40,"06u":40,"0810u":40,"08823112e":5,"0957m":40,"0\u53f7\u8bad\u7ec3\u8282\u70b9\u662f\u4e3b\u8bad\u7ec3\u8282\u70b9":43,"0\u5c42\u5e8f\u5217":31,"0\u8868\u793a\u7b2c\u4e00\u6b21\u7ecf\u8fc7\u8bad\u7ec3\u96c6":61,"0ab":9,"0b1":29,"0rc1":[23,27],"0rc2":23,"10000\u5f20\u4f5c\u4e3a\u6d4b\u8bd5\u96c6":54,"10007_10":61,"10014_7":61,"100m":24,"101\u5c42\u548c152\u5c42\u7684\u7f51\u7edc\u7ed3\u6784\u4e2d":55,"101\u5c42\u548c152\u5c42\u7684\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6\u53ef\u53c2\u7167":55,"101\u5c42\u7f51\u7edc\u6a21\u578b":55,"10\u4e2d\u7684\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6":54,"10\u6570\u636e\u96c6":54,"10\u6570\u636e\u96c6\u5305\u542b60000\u5f2032x32\u7684\u5f69\u8272\u56fe\u7247":54,"10\u6570\u636e\u96c6\u7684\u5b98\u65b9\u7f51\u5740":54,"10\u6570\u636e\u96c6\u8bad\u7ec3\u4e00\u4e2a\u5377\u79ef\u795e\u7ecf\u7f51\u7edc":54,"1150u":40,"11\u5b9e\u73b0\u4e86c":21,"11e6":48,"12194102e":5,"124n":40,"128\u7ef4\u548c256\u7ef4":53,"13m":48,"1490u":40,"14\u6570\u636e\u96c6":62,"14\u6570\u636e\u96c6\u4e0a\u5f97\u5230\u826f\u597d\u8868\u73b0\u7684\u8bad\u7ec3\u8fc7\u7a0b":62,"14\u8fd9\u79cd\u5199\u6cd5\u5c06\u4f1a\u6d4b\u8bd5\u6a21\u578b":45,"152\u5c42\u7f51\u7edc\u6a21\u578b":55,"15501715e":5,"1550u":40,"15\u884c":32,"1636k":62,"16u":40,"173m":55,"173n":40,"1770u":40,"18\u5c81\u4ee5\u4e0b":58,"18e457ce3d362ff5f3febf8e7f85ffec852f70f3b629add10aed84f930a68750":48,"197u":40,"1\u7684\u5c42\u4e4b\u5916":45,"1\u7a00\u758f\u6570\u636e":37,"1\u8f6e\u5b58\u50a8\u7684\u6240\u6709\u6a21\u578b":45,"1\u9664\u4ee5batch":54,"1m\u6570\u636e\u96c6\u4e2d":59,"1m\u7684\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6\u5728\u76ee\u5f55":59,"200\u6570\u636e\u96c6\u4e0a\u4f7f\u7528vgg\u6a21\u578b\u8bad\u7ec3\u4e00\u4e2a\u9e1f\u7c7b\u56fe\u7247\u5206\u7c7b\u6a21\u578b":54,"210u":40,"211839e770f7b538e2d8":10,"215n":40,"228u":40,"234m":55,"24\u5c81":58,"2520u":40,"25639710e":5,"25k":57,"2680u":40,"26\u884c":32,"27787406e":5,"279n":40,"27m":40,"285m":40,"2863m":40,"28\u7684\u56fe\u7247\u50cf\u7d20\u7070\u5ea6\u503c":3,"28\u7ef4\u7684\u7a20\u5bc6\u6d6e\u70b9\u6570\u5411\u91cf\u548c\u4e00\u4e2a":3,"28m":40,"2977m":40,"29997\u4e2a\u6700\u9ad8\u9891\u5355\u8bcd\u548c3\u4e2a\u7279\u6b8a\u7b26\u53f7":62,"2\u4e09\u7c7b\u7684\u6bd4\u4f8b\u4e3a":24,"2\u4e2a\u6d6e\u70b9\u6570":25,"2\u5206\u522b\u4ee3\u88683\u4e2a\u8282\u70b9\u7684trainer":49,"2\u610f\u5473\u77400\u53f7\u548c1\u53f7gpu\u5c06\u4f1a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97fc1\u548cfc2\u5c42":45,"2\u8fd9\u51e0\u4e2a\u76ee\u5f55\u8868\u793apaddlepaddle\u8282\u70b9\u4e0etrain":49,"2nd":9,"302n":40,"30u":40,"3206326\u4e2a\u8bcd\u548c4\u4e2a\u7279\u6b8a\u6807\u8bb0":53,"32777140e":5,"328n":40,"32\u7ef4":53,"32u":40,"32x32":13,"331n":40,"3320u":40,"34\u5c81":58,"35\u65f6":62,"36540484e":5,"36u":40,"3710m":40,"3768m":40,"387u":40,"38u":40,"3920u":40,"39u":40,"3\u4e2a\u7279\u6b8a\u7b26\u53f7":62,"3\u53f7gpu":24,"4035m":40,"4090u":40,"4096mb":43,"4279m":40,"43630644e":5,"43u":40,"448a5b355b84":48,"44\u5c81":58,"4560u":40,"4563m":40,"45u":40,"4650u":40,"4726m":40,"473m":48,"48565123e":5,"48684503e":5,"49316648e":5,"49\u5c81":58,"4gb":43,"500\u4e2atxt\u6587\u4ef6":61,"500m":24,"50\u5c42":55,"50\u5c42\u7f51\u7edc\u6a21\u578b":55,"51111044e":5,"514u":40,"525n":40,"526u":40,"53018653e":5,"536u":40,"5460u":40,"5470u":40,"54u":40,"55\u5c81":58,"55g":62,"5690m":40,"573u":40,"578n":40,"5798m":40,"586u":40,"58s":48,"5969m":40,"5\u4e2a\u6d4b\u8bd5\u6837\u4f8b\u548c2\u4e2a\u751f\u6210\u5f0f\u6837\u4f8b":53,"5\u5230\u672c\u5730\u73af\u5883\u4e2d":29,"6080u":40,"6082v4":9,"6140u":40,"6305m":40,"639u":40,"64\u7ef4":53,"655u":40,"6780u":40,"6810u":40,"682u":40,"6970u":40,"6\u4e07\u4ebf\u6b21\u6d6e\u70b9\u8fd0\u7b97\u6bcf\u79d2":40,"6\u4e2a\u8282\u70b9":41,"6\u5143\u4e0a\u4e0b\u6587\u4f5c\u4e3a\u8f93\u5165\u5c42":53,"704u":40,"70634608e":5,"7090u":40,"72296313e":5,"72u":40,"73u":40,"75u":40,"760u":40,"767u":40,"783n":40,"784u":40,"78m":40,"7eamaa":13,"7kb":48,"8250u":40,"8300u":40,"830n":40,"849m":40,"85625684e":5,"861u":40,"864k":62,"8661m":40,"877\u4e2a\u88ab\u5411\u91cf\u5316\u7684\u8bcd":53,"877\u884c":53,"892m":40,"901n":40,"90u":40,"918u":40,"9247m":40,"924n":40,"9261m":40,"93137714e":5,"9330m":40,"94u":40,"9530m":40,"96644767e":5,"983m":40,"988u":40,"997u":40,"99982715e":5,"99m":55,"99u":40,"9\u4e2d\u7684\u4e00\u4e2a\u6570\u5b57":3,"9f18":48,"\u0233":25,"\u03b5":25,"\u4e00":32,"\u4e00\u4e2a":46,"\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a0\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u4f8b\u5b50\u662f\u623f\u4ea7\u4f30\u503c":25,"\u4e00\u4e2a\u5178\u578b\u7684\u795e\u7ecf\u7f51\u7edc\u5982\u4e0b\u56fe\u6240\u793a":54,"\u4e00\u4e2a\u5206\u5e03\u5f0f\u4f5c\u4e1a\u91cc\u5305\u62ec\u82e5\u5e72trainer\u8fdb\u7a0b\u548c\u82e5\u5e72paramet":46,"\u4e00\u4e2a\u5206\u5e03\u5f0f\u7684\u5b58\u50a8\u7cfb\u7edf":47,"\u4e00\u4e2a\u5206\u5e03\u5f0fpaddle\u8bad\u7ec3\u4efb\u52a1\u4e2d\u7684\u6bcf\u4e2a\u8fdb\u7a0b\u90fd\u53ef\u4ee5\u4ececeph\u8bfb\u53d6\u6570\u636e":48,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8fdb\u5165":34,"\u4e00\u4e2a\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u5305\u542b\u5982\u4e0b\u5c42":54,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8fdb\u5165":34,"\u4e00\u4e2a\u53cc\u5c42rnn\u7531\u591a\u4e2a\u5355\u5c42rnn\u7ec4\u6210":34,"\u4e00\u4e2a\u53ef\u8c03\u7528\u7684\u51fd\u6570":34,"\u4e00\u4e2a\u57fa\u672c\u7684\u5e94\u7528\u573a\u666f\u662f\u533a\u5206\u7ed9\u5b9a\u6587\u672c\u7684\u8912\u8d2c\u4e24\u6781\u6027":61,"\u4e00\u4e2a\u6216\u591a\u4e2a":47,"\u4e00\u4e2a\u6570\u636e\u96c6\u5927\u90e8\u5206\u5e8f\u5217\u957f\u5ea6\u662f100":24,"\u4e00\u4e2a\u6587\u4ef6":3,"\u4e00\u4e2a\u662f\u6d6e\u70b9\u8ba1\u7b97\u91cf":40,"\u4e00\u4e2a\u662f\u76f4\u63a5\u628a\u4e00\u4e2a\u5bb9\u5668\u8f6c\u6362\u6210\u955c\u50cf":27,"\u4e00\u4e2a\u72ec\u7acb\u7684\u5143\u7d20":31,"\u4e00\u4e2a\u72ec\u7acb\u7684\u8bcd\u8bed":31,"\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u6a21\u578b\u7531\u5927\u91cf\u7684\u53c2\u6570\u7ec4\u6210":18,"\u4e00\u4e2a\u7b80\u5355\u7684\u4f8b\u5b50\u5982":61,"\u4e00\u4e2a\u7b80\u5355\u7684\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6\u4e3a":46,"\u4e00\u4e2a\u7ec8\u7aef\u8fd0\u884cvi":27,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad\u90e8\u5206\u628a\u8f93\u5165\u8f6c\u5316\u4e3a\u76f8\u5e94\u7684\u8f93\u51fa":37,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u53c2\u6570\u662f\u5728":37,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684c":37,"\u4e00\u4e2a\u91cd\u8981\u7684\u95ee\u9898\u662f\u9009\u62e9\u6b63\u786e\u7684learning_r":24,"\u4e00\u4e2agpu\u8bbe\u5907\u4e0a\u4e0d\u5141\u8bb8\u914d\u7f6e\u591a\u4e2a\u6a21\u578b":43,"\u4e00\u4e2alabel":32,"\u4e00\u4e2alogging\u5bf9\u8c61":3,"\u4e00\u4e2amemory\u5305\u542b":35,"\u4e00\u4e2apass\u610f\u5473\u7740paddlepaddle\u8bad\u7ec3\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u6837\u672c\u88ab\u904d\u5386\u4e00\u6b21":60,"\u4e00\u4e2apass\u8868\u793a\u8fc7\u4e00\u904d\u6240\u6709\u8bad\u7ec3\u6837\u672c":57,"\u4e00\u4e2apod\u4e2d\u7684\u6240\u6709\u5bb9\u5668\u4f1a\u88ab\u8c03\u5ea6\u5230\u540c\u4e00\u4e2anode\u4e0a":47,"\u4e00\u4e2apserver\u8fdb\u7a0b\u5171\u7ed1\u5b9a\u591a\u5c11\u7aef\u53e3\u7528\u6765\u505a\u7a00\u758f\u66f4\u65b0":46,"\u4e00\u4e9b\u60c5\u51b5\u4e0b":46,"\u4e00\u4e9b\u968f\u673a\u5316\u566a\u58f0\u6dfb\u52a0\u90fd\u5e94\u8be5\u5728dataprovider\u4e2d\u5b8c\u6210":46,"\u4e00\u4eba":32,"\u4e00\u53e5\u8bdd\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":34,"\u4e00\u53f0\u673a\u5668\u4e0a\u9762\u7684\u7ebf\u7a0b\u6570\u91cf":59,"\u4e00\u65e9":32,"\u4e00\u662fbatch":24,"\u4e00\u6761\u6837\u672c":3,"\u4e00\u6837\u8bbe\u7f6e":41,"\u4e00\u6b21\u4f5c\u4e1a\u79f0\u4e3a\u4e00\u4e2ajob":47,"\u4e00\u6b21\u6027\u676f\u5b50":32,"\u4e00\u6b21yield\u8c03\u7528":3,"\u4e00\u79cd\u5e38\u7528\u7684\u505a\u6cd5\u662f\u7528\u5b66\u4e60\u7684\u6a21\u578b\u5bf9\u53e6\u5916\u4e00\u7ec4\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u9884\u6d4b":25,"\u4e00\u7bc7\u8bba\u6587":62,"\u4e00\u7ea7\u76ee\u5f55":[61,62],"\u4e00\u81f4":[31,32],"\u4e00\u822c\u4e0d\u5141\u8bb8\u518d\u4ece":23,"\u4e00\u822c\u4ece":36,"\u4e00\u822c\u5728paddlepaddle\u4e2d":32,"\u4e00\u822c\u60c5\u51b5\u4e0b":[2,25],"\u4e00\u822c\u63a8\u8350\u8bbe\u7f6e\u6210true":3,"\u4e00\u822c\u662f\u5c01\u88c5\u4e86\u8bb8\u591a\u590d\u6742\u64cd\u4f5c\u7684\u96c6\u5408":46,"\u4e00\u822c\u662f\u7531\u4e8e\u76f4\u63a5\u4f20\u9012\u5927\u5b57\u5178\u5bfc\u81f4\u7684":24,"\u4e00\u822c\u6765\u8bf4":35,"\u4e00\u822c\u800c\u8a00":62,"\u4e00\u822c\u8868\u793a":32,"\u4e00\u884c\u4e3a\u4e00\u4e2a\u6837\u672c":57,"\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":3,"\u4e09\u7ea7\u76ee\u5f55":[61,62],"\u4e0a":36,"\u4e0a\u4e0b\u6587\u5927\u5c0f\u8bbe\u7f6e\u4e3a1\u7684\u4e00\u4e2a\u6837\u672c\u7684\u7279\u5f81\u5982\u4e0b":60,"\u4e0a\u53d1\u8868\u7684\u8bc4\u8bba\u5206\u6210\u6b63\u9762\u8bc4\u8bba\u548c\u8d1f\u9762\u8bc4\u8bba\u4e24\u7c7b":61,"\u4e0a\u56fe\u4e2d\u865a\u7ebf\u7684\u8fde\u63a5":32,"\u4e0a\u56fe\u63cf\u8ff0\u4e86\u4e00\u4e2a3\u8282\u70b9\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3\u573a\u666f":49,"\u4e0a\u7f51":32,"\u4e0a\u8fd0\u884c":27,"\u4e0a\u8ff0\u4ee3\u7801\u5c06bias\u5168\u90e8\u521d\u59cb\u5316\u4e3a1":24,"\u4e0a\u8ff0\u7684\u4ee3\u7801\u7247\u6bb5\u5305\u542b\u4e86\u4e24\u79cd\u65b9\u6cd5":40,"\u4e0a\u8ff0\u811a\u672c\u4f7f\u7528":41,"\u4e0b":54,"\u4e0b\u56fe\u4e2d\u5c31\u5c55\u793a\u4e86\u4e00\u4e9b\u5173\u4e8e\u5185\u5b58\u6570\u636e\u8fc1\u5f99\u548c\u8ba1\u7b97\u8d44\u6e90\u5229\u7528\u7387\u7684\u5efa\u8bae":40,"\u4e0b\u56fe\u5c55\u793a\u4e86\u6240\u6709\u7684\u56fe\u7247\u7c7b\u522b":54,"\u4e0b\u56fe\u5c55\u793a\u4e86\u65f6\u95f4\u6269\u5c55\u76842\u5c42":60,"\u4e0b\u56fe\u5c55\u793a\u7684\u662f\u57fa\u4e8e\u6b8b\u5dee\u7684\u8fde\u63a5\u65b9\u5f0f":55,"\u4e0b\u56fe\u63cf\u8ff0\u4e86\u7528\u6237\u4f7f\u7528\u6846\u56fe":46,"\u4e0b\u56fe\u662f\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u793a\u610f\u56fe":37,"\u4e0b\u6587\u4ee5nlp\u4efb\u52a1\u4e3a\u4f8b":34,"\u4e0b\u6587\u4f7f\u7528":49,"\u4e0b\u6587\u5c31\u662f\u7528job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u8fdb\u884c\u8bad\u7ec3":48,"\u4e0b\u6b21":32,"\u4e0b\u7684":49,"\u4e0b\u8868\u5c55\u793a\u4e86batch":55,"\u4e0b\u8f7d\u5b8c\u6570\u636e\u540e":48,"\u4e0b\u8f7d\u5b8c\u76f8\u5173\u5b89\u88c5\u5305\u540e":29,"\u4e0b\u8f7d\u6570\u636e\u96c6":54,"\u4e0b\u8f7dwmt":62,"\u4e0b\u9762\u4e3e\u4e2a\u7b80\u5355\u7684\u4f8b\u5b50":40,"\u4e0b\u9762\u4ecb\u7ecd\u9884\u5904\u7406\u8fc7\u7a0b\u5177\u4f53\u7684\u6b65\u9aa4":59,"\u4e0b\u9762\u5148\u7b80\u8981\u4ecb\u7ecd\u4e00\u4e0b\u672c\u6587\u7528\u5230\u7684\u51e0\u4e2akubernetes\u6982\u5ff5":47,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u6570\u636e\u6e90\u914d\u7f6e":46,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u67d0\u4e00\u7c7b\u6587\u4ef6\u7684\u5b9e\u73b0\u65b9\u5f0f":21,"\u4e0b\u9762\u5217\u51fa\u4e86":35,"\u4e0b\u9762\u5217\u51fa\u4e86\u5168\u8fde\u63a5\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":37,"\u4e0b\u9762\u5c06\u5206\u522b\u4ecb\u7ecd\u8fd9\u4e24\u90e8\u5206":46,"\u4e0b\u9762\u5c31\u6839\u636e\u8fd9\u51e0\u4e2a\u6b65\u9aa4\u5206\u522b\u4ecb\u7ecd":49,"\u4e0b\u9762\u6211\u4eec\u7ed9\u51fa\u4e86\u4e00\u4e2a\u914d\u7f6e\u793a\u4f8b":54,"\u4e0b\u9762\u662f\u4e00\u4e2a\u8bef\u5dee\u66f2\u7ebf\u56fe\u7684\u793a\u4f8b":54,"\u4e0b\u9762\u662fcifar":54,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u5b9e\u73b0\u4e86":37,"\u4e0b\u9762\u7684\u4f8b\u5b50\u4f7f\u7528\u4e86":55,"\u4e0b\u9762\u7684\u4f8b\u5b50\u540c\u6837\u4f7f\u7528\u4e86":55,"\u4e0b\u9762\u7ed9\u51fa\u4e86\u4e00\u4e2a\u4f8b\u5b50":37,"\u4e0b\u9762\u811a\u672c\u7b26\u5408paddlepaddle\u671f\u5f85\u7684\u8bfb\u53d6\u6570\u636e\u7684python\u7a0b\u5e8f\u7684\u6a21\u5f0f":25,"\u4e0b\u9762\u8fd9\u4e9blayer\u80fd\u591f\u63a5\u53d7\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165":31,"\u4e0d":32,"\u4e0d\u4e00\u5b9a\u548c\u65f6\u95f4\u6709\u5173\u7cfb":3,"\u4e0d\u4f1a\u518d\u4ece":24,"\u4e0d\u4f7f\u7528\u9759\u6001\u5e93":20,"\u4e0d\u4f7f\u7528\u989d\u5916\u7a7a\u95f4":37,"\u4e0d\u4f7f\u7528c":20,"\u4e0d\u4f7f\u7528swig":20,"\u4e0d\u5305\u542b\u5728\u5b57\u5178\u4e2d\u7684\u5355\u8bcd":62,"\u4e0d\u540c":60,"\u4e0d\u540c\u4e3b\u673a":47,"\u4e0d\u540c\u4ea7\u54c1":61,"\u4e0d\u540c\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u6570\u636e\u5927\u5c0f\u7684\u6700\u5927\u503c\u4e0e\u6700\u5c0f\u503c\u7684\u6bd4\u7387":43,"\u4e0d\u540c\u5c42\u7684\u7279\u5f81\u7531\u5206\u53f7":55,"\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u662f\u4e0d\u540c\u7684":35,"\u4e0d\u540c\u7248\u672c\u7684\u7f16\u8bd1\u5668\u4e4b\u95f4":20,"\u4e0d\u540c\u7684\u4f18\u5316\u7b97\u6cd5\u9700\u8981\u4f7f\u7528\u4e0d\u540c\u5927\u5c0f\u7684\u5185\u5b58":24,"\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":49,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":3,"\u4e0d\u540c\u7a7a\u95f4\u7684\u8d44\u6e90\u540d\u53ef\u4ee5\u91cd\u590d":47,"\u4e0d\u540c\u8bed\u8a00\u7684\u63a5\u53e3\u9002\u5e94\u4e0d\u540c\u8bed\u8a00\u7684\u7279\u6027":20,"\u4e0d\u540c\u8f93\u5165\u542b\u6709\u7684\u5b50\u53e5":34,"\u4e0d\u540c\u8f93\u5165\u5e8f\u5217\u542b\u6709\u7684\u8bcd\u8bed\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":34,"\u4e0d\u540cdataprovider\u5bf9\u6bd4\u5982\u4e0b":32,"\u4e0d\u540cpod\u4e4b\u95f4\u53ef\u4ee5\u901a\u8fc7ip\u5730\u5740\u8bbf\u95ee":47,"\u4e0d\u542b\u53ef\u5b66\u4e60\u53c2\u6570":46,"\u4e0d\u5728":21,"\u4e0d\u5c11":32,"\u4e0d\u5d4c\u5165\u5176\u4ed6\u8bed\u8a00\u89e3\u91ca\u5668":20,"\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":20,"\u4e0d\u5e94\u8be5\u88ab\u62c6\u89e3":34,"\u4e0d\u5fc5\u518d\u5c06\u4efb\u610f\u957f\u5ea6\u6e90\u8bed\u53e5\u4e2d\u7684\u6240\u6709\u4fe1\u606f\u538b\u7f29\u81f3\u4e00\u4e2a\u5b9a\u957f\u7684\u5411\u91cf\u4e2d":62,"\u4e0d\u6307\u5b9a\u65f6":34,"\u4e0d\u63d0\u4f9b\u5206\u5e03\u5f0f\u5b58\u50a8":47,"\u4e0d\u652f\u6301":60,"\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":3,"\u4e0d\u663e\u793a\u7684\u5199\u6bcf\u4e2a\u7c7b\u5177\u4f53\u5305\u542b\u4ec0\u4e48":20,"\u4e0d\u6ee1\u8db3\u94a9\u5b50\u7684":36,"\u4e0d\u7f13\u5b58\u4efb\u4f55\u6570\u636e":3,"\u4e0d\u80fd\u88ab\u63d0\u4ea4\u5230":36,"\u4e0d\u8fc7":32,"\u4e0d\u8fdc":32,"\u4e0d\u9519":32,"\u4e0d\u9700\u8981\u5bf9\u5e8f\u5217\u6570\u636e\u8fdb\u884c\u4efb\u4f55\u9884\u5904\u7406":35,"\u4e0e":[49,53,62],"\u4e0e\u529f\u80fd\u5206\u652f\u4e0d\u540c\u7684\u662f":23,"\u4e0e\u5355\u5c42rnn\u7684\u914d\u7f6e\u7c7b\u4f3c":32,"\u4e0e\u53ef\u80fd\u6709\u7684":23,"\u4e0e\u5728":60,"\u4e0e\u672c\u5730\u8bad\u7ec3\u76f8\u540c":41,"\u4e0e\u6b64\u4e0d\u540c\u7684\u662f":49,"\u4e0e\u7ffb\u8bd1":62,"\u4e0e\u8bad\u7ec3\u4e0d\u540c":61,"\u4e0e\u8bad\u7ec3\u6a21\u578b\u4e0d\u540c\u7684\u662f":62,"\u4e0e\u8fd9\u4e2a\u8bad\u7ec3\u6570\u636e\u4ea4\u4e92\u7684layer":24,"\u4e0eimdb\u7f51\u7ad9\u63d0\u4f9b\u7684\u4e00\u81f4":58,"\u4e0ejob":49,"\u4e0etime":58,"\u4e14":32,"\u4e14\u589e\u52a0\u4e00\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00":20,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":3,"\u4e14\u652f\u6301\u90e8\u7f72\u5230":47,"\u4e14\u6bcf\u4e2a\u53e5\u5b50\u8868\u793a\u4e3a\u5bf9\u5e94\u7684\u8bcd\u8868\u7d22\u5f15\u6570\u7ec4":32,"\u4e14\u8c03\u7528\u65f6\u4e0d\u80fd\u629b\u51fa\u5f02\u5e38\u6216\u51fa\u73b0\u8fd0\u884c\u65f6\u9519\u8bef":21,"\u4e14\u9ed8\u8ba4\u5728\u8bad\u7ec3\u96c6\u4e0a\u6784\u5efa\u5b57\u5178":61,"\u4e14c99\u652f\u6301bool\u7c7b\u578b\u548c\u5b9a\u957f\u6574\u6570":20,"\u4e14c99\u76f8\u5bf9\u4e8ec11\u4f7f\u7528\u66f4\u52a0\u5e7f\u6cdb":20,"\u4e24":32,"\u4e24\u4e2a\u5217\u8868\u6587\u4ef6":41,"\u4e24\u4e2a\u5b50\u76ee\u5f55\u4e0b":38,"\u4e24\u4e2a\u5d4c\u5957\u7684":34,"\u4e24\u4e2a\u64cd\u4f5c":40,"\u4e24\u4e2a\u6587\u4ef6\u5939":54,"\u4e24\u4e2a\u8f93\u5165\u7279\u5f81\u5728\u8fd9\u4e2a\u6d41\u7a0b\u4e2d\u8d77\u7740\u81f3\u5173\u91cd\u8981\u7684\u4f5c\u7528":60,"\u4e24\u4e2a\u8f93\u5165\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u4e5f\u5e76\u4e0d\u76f8\u540c":32,"\u4e24\u4e2a\u90e8\u5206":38,"\u4e24\u79cd\u7c7b\u522b":57,"\u4e24\u8005\u5747\u4e3a\u7eaf\u6587\u672c\u6587\u4ef6":2,"\u4e2a":57,"\u4e2a\u5185\u5b58\u6c60\u5b9e\u9645\u4e0a\u51b3\u5b9a\u4e86shuffle\u7684\u7c92\u5ea6":24,"\u4e2a\u5355\u8bcd":62,"\u4e2a\u6027\u5316\u63a8\u8350":[23,56],"\u4e2a\u6279\u6b21\u540e\u6253\u5370\u4e00\u4e2a":59,"\u4e2a\u6279\u6b21\u7684\u53c2\u6570\u5e73\u5747\u503c\u8fdb\u884c\u6d4b\u8bd5":43,"\u4e2a\u6a21\u578b\u6d4b\u8bd5\u6570\u636e":43,"\u4e2d":[20,21,24,37,46,49,54,57,59,60,61],"\u4e2d\u4e0d\u8981\u6dfb\u52a0\u5927\u6587\u4ef6\u7b49":36,"\u4e2d\u4ecb\u7ecd\u7684\u65b9\u6cd5":53,"\u4e2d\u4efb\u610f\u7b2ci\u884c\u7684\u53e5\u5b50\u4e4b\u95f4\u90fd\u5fc5\u987b\u6709\u7740\u4e00\u4e00\u5bf9\u5e94\u7684\u5173\u7cfb":62,"\u4e2d\u4efb\u610f\u7b2ci\u884c\u7684\u53e5\u5b50\u4e4b\u95f4\u90fd\u6709\u7740\u4e00\u4e00\u5bf9\u5e94\u7684\u5173\u7cfb":62,"\u4e2d\u5173\u4e8e\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc\u7684\u4ecb\u7ecd":32,"\u4e2d\u5199\u5165json\u5185\u5bb9":18,"\u4e2d\u5305\u542b\u4e86\u8bad\u7ec3\u6a21\u578b\u7684\u57fa\u672c\u547d\u4ee4":57,"\u4e2d\u5305\u542b\u5982\u4e0b\u8868\u6240\u793a\u76843\u4e2a\u6587\u4ef6\u5939":62,"\u4e2d\u5355\u5143\u6d4b\u8bd5\u7684\u4e00\u90e8\u5206":36,"\u4e2d\u5b89\u88c5":41,"\u4e2d\u5b8c\u5168\u4e00\u81f4":20,"\u4e2d\u5b8c\u6210":61,"\u4e2d\u5b9a\u4e49":35,"\u4e2d\u5b9a\u4e49\u4f7f\u7528\u54ea\u79cddataprovid":2,"\u4e2d\u5b9a\u4e49\u548c\u4f7f\u7528":34,"\u4e2d\u5b9e\u73b0\u7684\u7ed3\u6784\u4f53":21,"\u4e2d\u5bfc\u51fa\u9884\u5b9a\u4e49\u7684\u7f51\u7edc":61,"\u4e2d\u5c55\u793a\u4e86\u5982\u4f55\u4f7f\u7528python\u6765\u63d0\u53d6\u7279\u5f81":55,"\u4e2d\u6307\u5b9a":43,"\u4e2d\u6307\u5b9a\u7684\u540d\u5b57":45,"\u4e2d\u6307\u5b9a\u7684\u5c42\u987a\u5e8f\u4e00\u81f4":55,"\u4e2d\u63d0\u51fa\u7684resnet\u7f51\u7edc\u7ed3\u6784\u57282015\u5e74imagenet\u5927\u89c4\u6a21\u89c6\u89c9\u8bc6\u522b\u7ade\u8d5b":55,"\u4e2d\u641c\u7d22\u8fd9\u51e0\u4e2a\u5e93":26,"\u4e2d\u6587\u6587\u6863":38,"\u4e2d\u6587\u6587\u6863\u76ee\u5f55":38,"\u4e2d\u6587\u7ef4\u57fa\u767e\u79d1\u9875\u9762":32,"\u4e2d\u6709\u8bb8\u591a\u7684\u7279\u5f81":58,"\u4e2d\u6bcf\u4e2apod\u7684ip\u5730\u5740":49,"\u4e2d\u6bcf\u5c42\u7684\u6570\u503c\u7edf\u8ba1":43,"\u4e2d\u7684":55,"\u4e2d\u7684\u4e00\u884c":[3,36],"\u4e2d\u7684\u5185\u5bb9":60,"\u4e2d\u7684\u5bf9\u5e94\u5206\u652f\u5373\u53ef":36,"\u4e2d\u7684\u63a5\u53e3":59,"\u4e2d\u7684\u6570\u636e":55,"\u4e2d\u7684\u6570\u636e\u662f\u5426\u4e3a\u5e8f\u5217\u6a21\u5f0f":59,"\u4e2d\u7684\u6570\u636e\u8fdb\u884c\u9884\u6d4b":55,"\u4e2d\u7684\u6570\u6910\u96c6\u7684\u7ed3\u6784\u5982\u4e0b":61,"\u4e2d\u7684\u6bcf\u4e00\u884c\u547d\u4ee4":59,"\u4e2d\u7684\u751f\u6210\u7ed3\u679c\u5982\u4e0b\u6240\u793a":62,"\u4e2d\u7684\u7528\u6237\u8bc1\u4e66":47,"\u4e2d\u7684\u7b2ci\u884c":62,"\u4e2d\u7684\u8bf4\u660e":3,"\u4e2d\u7684\u8fd9\u4e9b\u6570\u636e\u6587\u4ef6":58,"\u4e2d\u770b\u5230\u4e0b\u9762\u7684\u6587\u4ef6":61,"\u4e2d\u83b7\u53d6":49,"\u4e2d\u8ba4\u771f\u8bbe\u7f6e":41,"\u4e2d\u8bbe\u7f6e":41,"\u4e2d\u8bbe\u7f6e\u7684\u6240\u6709\u8282\u70b9":41,"\u4e2d\u8be6\u7ec6\u4ecb\u7ecd":37,"\u4e2d\u8bfb\u53d6":3,"\u4e2d\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":35,"\u4e2d\u914d\u7f6e\u7684\u6548\u679c\u4e00\u81f4":3,"\u4e34\u65f6\u53d8\u91cf\u7b49\u7b49":24,"\u4e3a":[3,35],"\u4e3a0":3,"\u4e3a\u4e86\u4f7f\u7528\u63d0\u524d\u7f16\u5199\u7684\u811a\u672c":61,"\u4e3a\u4e86\u4fdd\u8bc1\u6548\u7387":37,"\u4e3a\u4e86\u4fdd\u8bc1gpu\u9a71\u52a8\u80fd\u591f\u5728\u955c\u50cf\u91cc\u9762\u6b63\u5e38\u8fd0\u884c":27,"\u4e3a\u4e86\u5145\u5206\u7684\u968f\u673a\u6253\u4e71\u8bad\u7ec3\u96c6":61,"\u4e3a\u4e86\u5b8c\u6210\u5206\u5e03\u5f0f\u673a\u5668\u5b66\u4e60\u8bad\u7ec3\u4efb\u52a1":47,"\u4e3a\u4e86\u5c01\u88c5\u80fd\u591f\u6b63\u786e\u5de5\u4f5c":37,"\u4e3a\u4e86\u63cf\u8ff0\u65b9\u4fbf":34,"\u4e3a\u4e86\u65b9\u4fbf\u5728\u56fd":27,"\u4e3a\u4e86\u65b9\u4fbf\u5927\u5bb6":36,"\u4e3a\u4e86\u65b9\u4fbf\u8d77\u89c1":41,"\u4e3a\u4e86\u66b4\u9732\u7684\u63a5\u53e3\u5c3d\u91cf\u7b80\u5355":21,"\u4e3a\u4e86\u66f4\u7075\u6d3b\u7684\u914d\u7f6e":46,"\u4e3a\u4e86\u6ee1\u8db3\u8bad\u7ec3":41,"\u4e3a\u4e86\u7528\u6237\u80fd\u591f\u7075\u6d3b\u7684\u5904\u7406\u6570\u636e":46,"\u4e3a\u4e86\u8fbe\u5230\u6027\u80fd\u6700\u4f18":40,"\u4e3a\u4e86\u8fd8\u539f":25,"\u4e3a\u4e86\u907f\u514d\u7528\u6237\u76f4\u63a5\u5199\u590d\u6742\u7684protobuf":46,"\u4e3a\u4f8b":57,"\u4e3a\u4f8b\u521b\u5efa\u5206\u5e03\u5f0f\u7684\u5355\u8fdb\u7a0b\u8bad\u7ec3":41,"\u4e3a\u4f8b\u8fdb\u884c\u9884\u6d4b":57,"\u4e3a\u53c2\u6570\u77e9\u9635\u7684\u5bbd\u5ea6":24,"\u4e3a\u60a8\u505a\u6027\u80fd\u8c03\u4f18\u63d0\u4f9b\u4e86\u65b9\u5411":40,"\u4e3a\u60f3\u4fee\u6b63\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u7528\u6237\u63d0\u4f9b\u4e86\u5c06\u6587\u672c\u8bcd\u5411\u91cf\u6a21\u578b\u8f6c\u6362\u4e3a\u4e8c\u8fdb\u5236\u6a21\u578b\u7684\u547d\u4ee4":53,"\u4e3a\u65b9\u4fbf\u4f5c\u4e1a\u542f\u52a8\u63d0\u4f9b\u4e86\u4e24\u4e2a\u72ec\u7279\u7684\u547d\u4ee4\u9009\u9879":41,"\u4e3a\u6b64":48,"\u4e3a\u8f93\u51fa\u5206\u914d\u5185\u5b58":37,"\u4e3a\u96c6\u7fa4\u4f5c\u4e1a\u8bbe\u7f6e\u989d\u5916\u7684":41,"\u4e3ajson\u6216yaml\u683c\u5f0f":59,"\u4e3aoutput_\u7533\u8bf7\u5185\u5b58":37,"\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":43,"\u4e3b\u8981\u5305\u62ec\u4ee5\u4e0b\u4e94\u4e2a\u6b65\u9aa4":5,"\u4e3b\u8981\u539f\u56e0":32,"\u4e3b\u8981\u539f\u56e0\u5305\u62ec\u4e24\u4e2a\u65b9\u9762":24,"\u4e3b\u8981\u539f\u56e0\u662f\u589e\u52a0\u4e86\u521d\u59cb\u5316\u673a\u5236":3,"\u4e3b\u8981\u6765\u81ea\u5317\u7f8e\u6d32":54,"\u4e3b\u8981\u7531layer\u7ec4\u6210":46,"\u4e3b\u8981\u7684\u89e3\u51b3\u529e\u6cd5\u662f\u51cf\u5c0f\u5b66\u4e60\u5f8b\u6216\u8005\u5bf9\u6570\u636e\u8fdb\u884c\u5f52\u4e00\u5316\u5904\u7406":24,"\u4e3b\u8981\u804c\u8d23\u5728\u4e8e\u5c06\u8bad\u7ec3\u6570\u636e\u4f20\u5165\u5185\u5b58\u6216\u8005\u663e\u5b58":57,"\u4e3e\u4e00\u4e2a\u4f8b\u5b50":24,"\u4e3e\u4f8b":24,"\u4e3e\u4f8b\u8bf4\u660e":32,"\u4e4b\u524d":36,"\u4e4b\u524d\u914d\u7f6e\u6587\u4ef6\u4e2d":57,"\u4e4b\u540e":[25,37],"\u4e4b\u540e\u4f60\u4f1a\u5f97\u5230\u8bad\u7ec3":41,"\u4e4b\u540e\u4f7f\u7528":37,"\u4e4b\u540e\u4f7f\u7528\u77e9\u9635\u8fd0\u7b97\u51fd\u6570\u6765\u8ba1\u7b97":37,"\u4e4b\u540e\u521d\u59cb\u5316\u6240\u6709\u7684\u6743\u91cd\u77e9\u9635":37,"\u4e4b\u540e\u5b9a\u4e49\u7684":54,"\u4e4b\u5916\u7684\u6240\u6709\u5934\u6587\u4ef6":21,"\u4e4b\u95f4\u7684\u8ddd\u79bb":25,"\u4e4b\u95f4\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":34,"\u4e58\u4e0a\u8f93\u51fa\u7684\u68af\u5ea6":37,"\u4e58\u9664\u7b49\u65f6\u5019":24,"\u4e5d\u4e2a":60,"\u4e5f":32,"\u4e5f\u4e0d\u4f7f\u7528\u5176\u4ed6\u52a8\u6001\u5e93":20,"\u4e5f\u4e0d\u5b58\u5728\u4e00\u4e2asubseq\u76f4\u63a5\u751f\u6210\u4e0b\u4e00\u4e2asubseq\u7684\u60c5\u51b5":34,"\u4e5f\u4e0d\u5e94\u8be5\u62a5\u9519":21,"\u4e5f\u4e0d\u751f\u6210":21,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":36,"\u4e5f\u53ef\u4ee5\u53bb\u6389\u8fd9\u4e9b\u8bc1\u4e66\u7684\u914d\u7f6e":47,"\u4e5f\u53ef\u4ee5\u5728\u5f00\u53d1\u955c\u50cf\u4e2d\u542f\u52a8\u4e00\u4e2asshd\u670d\u52a1":27,"\u4e5f\u53ef\u4ee5\u662f\u4e00\u4e2a\u8bcd\u8bed":34,"\u4e5f\u53ef\u4ee5\u8bf4\u662f\u67d0\u4e9b\u7279\u5b9a\u6307\u4ee4\u7684\u4f7f\u7528\u60c5\u51b5":40,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539":49,"\u4e5f\u53ef\u4ee5\u901a\u8fc7saving_period_by_batches\u8bbe\u7f6e\u6bcf\u9694\u591a\u5c11batch\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":57,"\u4e5f\u53ef\u4ee5\u914d\u7f6e\u4e0d\u540c\u7684\u91cd\u8bd5\u673a\u5236":47,"\u4e5f\u5c31\u662f":36,"\u4e5f\u5c31\u662f\u5c06\u8bcd\u5411\u91cf\u6a21\u578b\u8fdb\u4e00\u6b65\u6f14\u5316\u4e3a\u4e09\u4e2a\u65b0\u6b65\u9aa4":57,"\u4e5f\u5c31\u662f\u81ea\u5df1\u7528\u6237\u540d\u4e0b\u7684":36,"\u4e5f\u5c31\u662f\u8bf4":[43,45,53],"\u4e5f\u5f97\u5230\u4e00\u4e2a\u7528\u6237\u7279\u5f81":59,"\u4e5f\u63cf\u8ff0\u4e86\u5bb9\u5668\u9700\u8981\u4f7f\u7528\u7684\u5b58\u50a8\u5377\u6302\u8f7d\u7684\u60c5\u51b5":49,"\u4e5f\u652f\u6301cpu\u7684\u6027\u80fd\u5206\u6790":40,"\u4e5f\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":32,"\u4e5f\u662f\u5bb9\u5668\u4e0enode\u4e4b\u95f4\u5171\u4eab\u6587\u4ef6\u7684\u65b9\u5f0f":47,"\u4e5f\u662fdecoder\u5faa\u73af\u5c55\u5f00\u7684\u4f9d\u636e":34,"\u4e5f\u662fpaddlepaddle\u6240\u80fd\u591f\u4fdd\u8bc1\u7684shuffle\u7c92\u5ea6":3,"\u4e5f\u6ca1\u7528":24,"\u4e5f\u79f0\u4e3arnn\u6a21\u578b":57,"\u4e5f\u79f0\u4f5c":46,"\u4e5f\u8bb8\u662f\u56e0\u4e3a\u9700\u8981\u5b89\u88c5":54,"\u4e5f\u9700\u8981\u4e24\u6b21\u968f\u673a\u9009\u62e9\u5230\u76f8\u540cgenerator\u7684\u65f6\u5019":3,"\u4e66\u5199":20,"\u4e7e":32,"\u4e86":32,"\u4e86\u89e3\u60a8\u7684\u786c\u4ef6":40,"\u4e86\u89e3\u66f4\u591a\u7ec6\u8282":35,"\u4e86\u89e3\u66f4\u591a\u8be6\u7ec6\u4fe1\u606f":35,"\u4e8c\u7ea7\u76ee\u5f55":[61,62],"\u4e8c\u7ef4\u77e9\u9635":55,"\u4e8c\u8005\u8bed\u610f\u4e0a\u5b8c\u5168\u4e00\u81f4":32,"\u4e8c\u8fdb\u5236":53,"\u4e92\u76f8\u901a\u4fe1":47,"\u4e92\u8054\u7f51\u7535\u5f71\u6570\u636e\u5e93":61,"\u4e94\u661f\u7ea7":32,"\u4e9a\u9a6c\u900a":61,"\u4ea4\u901a":32,"\u4ea4\u901a\u4fbf\u5229":32,"\u4eab\u53d7\u60a8\u7684\u65c5\u7a0b":27,"\u4eba\u8138\u8bc6\u522b":19,"\u4ec0\u4e48":59,"\u4ec5\u4ec5\u4f7f\u7528":20,"\u4ec5\u4ec5\u662f\u4e00\u4e9b\u5173\u952e\u8bcd":61,"\u4ec5\u4ec5\u662f\u4e24\u4e2a\u5168\u8fde\u63a5\u5c42":59,"\u4ec5\u4ec5\u662f\u7b80\u5355\u7684\u5d4c\u5165":59,"\u4ec5\u5305\u542b\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6570\u6910\u96c6":61,"\u4ec5\u5728\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u65f6\u6709\u6548":37,"\u4ec5\u5bf9\u7a00\u758f\u6570\u636e\u6709\u6548":37,"\u4ec5\u9700\u8981\u77e5\u9053\u5982\u4f55\u4ece":3,"\u4ecb\u7ecd\u4e86\u4e00\u79cd\u901a\u8fc7ssh\u8fdc\u7a0b\u5206\u53d1\u4efb\u52a1":49,"\u4ecb\u7ecd\u5206\u5e03\u5f0f\u8bad\u7ec3\u4e4b\u524d":47,"\u4ecb\u7ecdpaddlepaddle\u7684\u57fa\u672c\u4f7f\u7528\u65b9\u6cd5":57,"\u4ece":[23,40,60],"\u4ece0\u5230num":43,"\u4ece\u4e00\u4e2aword\u751f\u6210\u4e0b\u4e00\u4e2aword":34,"\u4ece\u5185\u6838\u51fd\u6570\u7684\u89d2\u5ea6":40,"\u4ece\u56fe\u4e2d\u53ef\u4ee5\u770b\u5230":25,"\u4ece\u5916\u90e8\u7f51\u7ad9\u4e0a\u4e0b\u8f7d\u7684\u539f\u59cb\u6570\u6910\u96c6":61,"\u4ece\u5927\u5230\u5c0f":62,"\u4ece\u6570\u636e\u63d0\u4f9b\u7a0b\u5e8f\u52a0\u8f7d\u5b9e\u4f8b":60,"\u4ece\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u770b":32,"\u4ece\u6bcf\u4e2a\u5355\u8bcd\u5de6\u53f3\u4e24\u7aef\u5206\u522b\u83b7\u53d6k\u4e2a\u76f8\u90bb\u7684\u5355\u8bcd":57,"\u4ece\u78c1\u76d8\u6587\u4ef6\u4e2d\u52a0\u8f7duuid\u6587\u4ef6\u540d\u7684\u68c0\u67e5\u70b9\u5feb\u7167\u6587\u4ef6":18,"\u4ece\u7b2c0\u4e2a\u8bc4\u4f30\u5230\u5f53\u524d\u8bc4\u4f30\u4e2d":62,"\u4ece\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u7684\u5e73\u5747\u635f\u5931":61,"\u4ece\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u7684\u5e73\u5747cost":62,"\u4ece\u800c\u53ef\u4ee5\u505a\u4e00\u4e9b\u4e0e\u8ba1\u7b97\u91cd\u53e0\u7684\u5de5\u4f5c":37,"\u4ece\u800c\u53ef\u4ee5\u62df\u5408\u4efb\u610f\u7684\u51fd\u6570\u6765\u5b66\u4e60\u590d\u6742\u7684\u6570\u636e\u5173\u7cfb":25,"\u4ece\u800c\u751f\u6210\u591a\u4e2agener":3,"\u4ece\u800c\u80fd\u591f\u88abpaddlepaddl":57,"\u4ece\u800c\u9632\u6b62\u8fc7\u62df\u5408":2,"\u4ece\u8be5\u94fe\u63a5":62,"\u4ece\u8bed\u4e49\u4e0a\u770b":34,"\u4ece\u8f93\u5165\u6570\u636e\u4e0a\u770b":32,"\u4ece\u8f93\u51fa\u65e5\u5fd7\u53ef\u4ee5\u770b\u5230":25,"\u4ece\u9884\u8bad\u7ec3\u6a21\u578b\u4e2d":53,"\u4eceetcd\u4e2d\u8bfb\u53d6\u8282\u70b9":18,"\u4ecestart":43,"\u4ecetest":62,"\u4ed3\u5e93\u7684\u8fdc\u7a0b\u4e3b\u673a":36,"\u4ed4\u7ec6\u89c2\u5bdf":55,"\u4ed6\u4e3b\u8981\u5305\u542b\u4e86\u5b9e\u9645\u66b4\u9732\u7684\u7c7b\u578b\u7ed3\u6784":21,"\u4ed6\u4eec\u5206\u522b\u662f":32,"\u4ed6\u4eec\u5728paddle\u7684\u6587\u6863\u548capi\u4e2d\u662f\u4e00\u4e2a\u6982\u5ff5":32,"\u4ed6\u4eec\u63d0\u51fa\u6b8b\u5dee\u5b66\u4e60\u7684\u6846\u67b6\u6765\u7b80\u5316\u7f51\u7edc\u7684\u8bad\u7ec3":55,"\u4ed6\u662f\u5c06":21,"\u4ed6\u7684\u76ee\u6807\u662f\u4f7f\u7528c":20,"\u4ee3\u66ff":49,"\u4ee3\u7801":59,"\u4ee3\u7801\u4e2d9":32,"\u4ee3\u7801\u5982\u4e0b":35,"\u4ee3\u7801\u6ce8\u91ca\u8bf7\u9075\u5b88":36,"\u4ee3\u7801\u751f\u6210\u7684\u7b26\u53f7\u53ef\u80fd\u4e0d\u4e00\u81f4":20,"\u4ee3\u8868\u5bbf\u4e3b\u673a\u76ee\u5f55":49,"\u4ee3\u8868\u7f16\u53f7":59,"\u4ee3\u8868\u8fd9\u4e2ashard\u7684\u6700\u5927index":19,"\u4ee3\u8868shard\u7684index":19,"\u4ee5\u4e0a":36,"\u4ee5\u4e0a\u4ee3\u7801\u7684reader\u8f93\u51fa\u7684data":19,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u4e0b\u751f\u6210100\u4e2a\u6587\u4ef6":19,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u6309\u7167\u751f\u6210\u7684dockerfile\u628a\u751f\u6210\u7684\u7a0b\u5e8f\u62f7\u8d1d\u5230\u751f\u4ea7\u955c\u50cf\u4e2d\u5e76\u505a\u76f8\u5e94\u7684\u914d\u7f6e":27,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u7f16\u8bd1paddlepaddl":27,"\u4ee5\u4e0a\u65b9\u6cd5\u5728gpu\u955c\u50cf\u91cc\u4e5f\u80fd\u7528":27,"\u4ee5\u4e0b":[19,59],"\u4ee5\u4e0b\u4ee3\u7801\u6bb5\u5b9a\u4e49\u4e86\u4e09\u4e2a\u8f93\u5165":35,"\u4ee5\u4e0b\u4ee3\u7801\u7247\u6bb5\u5b9a\u4e49":35,"\u4ee5\u4e0b\u6211\u4eec\u7ffb\u8bd1\u6570\u636e\u96c6\u7f51\u7ad9\u4e2dreadme\u6587\u4ef6\u7684\u63cf\u8ff0":58,"\u4ee5\u4e0b\u6307\u4ee4\u80fd\u68c0\u67e5linux\u7535\u8111\u662f\u5426\u652f\u6301avx":27,"\u4ee5\u4e0b\u6559\u7a0b\u5c06\u6307\u5bfc\u60a8\u63d0\u4ea4\u4ee3\u7801":36,"\u4ee5\u4e0b\u662f\u5bf9\u4e0a\u8ff0\u6570\u636e\u52a0\u8f7d\u7684\u89e3\u91ca":57,"\u4ee5\u4e0b\u6b65\u9aa4\u57fa\u4e8e":41,"\u4ee5\u4e0b\u793a\u8303\u5982\u4f55\u4f7f\u7528\u9884\u8bad\u7ec3\u7684\u4e2d\u6587\u5b57\u5178\u548c\u8bcd\u5411\u91cf\u8fdb\u884c\u77ed\u8bed\u6539\u5199":53,"\u4ee5\u4e0b\u9009\u9879\u5fc5\u987b\u5728":41,"\u4ee5\u4ea4\u4e92\u5bb9\u5668\u65b9\u5f0f\u8fd0\u884c\u5f00\u53d1\u955c\u50cf":27,"\u4ee5\u4fbf\u7528\u6237":41,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u7684\u6b63\u786e\u8ba1\u7b97":37,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u8ba1\u7b97\u7684\u6b63\u786e\u6027":37,"\u4ee5\u5206\u7c7b\u6765\u81ea":61,"\u4ee5\u53ca":37,"\u4ee5\u53ca\u4f7f\u7528\u5b50\u5e8f\u5217\u6765\u5b9a\u4e49\u5206\u7ea7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":35,"\u4ee5\u53ca\u53cc\u5c42\u5e8f\u5217":31,"\u4ee5\u53ca\u5728wmt":62,"\u4ee5\u53ca\u5982\u4f55\u5728\u5c42\u4e4b\u95f4\u8fdb\u884c\u8fde\u63a5":54,"\u4ee5\u53ca\u6570\u636e\u8bfb\u53d6\u51fd\u6570":46,"\u4ee5\u53ca\u751f\u6210\u521b\u5efa\u751f\u4ea7\u955c\u50cf\u7684dockerfil":27,"\u4ee5\u53ca\u8ba1\u7b97\u903b\u8f91\u5728\u5e8f\u5217\u4e0a\u7684\u5faa\u73af\u5c55\u5f00":34,"\u4ee5\u53ca\u8f93\u5165\u7684\u68af\u5ea6":37,"\u4ee5\u53canumpi":19,"\u4ee5\u53capaddle\u5982\u4f55\u5904\u7406\u591a\u79cd\u7c7b\u578b\u7684\u8f93\u5165":59,"\u4ee5\u53carelu":37,"\u4ee5\u76f8\u5bf9\u8def\u5f84\u5f15\u7528":2,"\u4ee5\u83b7\u5f97\u66f4\u597d\u7684\u7f51\u7edc\u6027\u80fd":41,"\u4ee5\u9017\u53f7":53,"\u4ee5\u9017\u53f7\u95f4\u9694":43,"\u4ef7\u683c":32,"\u4efb\u52a1":59,"\u4efb\u52a1\u6765\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":41,"\u4efb\u610f\u5c06\u4e00\u4e9b\u6570\u636e\u7ec4\u5408\u6210\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":32,"\u4efb\u610f\u65f6\u523b\u53ea\u53ef\u80fd\u540c\u65f6\u6709\u4e00\u53f0\u670d\u52a1\u5668\u6545\u969c":18,"\u4f18\u5316":61,"\u4f18\u5316\u5668\u5219\u7528\u94fe\u5f0f\u6cd5\u5219\u6765\u5bf9\u6bcf\u4e2a\u53c2\u6570\u8ba1\u7b97\u635f\u5931\u51fd\u6570\u7684\u68af\u5ea6":37,"\u4f18\u5316\u7b97\u6cd5":46,"\u4f1a\u5148\u8fdb\u884c\u53c2\u6570\u7684\u521d\u59cb\u5316\u4e0e\u89e3\u6790":49,"\u4f1a\u5171\u4eab\u53c2\u6570":24,"\u4f1a\u52a0\u8f7d\u4e0a\u4e00\u8f6e\u7684\u53c2\u6570":43,"\u4f1a\u53d8\u6210\u8bcd\u8868\u4e2d\u7684\u4f4d\u7f6e":32,"\u4f1a\u542f\u52a8pserver\u4e0etrainer\u8fdb\u7a0b":49,"\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u751f\u6210\u4e24\u4e2a\u5b50\u76ee\u5f55":38,"\u4f1a\u5bf9\u6bcf\u4e00\u4e2a\u6fc0\u6d3b\u6682\u5b58\u4e00\u4e9b\u6570\u636e":24,"\u4f1a\u5bf9\u8fd9\u7c7b\u8f93\u5165\u8fdb\u884c\u62c6\u89e3":34,"\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7248\u672cpython\u5728\u4e00\u4e2a\u8fdb\u7a0b\u91cc\u7684bug":20,"\u4f1a\u5c06\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u62fc\u63a5":34,"\u4f1a\u5c06\u7b2c\u4e00\u4e2a":24,"\u4f1a\u6210\u4e3astep\u51fd\u6570\u7684\u8f93\u5165":34,"\u4f1a\u6254\u5230\u8fd9\u6761\u6570\u636e":3,"\u4f1a\u62a5\u9519":34,"\u4f1a\u6839\u636e\u547d\u4ee4\u884c\u53c2\u6570\u6307\u5b9a\u7684\u6d4b\u8bd5\u65b9\u5f0f":2,"\u4f1a\u6839\u636einput_types\u68c0\u67e5\u6570\u636e\u7684\u5408\u6cd5\u6027":3,"\u4f1a\u76f4\u63a5\u62a5\u9519\u9000\u51fa":20,"\u4f1a\u76f8\u5e94\u5730\u6539\u53d8\u8f93\u51fa\u7684\u5c3a\u5bf8":37,"\u4f1a\u81ea\u9002\u5e94\u5730\u4ece\u8fd9\u4e9b\u5411\u91cf\u4e2d\u9009\u62e9\u4e00\u4e2a\u5b50\u96c6\u51fa\u6765":62,"\u4f1a\u83b7\u53d6\u5f53\u524dnamespace\u4e0b\u7684\u6240\u6709pod":49,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":34,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u975e\u5e8f\u5217":34,"\u4f1a\u88abpickle\u5e8f\u5217\u5316\u6210\u5b57\u7b26\u4e32":19,"\u4f20\u5165":[3,19],"\u4f20\u5165\u4e0a\u4e00\u6b65\u89e3\u6790\u51fa\u6765\u7684\u6a21\u578b\u914d\u7f6e\u5c31\u53ef\u4ee5\u521b\u5efa\u4e00\u4e2a":5,"\u4f20\u5165\u9884\u6d4b\u6570\u636e":5,"\u4f20\u7ed9dataprovider\u7684\u67d0\u4e00\u4e2aargs\u8fc7\u5927":24,"\u4f20\u9012\u7ed9\u914d\u7f6e\u6587\u4ef6\u7684\u53c2\u6570":43,"\u4f46":21,"\u4f46\u4e0d\u66b4\u9732":21,"\u4f46\u4e0d\u7528\u4e8e\u8ba1\u7b97\u68af\u5ea6":37,"\u4f46\u4e0d\u9700\u8981\u63d0\u524d\u521b\u5efa":43,"\u4f46\u4e8e\u53cc\u5c42\u5e8f\u5217\u7684lstm\u6765\u8bf4":32,"\u4f46\u548c\u5355\u5c42rnn\u4e0d\u540c":32,"\u4f46\u5728\u8d77\u521d\u7684\u51e0\u8f6e\u8bad\u7ec3\u4e2d\u5b83\u4eec\u90fd\u5728\u5feb\u901f\u903c\u8fd1\u771f\u5b9e\u503c":25,"\u4f46\u5b50\u53e5\u542b\u6709\u7684\u8bcd\u8bed\u6570\u53ef\u4ee5\u4e0d\u76f8\u7b49":34,"\u4f46\u5c3d\u91cf\u8bf7\u4fdd\u6301\u7f16\u8bd1\u548c\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u662f\u540c\u4e00\u4e2a\u7248\u672c":26,"\u4f46\u5e76\u6ca1\u6709\u7ecf\u8fc7\u56de\u5f52\u6d4b\u8bd5":23,"\u4f46\u5e8f\u5217\u8f93\u51fa\u65f6":32,"\u4f46\u5f53\u8c03\u7528\u8fc7\u4e00\u6b21\u540e":3,"\u4f46\u6240\u6709fork\u7684\u7248\u672c\u5e93\u7684\u6240\u6709\u5206\u652f\u90fd\u76f8\u5f53\u4e8e\u7279\u6027\u5206\u652f":23,"\u4f46\u662f":[24,32],"\u4f46\u662f2008\u5e74\u4e4b\u524d\u751f\u4ea7\u7684\u65e7\u7535\u8111\u4e0d\u652f\u6301avx":27,"\u4f46\u662f\u4e5f\u6ca1\u6709\u5fc5\u8981\u5220\u9664\u65e0\u7528\u7684\u6587\u4ef6":41,"\u4f46\u662f\u53c8\u8fc7\u4e8e\u7410\u788e":21,"\u4f46\u662f\u5927\u90e8\u5206\u53c2\u6570\u662f\u4e3a\u5f00\u53d1\u8005\u63d0\u4f9b\u7684":42,"\u4f46\u662f\u5b50\u5e8f\u5217\u7684\u6570\u76ee\u5fc5\u987b\u4e00\u6837":32,"\u4f46\u662f\u6211\u4eec\u5e76\u4e0d\u63a8\u8350\u8fd9\u79cd\u65b9\u6cd5":27,"\u4f46\u662f\u652f\u6301avx\u6307\u4ee4\u96c6":36,"\u4f46\u662f\u65b9\u4fbf\u8c03\u8bd5\u548c\u6d4bbenchmark":26,"\u4f46\u662f\u6bcf\u4e2a\u6837\u672c\u4ec5\u5305\u542b\u51e0\u4e2a\u8bcd":45,"\u4f46\u662f\u7a81\u7136\u6709\u4e00\u4e2a10000\u957f\u7684\u5e8f\u5217":24,"\u4f46\u662f\u89e3\u91ca\u6027\u8bed\u8a00":20,"\u4f46\u662f\u8fd9\u4e2a\u503c\u4e0d\u53ef\u4ee5\u8c03\u7684\u8fc7\u5927":46,"\u4f46\u662f\u8fd9\u79cd\u65b9\u6cd5\u5728\u6bcf\u5c42\u53ea\u4fdd\u5b58\u9884\u8bbe\u6570\u91cf\u7684\u6700\u4f18\u72b6\u6001":62,"\u4f46\u662f\u8fdc\u672a\u5b8c\u5584":0,"\u4f46\u662f\u9690\u85cf\u5c42\u4e2d\u7684\u6bcf\u4e2a\u666e\u901a\u8282\u70b9\u88ab\u4e00\u4e2a\u8bb0\u5fc6\u5355\u5143\u66ff\u6362":61,"\u4f46\u662fbatch":24,"\u4f46\u6709\u503c\u7684\u5730\u65b9\u5fc5\u987b\u4e3a1":3,"\u4f46\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":3,"\u4f46\u8fd9\u4e2a\u5173\u7cfb\u53ef\u80fd\u4e0d\u6b63\u786e":3,"\u4f4d\u7f6e":32,"\u4f4f":32,"\u4f53\u88c1\u5b57\u5178":59,"\u4f53\u88c1\u5b57\u6bb5":59,"\u4f59\u5f26\u76f8\u4f3c\u5ea6\u56de\u5f52":59,"\u4f59\u5f26\u76f8\u4f3c\u5ea6\u5c42":59,"\u4f5c\u4e3a\u4e0b\u4e00\u4e2a\u5b50\u53e5memory\u7684\u521d\u59cb\u72b6\u6001":32,"\u4f5c\u4e3a\u4f8b\u5b50\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":35,"\u4f5c\u4e3a\u53c2\u6570\u7684id":24,"\u4f5c\u4e3a\u5f53\u524d\u65f6\u523b\u8f93\u5165":34,"\u4f5c\u4e3a\u6d88\u606f\u957f\u5ea6":46,"\u4f5c\u4e3a\u793a\u4f8b\u6570\u636e":58,"\u4f5c\u4e3a\u7c7b\u53e5\u67c4":20,"\u4f5c\u4e3a\u7ebf\u6027\u56de\u5f52\u7684\u8f93\u5165":25,"\u4f5c\u4e3a\u8f93\u51fa":35,"\u4f5c\u4e3a\u96c6\u7fa4\u8bad\u7ec3\u7684\u5de5\u4f5c\u7a7a\u95f4":41,"\u4f5c\u4e3aboot_layer\u4f20\u7ed9\u4e0b\u4e00\u4e2a\u5b50\u53e5\u7684memori":32,"\u4f5c\u5bb6":58,"\u4f5c\u7528":31,"\u4f60\u4e5f\u53ef\u4ee5\u4f7f\u7528\u8fd9\u4e09\u4e2a\u503c":55,"\u4f60\u4e5f\u53ef\u4ee5\u5148\u8df3\u8fc7\u672c\u6587\u7684\u89e3\u91ca\u73af\u8282":57,"\u4f60\u4e5f\u53ef\u4ee5\u7b80\u5355\u7684\u8fd0\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4":53,"\u4f60\u4e5f\u53ef\u4ee5\u901a\u8fc7\u5728\u547d\u4ee4\u884c\u53c2\u6570\u4e2d\u589e\u52a0\u4e00\u4e2a\u53c2\u6570\u5982":55,"\u4f60\u4e5f\u8bb8\u53ef\u4ee5\u5c1d\u8bd5\u66f4\u8001\u7684\u65b9\u6cd5":27,"\u4f60\u53ea\u9700\u5b8c\u6210":41,"\u4f60\u53ea\u9700\u8981\u5728\u547d\u4ee4\u884c\u8f93\u5165\u4ee5\u4e0b\u547d\u4ee4":57,"\u4f60\u53ea\u9700\u8981\u6309\u7167\u5982\u4e0b\u65b9\u5f0f\u7ec4\u7ec7\u6570\u636e":62,"\u4f60\u53ef\u4ee5\u4f7f\u7528":55,"\u4f60\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u811a\u672c\u4e0b\u8f7d":61,"\u4f60\u53ef\u4ee5\u4f7f\u7528\u5b83\u6765\u8bbe\u7f6e":41,"\u4f60\u53ef\u4ee5\u4f7f\u7528\u672c\u5730\u8bad\u7ec3\u4e2d\u7684\u76f8\u540c\u6a21\u578b\u6587\u4ef6\u8fdb\u884c\u96c6\u7fa4\u8bad\u7ec3":41,"\u4f60\u53ef\u4ee5\u5728\u4efb\u4f55\u65f6\u5019\u7528":59,"\u4f60\u53ef\u4ee5\u5bf9\u6a21\u578b\u8fdb\u884c\u8bc4\u4f30":59,"\u4f60\u53ef\u4ee5\u5c06\u7f51\u7edc\u914d\u7f6e\u6210\u67d0\u4e9b\u5c42\u4f7f\u7528gpu\u8ba1\u7b97":45,"\u4f60\u53ef\u4ee5\u6267\u884c\u4e0a\u8ff0\u547d\u4ee4\u6765\u4e0b\u8f7d\u6240\u6709\u7684\u6a21\u578b\u548c\u5747\u503c\u6587\u4ef6":55,"\u4f60\u53ef\u4ee5\u901a\u8fc7\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u5f97\u5230resnet\u7f51\u7edc\u7684\u7ed3\u6784\u53ef\u89c6\u5316\u56fe":55,"\u4f60\u53ef\u4ee5\u9884\u6d4b\u4efb\u4f55\u7528\u6237\u5bf9\u4e8e\u4efb\u4f55\u4e00\u90e8\u7535\u5f71\u7684\u8bc4\u4ef7":59,"\u4f60\u5c06\u4f1a\u770b\u5230\u4ee5\u4e0b\u7684\u6a21\u578b\u7ed3\u6784":53,"\u4f60\u5c06\u4f1a\u770b\u5230\u5982\u4e0b\u6d88\u606f":62,"\u4f60\u5c06\u4f1a\u770b\u5230\u5982\u4e0b\u7ed3\u679c":55,"\u4f60\u5c06\u4f1a\u770b\u5230\u7279\u5f81\u5b58\u50a8\u5728":55,"\u4f60\u5c06\u4f1a\u770b\u5230\u8fd9\u6837\u7684\u6d88\u606f":62,"\u4f60\u5c06\u5728\u76ee\u5f55":61,"\u4f60\u5c06\u770b\u5230\u5982\u4e0b\u7684\u4fe1\u606f":59,"\u4f60\u7684\u5de5\u4f5c\u7a7a\u95f4\u5e94\u5982\u4e0b\u6240\u793a":41,"\u4f60\u8fd8\u53ef\u4ee5\u5c06\u7528\u6237\u548c":41,"\u4f60\u9700\u8981\u4e00\u4e9b\u66f4\u590d\u6742\u7684\u5355\u5143\u6d4b\u8bd5\u6765\u4fdd\u8bc1\u4f60\u5b9e\u73b0\u7684\u7f51\u7edc\u5c42\u662f\u6b63\u786e\u7684":37,"\u4f60\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a\u8bbe\u5907\u7684id\u53f7":45,"\u4f60\u9700\u8981\u5728\u914d\u7f6ecmake\u65f6\u5c06":37,"\u4f60\u9700\u8981\u5b89\u88c5python\u7684\u7b2c\u4e09\u65b9\u5e93":59,"\u4f60\u9700\u8981\u628a\u8be5\u6587\u4ef6\u52a0\u5165":37,"\u4f7f\u5176\u8f6c\u53d8\u4e3a\u7ef4\u5ea6\u4e3ahidden_dim\u7684\u65b0\u5411\u91cf":57,"\u4f7f\u5f97":25,"\u4f7f\u5f97\u4e24\u4e2a\u5b57\u5178\u6709\u76f8\u540c\u7684\u4e0a\u4e0b\u6587":62,"\u4f7f\u5f97\u5355\u5143\u6d4b\u8bd5\u6709\u4e00\u4e2a\u5e72\u51c0\u7684\u73af\u5883":24,"\u4f7f\u5f97\u642d\u6a21\u578b\u65f6\u66f4\u65b9\u4fbf":37,"\u4f7f\u5f97\u6700\u7ec8\u5f97\u5230\u7684\u6a21\u578b\u51e0\u4e4e\u4e0e\u771f\u5b9e\u6a21\u578b\u4e00\u81f4":25,"\u4f7f\u7528":[21,23,24,27,32,34,35,37,40,43,46,57,60,61],"\u4f7f\u75280\u53f7\u548c1\u53f7gpu\u8ba1\u7b97fc2\u5c42":45,"\u4f7f\u75280\u53f7gpu\u8ba1\u7b97fc2\u5c42":45,"\u4f7f\u752810\u4e2a\u88c1\u526a\u56fe\u50cf\u5757":55,"\u4f7f\u75281\u53f7gpu\u8ba1\u7b97fc3\u5c42":45,"\u4f7f\u75282\u53f7\u548c3\u53f7gpu\u8ba1\u7b97fc3\u5c42":45,"\u4f7f\u7528\u4e00\u4e2a\u5c3a\u5ea6\u4e3a":37,"\u4f7f\u7528\u4e00\u4e2a\u8bcd\u524d\u4e24\u4e2a\u8bcd\u548c\u540e\u4e24\u4e2a\u8bcd":24,"\u4f7f\u7528\u4e0a\u6587\u521b\u5efa\u7684yaml\u6587\u4ef6\u521b\u5efakubernet":48,"\u4f7f\u7528\u4e0b\u9762\u547d\u4ee4":19,"\u4f7f\u7528\u4e86":46,"\u4f7f\u7528\u4e86\u540c\u6837\u7684parameter\u548cbia":24,"\u4f7f\u7528\u4e86\u57fa\u4e8e\u53e5\u6cd5\u7ed3\u6784\u7684\u9884\u5b9a\u4e49\u7279\u5f81\u6a21\u677f":60,"\u4f7f\u7528\u4e86avx\u6307\u4ee4\u96c6":29,"\u4f7f\u7528\u4ee5\u4e0b\u547d\u4ee4\u8fdb\u884c\u6a21\u578b\u8bad\u7ec3":53,"\u4f7f\u7528\u5177\u6709softmax\u6fc0\u6d3b\u7684\u5168\u8fde\u63a5\u524d\u9988\u5c42\u6765\u6267\u884c\u5206\u7c7b\u4efb\u52a1":61,"\u4f7f\u7528\u52a8\u6001\u5e93":20,"\u4f7f\u7528\u540c\u6837\u7684\u8bad\u7ec3\u6570\u636eblock":18,"\u4f7f\u7528\u591a\u5757\u663e\u5361\u8bad\u7ec3":24,"\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bad\u7ec3":24,"\u4f7f\u7528\u5982\u4e0b\u53c2\u6570":54,"\u4f7f\u7528\u5982\u4e0b\u547d\u4ee4":53,"\u4f7f\u7528\u5b66\u4e60\u5b8c\u6210\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u751f\u6210\u5e8f\u5217":35,"\u4f7f\u7528\u5bb9\u5668\u65b9\u5f0f\u8fd0\u884c\u8bad\u7ec3\u4efb\u52a1\u7684kubernet":49,"\u4f7f\u7528\u6211\u4eec\u4e4b\u524d\u6784\u9020\u7684\u955c\u50cf":48,"\u4f7f\u7528\u624b\u5de5\u6307\u5b9a\u7aef\u53e3\u6570\u91cf":46,"\u4f7f\u7528\u663e\u5361\u8bad\u7ec3":24,"\u4f7f\u7528\u667a\u80fd\u6307\u9488\u7684\u539f\u56e0\u662f":21,"\u4f7f\u7528\u6848\u4f8b":44,"\u4f7f\u7528\u7684":24,"\u4f7f\u7528\u76f8\u5bf9\u8def\u5f84\u7684\u5f15\u7528\u65b9\u5f0f":21,"\u4f7f\u7528\u8005\u4e0d\u9700\u8981\u5173\u5fc3":43,"\u4f7f\u7528\u8005\u53ea\u9700\u8981\u5173\u6ce8\u4e8e\u8bbe\u8ba1rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":34,"\u4f7f\u7528\u8005\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684python\u811a\u672c\u6765\u8bfb\u53d6\u53c2\u6570\u503c":55,"\u4f7f\u7528\u8005\u65e0\u9700\u5173\u5fc3\u8fd9\u4e2a\u53c2\u6570":43,"\u4f7f\u7528\u8005\u901a\u5e38\u65e0\u9700\u5173\u5fc3":43,"\u4f7f\u7528\u81ea\u52a8\u7684\u66ff\u8865\u6765\u66ff\u4ee3\u7ecf\u9a8c\u4e30\u5bcc\u7684\u4eba\u5de5\u8bc4\u5224":62,"\u4f7f\u7528\u8c13\u8bcd\u4e0a\u4e0b\u6587":60,"\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u5b8c\u6210\u5bf9\u65b0\u6570\u636e\u7684\u9884\u6d4b":18,"\u4f7f\u7528\u8fd9\u4e2a\u811a\u672c\u524d\u8bf7\u786e\u8ba4\u5df2\u7ecf\u5b89\u88c5\u4e86pillow\u53ca\u76f8\u5173\u4f9d\u8d56\u6a21\u5757":54,"\u4f7f\u7528\u8fd9\u79cd\u65b9\u5f0f":32,"\u4f7f\u7528\u8fdc\u7a0b\u7a00\u758f\u65b9\u5f0f\u8bad\u7ec3\u65f6":37,"\u4f7f\u7528\u968f\u673a\u68af\u5ea6\u4e0b\u964d":61,"\u4f7f\u7528\u9759\u6001\u5e93\u548c\u52a8\u6001\u5e93\u96be\u5ea6\u5dee\u4e0d\u591a":20,"\u4f7f\u7528\u9884\u8bad\u7ec3\u7684\u6807\u51c6\u683c\u5f0f\u8bcd\u5411\u91cf\u6a21\u578b":53,"\u4f7f\u7528args\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u8bbe\u7f6e":3,"\u4f7f\u7528c":21,"\u4f7f\u7528c99\u505a\u63a5\u53e3":20,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c11\u7684\u539f\u56e0\u662f":20,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c89":20,"\u4f7f\u7528checkgrad\u6a21\u5f0f\u65f6\u7684\u53c2\u6570\u53d8\u5316\u5927\u5c0f":43,"\u4f7f\u7528cpu\u4e24\u7ebf\u7a0b\u8ba1\u7b97fc4\u5c42":45,"\u4f7f\u7528cpu\u8ba1\u7b97fc4\u5c42":45,"\u4f7f\u7528cpu\u8bad\u7ec3":61,"\u4f7f\u7528init":45,"\u4f7f\u7528lstm\u4f5c\u4e3aencod":32,"\u4f7f\u7528max":54,"\u4f7f\u7528memory\u7684rnn\u5b9e\u73b0\u4fbf\u5982\u4e0b\u56fe\u6240\u793a":32,"\u4f7f\u7528model":45,"\u4f7f\u7528paddlepaddl":57,"\u4f7f\u7528python\u6570\u636e\u63d0\u4f9b\u5668":54,"\u4f7f\u7528rdma\u8fd8\u662ftcp\u4f20\u8f93\u534f\u8bae":43,"\u4f7f\u7528regress":23,"\u4f7f\u7528swig\u53ea\u652f\u6301cpython\u89e3\u91ca\u5668":20,"\u4f7f\u7528swig\u9700\u8981\u591a\u8bed\u8a00\u7ed1\u5b9a\u7684\u5f00\u53d1\u4eba\u5458\u719f\u7ec3\u638c\u63e1swig\u914d\u7f6e":20,"\u4f7f\u7528void":20,"\u4f7f\u8f93\u5165\u5c42\u5230\u9690\u85cf\u5c42\u7684\u795e\u7ecf\u5143\u662f\u5168\u90e8\u8fde\u63a5\u7684":54,"\u4f86":32,"\u4f8b\u5982":[3,20,21,23,24,26,32,35,37,40,41,42,43,45,46,49,55,57,59,61],"\u4f8b\u5982\u4e0a\u6587\u7684pod":47,"\u4f8b\u5982\u4e0a\u9762\u7684":25,"\u4f8b\u5982\u4ee5\u592a\u7f51\u7684":41,"\u4f8b\u5982\u4f7f\u7528":24,"\u4f8b\u5982\u586b\u5145":35,"\u4f8b\u5982\u5bf9\u4e8ejava\u6216\u8005python":20,"\u4f8b\u5982\u5bf9\u4e8ejava\u6765\u8bf4":20,"\u4f8b\u5982\u5bf9\u4e8epython":20,"\u4f8b\u5982\u5c06\u7b2c\u4e00\u6761\u6570\u636e\u8f6c\u5316\u4e3a":32,"\u4f8b\u5982\u6587\u672c\u5206\u7c7b\u4e2d":32,"\u4f8b\u5982\u672c\u4f8b\u4e2d\u7684\u4e24\u4e2a\u7279\u5f81":32,"\u4f8b\u5982\u673a\u5668\u4e0a\u67094\u5757gpu":24,"\u4f8b\u5982\u7b2c300\u4e2apass\u7684\u6a21\u578b\u4f1a\u88ab\u4fdd\u5b58\u5728":54,"\u4f8b\u5982c":20,"\u4f8b\u5982hostpath":47,"\u4f8b\u5982java\u4e0epython\u7684\u9519\u8bef\u5904\u7406\u662f\u76f4\u63a5\u6254\u51fa\u6765except":20,"\u4f8b\u5982output\u76ee\u5f55\u4e0b\u5c31\u5b58\u653e\u4e86\u8f93\u51fa\u7ed3\u679c":49,"\u4f8b\u5982python\u53ef\u4ee5\u4f7f\u7528":20,"\u4f8b\u5982python\u7684":20,"\u4f8b\u5982sigmoid":37,"\u4f8b\u5982sigmoid\u53d8\u6362":57,"\u4f8b\u5b50\u4e2d\u662f":37,"\u4f8b\u5b50\u4e2d\u662f0":37,"\u4f8b\u5b50\u4e2d\u662f100":37,"\u4f8b\u5b50\u4e2d\u662f4096":37,"\u4f8b\u5b50\u4e2d\u662f8192":37,"\u4f8b\u5b50\u4e2d\u662ffc":37,"\u4f8b\u5b50\u4e2d\u662fsoftmax":37,"\u4f8b\u5b50\u4f7f\u7528":47,"\u4f9bpaddlepaddle\u52a0\u8f7d":43,"\u4f9d\u636e\u5206\u7c7b\u9519\u8bef\u7387\u83b7\u5f97\u6700\u4f73\u6a21\u578b\u8fdb\u884c\u6d4b\u8bd5":61,"\u4f9d\u6b21\u7c7b\u63a8":23,"\u4f9d\u8d56\u4e8epython\u7684":54,"\u4fbf\u4e8e\u5b58\u50a8\u8d44\u6e90\u7ba1\u7406\u548cpod\u5f15\u7528":47,"\u4fbf\u4e8e\u672c\u5730\u9a8c\u8bc1\u548c\u6d4b\u8bd5":47,"\u4fbf\u4e8e\u7528\u6237\u6d4f\u89c8c":27,"\u4fbf\u5229":32,"\u4fbf\u548c\u5355\u5c42rnn\u914d\u7f6e\u4e2d\u7684":32,"\u4fbf\u5b9c":32,"\u4fbf\u662f\u5c06\u9759\u6001\u5e93\u52a0\u5165jvm\u4e2d":20,"\u4fdd\u5b58\u6a21\u578b\u53c2\u6570\u7684\u76ee\u5f55":43,"\u4fdd\u5b58\u751f\u6210\u7ed3\u679c\u7684\u6587\u4ef6":62,"\u4fdd\u5b58\u7f51\u7edc\u5c42\u8f93\u51fa\u7ed3\u679c\u7684\u76ee\u5f55":43,"\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c\u7684\u6587\u4ef6\u540d":43,"\u4fdd\u6301\u5bbd\u9ad8\u6bd4\u7f29\u653e\u5230\u77ed\u8fb9\u4e3a256":55,"\u4fe1\u53f7\u6765\u81ea\u52a8\u7ec8\u6b62\u5b83\u542f\u52a8\u7684\u6240\u6709\u8fdb\u7a0b":41,"\u4fee\u590d\u6240\u6709bug\u540e":23,"\u4fee\u590ddocker\u7f16\u8bd1\u955c\u50cf\u95ee\u9898":23,"\u4fee\u590dubuntu":23,"\u4fee\u6539":[47,48],"\u4fee\u6539\u542f\u52a8\u811a\u672c\u540e":48,"\u4fee\u6539\u6210\u66f4\u5feb\u7684\u7248\u672c":40,"\u4fee\u6539\u6587\u6863":39,"\u503c\u5f97\u6ce8\u610f\u7684\u662f":[32,36],"\u503c\u5f97\u6df1\u5165\u5206\u6790":40,"\u503c\u7c7b\u578b":45,"\u5047\u5982\u6211\u4eec\u662f\u4e09\u5206\u7c7b\u95ee\u9898":24,"\u5047\u8bbe":37,"\u5047\u8bbe\u53d8\u91cf":25,"\u5047\u8bbe\u60a8\u5df2\u7ecf\u5b8c\u6210\u4e86\u4e00\u4e2aai\u8bad\u7ec3\u7684python\u7a0b\u5e8f":27,"\u5047\u8bbe\u635f\u5931\u51fd\u6570\u662f":37,"\u5047\u8bbe\u8bcd\u5411\u91cf\u7ef4\u5ea6\u4e3a32":53,"\u504f\u7f6e\u53c2\u6570":55,"\u504f\u7f6e\u53c2\u6570\u7684\u5927\u5c0f":37,"\u505a\u5982\u4e0b\u51e0\u4e2a\u64cd\u4f5c":23,"\u505a\u63a5\u53e3":20,"\u505c\u6b62\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":18,"\u505c\u6b62\u52a0\u8f7d\u6570\u636e":43,"\u505c\u7535":32,"\u50cf":27,"\u513f\u7ae5\u7247":58,"\u5143\u7d20":31,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684\u8f93\u5165\u4fe1\u606f":31,"\u5148\u4f7f\u7528\u547d\u4ee4":46,"\u5148\u5b9e\u73b0\u6a21\u578b\u63a8\u65ad\u7684api":21,"\u5148\u8c03\u7528initializer\u51fd\u6570":57,"\u5168\u5bb6":32,"\u5168\u8fde\u63a5\u5c42":[25,46,53,54,59],"\u5168\u8fde\u63a5\u5c42\u4ee5\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":37,"\u5168\u8fde\u63a5\u5c42\u5c06\u7535\u5f71\u7684\u6bcf\u4e2a\u7279\u5f81\u7ed3\u5408\u6210\u4e00\u4e2a\u7535\u5f71\u7279\u5f81":59,"\u5168\u8fde\u63a5\u5c42\u6743\u91cd":55,"\u5168\u8fde\u63a5\u5c42\u6ca1\u6709\u7f51\u7edc\u5c42\u914d\u7f6e\u7684\u8d85\u53c2\u6570":37,"\u5168\u8fde\u63a5\u5c42\u7684\u5b9e\u73b0\u4f4d\u4e8e":37,"\u5168\u8fde\u63a5\u5c42\u7684\u6bcf\u4e2a\u8f93\u51fa\u90fd\u8fde\u63a5\u5230\u4e0a\u4e00\u5c42\u7684\u6240\u6709\u7684\u795e\u7ecf\u5143\u4e0a":37,"\u5168\u8fde\u63a5\u5c42python\u5c01\u88c5\u7684\u4f8b\u5b50\u4e2d\u5305\u542b\u4e0b\u9762\u51e0\u6b65":37,"\u516b\u4e2a\u7279\u5f81\u5206\u522b\u8f6c\u6362\u4e3a\u5411\u91cf":60,"\u516c\u5f0f":27,"\u516c\u94a5\u5199\u5165":41,"\u516d\u4e2a\u7279\u5f81\u548c\u6807\u7b7e\u90fd\u662f\u7d22\u5f15\u69fd":60,"\u5171\u4eab\u4efb\u52a1\u4e2d\u8bbe\u7f6e\u7684\u6570\u636e\u8fdb\u884c\u8bad\u7ec3\u548c\u6d4b\u8bd5":60,"\u5171\u4eab\u5b58\u50a8\u6302\u5728\u7684\u8def\u5f84":49,"\u5171\u670932":53,"\u5173\u4e8e\u5982\u4f55\u5b9a\u4e49\u7f51\u7edc\u4e2d\u7684\u5c42":54,"\u5173\u4e8e\u65f6\u95f4\u5e8f\u5217":32,"\u5173\u4e8e\u6784\u5efa\u548c\u6d4b\u8bd5\u7684\u66f4\u591a\u4fe1\u606f":36,"\u5173\u4e8epaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":49,"\u5173\u4e8eunbound":34,"\u5173\u4e8evgg\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u7684\u63cf\u8ff0\u53ef\u4ee5\u53c2\u8003":54,"\u5176\u4e0b\u5b50\u6587\u4ef6\u5939\u7684\u7ed3\u6784\u5982\u4e0b":54,"\u5176\u4e2d":[3,20,23,24,25,35,37,46,53,54,55],"\u5176\u4e2d156\u548c285\u662f\u8fd9\u4e9b\u56fe\u50cf\u7684\u5206\u7c7b\u6807\u7b7e":55,"\u5176\u4e2d50000\u5f20\u56fe\u7247\u4f5c\u4e3a\u8bad\u7ec3\u96c6":54,"\u5176\u4e2d\u5206\u522b\u5305\u542b\u4e86cifar":54,"\u5176\u4e2d\u5305\u542b6":58,"\u5176\u4e2d\u5305\u542b\u4e86200\u79cd\u9e1f\u7c7b\u7684\u7167\u7247":54,"\u5176\u4e2d\u5305\u542b\u7b97\u6cd5\u548c\u7f51\u7edc\u914d\u7f6e":61,"\u5176\u4e2d\u5305\u62ec\u51fd\u6570":60,"\u5176\u4e2d\u5b9a\u4e49\u4e86\u6a21\u578b\u67b6\u6784\u548csolver\u914d\u7f6e":62,"\u5176\u4e2d\u6570\u636e\u6e90\u914d\u7f6e\u4e0edataprovider\u7684\u5173\u7cfb\u662f":46,"\u5176\u4e2d\u6587\u672c\u8f93\u5165\u7c7b\u578b\u5b9a\u4e49\u4e3a\u6574\u6570\u65f6\u5e8f\u7c7b\u578binteger_value_sequ":57,"\u5176\u4e2d\u6bcf\u4e00\u884c\u5bf9\u5e94\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u5730\u5740":2,"\u5176\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u53cc\u5c42\u5e8f\u5217\u4e2d\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a":31,"\u5176\u4e2d\u6bcf\u4e2a\u5411\u91cf\u5bf9\u5e94\u8f93\u5165\u8bed\u53e5\u4e2d\u7684\u4e00\u4e2a\u5143\u7d20":62,"\u5176\u4e2d\u6bcf\u6761pass\u82b1\u8d39\u4e867\u4e2a\u5c0f\u65f6":62,"\u5176\u4e2d\u6bcf\u884c\u6570\u636e\u4ee3\u8868\u4e00\u5f20\u56fe\u7247":3,"\u5176\u4e2d\u8be6\u7ec6\u8bf4\u660e\u4e86\u6a21\u578b\u67b6\u6784":62,"\u5176\u4e2d\u8f93\u5165\u56fe\u50cf\u7684\u989c\u8272\u901a\u9053\u987a\u5e8f\u4e3a":55,"\u5176\u4e2dbeam":62,"\u5176\u4e2dcheckgrad\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":43,"\u5176\u4e2dmean\u548cstd\u662f\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u53c2\u6570":43,"\u5176\u4e2dvalue\u5373\u4e3asoftmax\u5c42\u7684\u8f93\u51fa":5,"\u5176\u4ed6":58,"\u5176\u4ed6\u4eba\u5f88\u5bb9\u6613\u770b\u61c2\u955c\u50cf\u751f\u6210\u8fc7\u7a0b":27,"\u5176\u4ed6\u516d\u884c\u5217\u51fa\u4e86\u96c6\u675f\u641c\u7d22\u7684\u7ed3\u679c":62,"\u5176\u4ed6\u5185\u5b58\u6742\u9879":24,"\u5176\u4ed6\u5185\u5b58\u6742\u9879\u662f\u6307paddlepaddle\u672c\u8eab\u6240\u7528\u7684\u4e00\u4e9b\u5185\u5b58":24,"\u5176\u4ed6\u51fd\u6570\u5747\u8fd4\u56de":21,"\u5176\u4ed6\u53c2\u6570\u4f7f\u7528":3,"\u5176\u4ed6\u53c2\u6570\u8bf7\u53c2\u8003":57,"\u5176\u4ed6\u6240\u6709\u5c42\u90fd\u4f1a\u4f7f\u7528gpu\u8ba1\u7b97":45,"\u5176\u4ed6\u7528\u6237\u7684fork\u7248\u672c\u5e93\u5e76\u4e0d\u9700\u8981\u4e25\u683c\u9075\u5b88":23,"\u5176\u4ed6\u9ad8\u7ea7\u529f\u80fd\u5305\u62ec\u5b9a\u4e49\u591a\u4e2amemori":35,"\u5176\u4f1a\u81ea\u52a8\u88ab\u52a0\u5165\u7f16\u8bd1\u5217\u8868":37,"\u5176\u4f59\u884c\u662f":53,"\u5176\u4f5c\u7528\u662f\u5c06\u6570\u636e\u4f20\u5165\u5185\u5b58\u6216\u663e\u5b58":2,"\u5176\u5177\u4f53\u8bf4\u660e\u4e86\u5b57\u6bb5\u7c7b\u578b\u548c\u6587\u4ef6\u540d\u79f0":59,"\u5176\u5185\u90e8\u7684\u6587\u4ef6\u4e5f\u4f1a\u968f\u4e4b\u6d88\u5931":47,"\u5176\u5305\u62ec\u4e24\u4e2a\u51fd\u6570":57,"\u5176\u53c2\u6570\u5982\u4e0b":3,"\u5176\u5b83\u90e8\u5206\u548c\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u7ed3\u6784\u4e00\u81f4":57,"\u5176\u5b83layer\u7684\u8f93\u51fa":34,"\u5176\u5b9e\u4e5f\u662f\u548c\u6bcf\u4e2amini":24,"\u5176\u63d0\u4f9b\u5e94\u7528\u90e8\u7f72":47,"\u5176\u6b21":[3,32,57],"\u5176\u76ee\u7684\u662f\u5728\u7ed9\u5b9a\u7684\u8f93\u5165\u53e5\u5b50\u4e2d\u53d1\u73b0\u6bcf\u4e2a\u8c13\u8bcd\u7684\u8c13\u8bcd\u8bba\u5143\u7ed3\u6784":60,"\u5176\u8bf4\u660e\u5982\u4e0b":32,"\u5176\u8f93\u5165\u53c2\u6570\u5982\u4e0b":62,"\u5176\u8f93\u51fa\u88ab\u7528\u4f5cmemory\u7684\u521d\u59cb\u503c":35,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":[21,24],"\u5177\u4f53\u539f\u56e0\u53c2\u8003":21,"\u5177\u4f53\u53ef\u4ee5\u53c2\u8003":[3,37],"\u5177\u4f53\u53ef\u53c2\u8003\u6587\u6863":34,"\u5177\u4f53\u5982\u4e0b":27,"\u5177\u4f53\u60c5\u51b5\u56e0\u4eba\u800c\u5f02":40,"\u5177\u4f53\u64cd\u4f5c\u5982\u4e0b":24,"\u5177\u4f53\u6d41\u7a0b\u5982\u4e0b":57,"\u5177\u4f53\u7684\u4f7f\u7528\u65b9\u6cd5\u8bf7\u53c2\u8003":46,"\u5177\u4f53\u7684\u683c\u5f0f\u8bf4\u660e":3,"\u5177\u4f53\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":24,"\u5177\u4f53\u8ba1\u7b97\u662f\u901a\u8fc7\u5185\u90e8\u7684":46,"\u5177\u4f53\u8bf7\u53c2\u7167\u793a\u4f8b":55,"\u5177\u4f53\u8bf7\u53c2\u8003":[3,21,36],"\u5177\u4f53\u8bf7\u89c1":36,"\u5177\u6709\u76f8\u540c\u7684\u7ed3\u679c\u4e86":32,"\u5177\u6709\u81ea\u5faa\u73af\u8fde\u63a5\u7684\u795e\u7ecf\u5143":61,"\u517c\u5907\u6613\u7528\u6027":0,"\u5185":35,"\u5185\u5b58":40,"\u5185\u5b58\u5bb9\u9650\u9608\u503c":43,"\u5185\u5bb9":57,"\u5185\u5bb9\u5982\u4e0b":48,"\u5185\u5c42inner_step\u7684recurrent_group\u548c\u5355\u5c42\u5e8f\u5217\u7684\u51e0\u4e4e\u4e00\u6837":32,"\u5185\u5df2\u7ecf\u5305\u542bpaddlepaddle\u7684\u6267\u884c\u7a0b\u5e8f\u4f46\u662f\u8fd8\u6ca1\u4e0a\u8ff0\u529f\u80fd":49,"\u5185\u7684\u5f00\u53d1\u8005\u4e0b\u8f7ddocker\u955c\u50cf":27,"\u5185\u90e8":49,"\u5185\u90e8\u9a71\u52a8python\u89e3\u91ca\u5668\u8fdb\u884c\u6a21\u578b\u914d\u7f6e\u89e3\u6790\u548c\u6570\u636e\u8bfb\u53d6":20,"\u518d\u4e3apaddle\u7684\u8bad\u7ec3\u8fc7\u7a0b\u63d0\u4f9b\u6587\u4ef6\u5217\u8868":59,"\u518d\u4f20\u5165\u7ed9train":46,"\u518d\u5728\u6bcf\u4e00\u4e2aapi\u4e2d\u81ea\u5df1\u68c0\u67e5\u7c7b\u578b":20,"\u518d\u57fa\u4e8e":23,"\u518d\u5bf9\u6bcf\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u8fdb\u884c\u5904\u7406":32,"\u518d\u5bf9\u6bcf\u4e00\u53e5\u8bdd\u7684\u7f16\u7801\u5411\u91cf\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u6bb5\u843d\u7684\u5411\u91cf":32,"\u518d\u5bf9\u8fd9\u4e2a\u6bb5\u843d\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":32,"\u518d\u6307\u5b9a":26,"\u518d\u6b21\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":40,"\u518d\u7528\u8fd9\u4e2a\u68af\u5ea6\u53bb\u548c":37,"\u518d\u901a\u8fc7\u51fd\u6570":49,"\u5192\u9669\u7247":58,"\u5197\u4f59\u7b49\u529f\u80fd":47,"\u5199\u4ee3\u7801":20,"\u5199\u5165\u5feb\u7167\u6570\u636e":18,"\u5199\u5b8c\u6a21\u578b\u914d\u7f6e\u4e4b\u540e":62,"\u5199\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5\u662f\u4e00\u4e2a\u9a8c\u8bc1\u65b0\u5b9e\u73b0\u7684\u5c42\u662f\u5426\u6b63\u786e\u7684\u76f8\u5bf9\u7b80\u5355\u7684\u529e\u6cd5":37,"\u519c\u6c11":58,"\u51c6\u5907":32,"\u51c6\u5907\u597d\u6570\u636e":59,"\u51c6\u5907\u7528\u6765\u5b66\u4e60\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u5e8f\u5217\u6570\u636e":35,"\u51c6\u5907\u9884\u6d4b\u6570\u636e":5,"\u51cf\u5c0f\u5e8f\u5217\u7684\u957f\u5ea6":24,"\u51cf\u5c0f\u8fd9\u4e2a\u5185\u5b58\u6c60\u5373\u53ef\u51cf\u5c0f\u5185\u5b58\u5360\u7528":24,"\u51cf\u5c0fbatch":24,"\u51fa\u53bb\u73a9":32,"\u51fa\u5dee":32,"\u51fa\u6765":32,"\u51fa\u73b0\u4ee5\u4e0b\u9519\u8bef":24,"\u51fa\u73b0\u8fd9\u4e2a\u95ee\u9898\u7684\u4e3b\u8981\u539f\u56e0\u662f":24,"\u51fd\u6570":[3,25,35,37,40,60,61],"\u51fd\u6570\u4e2d":35,"\u51fd\u6570\u4e2d\u4f7f\u7528":3,"\u51fd\u6570\u4e2d\u8bbe\u7f6e\u7684":41,"\u51fd\u6570\u5047\u8bbe":35,"\u51fd\u6570\u52a0\u5230\u4ee3\u7801\u4e2d":40,"\u51fd\u6570\u5373\u53ef\u5b8c\u6210\u8f6c\u6362":19,"\u51fd\u6570\u53ea\u5173\u6ce8\u4e8ernn\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8ba1\u7b97":34,"\u51fd\u6570\u540d\u4e3a":21,"\u51fd\u6570\u547d\u540d":20,"\u51fd\u6570\u5c06\u8fd4\u56de\u4e09\u4e2a\u6574\u6570\u5217\u8868":35,"\u51fd\u6570\u5c31\u662f\u6839\u636e\u8be5\u673a\u5236\u914d\u7f6e\u7684":3,"\u51fd\u6570\u5f97\u5230\u7684\u68af\u5ea6\u53bb\u5bf9\u6bd4":37,"\u51fd\u6570\u5fc5\u987b\u5148\u8c03\u7528\u57fa\u7c7b\u4e2d\u7684\u51fd\u6570":37,"\u51fd\u6570\u5fc5\u987b\u8fd4\u56de\u4e00\u4e2a\u6216\u591a\u4e2alayer\u7684\u8f93\u51fa":34,"\u51fd\u6570\u6307\u51fa\u4e86\u5728\u8bad\u7ec3\u65f6\u9700\u8981\u4ece\u53c2\u6570\u670d\u52a1\u5668\u53d6\u51fa\u7684\u884c":37,"\u51fd\u6570\u6765\u5c06\u4fe1\u606f\u8f93\u51fa\u5230\u754c\u9762\u4e2d":40,"\u51fd\u6570\u67e5\u8be2\u8f6f\u4ef6\u5305\u76f8\u5173api\u8bf4\u660e":5,"\u51fd\u6570\u7684":3,"\u51fd\u6570\u7684\u5b9e\u73b0\u662f\u6b63\u786e\u7684":37,"\u51fd\u6570\u7684\u5f00\u5934\u5fc5\u987b\u8c03\u7528":37,"\u5206\u4e3a\u597d\u8bc4":57,"\u5206\u522b\u4e3a":53,"\u5206\u522b\u4e3atrain":62,"\u5206\u522b\u4ece\u8bcd\u8bed\u548c\u53e5\u5b50\u7ea7\u522b\u7f16\u7801\u8f93\u5165\u6570\u636e":34,"\u5206\u522b\u4f7f\u7528\u5355\u53cc\u5c42rnn\u4f5c\u4e3a\u7f51\u7edc\u914d\u7f6e\u7684\u6a21\u578b":32,"\u5206\u522b\u5305\u542b\u4e86\u6cd5\u8bed\u5230\u82f1\u8bed\u7684\u5e73\u884c\u8bed\u6599\u5e93\u7684\u8bad\u7ec3\u6570\u636e":62,"\u5206\u522b\u5b9a\u4e49\u5b50\u53e5\u7ea7\u522b\u548c\u8bcd\u8bed\u7ea7\u522b\u4e0a\u9700\u8981\u5b8c\u6210\u7684\u8fd0\u7b97":34,"\u5206\u522b\u5bf9\u5e94\u4e8e\u53d8\u91cf":25,"\u5206\u522b\u662f":31,"\u5206\u522b\u662frnn\u72b6\u6001\u548c\u8f93\u5165\u7684\u53d8\u6362\u77e9\u9635":35,"\u5206\u522b\u662fsentences\u548clabel":32,"\u5206\u522b\u662fwords\u548clabel":32,"\u5206\u522b\u8ba1\u7b97\u6bcf\u4e2a\u53c2\u6570\u7684\u68af\u5ea6":37,"\u5206\u522b\u8fdb\u884c\u5e8f\u5217\u64cd\u4f5c":32,"\u5206\u5272":[58,60],"\u5206\u5272\u6587\u4ef6\u7684\u65b9\u6cd5\u662f":59,"\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":18,"\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":47,"\u5206\u6210\u4e24\u90e8\u5206":3,"\u5206\u652f":[23,36],"\u5206\u652f\u4e00\u65e6\u5efa\u7acb":23,"\u5206\u652f\u4e0a":36,"\u5206\u652f\u4e0a\u521b\u5efa\u65b0\u5206\u652f":36,"\u5206\u652f\u4e2d":23,"\u5206\u652f\u4e3a\u5f00\u53d1":23,"\u5206\u652f\u4e3a\u6bcf\u4e00\u6b21release\u65f6\u5efa\u7acb\u7684\u4e34\u65f6\u5206\u652f":23,"\u5206\u652f\u4e3a\u7a33\u5b9a":23,"\u5206\u652f\u529f\u80fd\u7684\u5c01\u95ed":23,"\u5206\u652f\u5408\u5165":23,"\u5206\u652f\u5408\u5165master\u5206\u652f":23,"\u5206\u652f\u540c\u6b65\u4e3b\u7248\u672c\u5e93\u7684":23,"\u5206\u652f\u540d":36,"\u5206\u652f\u540d\u4e3a":23,"\u5206\u652f\u5b58\u5728\u7684\u65f6\u5019":23,"\u5206\u652f\u6d3e\u751f\u51fa\u65b0\u7684\u5206\u652f":23,"\u5206\u652f\u7684\u7248\u672c\u90fd\u662f\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5\u548c\u56de\u5f52\u6d4b\u8bd5\u7684\u7248\u672c":23,"\u5206\u652f\u7684\u7248\u672c\u90fd\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5":23,"\u5206\u652f\u89c4\u8303":36,"\u5206\u6790\u5f97\u5230\u7684\u4fe1\u606f\u7528\u4e8e\u534f\u52a9\u8fdb\u884c\u7a0b\u5e8f\u7684\u4f18\u5316":40,"\u5206\u7247":18,"\u5206\u7c7b\u6210\u6b63\u9762\u60c5\u7eea\u548c\u8d1f\u9762\u60c5\u7eea\u4e24\u7c7b":3,"\u5206\u7c7b\u8bef\u5dee\u662f0":61,"\u5206\u7c7b\u9519\u8bef\u7387\u548c\u6a21\u578b\u5927\u5c0f\u7531\u4e0b\u8868\u7ed9\u51fa":55,"\u5206\u8bcd\u5e8f\u5217\u7684\u5f00\u59cb":53,"\u5206\u8bcd\u5e8f\u5217\u7684\u7ed3\u675f":53,"\u5206\u8bcd\u98ce\u683c\u5982\u4e0b":53,"\u5206\u914d\u5230\u5f53\u524d\u6570\u636e\u5757\u6837\u672c\u6570\u7684\u56db\u5206\u4e4b\u4e00":43,"\u5206\u9694":[53,59],"\u5206\u9694\u7b26\u4e3a":58,"\u5207\u6362\u5230":36,"\u5207\u6362\u5230\u6240\u5efa\u5206\u652f":36,"\u5217\u8868":59,"\u5217\u8868\u5982\u4e0b":3,"\u5219\u4e0d\u5728\u4e4e\u5185\u5b58\u6682\u5b58\u591a\u5c11\u6761\u6570\u636e":3,"\u5219\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":37,"\u5219\u4f1a\u9884\u5148\u8bfb\u53d6\u5168\u90e8\u6570\u636e\u5230\u5185\u5b58\u4e2d":3,"\u5219\u4f1a\u9ed8\u8ba4\u751f\u6210\u4e00\u4e2alist\u6587\u4ef6":46,"\u5219\u4f7f\u7528\u533a\u57df\u6807\u8bb0":60,"\u5219\u4f7f\u7528\u540c\u6b65\u8bad\u7ec3":43,"\u5219\u4f7f\u7528\u542f\u52a8\u53c2\u6570\u5b9a\u4e49\u7684\u521d\u59cb\u5316\u65b9\u6cd5\u521d\u59cb\u5316\u53c2\u6570":18,"\u5219\u4f7f\u7528\u8be5\u53c2\u6570\u4f5c\u4e3a\u9ed8\u8ba4\u503c":43,"\u5219\u5148\u505a\u5d4c\u5165":59,"\u5219\u53ef\u4ee5\u50cf":41,"\u5219\u5b57\u4e0e\u5b57\u4e4b\u95f4\u7528\u7a7a\u683c\u5206\u9694":57,"\u5219\u5ffd\u7565":18,"\u5219\u603b\u4f1a\u663e\u793a\u963b\u9694\u6458\u8981\u4fe1\u606f":43,"\u5219\u628a\u53e6\u4e00\u4e2a\u6162\u901f\u7684kill\u6389":18,"\u5219\u63a8\u8350\u5927\u4e8e\u8bad\u7ec3\u65f6batch":3,"\u5219\u662f\u5e26gui\u7684nvidia\u53ef\u89c6\u5316\u6027\u80fd\u5206\u6790\u5de5\u5177":40,"\u5219\u663e\u793a\u963b\u9694\u6027\u80fd\u7684\u6458\u8981\u4fe1\u606f":43,"\u5219\u76f4\u63a5\u5f15\u5165\u53e6\u4e00\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5219\u9700\u8981\u4f7f\u7528\u7b49\u4e8e\u6743\u91cd\u53c2\u6570\u89c4\u6a21\u5927\u7ea65\u500d\u7684\u5185\u5b58":24,"\u5219\u9700\u8981\u56de\u6eda\u5230\u4e0a\u4e00\u4e2a\u68c0\u67e5\u70b9":18,"\u5219\u9700\u8981\u914d\u7f6e":47,"\u521b\u5efa\u4e00\u4e2akubernet":49,"\u521b\u5efa\u548c\u53d1\u5e03\u81ea\u5df1\u7684ai\u7a0b\u5e8f\u955c\u50cf":27,"\u521b\u5efa\u5e76\u5207\u6362\u5230\u65b0\u5206\u652f":36,"\u521b\u5efa\u6210\u529f\u540e":49,"\u521b\u5efa\u8bad\u7ec3\u6570\u636e\u7684":62,"\u521b\u5efagener":3,"\u521d\u59cb\u5316\u4e4b\u540e":5,"\u521d\u59cb\u5316\u504f\u7f6e\u5411\u91cf":37,"\u521d\u59cb\u5316\u65f6\u8c03\u7528\u7684\u51fd\u6570":3,"\u521d\u59cb\u5316\u6743\u91cd\u8868":37,"\u521d\u59cb\u5316\u6a21\u578b\u7684\u8def\u5f84":43,"\u521d\u59cb\u5316\u6a21\u578b\u7684\u8def\u5f84\u914d\u7f6e\u4e3a":53,"\u521d\u59cb\u5316\u7236\u7c7b":37,"\u521d\u59cb\u5316biases_":37,"\u521d\u59cb\u5316paddlepaddle\u73af\u5883":5,"\u521d\u59cb\u72b6\u6001":34,"\u5220\u9664":36,"\u5220\u9664\u78c1\u76d8\u76ee\u5f55\u4e2d\u4e0d\u662f\u5f53\u524duuid\u7684\u5feb\u7167\u6587\u4ef6":18,"\u5229\u7528\u5206\u5e03\u5f0f\u8bad\u7ec3\u9a7e\u9a6d\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":24,"\u5229\u7528\u5355\u8bcdid\u67e5\u627e\u8be5\u5355\u8bcd\u5bf9\u5e94\u7684\u8fde\u7eed\u5411\u91cf":57,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90\u53ef\u4ee5\u5206\u4e3a\u4e00\u4e0b\u51e0\u4e2a\u65b9\u5f0f\u6765\u8fdb\u884c":24,"\u5229\u7528\u6b63\u5219\u8868\u8fbe\u5f0f\u6765\u89e3\u6790\u8be5\u7279\u5f81":59,"\u5229\u7528\u8bad\u7ec3\u96c6\u751f\u6210\u7684\u5b57\u5178":61,"\u5229\u7528\u8fd9\u79cd\u7279\u6027":34,"\u5229\u7528\u903b\u8f91\u56de\u5f52\u6a21\u578b\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":57,"\u5229\u7528kubernetes\u80fd\u65b9\u4fbf\u5730\u7ba1\u7406\u8de8\u673a\u5668\u8fd0\u884c\u5bb9\u5668\u5316\u7684\u5e94\u7528":47,"\u5229\u843d":32,"\u5230":[18,24,35],"\u5230\u6240\u6709\u8282\u70b9\u800c\u4e0d\u7528\u5bc6\u7801":41,"\u5230\u672c\u5730":36,"\u5230\u6b64":27,"\u5230\u76ee\u524d\u4e3a\u6b62":60,"\u5236\u4f5c\u65b0\u955c\u50cf\u6765\u5b8c\u6210\u4ee5\u4e0a\u7684\u5de5\u4f5c":49,"\u5236\u4f5cpaddlepaddle\u5f00\u53d1\u955c\u50cf":27,"\u5236\u4f5cpaddlepaddle\u751f\u4ea7\u955c\u50cf":27,"\u5236\u4f5cpaddlepaddle\u955c\u50cf":49,"\u5237\u7259":32,"\u524d\u4e00\u7bc7\u6587\u7ae0\u4ecb\u7ecd\u4e86\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673apaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":49,"\u524d\u4e09\u884cimport\u4e86\u5b9a\u4e49network":62,"\u524d\u53f0":32,"\u524d\u5411\u4f20\u64ad":37,"\u524d\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u5165":37,"\u524d\u5411\u548c\u540e\u5411":37,"\u524d\u81ea\u52a8\u68c0\u67e5\u4e00\u4e9b\u57fa\u672c\u4e8b\u5b9c":36,"\u5269\u4e0b\u7684pass\u4f1a\u76f4\u63a5\u4ece\u5185\u5b58\u91cc":3,"\u529f\u80fd\u7684\u6b63\u786e\u6027\u5305\u62ec\u9a8c\u8bc1paddle\u76ee\u524d\u7684":23,"\u52a0\u4e0a\u504f\u7f6e\u5411\u91cf":37,"\u52a0\u4e86l2\u6b63\u5219\u548c\u68af\u5ea6\u622a\u65ad":57,"\u52a0\u5165":40,"\u52a0\u6743\u548c\u7528\u6765\u751f\u6210":35,"\u52a0\u6743\u7f16\u7801\u5411\u91cf":35,"\u52a0\u8f7d\u6570\u636e":60,"\u52a0\u8f7d\u6a21\u578b":60,"\u52a0\u8f7d\u6a21\u578b\u53c2\u6570":62,"\u52a0\u8f7dtest":43,"\u52a0\u901fpaddlepaddle\u8bad\u7ec3\u53ef\u4ee5\u8003\u8651\u4ece\u4ee5\u4e0b\u51e0\u4e2a\u65b9\u9762":24,"\u52a8\u4f5c\u7247":58,"\u52a8\u6001\u5e93":20,"\u52a8\u753b\u7247":58,"\u52a8\u8bcd":60,"\u52a9\u624b":37,"\u5305\u542b12":61,"\u5305\u542b20\u4e2a\u8bad\u7ec3\u6837\u4f8b":53,"\u5305\u542b3\u4e2a\u5c5e\u6027":53,"\u5305\u542b50":61,"\u5305\u542b\u4e86\u67d0\u79cd\u7c7b\u578b\u7684\u7c7b\u578b\u5b9a\u4e49\u548c\u66b4\u9732\u7684\u5168\u90e8\u51fd\u6570":21,"\u5305\u542b\u7684\u6240\u6709\u4f9d\u8d56\u5047\u8bbe\u90fd\u53ef\u4ee5\u5728paddle\u7684\u8fd0\u884c\u5bb9\u5668\u4e2d":27,"\u5305\u5e76\u91cd\u65b0\u7f16\u8bd1paddlepaddl":24,"\u5305\u62ec":[19,43,57,60,62],"\u5305\u62ec\u4e86\u56fe\u50cf\u7684\u5377\u79ef":46,"\u5305\u62ec\u4ee5\u4e0b\u4e24\u79cd":3,"\u5305\u62ec\u53d1\u884c\u65f6\u95f4":58,"\u5305\u62ec\u5b57\u7b26\u4e32\u5206\u914d":24,"\u5305\u62ec\u5b66\u4e60\u7387":46,"\u5305\u62ec\u6570\u636e\u8f93\u5165":25,"\u5305\u62ec\u6743\u91cdw\u548c\u504f\u7f6eb":18,"\u5305\u62ec\u751f\u6210cpu":26,"\u5305\u62ec\u7b80\u5355\u7684":57,"\u5305\u62ecbool":45,"\u5305\u62ecdocker\u955c\u50cf":28,"\u5305\u62eclinux":27,"\u5305\u662f\u6700\u65b0\u7684":24,"\u5305\u6bd4\u8f83\u8001":24,"\u5305\u7684\u65b9\u6cd5\u662f":24,"\u533a\u522b\u662f\u540c\u65f6\u5904\u7406\u4e86\u4e24\u4e2a\u8f93\u5165":32,"\u533a\u522b\u662frnn\u4f7f\u7528\u4e24\u5c42\u5e8f\u5217\u6a21\u578b":32,"\u533b\u751f":58,"\u533b\u7597\u4fdd\u5065":58,"\u5341\u4e00":32,"\u5347\u5e8f\u6392\u5217":62,"\u534e\u6da6\u4e07\u5bb6":32,"\u534f\u540c\u5b8c\u6210releas":23,"\u5355\u4e2a\u503c":19,"\u5355\u4f4d\u662fmb":43,"\u5355\u5143\u6d4b\u8bd5\u4f1a\u5f15\u7528site":24,"\u5355\u5143\u6d4b\u8bd5checkgrad_ep":42,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u53e5\u5b50\u662f\u4e00\u6837\u7684":32,"\u5355\u53cc\u5c42rnn":33,"\u5355\u53d8\u91cf\u7684\u7ebf\u6027\u56de\u5f52":25,"\u5355\u5c42":34,"\u5355\u5c42\u4e0d\u7b49\u957frnn":32,"\u5355\u5c42\u548c\u53cc\u5c42\u5e8f\u5217\u7684\u4f7f\u7528\u548c\u793a\u4f8b2\u4e2d\u7684\u793a\u4f8b\u7c7b\u4f3c":32,"\u5355\u5c42\u5e8f\u5217":31,"\u5355\u5c42\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20":31,"\u5355\u5c42\u5e8f\u5217\u7b2ci\u4e2a\u5143\u7d20":31,"\u5355\u5c42\u6216\u53cc\u5c42":31,"\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":32,"\u5355\u5c42rnn":[32,34],"\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":32,"\u5355\u673a\u6a21\u5f0f\u7528\u547d\u4ee4":46,"\u5355\u673a\u8bad\u7ec3\u901a\u5e38\u53ea\u5305\u62ec\u4e00\u4e2atrainer\u8fdb\u7a0b":46,"\u5355\u673acpu\u8bad\u7ec3":24,"\u5355\u673agpu\u8bad\u7ec3":24,"\u5355\u6b65\u51fd\u6570":35,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u5728":35,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u90fd\u975e\u5e38\u7b80\u5355":35,"\u5355\u6b65\u51fd\u6570\u7684\u5b9e\u73b0\u5982\u4e0b\u6240\u793a":35,"\u5355\u70b9\u6545\u969c":18,"\u5355\u8fdb\u5355\u51fa":34,"\u5360\u4f4d\u7b26":53,"\u536b\u751f":32,"\u5373":[21,24,25,38,49,57,61],"\u5373\u4e00\u4e2a\u5c06\u5355\u8bcd\u5b57\u7b26\u4e32\u6620\u5c04\u5230\u5355\u8bcdid\u7684\u5b57\u5178":3,"\u5373\u4e0a\u8ff0\u4ee3\u7801\u4e2d\u7684\u7b2c19\u884c":32,"\u5373\u4e0d\u8981\u5c06\u6bcf\u4e00\u4e2a\u6837\u672c\u90fd\u653e\u5165train":3,"\u5373\u4e0d\u9700\u8981\u4f7f\u7528memori":32,"\u5373\u4e3a\u4e00\u4e2a\u65f6\u95f4\u6b65":32,"\u5373\u4e3a\u5355\u5c42rnn\u5e8f\u5217\u7684\u4f7f\u7528\u4ee3\u7801":32,"\u5373\u4e3a\u65f6\u95f4\u5e8f\u5217\u7684\u8f93\u5165":32,"\u5373\u4e3a\u8fd9\u4e2a\u53cc\u5c42rnn\u7684\u7f51\u7edc\u7ed3\u6784":32,"\u5373\u4e3a\u8fd9\u4e2a\u6570\u636e\u6587\u4ef6\u7684\u540d\u5b57":3,"\u5373\u4e8c\u7ef4\u6570\u7ec4":32,"\u5373\u4f7f\u7528":21,"\u5373\u4f7f\u7528\u6237\u76f4\u63a5\u5f15\u7528\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5373\u4f7f\u95f4\u9694\u5f88\u5c0f":43,"\u5373\u4f7fc":21,"\u5373\u4f7fprocess\u51fd\u6570\u91cc\u9762\u53ea\u6709\u4e00\u4e2ayield":3,"\u5373\u4f8b\u5982":21,"\u5373\u4fbf\u8bbe\u7f6e":24,"\u5373\u4fbfpaddl":21,"\u5373\u521d\u59cb\u72b6\u6001\u4e3a0":34,"\u5373\u5305\u542b\u65f6\u95f4\u6b65\u4fe1\u606f":3,"\u5373\u5355\u65f6\u95f4\u6b65\u6267\u884c\u7684\u51fd\u6570":35,"\u5373\u53cc\u5411lstm\u548c\u4e09\u5c42\u5806\u53e0lstm":61,"\u5373\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":32,"\u5373\u53cc\u5c42rnn\u7684\u6bcf\u4e2a\u72b6\u6001":34,"\u5373\u53ef":25,"\u5373\u53ef\u4ee5\u6781\u5927\u7684\u52a0\u901f\u6570\u636e\u8f7d\u5165\u6d41\u7a0b":24,"\u5373\u53ef\u8fdb\u884c\u4e0b\u4e00\u6b65\u64cd\u4f5c":38,"\u5373\u5728\u53cc\u5c42\u5e8f\u5217\u7684\u539f\u59cb\u6570\u636e\u4e2d":32,"\u5373\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":24,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":3,"\u5373\u5b8c\u6210\u67d0\u4e00\u4e2a\u4efb\u52a1\u7684\u6700\u5c11\u51fd\u6570":21,"\u5373\u5bf9\u7b2c\u4e09\u6b65\u8fdb\u884c\u66ff\u6362":57,"\u5373\u5c06\u4e00\u6bb5\u82f1\u6587\u6587\u672c\u6570\u636e":3,"\u5373\u5c06\u4e00\u6bb5\u8bdd\u8fdb\u884c\u5206\u7c7b":32,"\u5373\u5f53\u524d\u65f6\u95f4\u6b65\u4e0b\u7684\u795e\u7ecf\u7f51\u7edc\u4f9d\u8d56\u524d\u4e00\u4e2a\u65f6\u95f4\u6b65\u795e\u7ecf\u7f51\u7edc\u4e2d\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u8f93\u51fa":32,"\u5373\u6211\u4eec\u7684\u8bad\u7ec3\u76ee\u6807":25,"\u5373\u628a\u5355\u5c42rnn\u751f\u6210\u540e\u7684subseq\u7ed9\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42seq":34,"\u5373\u6574\u4e2a\u53cc\u5c42group\u662f\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":32,"\u5373\u6574\u4e2a\u8f93\u5165\u5e8f\u5217":31,"\u5373\u6574\u6570\u6570\u7ec4":32,"\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":32,"\u5373\u662f\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684\u7f51\u7edc\u8fde\u63a5":32,"\u5373\u66b4\u9732":21,"\u5373\u6b63\u9762\u548c\u8d1f\u9762":61,"\u5373\u6b63\u9762\u8bc4\u4ef7\u6807\u7b7e\u548c\u8d1f\u9762\u8bc4\u4ef7\u6807\u7b7e":61,"\u5373\u7279\u5f81\u7684\u6570\u7ec4":32,"\u5373\u7f51\u5361\u540d":49,"\u5373\u82e5\u5e72\u6570\u636e\u6587\u4ef6\u8def\u5f84\u7684\u67d0\u4e00\u4e2a":3,"\u5373\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0nan\u6216\u8005inf":24,"\u5373\u8bbe\u7f6e":24,"\u5373\u8fd9\u4e2a\u52a8\u6001\u5e93\u662f\u4e0d\u4f9d\u8d56\u4e8e\u5176\u4ed6\u4efb\u4f55\u6587\u4ef6\u7684":20,"\u5373define_py_data_sources2\u5e94\u6539\u4e3a":24,"\u5373input":34,"\u5373rnn\u4e4b\u95f4\u6709\u4e00\u6b21\u5d4c\u5957\u5173\u7cfb":32,"\u5377\u79ef\u5c42":54,"\u5377\u79ef\u5c42\u6743\u91cd":55,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fa8\u8bc6\u56fe\u7247\u4e2d\u7684\u4e3b\u4f53":54,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u5728\u56fe\u7247\u5206\u7c7b\u4e0a\u6709\u7740\u60ca\u4eba\u7684\u6027\u80fd":54,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u662f\u4e00\u79cd\u4f7f\u7528\u5377\u79ef\u5c42\u7684\u524d\u5411\u795e\u7ecf\u7f51\u7edc":54,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u80fd\u591f\u5f88\u597d\u7684\u8868\u793a\u8fd9\u4e24\u7c7b\u4fe1\u606f":54,"\u5377\u79ef\u7f51\u7edc\u662f\u4e00\u79cd\u7279\u6b8a\u7684\u4ece\u8bcd\u5411\u91cf\u8868\u793a\u5230\u53e5\u5b50\u8868\u793a\u7684\u65b9\u6cd5":57,"\u5378\u8f7dpaddlepaddle\u5305":24,"\u538b\u7f29\u6210\u4e00\u4e2a\u5411\u91cf":32,"\u539f\u56e0\u5728\u4e8e\u6ca1\u6709\u628a\u673a\u5668\u4e0acuda\u76f8\u5173\u7684\u9a71\u52a8\u548c\u5e93\u6620\u5c04\u5230\u5bb9\u5668\u5185\u90e8":24,"\u539f\u56e0\u662f\u672a\u8bbe\u7f6ecuda\u8fd0\u884c\u65f6\u73af\u5883\u53d8\u91cf":29,"\u53bb\u8fc7":32,"\u53c2\u6570":[3,7,8,9,10,11,12,13,15,20,37,42,49,53,55,61],"\u53c2\u6570\u5171\u4eab\u7684\u914d\u7f6e\u793a\u4f8b\u4e3a":24,"\u53c2\u6570\u521d\u59cb\u5316\u8def\u5f84":60,"\u53c2\u6570\u5373\u53ef":61,"\u53c2\u6570\u540d":55,"\u53c2\u6570\u6570\u91cf":57,"\u53c2\u6570\u670d\u52a1\u5668":42,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":43,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u76d1\u542c\u7aef\u53e3":43,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u7f51\u7edc\u8bbe\u5907\u540d\u79f0":43,"\u53c2\u6570\u670d\u52a1\u5668\u7684ip\u5730\u5740":43,"\u53c2\u6570\u670d\u52a1\u5668\u7a00\u758f\u66f4\u65b0\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":43,"\u53c2\u6570\u6765\u63a7\u5236\u7f13\u5b58\u65b9\u6cd5":24,"\u53c2\u6570\u6982\u8ff0":44,"\u53c2\u6570\u7684\u89e3\u6790":49,"\u53c2\u6570\u7ef4\u5ea6":53,"\u53c2\u6570\u884c":53,"\u53c2\u6570\u8bbe\u7f6e\u4e86\u5916\u5c42":32,"\u53c2\u6570\u9700\u8981\u5b9e\u73b0":35,"\u53c2\u7167\u4e0a\u8ff0\u6b65\u9aa4\u66f4\u65b0":36,"\u53c2\u8003":[20,47],"\u53c2\u8003\u5f3a\u8c03\u90e8\u5206":40,"\u53c2\u8003\u6587\u732e":62,"\u53c2\u8003\u65f6\u95f4\u5e8f\u5217":32,"\u53c2\u8003\u955c\u50cf\u7684":49,"\u53c8":32,"\u53c8\u662f\u4e00\u4e2a\u5355\u5c42\u7684\u5e8f\u5217":31,"\u53c8\u8981\u4fdd\u8bc1\u6570\u636e\u662f\u968f\u673a\u7684":24,"\u53ca":37,"\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u9690\u85cf\u72b6\u6001":35,"\u53cc\u5c42":34,"\u53cc\u5c42\u4e0d\u7b49\u957frnn":32,"\u53cc\u5c42\u5e8f\u5217":31,"\u53cc\u5c42\u5e8f\u5217\u6216\u5355\u5c42\u5e8f\u5217":31,"\u53cc\u5c42\u5e8f\u5217\u6570\u636e\u4e00\u5171\u67094\u4e2a\u6837\u672c":32,"\u53cc\u5c42\u5e8f\u5217\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u5e8f\u5217":31,"\u53cc\u5c42\u5e8f\u5217\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":34,"\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u4e2d\u6bcf\u4e2a\u5143\u7d20":31,"\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"\u53cc\u5c42\u6216\u8005\u5355\u5c42":31,"\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u7684\u4ee3\u7801":32,"\u53cc\u5c42rnn":34,"\u53cc\u5c42rnn\u6570\u636e\u968f\u610f\u52a0\u4e86\u4e00\u4e9b\u9694\u65ad":32,"\u53cc\u5c42rnn\u987e\u540d\u601d\u4e49":32,"\u53cc\u7f13\u51b2":46,"\u53cc\u8fdb\u5355\u51fa":34,"\u53cc\u8fdb\u53cc\u51fa":34,"\u53cd\u4e4b\u5219":60,"\u53cd\u5411\u4f20\u64ad":37,"\u53cd\u5411\u4f20\u64ad\u6839\u636e\u8f93\u51fa\u7684\u68af\u5ea6":37,"\u53d1\u5e03":27,"\u53d1\u5e03\u5230dockerhub":23,"\u53d1\u5e03\u5230github":23,"\u53d1\u6563\u5230\u4e86\u4e00\u4e2a\u6570\u503c\u7279\u522b\u5927\u7684\u5730\u65b9":24,"\u53d1\u884c\u548c\u7ef4\u62a4":36,"\u53d1\u9001\u53c2\u6570\u7684\u7aef\u53e3\u53f7":43,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230cuda\u5de5\u5177\u94fe":26,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gtest":26,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230swig":26,"\u53d8\u6362\u77e9\u9635":37,"\u53d8\u91cf\u6765\u8bbe\u7f6e\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761":3,"\u53e3\u5934":32,"\u53e3\u7edf\u8ba1\u5b66\u4fe1\u606f\u7684\u7528\u6237\u624d\u88ab\u5305\u542b\u5728\u6570\u636e\u96c6\u4e2d":58,"\u53e5\u5b50":61,"\u53e5\u5b50\u4e2d\u7684\u7ec4\u5757\u5c06\u4f1a\u626e\u6f14\u67d0\u4e9b\u8bed\u4e49\u89d2\u8272":60,"\u53e5\u5b50\u8868\u793a\u7684\u8ba1\u7b97\u66f4\u65b0\u4e3a\u4e24\u6b65":57,"\u53e6\u4e00\u4e2a\u4f8b\u5b50\u662f\u901a\u8fc7\u5206\u6790\u6bcf\u65e5twitter\u535a\u5ba2\u7684\u6587\u672c\u5185\u5bb9\u6765\u9884\u6d4b\u80a1\u7968\u53d8\u52a8":61,"\u53e6\u4e00\u4e2a\u597d\u5904\u662f\u6211\u4eec\u53ef\u4ee5\u628apaddlepaddle\u5bb9\u5668\u8fd0\u884c\u5728\u8fdc\u7a0b\u670d\u52a1\u5668\u4e0a":27,"\u53e6\u4e00\u4e2a\u662f\u5185\u5b58\u64cd\u4f5c\u91cf":40,"\u53e6\u4e00\u4e2a\u662f\u521b\u5efadockerfile\u5e76\u8fd0\u884cdock":27,"\u53e6\u4e00\u4e2a\u662f\u6bcf\u6761\u5e8f\u5217":24,"\u53e6\u4e00\u4e2a\u7ec8\u7aef\u8fd0\u884cpython":27,"\u53e6\u4e00\u65b9\u9762":61,"\u53e6\u4e00\u79cd\u65b9\u5f0f\u662f\u5c06\u7f51\u7edc\u5c42\u5212\u5206\u5230\u4e0d\u540c\u7684gpu\u4e0a\u53bb\u8ba1\u7b97":45,"\u53e6\u5916":[32,46],"\u53e6\u5916\u4e24\u4e2a\u5206\u522b\u662f\u6ed1\u52a8\u5747\u503c\u548c\u65b9\u5dee":55,"\u53e6\u5916\u7a00\u758f\u66f4\u65b0\u7684\u7aef\u53e3\u5982\u679c\u592a\u5927\u7684\u8bdd":46,"\u53ea\u4f5c\u4e3aread":34,"\u53ea\u4fdd\u5b58\u6700\u540e\u4e00\u8f6e\u7684\u53c2\u6570":43,"\u53ea\u5141\u8bb8\u6574\u6570\u7684\u661f\u7ea7":58,"\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":26,"\u53ea\u5bf9\u7279\u6b8a\u5728\u7ebf\u7cfb\u7edf\u8003\u8651\u4e24\u53f0\u4ee5\u4e0a\u540c\u65f6\u6545\u969c\u7684\u5bb9\u707e":18,"\u53ea\u622a\u53d6\u4e2d\u5fc3\u65b9\u5f62\u7684\u56fe\u50cf\u533a\u57df":55,"\u53ea\u662f\u53cc\u5c42\u5e8f\u5217\u5c06\u5176\u53c8\u505a\u4e86\u5b50\u5e8f\u5217\u5212\u5206":32,"\u53ea\u662f\u5c06\u53e5\u5b50\u7528\u8fde\u7eed\u5411\u91cf\u8868\u793a\u66ff\u6362\u4e3a\u7528\u7a00\u758f\u5411\u91cf\u8868\u793a":57,"\u53ea\u662f\u8bf4\u660e\u6570\u636e\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684":3,"\u53ea\u662f\u8bf7\u4e0d\u8981\u5fd8\u8bb0\u63d0\u524d\u5728\u7269\u7406\u673a\u4e0a\u5b89\u88c5gpu\u6700\u65b0\u9a71\u52a8":27,"\u53ea\u66b4\u9732\u6982\u5ff5\u7684\u63a5\u53e3":21,"\u53ea\u6709":32,"\u53ea\u67092\u4e2a\u914d\u7f6e\u4e0d\u4e00\u6837":53,"\u53ea\u6709\u542b\u6709\u4eba":58,"\u53ea\u6709\u5f53\u8bbe\u7f6e\u4e86spars":43,"\u53ea\u7528\u4e8e\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d\u6307\u5b9a\u8f93\u5165\u6570\u636e":34,"\u53ea\u80fd\u6d4b\u8bd5\u5355\u4e2a\u6a21\u578b":45,"\u53ea\u80fd\u8c03\u7528paddle\u7684\u52a8\u6001\u5e93":20,"\u53ea\u8981\u4e00\u7cfb\u5217\u7279\u5f81\u6570\u636e\u4e2d\u7684":32,"\u53ea\u8981\u51fa\u73b0\u6d6e\u70b9\u6570\u5f02\u5e38":24,"\u53ea\u8981\u5728docker\u91cc\u542f\u52a8paddlepaddle\u7684\u65f6\u5019\u7ed9\u5b83\u4e00\u4e2a\u540d\u5b57":27,"\u53ea\u8bfbmemory\u8f93\u5165":34,"\u53ea\u9488\u5bf9\u5185\u5b58":24,"\u53ea\u9700\u4e2d\u65ad":41,"\u53ea\u9700\u4f7f\u7528":41,"\u53ea\u9700\u5220\u9664\u6700\u540e\u4e00\u884c\u4e2d\u7684\u6ce8\u91ca\u5e76\u628a":61,"\u53ea\u9700\u5728linux\u4e0b\u8fd0\u884c\u5982\u4e0b\u547d\u4ee4":62,"\u53ea\u9700\u7528\u4f60\u5b9a\u4e49\u7684\u76ee\u5f55\u4fee\u6539":41,"\u53ea\u9700\u77e5\u9053\u8fd9\u662f\u4e00\u4e2a\u6807\u8bb0\u5c5e\u6027\u7684\u65b9\u6cd5\u5c31\u53ef\u4ee5\u4e86":3,"\u53ea\u9700\u8981":35,"\u53ea\u9700\u8981\u4e00\u884c\u4ee3\u7801\u5c31\u53ef\u4ee5\u8c03\u7528\u8fd9\u4e2apydataprovider2":3,"\u53ea\u9700\u8981\u5728\u51fd\u6570\u4e2d\u8c03\u7528\u591a\u6b21yield\u5373\u53ef":3,"\u53ea\u9700\u8981\u6062\u590d\u8fd9\u53f0\u8282\u70b9":18,"\u53ea\u9700\u8981\u7b80\u5355\u5730\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4":62,"\u53ea\u9700\u8981\u7b80\u5355\u7684\u8fd0\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5373\u53ef":59,"\u53ea\u9700\u8981\u8fd0\u884c":59,"\u53ef\u4ee5":[32,36,41],"\u53ef\u4ee5\u4ee5\u540e\u53f0\u8fdb\u7a0b\u65b9\u5f0f\u8fd0\u884c\u5bb9\u5668":27,"\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":3,"\u53ef\u4ee5\u4f30\u8ba1\u51fa\u5982\u679c\u6a21\u578b\u91c7\u7528\u4e0d\u53d8\u7684\u8f93\u51fa\u6700\u5c0f\u7684cost0\u662f\u591a\u5c11":24,"\u53ef\u4ee5\u4f7f\u7528":[24,46],"\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4":29,"\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u4ee3\u7801":24,"\u53ef\u4ee5\u4f7f\u7528\u8be5\u53c2\u6570":43,"\u53ef\u4ee5\u4f7f\u7528kubernetes\u7684\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":49,"\u53ef\u4ee5\u4f7f\u7528python\u7684":5,"\u53ef\u4ee5\u51cf\u5c0f\u7cfb\u7edf\u590d\u6742\u6027":18,"\u53ef\u4ee5\u51cf\u5c11\u7f13\u5b58\u6c60\u7684\u5927\u5c0f":24,"\u53ef\u4ee5\u521b\u5efa\u4e00\u4e2a":48,"\u53ef\u4ee5\u53c2\u7167\u4e0b\u9762\u7684\u547d\u4ee4\u8fdb\u884c\u5b89\u88c5":54,"\u53ef\u4ee5\u53c2\u8003":[32,35,46,47,49,62],"\u53ef\u4ee5\u53c2\u8003\u4fdd\u5b58\u5728":53,"\u53ef\u4ee5\u542f\u52a8":46,"\u53ef\u4ee5\u542f\u52a8\u4e00\u4e2atrainer\u8fdb\u7a0b":46,"\u53ef\u4ee5\u542f\u52a8\u5206\u5e03\u5f0f\u4f5c\u4e1a":46,"\u53ef\u4ee5\u544a\u8bc9\u60a8\u67d0\u4e2a\u64cd\u4f5c\u5230\u5e95\u82b1\u4e86\u591a\u957f\u65f6\u95f4":40,"\u53ef\u4ee5\u5728":27,"\u53ef\u4ee5\u5728\u4efb\u4f55\u673a\u5668\u4e0a\u6267\u884c\u7684":20,"\u53ef\u4ee5\u5728\u5171\u4eab\u5b58\u50a8\u4e0a\u67e5\u770b\u8f93\u51fa\u7684\u65e5\u5fd7\u548c\u6a21\u578b":49,"\u53ef\u4ee5\u5728\u5f88\u5927\u7a0b\u5ea6\u4e0a\u6d88\u9664\u6b67\u4e49":60,"\u53ef\u4ee5\u5728\u7f51\u7ad9\u4e0a\u627e\u5230":60,"\u53ef\u4ee5\u5728\u8fd9\u4e2a":36,"\u53ef\u4ee5\u5728kubernetes\u4e2d\u6309\u7167":47,"\u53ef\u4ee5\u5c06\u67d0\u4e00\u4e2a\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovider2":3,"\u53ef\u4ee5\u5c06\u78c1\u76d8\u4e0a\u67d0\u4e2a\u76ee\u5f55\u5171\u4eab\u7ed9\u7f51\u7edc\u4e2d\u5176\u4ed6\u673a\u5668\u8bbf\u95ee":47,"\u53ef\u4ee5\u5c06memory\u7406\u89e3\u4e3a\u4e00\u4e2a\u65f6\u5ef6\u64cd\u4f5c":34,"\u53ef\u4ee5\u5e2e\u60a8\u63d0\u4f9b\u4e00\u4e9b\u5b9a\u4f4d\u6027\u80fd\u74f6\u9888\u7684\u5efa\u8bae":40,"\u53ef\u4ee5\u5feb\u901f\u8fed\u4ee3":27,"\u53ef\u4ee5\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u7f16\u8bd1\u751f\u6210\u6587\u6863":38,"\u53ef\u4ee5\u628a\u672c\u5730\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5b58\u50a8\u96c6\u7fa4\u4e2d":19,"\u53ef\u4ee5\u6307\u5b9a\u54ea\u4e00\u4e2a\u8f93\u5165\u548c\u8f93\u51fa\u5e8f\u5217\u4fe1\u606f\u4e00\u81f4":32,"\u53ef\u4ee5\u6309\u5982\u4e0b\u7684\u7ed3\u6784\u6765\u51c6\u5907\u6570\u6910":61,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":[31,34],"\u53ef\u4ee5\u662f\u4e00\u4e2a\u975e\u5e8f\u5217":34,"\u53ef\u4ee5\u662f\u4ee5\u4e0b\u51e0\u79cd":37,"\u53ef\u4ee5\u663e\u793a\u5730\u6307\u5b9a\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7528\u4e8e\u521d\u59cb\u5316memori":34,"\u53ef\u4ee5\u6709\u4ee5\u4e0b\u4e24\u79cd":34,"\u53ef\u4ee5\u6709\u53ef\u5b66\u4e60\u7684\u53c2\u6570":46,"\u53ef\u4ee5\u6709\u6548\u51cf\u5c0f\u7f51\u7edc\u7684\u963b\u585e":43,"\u53ef\u4ee5\u6709\u6548\u7684\u907f\u514dparamet":18,"\u53ef\u4ee5\u67e5\u770b":49,"\u53ef\u4ee5\u67e5\u770b\u6b64pod\u8fd0\u884c\u7684\u5bbf\u4e3b\u673a":48,"\u53ef\u4ee5\u6d4b\u8bd5\u591a\u4e2a\u6a21\u578b":45,"\u53ef\u4ee5\u7528\u4e8e\u4ece\u5b98\u65b9\u7f51\u7ad9\u4e0a\u4e0b\u8f7dcifar":54,"\u53ef\u4ee5\u7528\u4e8e\u5c0f\u91cf\u6570\u636e\u7684\u9a8c\u8bc1":47,"\u53ef\u4ee5\u7528\u4e8e\u63a5\u6536\u548cpydataprovider2\u4e00\u6837\u7684\u8f93\u5165\u6570\u636e\u5e76\u8f6c\u6362\u6210\u9884\u6d4b\u63a5\u53e3\u6240\u9700\u7684\u6570\u636e\u7c7b\u578b":5,"\u53ef\u4ee5\u7528\u4ee5\u4e0b\u6307\u4ee4":19,"\u53ef\u4ee5\u7528\u5982\u4e0b\u547d\u4ee4":36,"\u53ef\u4ee5\u7528\u6765\u8ba1\u7b97cpu\u51fd\u6570\u6216cuda\u5185\u6838\u7684\u65f6\u95f4\u6d88\u8017":40,"\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u975e\u5e8f\u5217\u8f93\u5165":31,"\u53ef\u4ee5\u7cbe\u786e\u8bf4\u660e\u4e00\u4e2a\u957f\u8017\u65f6\u64cd\u4f5c\u7684\u5177\u4f53\u539f\u56e0":40,"\u53ef\u4ee5\u7ee7\u7eed\u5728\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f\u63d0\u4ea4\u4ee3\u7801":23,"\u53ef\u4ee5\u7f16\u5199":27,"\u53ef\u4ee5\u8003\u8651\u4f7f\u7528\u4e00\u4e9b\u4f18\u5316\u7b97\u6cd5":24,"\u53ef\u4ee5\u8bbe\u7f6e":54,"\u53ef\u4ee5\u8fd0\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u751f\u6210":59,"\u53ef\u4ee5\u8fd0\u884c\u811a\u672ctrain":54,"\u53ef\u4ee5\u9009\u62e9\u662f\u5426\u4f7f\u7528\u53c2\u6570":45,"\u53ef\u4ee5\u901a\u8fc7":47,"\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539\u8fd9\u4e24\u4e2a\u51fd\u6570\u6765\u5b9e\u73b0\u590d\u6742\u7684\u7f51\u7edc\u914d\u7f6e":35,"\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528":5,"\u53ef\u4ee5\u901a\u8fc7\u9636\u6bb5\u6027\u7684\u4fdd\u5b58\u6bcf\u4e2aparamet":18,"\u53ef\u4ee5\u901a\u8fc7show_parameter_stats_period\u8bbe\u7f6e\u6253\u5370\u53c2\u6570\u4fe1\u606f\u7b49":57,"\u53ef\u4ee5\u901a\u8fc7volume\u6216persistentvolume\u6302\u8f7d\u5b58\u50a8\u7a7a\u95f4\u5230\u6bcf\u4e2a\u5bb9\u5668\u4e2d":19,"\u53ef\u7528\u4e8e\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u89e3\u6790\u8fd9\u4e9b\u53c2\u6570":45,"\u53ef\u7528\u5728\u6d4b\u8bd5\u6216\u8bad\u7ec3\u65f6\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b":57,"\u53ef\u80fd\u4f1a\u5bfc\u81f4\u51fa\u9519":49,"\u53ef\u80fd\u4f1a\u9020\u6210\u7f51\u7edc\u62e5\u585e":18,"\u53ef\u80fd\u7684\u4ee3\u7801\u4e3a":24,"\u53ef\u80fd\u7684\u539f\u56e0\u662f":24,"\u53ef\u80fd\u7684\u53c2\u6570\u662f":46,"\u53ef\u80fd\u7684\u60c5\u51b5\u4e0b":40,"\u53ef\u9009":[3,37],"\u53f3\u8fb9\u662f":55,"\u5403":32,"\u5403\u996d":32,"\u5404\u65b9\u9762":32,"\u5404\u9879\u53c2\u6570\u7684\u8be6\u7ec6\u8bf4\u660e\u53ef\u4ee5\u5728\u547d\u4ee4\u884c\u53c2\u6570\u76f8\u5173\u6587\u6863\u4e2d\u627e\u5230":54,"\u5408":32,"\u5408\u5e76":62,"\u5408\u5e76\u6bcf\u4e2a":62,"\u5408\u7406":32,"\u540c\u65f6":[24,40],"\u540c\u65f6\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":26,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u52a0\u901f\u5f00\u59cb\u8bad\u7ec3\u524d\u6570\u636e\u8f7d\u5165\u7684\u8fc7\u7a0b":24,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u901a\u8fc7":36,"\u540c\u65f6\u4e5f\u80fd\u591f\u5f15\u5165\u66f4\u52a0\u590d\u6742\u7684\u8bb0\u5fc6\u673a\u5236":34,"\u540c\u65f6\u4f1a\u8ba1\u7b97\u5206\u7c7b\u51c6\u786e\u7387":57,"\u540c\u65f6\u4f60\u53ef\u4ee5\u4f7f\u7528":55,"\u540c\u65f6\u4f7f\u7528\u4e86l2\u6b63\u5219":57,"\u540c\u65f6\u5176\u5185\u90e8\u5b9e\u73b0\u53ef\u4ee5\u907f\u514d\u7eafcpu\u7248\u672cpaddlepaddle\u5728\u6267\u884c\u672c\u8bed\u53e5\u65f6\u53d1\u751f\u5d29\u6e83":40,"\u540c\u65f6\u518d\u5c06":23,"\u540c\u65f6\u53ef\u4ee5\u4f7f\u7528\u6237\u53ea\u5173\u6ce8\u5982\u4f55\u4ece\u6587\u4ef6\u4e2d\u8bfb\u53d6\u6bcf\u4e00\u6761\u6570\u636e":3,"\u540c\u65f6\u5728\u5185\u5b58\u91cc\u76f4\u63a5\u968f\u5373\u9009\u53d6\u6570\u636e\u6765\u505ashuffl":24,"\u540c\u65f6\u5c06\u53c2\u6570\u521d\u59cb\u5316\u4e3a":24,"\u540c\u65f6\u6211\u4eec\u5e0c\u671b\u5e7f\u5927\u5f00\u53d1\u8005\u79ef\u6781\u63d0\u4f9b\u53cd\u9988\u548c\u8d21\u732e\u6e90\u4ee3\u7801":0,"\u540c\u65f6\u63d0\u8d77":23,"\u540c\u65f6\u6b22\u8fce\u8d21\u732e\u66f4\u591a\u7684\u5b89\u88c5\u5305":28,"\u540c\u65f6\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":45,"\u540c\u65f6\u8bbe\u7f6e\u5185\u5b58\u7f13\u5b58\u529f\u80fd":24,"\u540c\u65f6\u8bbe\u7f6e\u5b83\u7684input_types\u5c5e\u6027":3,"\u540c\u65f6\u9884\u6d4b\u7f51\u7edc\u901a\u5e38\u76f4\u63a5\u8f93\u51fa\u6700\u540e\u4e00\u5c42\u7684\u7ed3\u679c\u800c\u4e0d\u662f\u50cf\u8bad\u7ec3\u7f51\u7edc\u4e00\u6837\u518d\u63a5\u4e00\u5c42cost":5,"\u540c\u6837\u4e5f\u53ef\u4ee5\u5728\u6d4b\u8bd5\u6a21\u5f0f\u4e2d\u6307\u5b9a\u6a21\u578b\u8def\u5f84":43,"\u540c\u6837\u529f\u80fd\u7684":46,"\u540c\u6837\u53ef\u4ee5\u6269\u5c55\u5230\u53cc\u5c42\u5e8f\u5217\u7684\u5904\u7406\u4e0a":34,"\u540c\u6b65\u6267\u884c\u64cd\u4f5c\u7684\u7ebf\u7a0b\u6570":43,"\u540d\u5b57\u4fee\u9970":20,"\u540d\u79f0":57,"\u540e":[24,26,36,49,61],"\u540e\u5411\u4f20\u64ad":37,"\u540e\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u51fa\u7684\u68af\u5ea6":37,"\u540e\u9762\u8fde\u5168\u8fde\u63a5\u5c42\u548csoftmax\u5c42":61,"\u5411\u6307\u5b9a\u7684\u76ee\u5f55\u4e2d\u4e00\u4e2a\u65b0\u7684\u6587\u4ef6":18,"\u5411\u91cfenable_parallel_vector":42,"\u5411paddle\u7684\u4e3b\u7248\u672c\u5e93\u63d0\u4ea4":23,"\u5426":26,"\u5426\u5219":[2,41,59],"\u5426\u5219\u4f1a\u628a":36,"\u5426\u5219\u4f60\u9700\u8981\u81ea\u5df1\u4e0b\u8f7d":62,"\u5426\u5219\u4f7f\u7528\u591a\u673a\u8bad\u7ec3":43,"\u5426\u5219\u4f7f\u7528cpu\u6a21\u5f0f":43,"\u5426\u5219\u4f7f\u7528gpu":45,"\u5426\u5219\u5b83\u4ee5\u4e00\u4e2a\u5e8f\u5217\u8f93\u5165":35,"\u5426\u5219\u5f97\u628apaddle\u9759\u6001\u5e93\u94fe\u63a5\u5230\u89e3\u91ca\u5668\u91cc":20,"\u5426\u5219\u9891\u7e41\u7684\u591a\u8282\u70b9\u5de5\u4f5c\u7a7a\u95f4\u90e8\u7f72\u53ef\u80fd\u4f1a\u5f88\u9ebb\u70e6":41,"\u5426\u5b9a":60,"\u542b\u4e49":[55,61],"\u542b\u53ef\u5b66\u4e60\u53c2\u6570":46,"\u542b\u6709":58,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u548c\u5b50\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":37,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u6574\u6570":37,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":37,"\u542f\u52a8\u4e00\u4e2a\u65b0\u7684\u7ebf\u7a0b\u5f00\u59cb\u4fdd\u5b58\u68c0\u67e5\u70b9":18,"\u542f\u52a8\u4e00\u4e2apserver\u8fdb\u7a0b":46,"\u542f\u52a8\u4e4b\u540e":46,"\u542f\u52a8\u5bb9\u5668\u5f00\u59cb\u8bad\u7ec3":49,"\u542f\u52a8\u5e76\u884c\u5411\u91cf\u7684\u9608\u503c":43,"\u542f\u52a8\u5feb\u901f\u5e94\u7b54":43,"\u542f\u7528\u68af\u5ea6\u53c2\u6570\u7684\u9608\u503c":43,"\u5440":32,"\u544a\u8bc9paddle\u54ea\u4e2a\u6587\u4ef6\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u914d\u7f6e\u6587\u4ef6":59,"\u544a\u8bc9paddle\u5c06\u6a21\u578b\u4fdd\u5b58\u5728":59,"\u5468\u56f4":32,"\u547d\u4ee4":41,"\u547d\u4ee4\u4e3a":48,"\u547d\u4ee4\u521b\u5efa\u65b0\u955c\u50cf":48,"\u547d\u4ee4\u53ef\u4ee5\u8bbe\u7f6e":26,"\u547d\u4ee4\u6307\u5b9a\u7684\u53c2\u6570\u4f1a\u4f20\u5165\u7f51\u7edc\u914d\u7f6e\u4e2d":57,"\u547d\u4ee4\u884c\u53c2\u6570\u6587\u6863":57,"\u547d\u4ee4\u8bbe\u7f6e\u8be5\u7c7b\u7f16\u8bd1\u9009\u9879":26,"\u547d\u4ee4\u9009\u9879\u5e76\u4e14":41,"\u547d\u540d\u4e3a":36,"\u547d\u540d\u7a7a\u95f4":47,"\u547d\u540d\u7a7a\u95f4\u4e3b\u8981\u4e3a\u4e86\u5bf9\u8c61\u8fdb\u884c\u903b\u8f91\u4e0a\u7684\u5206\u7ec4\u4fbf\u4e8e\u7ba1\u7406":47,"\u548c":[19,20,21,23,24,25,26,32,35,36,37,38,40,41,45,46,47,53,54,57,59,62],"\u548c\u4e00\u4e2a\u5df2\u7ecf\u5206\u8bcd\u540e\u7684\u53e5\u5b50":32,"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":3,"\u548c\u4e2d\u6587\u6587\u6863":38,"\u548c\u4e4b\u524d\u51cf\u5c0f\u901a\u8fc7\u51cf\u5c0f\u7f13\u5b58\u6c60\u6765\u51cf\u5c0f\u5185\u5b58\u5360\u7528\u7684\u539f\u7406\u4e00\u81f4":24,"\u548c\u504f\u7f6e\u5411\u91cf":37,"\u548c\u533a\u57df\u6807\u8bb0":60,"\u548c\u53cc\u5c42\u5e8f\u5217\u542b\u6709subseq":31,"\u548c\u5728":3,"\u548c\u5bf9\u8c61\u5b58\u50a8api":47,"\u548c\u5dee\u8bc4":57,"\u548c\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":31,"\u548c\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165":35,"\u548c\u68af\u5ea6\u622a\u65ad":57,"\u548c\u6a21\u578b\u8def\u5f84":61,"\u548c\u6c60\u5316":46,"\u548c\u771f\u5b9e":25,"\u548c\u793a\u4f8b2\u4e2d\u7684\u914d\u7f6e\u7c7b\u4f3c":32,"\u548c\u79bb\u7ebf\u6570\u636e\u7684\u65b9\u5f0f":19,"\u548c\u7b2c6\u884c\u7684":62,"\u548c\u90e8\u5206layer":34,"\u548cadam\u5b66\u4e60\u65b9\u6cd5":62,"\u548cargument":60,"\u548ccudnn":29,"\u548cpool":31,"\u548cpython\u63a5\u53e3\u6765\u63d0\u53d6\u7279\u5f81":55,"\u54c1\u8d28":32,"\u54ea\u4e2atrainer\u5148\u5b8c\u6210block\u7684\u8bad\u7ec3":18,"\u54ea\u4e9b\u4e0d\u662f":32,"\u552f\u4e00\u9700\u8981\u505a\u7684\u662f\u5c06\u76f8\u5e94\u7c7b\u578b\u8bbe\u7f6e\u4e3a\u8f93\u5165":35,"\u5546\u52a1":32,"\u554a":32,"\u559c\u5267\u7247":58,"\u5668":57,"\u56db\u79cd\u6570\u636e\u7c7b\u578b":3,"\u56de\u5f52\u8bef\u5dee\u4ee3\u4ef7\u5c42":25,"\u56e0\u4e3a\u5168\u8fde\u63a5\u5c42\u7684\u6fc0\u6d3b\u53ef\u4ee5\u662fsoftmax":37,"\u56e0\u4e3a\u5176\u4e3a\u8d1f\u8d23\u63d0\u4f9bgradient":46,"\u56e0\u4e3a\u5355\u4e2a\u8c13\u8bcd\u4e0d\u80fd\u7cbe\u786e\u5730\u63cf\u8ff0\u8c13\u8bcd\u4fe1\u606f":60,"\u56e0\u4e3a\u53c2\u6570":45,"\u56e0\u4e3a\u589e\u52a0\u8fd9\u4e2a\u503c":46,"\u56e0\u4e3a\u5b83\u4eec\u7684\u8ba1\u7b97\u6548\u7387\u6bd4":35,"\u56e0\u4e3a\u5b83\u6bd4":35,"\u56e0\u4e3a\u5b98\u65b9\u955c\u50cf":49,"\u56e0\u4e3a\u5bb9\u5668\u5185\u7684\u6587\u4ef6\u90fd\u662f\u6682\u65f6\u5b58\u5728\u7684":47,"\u56e0\u4e3a\u8be5\u6587\u4ef6\u53ef\u9002\u7528\u4e8e\u9884\u6d4b":54,"\u56e0\u4e3a\u8fd9\u6837\u505a\u4e5f\u6ca1\u6cd5\u4fdd\u8bc1\u6d88\u9664\u968f\u673a\u6027":18,"\u56e0\u4e3adocker\u80fd\u5728\u6240\u6709\u4e3b\u8981\u64cd\u4f5c\u7cfb\u7edf":27,"\u56e0\u4e3apaddlepaddle\u7684v2":38,"\u56e0\u4e3apython\u7684\u641c\u7d22\u8def\u5f84\u662f\u4f18\u5148\u5df2\u7ecf\u5b89\u88c5\u7684python\u5305":24,"\u56e0\u4e3aswig\u5728\u7b2c\u4e09\u65b9\u8bed\u8a00\u4e2d\u66b4\u9732\u7684\u51fd\u6570\u540d":20,"\u56e0\u6b64":[2,3,32,34,37,46],"\u56e0\u6b64\u4f7f\u7528":3,"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":32,"\u56e0\u6b64\u53ef\u4ee5\u4f7f\u7528\u8be5\u9009\u9879":53,"\u56e0\u6b64\u53ef\u80fd\u4f1a\u6709\u4e00\u4e9b\u9519\u8bef\u548c\u4e0d\u4e00\u81f4\u53d1\u751f":58,"\u56e0\u6b64\u5982\u679c\u8fd9\u4e2a\u811a\u672c\u8fd0\u884c\u5931\u8d25":54,"\u56e0\u6b64\u5b83\u662finteger_value_sub_sequ":32,"\u56e0\u6b64\u6211\u4eec\u91c7\u7528\u8f93\u51fa\u7684\u52a0\u6743\u548c":37,"\u56e0\u6b64\u6709\u4e24\u79cd\u89e3\u51b3\u65b9\u6848":3,"\u56e0\u6b64\u7528\u6237\u5e76\u4e0d\u9700\u8981\u5173\u5fc3\u5b83\u4eec":42,"\u56e0\u6b64\u8be5\u5c42\u4e2d\u6ca1\u6709\u504f\u7f6e":55,"\u56e0\u6b64\u9519\u8bef\u7684\u4f7f\u7528\u4e8c\u8fdb\u5236\u53d1\u884c\u7248\u53ef\u80fd\u4f1a\u5bfc\u81f4\u8fd9\u79cd\u9519\u8bef":24,"\u56e0\u6b64init_hook\u5c3d\u91cf\u4f7f\u7528":3,"\u56e2\u8d2d\u7f51\u7ad9":61,"\u56fe":[55,61],"\u56fe2\u662f\u53cc\u5411lstm\u7f51\u7edc":61,"\u56fe3\u662f\u4e09\u5c42lstm\u7ed3\u6784":61,"\u56fe\u4e2d\u6bcf\u4e2a\u7070\u8272\u65b9\u5757\u662f\u4e00\u53f0\u673a\u5668":46,"\u56fe\u50cf\u5206\u7c7b":[23,56],"\u56fe\u50cf\u5927\u5c0f\u4e3a3":55,"\u56fe\u50cf\u63cf\u8ff0":62,"\u56fe\u7247\u5206\u4e3a10\u7c7b":54,"\u56fe\u7684\u5e95\u90e8\u662fword":61,"\u56fe\u8868":[27,61],"\u5728":[3,21,23,29,31,32,35,36,41,46,55,57,58,60],"\u5728\u4e00\u4e2a\u4e0d\u53ef\u4e2d\u65ad\u5e76\u7f3a\u5c11\u5907\u4efd\u7684\u8bad\u7ec3\u4efb\u52a1\u4e2d":18,"\u5728\u4e00\u4e2a\u529f\u80fd\u9f50\u5168\u7684kubernetes\u673a\u7fa4\u91cc":48,"\u5728\u4e00\u4e2a\u53c2\u6570\u7684\u68af\u5ea6\u88ab\u66f4\u65b0\u540e":37,"\u5728\u4e00\u4e2a\u5468\u671f\u5185\u6d4b\u8bd5\u6240\u6709\u6570\u636e":60,"\u5728\u4e00\u8f6e\u4e2d\u6bcfsave":43,"\u5728\u4e0a\u56fe\u4e2d\u663e\u793a\u4e86\u5728\u4e00\u4e2a\u5b9e\u9645\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u5e94\u7528":19,"\u5728\u4e0a\u9762\u4ee3\u7801\u4e2d":32,"\u5728\u4e0b\u4e00\u7bc7\u4e2d":48,"\u5728\u4e0b\u9762\u4f8b\u5b50\u91cc":57,"\u5728\u4e0b\u9762\u7684\u4f8b\u5b50\u4e2d":54,"\u5728\u4e0d\u540c\u64cd\u4f5c\u7cfb\u7edf":47,"\u5728\u4e0d\u540c\u7684\u5e94\u7528\u91cc":46,"\u5728\u4e4b\u540e\u7684":24,"\u5728\u4efb\u610f\u957f\u5ea6\u8bed\u53e5\u7ffb\u8bd1\u7684\u573a\u666f\u4e0b\u90fd\u53ef\u4ee5\u89c2\u5bdf\u5230\u5176\u6548\u679c\u7684\u63d0\u5347":62,"\u5728\u4f7f\u7528\u5b83\u4e4b\u524d\u8bf7\u5b89\u88c5paddlepaddle\u7684python":61,"\u5728\u5168\u8fde\u63a5\u5c42\u4e2d":37,"\u5728\u51c6\u5907\u53d1\u8d77":36,"\u5728\u51fa\u73b0\u5355\u70b9\u6545\u969c\u65f6":18,"\u5728\u51fd\u6570":49,"\u5728\u5206\u5e03\u5f0f\u73af\u5883\u4e2d\u6d4b\u8bd5":43,"\u5728\u5206\u5e03\u5f0f\u8bad\u7ec3\u4e2d":43,"\u5728\u5355\u5c42\u6570\u636e\u7684\u57fa\u7840\u4e0a":32,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u52a0\u8f7d\u548c\u4fdd\u5b58\u53c2\u6570":43,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u7ec8\u7aef\u6bcflog":43,"\u5728\u53cc\u5c42rnn\u4e2d\u7684\u7ecf\u5178\u60c5\u51b5\u662f\u5c06\u5185\u5c42\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u6570\u636e":32,"\u5728\u53cd\u5411\u4f20\u9012\u7684\u65f6\u5019":24,"\u5728\u53d8\u6362\u65f6\u9700\u8981\u5c06\u8f93\u5165\u5e8f\u5217\u4f20\u5165":32,"\u5728\u5404\u4e2a\u673a\u5668\u4e0a\u8fd0\u884c\u5982\u4e0b\u547d\u4ee4":46,"\u5728\u540c\u4e00\u4e2a\u547d\u540d\u7a7a\u95f4\u4e2d":47,"\u5728\u542f\u52a8job\u4e4b\u524d":49,"\u5728\u56fd\u5185":27,"\u5728\u58f0\u660edataprovider\u7684\u65f6\u5019\u4f20\u5165dictionary\u4f5c\u4e3a\u53c2\u6570":3,"\u5728\u591acpu\u8bad\u7ec3\u65f6\u5171\u4eab\u8be5\u53c2\u6570":43,"\u5728\u5b9e\u73b0\u8fc7\u7a0b\u4e2d":21,"\u5728\u5bb9\u5668\u521b\u5efa\u540e":49,"\u5728\u5bf9\u5bb9\u5668\u7684\u63cf\u8ff0":49,"\u5728\u5c42\u4e2d\u6307\u5b9a":45,"\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d":34,"\u5728\u5f00\u59cb\u8bad\u7ec3\u4e4b\u524d":54,"\u5728\u5f02\u6784\u96c6\u7fa4\u4e2d":18,"\u5728\u5f15\u5165\u5176\u4ed6\u7c7b\u578b\u7684\u5934\u6587\u4ef6\u65f6":21,"\u5728\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":37,"\u5728\u5f97\u5230":49,"\u5728\u5feb\u7167\u5199\u5165\u5b8c\u6210\u540e":18,"\u5728\u60a8\u7684\u5b9e\u9645\u73af\u5883\u4e2d":18,"\u5728\u6211\u4eec\u7684\u4f8b\u5b50\u4e2d":35,"\u5728\u6211\u4eec\u7684\u6d4b\u8bd5\u4e2d":61,"\u5728\u63d0\u4ea4":36,"\u5728\u6570\u636e\u52a0\u8f7d\u548c\u7f51\u7edc\u914d\u7f6e\u5b8c\u6210\u4e4b\u540e":57,"\u5728\u6570\u636e\u96c6\u53ef\u4ee5\u88ab\u8bad\u7ec3\u4e4b\u524d":19,"\u5728\u6587\u4ef6":59,"\u5728\u6587\u4ef6\u7684\u5f00\u59cb":46,"\u5728\u6709\u65b0\u7684\u5355\u8bcd\u6765\u4e34\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u9aa4\u5185":61,"\u5728\u672c\u4f8b\u4e2d":[32,36,45],"\u5728\u672c\u4f8b\u4e2d\u6ca1\u6709\u4f7f\u7528":3,"\u5728\u672c\u6559\u7a0b\u4e2d":[35,54],"\u5728\u672c\u6587\u4e2d":41,"\u5728\u672c\u6587\u4e2d\u4f7f\u7528\u7684":41,"\u5728\u672c\u6f14\u793a\u4e2d":61,"\u5728\u672c\u793a\u4f8b\u4e2d":[32,61],"\u5728\u672c\u8282\u4e2d":35,"\u5728\u6811\u7684\u6bcf\u4e00\u5c42\u4e0a":43,"\u5728\u6837\u4f8b\u4e2d":21,"\u5728\u6a21\u578b\u6587\u4ef6\u7684":41,"\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u901a\u8fc7":57,"\u5728\u6b64":[0,42,45],"\u5728\u6b64\u4e3a\u65b9\u4fbf\u5bf9\u6bd4\u4e0d\u540c\u7f51\u7edc\u7ed3\u6784":57,"\u5728\u6b64\u611f\u8c22":53,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":35,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u76f8\u7b49":32,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u957f":35,"\u5728\u6bcf\u4e2a\u673a\u5668\u4e2d":46,"\u5728\u6bcf\u4e2apod\u4e0a\u90fd\u901a\u8fc7volume\u65b9\u5f0f\u6302\u8f7d\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u7684\u4e00\u4e2a\u76ee\u5f55\u7528\u4e8e\u4fdd\u5b58\u8bad\u7ec3\u6570\u636e\u548c\u8f93\u51fa\u7ed3\u679c":49,"\u5728\u6bcf\u8bad\u7ec3":59,"\u5728\u6d4b\u8bd5\u9636\u6bb5":43,"\u5728\u6d4b\u8bd5\u9636\u6bb5\u5b83\u4eec\u5c06\u4f1a\u88ab\u52a0\u8f7d\u5230\u6a21\u578b\u4e2d":55,"\u5728\u6e90\u7801\u76ee\u5f55\u6811\u7684\u6839\u76ee\u5f55\u4e2d\u8fd0\u884c":36,"\u5728\u6f14\u793a\u4e2d":60,"\u5728\u7269\u7406\u673a\u4e0a\u624b\u52a8\u90e8\u7f72":47,"\u5728\u751f\u6210\u65f6":35,"\u5728\u751f\u6210\u8fc7\u7a0b\u4e2d":62,"\u5728\u751f\u6210\u8fc7\u7a0b\u4e2d\u6211\u4eec\u4f7f\u7528sgd\u8bad\u7ec3\u7b97\u6cd5":62,"\u5728\u7528\u6237\u4f7f\u7528c":21,"\u5728\u7528\u6237\u6587\u4ef6user":59,"\u5728\u7535\u5f71\u6587\u4ef6movi":59,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u4f7f\u7528attention\u7248\u672c\u7684gru\u7f16\u89e3\u7801\u7f51\u7edc":62,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u4f7f\u7528sgd\u8bad\u7ec3\u7b97\u6cd5":62,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u5b9a\u4e49\u4e86\u4e00\u4e2a\u5e8f\u5217\u5230\u5e8f\u5217\u7684\u751f\u6210\u6570\u636e":62,"\u5728\u793a\u4f8b\u4e2d\u6211\u4eec\u5b9a\u4e49\u4e86\u4e00\u4e2a\u5e8f\u5217\u5230\u5e8f\u5217\u7684\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6570\u636e":62,"\u5728\u7a0b\u5e8f\u5f00\u59cb\u9636\u6bb5":5,"\u5728\u7b2c\u4e00\u884c\u4e2d\u6211\u4eec\u8f7d\u5165\u7528\u4e8e\u5b9a\u4e49\u7f51\u7edc\u7684\u51fd\u6570":54,"\u5728\u7ebf\u6a21\u578b\u9884\u6d4b\u670d\u52a1":19,"\u5728\u7f16\u8bd1\u5b89\u88c5\u6b63\u786e\u4e4b\u540e":38,"\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":37,"\u5728\u7f51\u7edc\u914d\u7f6e\u91cc":3,"\u5728\u7ffb\u8bd1\u6cd5\u8bed\u53e5\u5b50\u4e4b\u524d":62,"\u5728\u811a\u672c":59,"\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d":31,"\u5728\u8bad\u7ec3\u4e2d":35,"\u5728\u8bad\u7ec3\u4e4b\u524d":49,"\u5728\u8bad\u7ec3\u4e86":59,"\u5728\u8bad\u7ec3\u4e86\u51e0\u4e2a\u8f6e\u6b21\u4ee5\u540e":59,"\u5728\u8bad\u7ec3\u5b8c\u6210\u540e":54,"\u5728\u8bad\u7ec3\u6570\u96c6\u4e0a\u8bad\u7ec3\u751f\u6210\u8bcd\u5411\u91cf\u5b57\u5178":53,"\u5728\u8bad\u7ec3\u65f6":48,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":[49,62],"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6bcfshow":43,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u8fdb\u884c\u6d4b\u8bd5":2,"\u5728\u8bc4\u5ba1\u8fc7\u7a0b\u4e2d":23,"\u5728\u8be5\u914d\u7f6e\u76847":32,"\u5728\u8bed\u8a00\u751f\u6210\u9886\u57df\u4e2d":62,"\u5728\u8d2d\u7269\u7f51\u7ad9\u4e0a":57,"\u5728\u8f6f\u4ef6\u5de5\u7a0b\u7684\u8303\u7574\u91cc":40,"\u5728\u8f93\u51fa\u7684\u8fc7\u7a0b\u4e2d":34,"\u5728\u8fd0\u884c":61,"\u5728\u8fd9\u4e2a":23,"\u5728\u8fd9\u4e2a\u4efb\u52a1\u4e2d":62,"\u5728\u8fd9\u4e2a\u4f8b\u5b50\u4e2d":[25,61],"\u5728\u8fd9\u4e2a\u4f8b\u5b50\u91cc":[37,48],"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":32,"\u5728\u8fd9\u4e2a\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":20,"\u5728\u8fd9\u4e2a\u6559\u7a0b\u4e2d":40,"\u5728\u8fd9\u4e2a\u6a21\u578b\u4e2d":35,"\u5728\u8fd9\u4e2a\u793a\u4f8b\u4e2d":60,"\u5728\u8fd9\u4e2a\u9636\u6bb5\u7684\u4ee3\u7801\u6b63\u5728\u7ecf\u5386\u56de\u5f52\u6d4b\u8bd5":23,"\u5728\u8fd9\u4e9b\u5934\u6587\u4ef6\u4e2d":21,"\u5728\u8fd9\u4e9b\u6587\u4ef6\u4e2d":21,"\u5728\u8fd9\u4e9b\u7f51\u7edc\u4e2d":59,"\u5728\u8fd9\u4e9blayer\u4e2d":32,"\u5728\u8fd9\u6b65\u4efb\u52a1\u4e2d":61,"\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b":[35,37],"\u5728\u8fd9\u79cd\u7ed3\u6784\u4e2d":34,"\u5728\u8fd9\u7bc7\u6587\u6863\u91cc":48,"\u5728\u8fd9\u7bc7\u6587\u7ae0\u91cc":49,"\u5728\u8fd9\u91cc":34,"\u5728\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u5168\u8fde\u63a5\u5c42\u4f5c\u4e3a\u4f8b\u5b50\u6765\u5c55\u793a\u5b9e\u73b0\u65b0\u7f51\u7edc\u5c42\u6240\u9700\u8981\u7684\u56db\u4e2a\u6b65\u9aa4":37,"\u5728\u914d\u7f6e\u4e2d\u9700\u8981\u8bfb\u53d6\u5916\u90e8\u5b57\u5178":3,"\u5728\u914d\u7f6e\u6587\u4ef6\u4e2d\u7684":55,"\u5728\u91c7\u7528sgd":24,"\u5728\u96c6\u7fa4\u4e0a\u8bad\u7ec3\u4e00\u4e2a\u7a00\u758f\u6a21\u578b\u9700\u8981\u52a0\u4e0a\u4e0b\u9762\u7684\u53c2\u6570":45,"\u5728\u96c6\u7fa4\u8fd0\u884cmapreduc":19,"\u5728\u9884\u5904\u7406\u542b\u6709\u591a\u884c\u6570\u6910\u7684\u6587\u672c\u6587\u4ef6\u65f6\u53c2\u6570\u8bbe\u7f6e\u7a0d\u6709\u4e0d\u540c":61,"\u5728\u9884\u6d4b\u5e8f\u5217\u6216\u6bb5\u843d\u7684\u60c5\u611f\u4e2d\u8d77\u4e3b\u8981\u4f5c\u7528":61,"\u5728aws\u4e0a\u5feb\u901f\u90e8\u7f72\u96c6\u7fa4":47,"\u5728c":20,"\u5728c\u7684\u5934\u6587\u4ef6":20,"\u5728cephfs\u5b58\u50a8\u7cfb\u7edf\u4e2d\u7684\u516c\u5f00\u76ee\u5f55":19,"\u5728cub":54,"\u5728generator\u7684\u4e0a\u4e0b\u6587\u4e2d\u5c3d\u91cf\u7559\u4e0b\u975e\u5e38\u5c11\u7684\u53d8\u91cf\u5f15\u7528":3,"\u5728kubernetes\u4e0a\u8fd0\u884c\u7684\u4e0d\u540c\u7684\u8ba1\u7b97\u6846\u67b6":19,"\u5728kubernetes\u4e2d\u521b\u5efa\u7684\u6240\u6709\u8d44\u6e90\u5bf9\u8c61":47,"\u5728linux\u4e0b":62,"\u5728meta\u6587\u4ef6\u4e2d\u6709\u4e24\u79cd\u7279\u5f81":59,"\u5728movielen":59,"\u5728paddl":49,"\u5728paddle\u4e2d":45,"\u5728paddle\u4e4b\u4e0a\u8fd0\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u8f93\u51fa\u7684\u6a21\u578b\u4f1a\u63d0\u4f9b\u7ed9\u5728\u7ebf\u4eba\u8138\u8bc6\u522b\u7684\u5e94\u7528\u4f7f\u7528":19,"\u5728paddlepaddle\u4e2d":34,"\u5728paddlepaddle\u7684\u6587\u6863\u4e2d":32,"\u5728paddlepaddle\u91cc":25,"\u5728paramet":18,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49":34,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49memori":34,"\u5728trainer":45,"\u5730\u5740\u4e5f\u53ef\u4ee5\u4e3ahdfs\u6587\u4ef6\u8def\u5f84":2,"\u5730\u6bb5":32,"\u5730\u7406\u4f4d\u7f6e":32,"\u5730\u94c1\u7ad9":32,"\u5747\u4f1a\u88ab\u5b89\u88c5\u5230includ":21,"\u5747\u503c\u56fe\u50cf\u6587\u4ef6":55,"\u5747\u5300\u5206\u5e03":24,"\u5747\u5300\u5206\u5e03\u7684\u8303\u56f4\u662f":43,"\u5747\u662f\u5728":21,"\u5747\u6709\u4e09\u4e2a\u5b50\u5e8f\u5217":32,"\u5747\u6709\u4e24\u7ec4\u7279\u5f81":32,"\u57fa\u4e8e\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":34,"\u57fa\u4e8e\u5b57\u6bcd\u7684\u8bcd\u5d4c\u5165\u7279\u5f81":59,"\u57fa\u4e8epython\u7684\u6a21\u578b\u9884\u6d4b":5,"\u57fa\u4e8epython\u7684\u9884\u6d4b":[4,57],"\u57fa\u672c\u4e0a\u548cmnist\u6837\u4f8b\u4e00\u81f4":3,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":39,"\u57fa\u672c\u76f8\u540c":53,"\u586b\u5199":36,"\u589e\u52a0\u4e86\u4e00\u6761cd\u547d\u4ee4":48,"\u589e\u52a0\u5982\u4e0b\u53c2\u6570":45,"\u589e\u52a0\u68af\u5ea6\u68c0\u6d4b\u7684\u5355\u5143\u6d4b\u8bd5":37,"\u58f0\u660epython\u6570\u636e\u6e90":59,"\u5904\u7406\u5668\u6709\u4e24\u4e2a\u5173\u952e\u6027\u80fd\u9650\u5236":40,"\u5904\u7406\u6570\u636e\u7684python\u811a\u672c\u6587\u4ef6":57,"\u5904\u7406\u7684\u8f93\u5165\u5e8f\u5217\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u4e09\u79cd\u7c7b\u578b":34,"\u5904\u7406\u76f8\u4f3c\u5ea6\u56de\u5f52":59,"\u5904\u7406\u8fc7\u7a0b\u4e2d\u6570\u636e\u5b58\u50a8\u683c\u5f0f":54,"\u5904\u7406batch":46,"\u5907\u6ce8":40,"\u590d\u6742\u5ea6\u6216\u65f6\u95f4\u590d\u6742\u5ea6":40,"\u5916\u5c42memory\u662f\u4e00\u4e2a\u5143\u7d20":32,"\u5916\u5c42outer_step\u4e2d":32,"\u591a\u4e2a\u503c":19,"\u591a\u4e2ainput\u4ee5list\u65b9\u5f0f\u8f93\u5165":57,"\u591a\u4e2aparamet":18,"\u591a\u53e5\u8bdd\u8fdb\u4e00\u6b65\u6784\u6210\u4e86\u6bb5\u843d":34,"\u591a\u673a\u8bad\u7ec3":24,"\u591a\u673a\u8bad\u7ec3\u7684\u7ecf\u5178\u62d3\u6251\u7ed3\u6784\u5982\u4e0b":46,"\u591a\u7ebf\u7a0b\u7684\u6570\u636e\u8bfb\u53d6":3,"\u591a\u8f6e\u5bf9\u8bdd\u7b49\u66f4\u4e3a\u590d\u6742\u7684\u8bed\u8a00\u6570\u636e":34,"\u5927\u578b\u7535\u5f71\u8bc4\u8bba\u6570\u636e\u96c6":61,"\u5927\u591a\u6570\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u51fd\u6570":37,"\u5927\u591a\u6570\u5c42\u9700\u8981\u8bbe\u7f6e\u4e3a":37,"\u5927\u591a\u6570\u6210\u529f\u7684srl\u7cfb\u7edf\u662f\u5efa\u7acb\u5728\u67d0\u79cd\u5f62\u5f0f\u7684\u53e5\u6cd5\u5206\u6790\u7ed3\u679c\u4e4b\u4e0a\u7684":60,"\u5927\u591a\u6570\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u652f\u6301\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":37,"\u5927\u591a\u6570\u8bed\u8a00\u90fd\u652f\u6301\u4f7f\u7528c\u8bed\u8a00api":20,"\u5927\u5b66\u751f":58,"\u5927\u5bb6\u53ef\u4ee5\u901a\u8fc7\u5b83\u5236\u4f5c\u548c\u5206\u4eab\u5e26\u6709\u4ee3\u7801":27,"\u5927\u5c0f":41,"\u5929":32,"\u5929\u4e00\u5e7f\u573a":32,"\u5929\u4e00\u9601":32,"\u5929\u732b":61,"\u5934\u6587\u4ef6\u4e2d\u628a\u53c2\u6570\u5b9a\u4e49\u4e3a\u7c7b\u7684\u6210\u5458\u53d8\u91cf":37,"\u5934\u6587\u4ef6\u5982\u4e0b":37,"\u5947\u5e7b\u7247":58,"\u597d":32,"\u597d\u5403":32,"\u597d\u8bc4":57,"\u5982":[3,35,36,41,45,46],"\u59822":41,"\u5982\u4e0b":[3,59,61],"\u5982\u4e0b\u56fe\u6240\u793a":[32,40,54],"\u5982\u4e0b\u6240\u793a":[45,55,59],"\u5982\u4e0b\u662f\u4e00\u6bb5\u4f7f\u7528mnist":5,"\u5982\u4e0b\u8868\u683c":57,"\u5982\u4f55":59,"\u5982\u4f55\u5b58\u50a8\u7b49\u7b49":3,"\u5982\u4f55\u89e3\u6790\u8be5\u5730\u5740\u4e5f\u662f\u7528\u6237\u81ea\u5b9a\u4e49dataprovider\u65f6\u9700\u8981\u8003\u8651\u7684\u5730\u65b9":2,"\u5982\u4f55\u8d21\u732e":39,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":39,"\u5982\u4f55\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":57,"\u5982\u4fe1\u606f\u63d0\u53d6":60,"\u5982\u56fe2\u6240\u793a":61,"\u5982\u56fe\u4e2dtrainer":18,"\u5982\u5f62\u5bb9\u8bcd\u548c\u526f\u8bcd":61,"\u5982\u60f3\u4e86\u89e3\u66f4\u591a\u8be6\u7ec6\u7684\u89e3\u91ca":62,"\u5982\u672c\u4f8b\u4e2d":3,"\u5982\u672c\u4f8b\u7684":3,"\u5982\u679c\u4e00\u4e2a\u7f51\u7edc\u5c42\u9700\u8981\u914d\u7f6e\u7684\u8bdd":37,"\u5982\u679c\u4e0a\u9762\u4e24\u6b65\u51fa\u73b0\u9519\u8bef":18,"\u5982\u679c\u4e0b\u8f7d\u6210\u529f":55,"\u5982\u679c\u4e0d\u4e3a0":43,"\u5982\u679c\u4e0d\u4e86\u89e3":3,"\u5982\u679c\u4e0d\u5207\u8bcd":57,"\u5982\u679c\u4e0d\u6536\u655b":24,"\u5982\u679c\u4e0d\u662f\u5e8f\u5217":59,"\u5982\u679c\u4e3a":3,"\u5982\u679c\u4e3a0":43,"\u5982\u679c\u4e3afals":43,"\u5982\u679c\u4e3atrue":[3,43],"\u5982\u679c\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe\u7f6e":26,"\u5982\u679c\u4ed4\u7ec6\u8bbe\u7f6e\u7684\u8bdd":43,"\u5982\u679c\u4f20\u5165\u4e00\u4e2alist\u7684\u8bdd":46,"\u5982\u679c\u4f20\u5165\u5b57\u7b26\u4e32\u7684\u8bdd":46,"\u5982\u679c\u4f60\u4e0d\u9700\u8981\u8fd9\u4e2a\u64cd\u4f5c":61,"\u5982\u679c\u4f60\u53ea\u9700\u8981\u4f7f\u7528\u7b80\u5355\u7684rnn":35,"\u5982\u679c\u4f60\u5b89\u88c5gpu\u7248\u672c\u7684paddlepaddl":61,"\u5982\u679c\u4f60\u60f3\u4f7f\u7528\u8fd9\u4e9b\u7279\u6027":45,"\u5982\u679c\u4f60\u60f3\u8981\u4fdd\u5b58\u67d0\u4e9b\u5c42\u7684\u7279\u5f81\u56fe":43,"\u5982\u679c\u4f60\u60f3\u8fdb\u884c\u8bf8\u5982\u8bed\u4e49\u8f6c\u8ff0":62,"\u5982\u679c\u4f60\u6267\u884c\u5176\u5b83\u7684\u7528\u60c5\u611f\u5206\u6790\u6765\u5206\u7c7b\u6587\u672c\u7684\u4efb\u52a1":61,"\u5982\u679c\u4f60\u6b63\u5728\u5904\u7406\u5e8f\u5217\u6807\u8bb0\u4efb\u52a1":35,"\u5982\u679c\u4f60\u6ca1\u6709gpu\u73af\u5883":54,"\u5982\u679c\u4f60\u8981\u4e3a\u4e86\u6d4b\u8bd5\u800c\u589e\u52a0\u65b0\u7684\u6587\u4ef6":37,"\u5982\u679c\u4f7f\u7528":[41,53],"\u5982\u679c\u4f7f\u7528gpu\u7248\u672c\u7684paddlepaddl":29,"\u5982\u679c\u4f7f\u7528nvidia":27,"\u5982\u679c\u4f7f\u7528ssl\u8ba4\u8bc1":47,"\u5982\u679c\u4f7f\u7528swig\u6211\u4eec\u9700\u8981\u5c06\u5728interface\u6587\u4ef6\u91cc":20,"\u5982\u679c\u51fa\u73b0\u4ee5\u4e0bpython\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86\u7684\u60c5\u51b5":24,"\u5982\u679c\u53c2\u6570\u4fdd\u5b58\u4e0b\u6765\u7684\u6a21\u578b\u76ee\u5f55":24,"\u5982\u679c\u53c2\u6570\u6a21\u578b\u6587\u4ef6\u7f3a\u5931":53,"\u5982\u679c\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u672a\u8bbe\u7f6easync":43,"\u5982\u679c\u5728\u8bad\u7ec3\u671f\u95f4\u540c\u65f6\u53d1\u8d77\u53e6\u5916\u4e00\u4e2a\u8fdb\u7a0b\u8fdb\u884c\u6d4b\u8bd5":43,"\u5982\u679c\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6ebatch":43,"\u5982\u679c\u5728\u8bad\u7ec3nlp\u76f8\u5173\u6a21\u578b\u65f6":24,"\u5982\u679c\u5931\u8d25":23,"\u5982\u679c\u5b58\u5728\u67d0\u4e9btrainer\u6267\u884c\u901f\u5ea6\u8fc7\u6162\u4f1a\u5f71\u54cd\u6574\u4f53\u96c6\u7fa4\u7684\u901f\u5ea6":18,"\u5982\u679c\u5b83\u4f4d\u4e8e\u8c13\u8bcd\u4e0a\u4e0b\u6587\u533a\u57df\u4e2d":60,"\u5982\u679c\u5c06\u8fd9\u4e2a\u5185\u5b58\u6c60\u51cf\u5c0f":24,"\u5982\u679c\u5df2\u5b89\u88c5":60,"\u5982\u679c\u5df2\u7ecf\u6709pod\u8fd0\u884c":49,"\u5982\u679c\u5df2\u7ecf\u6b63\u5728\u6267\u884c\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":18,"\u5982\u679c\u5f00\u542f\u4f1a\u5bfc\u81f4\u8fd0\u884c\u7565\u6162":26,"\u5982\u679c\u60a8":27,"\u5982\u679c\u60a8\u60f3\u8981\u66f4\u6df1\u5165\u4e86\u89e3deep":27,"\u5982\u679c\u60a8\u6709\u597d\u7684\u5efa\u8bae\u6765":59,"\u5982\u679c\u60a8\u7684gpu\u7406\u8bba\u53ef\u4ee5\u8fbe\u52306":40,"\u5982\u679c\u60f3\u4e3a\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u8fd4\u56de\u591a\u6761\u6837\u672c":3,"\u5982\u679c\u60f3\u4f7f\u7528\u53ef\u89c6\u5316\u7684\u5206\u6790\u5668":40,"\u5982\u679c\u60f3\u5f88\u597d\u7684\u7406\u89e3\u7a0b\u5e8f\u7684\u884c\u4e3a":40,"\u5982\u679c\u60f3\u8981\u4e86\u89e3\u53cc\u5c42rnn\u5728\u5177\u4f53\u95ee\u9898\u4e2d\u7684\u4f7f\u7528":32,"\u5982\u679c\u60f3\u8981\u542f\u7528paddlepaddle\u7684\u5185\u7f6e\u5b9a\u65f6\u5668":40,"\u5982\u679c\u6211\u77e5\u9053\u5185\u6838\u82b1\u4e8610ms\u6765\u79fb\u52a81gb\u6570\u636e":40,"\u5982\u679c\u6267\u884c\u5931\u8d25":47,"\u5982\u679c\u6267\u884c\u6210\u529f":55,"\u5982\u679c\u63d0\u793a\u6b63\u786e":38,"\u5982\u679c\u63d0\u793a\u9519\u8bef":38,"\u5982\u679c\u6570\u636e\u6587\u4ef6\u5b58\u4e8e\u672c\u5730\u78c1\u76d8":2,"\u5982\u679c\u6570\u636e\u89c4\u6a21\u6bd4\u8f83\u5927":46,"\u5982\u679c\u6570\u6910\u83b7\u53d6\u6210\u529f":61,"\u5982\u679c\u662f\u4f7f\u7528\u975essl\u65b9\u5f0f\u8bbf\u95ee":47,"\u5982\u679c\u662f\u5176\u5b83\u7c7b\u578b":19,"\u5982\u679c\u662f\u5e8f\u5217":59,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165":34,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":34,"\u5982\u679c\u6709\u66f4\u590d\u6742\u7684\u4f7f\u7528":2,"\u5982\u679c\u6709\u9700\u8981\u4fee\u6539\u7684\u5730\u65b9":36,"\u5982\u679c\u6709bugfix\u7684\u884c\u4e3a":23,"\u5982\u679c\u672a\u8bbe\u7f6e":43,"\u5982\u679c\u672a\u8bbe\u7f6egpu":45,"\u5982\u679c\u67d0\u4e00\u4e2a\u7c7b\u578b\u9700\u8981\u5f15\u7528\u53e6\u4e00\u4e2a\u7c7b\u578b":21,"\u5982\u679c\u67d0\u4e00\u4e2apaddl":21,"\u5982\u679c\u67d0\u4e00\u4e2apaddle\u6982\u5ff5\u5fc5\u987b\u8981\u66b4\u9732":21,"\u5982\u679c\u67d0\u4e00\u5757\u6839\u672c\u5c31\u4e0d\u600e\u4e48\u8017\u65f6":40,"\u5982\u679c\u68c0\u67e5\u5230\u5206\u914d\u5728\u4e0d\u540c\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u53c2\u6570\u7684\u5206\u5e03\u4e0d\u5747\u5300\u6b21\u6570\u5927\u4e8echeck":43,"\u5982\u679c\u6ca1\u6709\u5b9a\u4e49memori":34,"\u5982\u679c\u6ca1\u6709\u8bbe\u7f6e":62,"\u5982\u679c\u6ca1\u6709\u8bbe\u7f6etest":2,"\u5982\u679c\u6d88\u606f\u6570\u636e\u592a\u5c0f":43,"\u5982\u679c\u6ee1\u8db3\u6761\u4ef6":18,"\u5982\u679c\u7528\u6237\u4e0d\u663e\u793a\u6307\u5b9a\u8fd4\u56de\u6570\u636e\u7684\u5bf9\u5e94\u5173\u7cfb":3,"\u5982\u679c\u7528\u6237\u60f3\u8981\u4e86\u89e3\u8be6\u7ec6\u7684\u6570\u636e\u96c6\u7684\u683c\u5f0f":53,"\u5982\u679c\u7528\u6237\u60f3\u8981\u81ea\u5b9a\u4e49\u521d\u59cb\u5316\u65b9\u5f0f":24,"\u5982\u679c\u7528\u6237\u8981\u628apaddle\u7684\u9759\u6001\u5e93":20,"\u5982\u679c\u771f\u60f3\u6316\u6398\u5185\u6838\u6df1\u5904\u7684\u67d0\u4e2a\u79d8\u5bc6":40,"\u5982\u679c\u7a0b\u5e8f\u5d29\u6e83\u4f60\u4e5f\u53ef\u4ee5\u624b\u52a8\u7ec8\u6b62":41,"\u5982\u679c\u7cfb\u7edf\u5b89\u88c5\u4e86\u591a\u4e2apython\u7248\u672c":24,"\u5982\u679c\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":37,"\u5982\u679c\u7f51\u7edc\u67b6\u6784\u7b80\u5355":35,"\u5982\u679c\u8981\u4f7f\u7528\u53cc\u5411lstm":61,"\u5982\u679c\u8981\u4f7f\u7528gpu":27,"\u5982\u679c\u8981\u542f\u7528gpu":41,"\u5982\u679c\u8981\u8fd0\u884c\u6240\u6709\u7684\u5355\u5143\u6d4b\u8bd5":36,"\u5982\u679c\u8bad\u7ec3\u4e00\u4e2apass":24,"\u5982\u679c\u8bad\u7ec3\u8fc7\u7a0b\u542f\u52a8\u6210\u529f\u7684\u8bdd":59,"\u5982\u679c\u8bad\u7ec3\u8fc7\u7a0b\u7684\u7684cost\u660e\u663e\u9ad8\u4e8e\u8fd9\u4e2a\u5e38\u6570\u8f93\u51fa\u7684cost":24,"\u5982\u679c\u8bbe\u7f6e":3,"\u5982\u679c\u8bbe\u7f6e\u8be5\u53c2\u6570":43,"\u5982\u679c\u8c03\u7528\u9759\u6001\u5e93\u53ea\u80fd\u5c06\u9759\u6001\u5e93\u4e0e\u89e3\u91ca\u5668\u94fe\u63a5":20,"\u5982\u679c\u8f93\u51fa\u662fno":27,"\u5982\u679c\u8fd0\u884c\u6210\u529f":[55,61],"\u5982\u679c\u96c6\u7fa4\u8282\u70b9\u6570\u91cf\u5c11":41,"\u5982\u679c\u9700\u8981\u5305\u542b\u66f4\u591a\u7684\u4f9d\u8d56":27,"\u5982\u679c\u9700\u8981\u6269\u5927\u77e9\u9635":37,"\u5982\u679c\u9700\u8981\u7f29\u51cf\u77e9\u9635":37,"\u5982\u679clearning_rate\u592a\u5927":24,"\u5982\u679clearning_rate\u592a\u5c0f":24,"\u5982\u679cpaddlepaddle\u5305\u5df2\u7ecf\u5728python\u7684sit":24,"\u5982\u679cparamet":18,"\u5982\u6bcf\u4e2a\u6587\u4ef6\u53ea\u6709\u4e00\u4e2a":36,"\u5982\u795e\u7ecf\u5143\u6fc0\u6d3b\u503c\u7b49":24,"\u5982\u8981build\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf":36,"\u5982\u9ad8\u4eae\u90e8\u5206":40,"\u5b50":32,"\u5b50\u53e5":34,"\u5b50\u53e5\u7684\u5355\u8bcd\u6570\u548c\u6307\u5b9a\u7684\u4e00\u4e2a\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":34,"\u5b57\u5178":62,"\u5b57\u5178\u4f1a\u5305\u542b\u8f93\u5165\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u5355\u8bcd":62,"\u5b57\u5178\u5171\u5305\u542b":53,"\u5b57\u5178\u6587\u4ef6":[60,61],"\u5b57\u5178\u91c7\u7528utf8\u7f16\u7801":53,"\u5b57\u5178imdb":61,"\u5b57\u6bb5\u4e2d":49,"\u5b57\u6bb5\u8868\u793a\u5bb9\u5668\u7684\u73af\u5883\u53d8\u91cf":49,"\u5b57\u6bb5\u8868\u793a\u8fd9\u4e2ajob\u4f1a\u540c\u65f6\u5f00\u542f3\u4e2apaddlepaddle\u8282\u70b9":49,"\u5b57\u7b26\u4e32":19,"\u5b58\u50a8":19,"\u5b58\u50a8\u5377":47,"\u5b58\u50a8\u5728\u8bb0\u5fc6\u5355\u5143\u533a\u5757\u7684\u5386\u53f2\u4fe1\u606f\u88ab\u66f4\u65b0\u7528\u6765\u8fed\u4ee3\u7684\u5b66\u4e60\u5355\u8bcd\u4ee5\u5408\u7406\u7684\u5e8f\u5217\u7a0b\u73b0":61,"\u5b58\u50a8\u6a21\u578b\u7684\u8def\u5f84":62,"\u5b58\u50a8\u7740\u7535\u5f71\u6216\u7528\u6237\u4fe1\u606f":59,"\u5b58\u5165settings\u5bf9\u8c61":3,"\u5b58\u5728\u6216\u66f4\u6539\u4e3a\u5176\u5b83\u6a21\u578b\u8def\u5f84":61,"\u5b66\u4e60\u6210\u672c\u9ad8":20,"\u5b66\u4e60\u7b97\u6cd5":25,"\u5b66\u672f":58,"\u5b81\u6ce2":32,"\u5b83\u4e0d\u4ec5\u80fd\u591f\u5904\u7406imdb\u6570\u636e":61,"\u5b83\u4eec\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4f5c\u4e3a\u7f51\u7edc\u7684\u51fa\u53e3":25,"\u5b83\u4eec\u7684\u5927\u5c0f\u662f":35,"\u5b83\u4eec\u7684\u6587\u4ef6\u540d\u662f":19,"\u5b83\u4eec\u8fd8\u53ef\u4ee5\u4f9b\u90a3\u4e9b\u8fd0\u884c\u66f4\u590d\u6742\u7684\u96c6\u7fa4\u7ba1\u7406\u7cfb\u7edf":41,"\u5b83\u4eec\u90fd\u662f\u5e8f\u5217":35,"\u5b83\u4f1a\u5728dataprovider\u521b\u5efa\u7684\u65f6\u5019\u6267\u884c":3,"\u5b83\u4f7f\u752850\u5c42\u7684resnet\u6a21\u578b\u6765\u5bf9":55,"\u5b83\u5305\u542b\u4ee5\u4e0b\u51e0\u6b65":37,"\u5b83\u5305\u542b\u4ee5\u4e0b\u53c2\u6570":37,"\u5b83\u5305\u542b\u56db\u4e2a\u7248\u672c":29,"\u5b83\u5305\u542b\u7684\u5c5e\u6027\u53c2\u6570\u5982\u4e0b":3,"\u5b83\u5305\u62ec\u4e86\u4e00\u4e2a\u53cc\u5411\u7684gru\u4f5c\u4e3a\u7f16\u7801\u5668\u548c\u89e3\u7801\u5668":62,"\u5b83\u53eb\u505a":35,"\u5b83\u53ef\u4ee5\u5728\u53e5\u5b50\u7ea7\u522b\u5229\u7528\u53ef\u6269\u5c55\u7684\u4e0a\u4e0b\u6587":61,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u51cf\u5c11\u5206\u53d1\u5ef6\u8fdf":41,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u6211\u4eec\u683c\u5f0f\u5316\u6e90\u4ee3\u7801":36,"\u5b83\u53ef\u4ee5\u6307\u6d4b\u91cf\u4e00\u4e2a\u7a0b\u5e8f\u7684\u7a7a\u95f4":40,"\u5b83\u53ef\u4ee5\u88ab\u5e94\u7528\u4e8e\u8fdb\u884c\u673a\u5668\u7ffb\u8bd1":62,"\u5b83\u53ef\u80fd\u6709\u4e0d\u6b62\u4e00\u4e2a\u6743\u91cd":37,"\u5b83\u540c\u65f6\u5b66\u4e60\u6392\u5217":62,"\u5b83\u548c\u6570\u636e\u4f20\u5165\u51fd\u6570\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570":3,"\u5b83\u5b58\u50a8\u5f53\u524d\u8282\u70b9\u6240\u6709\u8bad\u7ec3":41,"\u5b83\u5b9a\u4e49\u4e86":35,"\u5b83\u5b9a\u4e49\u4e86\u6a21\u578b\u53c2\u6570\u6539\u53d8\u7684\u89c4\u5219":25,"\u5b83\u5b9a\u4e49\u89e3\u7801\u7f51\u7edc\u7684":35,"\u5b83\u5c06\u88ab\u5206\u53d1\u5230":41,"\u5b83\u5c06\u8f93\u5165\u8bed\u53e5\u7f16\u7801\u4e3a\u5411\u91cf\u7684\u5e8f\u5217":62,"\u5b83\u5c06\u8fd4\u56de\u5982\u4e0b\u7684\u5b57\u5178":55,"\u5b83\u5c31\u4f1a\u5728\u6e90\u8bed\u53e5\u4e2d\u641c\u7d22\u51fa\u6700\u76f8\u5173\u4fe1\u606f\u7684\u4f4d\u7f6e\u7684\u96c6\u5408":62,"\u5b83\u652f\u6301\u591a\u7ebf\u7a0b\u66f4\u65b0":37,"\u5b83\u662finteger_value\u7c7b\u578b\u7684":32,"\u5b83\u662finteger_value_sequence\u7c7b\u578b\u7684":32,"\u5b83\u6709\u52a9\u4e8e\u5e2e\u52a9\u9891\u7e41\u4fee\u6539\u548c\u8bbf\u95ee\u5de5\u4f5c\u533a\u6587\u4ef6\u7684\u7528\u6237\u51cf\u5c11\u8d1f\u62c5":41,"\u5b83\u6a21\u62df\u4e86\u89e3\u7801\u7ffb\u8bd1\u8fc7\u7a0b\u4e2d\u5728\u6e90\u8bed\u53e5\u4e2d\u7684\u641c\u7d22":62,"\u5b83\u7684":35,"\u5b83\u7684\u6536\u655b\u901f\u5ea6\u6bd4":61,"\u5b83\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20":31,"\u5b83\u7684\u76ee\u7684\u662f\u9884\u6d4b\u5728\u4e00\u4e2a\u5e8f\u5217\u4e2d\u8868\u8fbe\u7684\u60c5\u611f\u6001\u5ea6":61,"\u5b83\u7684\u8f93\u5165\u4e0e\u7ecf\u8fc7\u5b66\u4e60\u7684\u53c2\u6570\u505a\u5185\u79ef\u5e76\u52a0\u4e0a\u504f\u7f6e":37,"\u5b83\u76f4\u63a5\u5b66\u4e60\u6bb5\u843d\u8868\u793a":61,"\u5b83\u80fd\u591f\u4ece\u8bcd\u7ea7\u5230\u5177\u6709\u53ef\u53d8\u4e0a\u4e0b\u6587\u957f\u5ea6\u7684\u4e0a\u4e0b\u6587\u7ea7\u522b\u6765\u603b\u7ed3\u8868\u793a":61,"\u5b83\u8bfb\u5165\u6570\u636e\u5e76\u5c06\u5b83\u4eec\u4f20\u8f93\u5230\u63a5\u4e0b\u6765\u7684\u7f51\u7edc\u5c42":25,"\u5b83\u8fd4\u56degen":62,"\u5b83\u8fd4\u56detrain":62,"\u5b83\u9700\u8981\u5728\u8fd9\u91cc\u6307\u5b9a":61,"\u5b83\u9996\u5148\u8c03\u7528\u57fa\u6784\u9020\u51fd\u6570":37,"\u5b89\u6392":32,"\u5b89\u88c5\u4e0e\u7f16\u8bd1":30,"\u5b89\u88c5\u5305\u7684\u4e0b\u8f7d\u5730\u5740\u662f":29,"\u5b89\u88c5\u540e\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":21,"\u5b89\u88c5\u597ddocker\u4e4b\u540e\u53ef\u4ee5\u4f7f\u7528\u6e90\u7801\u76ee\u5f55\u4e0b\u7684\u811a\u672c\u6784\u5efa\u6587\u6863":38,"\u5b89\u88c5\u5b8c\u6210\u540e":29,"\u5b89\u88c5\u6d41\u7a0b":57,"\u5b89\u88c5\u8be5\u8f6f\u4ef6\u5305\u5c31\u53ef\u4ee5\u5728python\u73af\u5883\u4e0b\u5b9e\u73b0\u6a21\u578b\u9884\u6d4b":5,"\u5b89\u88c5paddlepaddl":57,"\u5b89\u88c5pillow":54,"\u5b89\u9759":32,"\u5b8c\u6210":[36,46],"\u5b8c\u6210\u4efb\u610f\u7684\u8fd0\u7b97\u903b\u8f91":34,"\u5b8c\u6210\u540evolume\u4e2d\u7684\u6587\u4ef6\u5185\u5bb9\u5927\u81f4\u5982\u4e0b":49,"\u5b8c\u6210\u5f00\u53d1":27,"\u5b8c\u6210\u6570\u636e\u7684\u9884\u5904\u7406":19,"\u5b8c\u6210\u76f8\u5e94\u7684\u8ba1\u7b97":31,"\u5b8c\u6574\u6e90\u7801\u53ef\u53c2\u8003":24,"\u5b8c\u6574\u7684\u4ee3\u7801\u89c1":5,"\u5b8c\u6574\u7684\u53c2\u6570\u77e9\u9635\u88ab\u5206\u5e03\u5728\u4e0d\u540c\u7684\u53c2\u6570\u670d\u52a1\u5668\u4e0a":37,"\u5b8c\u6574\u7684\u6570\u636e\u63d0\u4f9b\u6587\u4ef6\u5728":35,"\u5b8c\u6574\u7684\u914d\u7f6e\u6587\u4ef6\u5728":35,"\u5b98\u65b9\u6587\u6863":26,"\u5b9a\u4e49\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185rnn\u5355\u5143\u5b8c\u6210\u7684\u8ba1\u7b97":34,"\u5b9a\u4e49\u4e00\u4e2apython\u7684":3,"\u5b9a\u4e49\u4e86\u4e00\u4e2a\u53ea\u8bfb\u7684memori":34,"\u5b9a\u4e49\u4e86\u7f51\u7edc\u7684\u6570\u636e\u69fd":60,"\u5b9a\u4e49\u4e86\u7f51\u7edc\u7ed3\u6784":54,"\u5b9a\u4e49\u4e86\u7f51\u7edc\u7ed3\u6784\u5e76\u4fdd\u5b58\u4e3a":25,"\u5b9a\u4e49\u5728\u5916\u5c42":34,"\u5b9a\u4e49\u5f02\u6b65\u8bad\u7ec3\u7684\u957f\u5ea6":43,"\u5b9a\u4e49\u6570\u636e\u6765\u6e90":25,"\u5b9a\u4e49\u6e90\u8bed\u53e5\u7684\u6570\u636e\u5c42":35,"\u5b9a\u4e49\u89e3\u7801\u5668\u7684memori":35,"\u5b9a\u4e49\u8bad\u7ec3\u6570\u6910\u548c\u6d4b\u8bd5\u6570\u6910\u63d0\u4f9b\u8005":61,"\u5b9a\u4e49\u8f93\u5165\u6570\u636e\u5927\u5c0f":46,"\u5b9a\u4e49\u8f93\u5165\u6570\u636e\u7684\u7c7b\u578b":25,"\u5b9a\u4e49\u8f93\u51fa\u51fd\u6570":35,"\u5b9a\u4e49\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5355\u6b65\u51fd\u6570":35,"\u5b9e\u4f8b\u5982\u4e0b":60,"\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":32,"\u5b9e\u73b0\u524d\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":37,"\u5b9e\u73b0\u5355\u6b65\u51fd\u6570":35,"\u5b9e\u73b0\u540e\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":37,"\u5b9e\u73b0\u6570\u636e\u8f93\u5165\u51fd\u6570":3,"\u5b9e\u73b0\u6784\u9020\u51fd\u6570":37,"\u5b9e\u73b0\u7b80\u5355":20,"\u5b9e\u73b0\u7ec6\u8282":37,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad":37,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u540e\u5411\u4f20\u64ad":37,"\u5b9e\u73b0\u8bcd\u8bed\u548c\u53e5\u5b50\u4e24\u4e2a\u7ea7\u522b\u7684\u53cc\u5c42rnn\u7ed3\u6784":34,"\u5b9e\u73b0\u8be5\u5c42\u7684c":37,"\u5b9e\u9645\u4e0a\u53ea\u6709":55,"\u5b9e\u9645\u4e0a\u662fcsv\u6587\u4ef6":58,"\u5ba2\u6237":32,"\u5ba2\u6237\u670d\u52a1":58,"\u5bb6":32,"\u5bb9\u5668":47,"\u5bb9\u5668\u4e0d\u4f1a\u4fdd\u7559\u5728\u8fd0\u884c\u65f6\u751f\u6210\u7684\u6570\u636e":47,"\u5bb9\u5668\u8fd0\u884c\u90fd\u8fd0\u884c":49,"\u5bbf\u4e3b\u673a\u76ee\u5f55":47,"\u5bc4\u5b58\u5668\u4f7f\u7528\u60c5\u51b5\u548c\u5171\u4eab\u5185\u5b58\u4f7f\u7528\u60c5\u51b5\u80fd\u8ba9\u6211\u4eec\u5bf9gpu\u7684\u6574\u4f53\u4f7f\u7528\u6709\u66f4\u597d\u7684\u7406\u89e3":40,"\u5bf9":32,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f01\u5411\u91cf":3,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f\u6d6e\u70b9\u5411\u91cf":3,"\u5bf9\u4e8e":35,"\u5bf9\u4e8e\u4e0d\u540c\u8bed\u8a00":20,"\u5bf9\u4e8e\u4e24\u79cd\u4e0d\u540c\u7684\u8f93\u5165\u6570\u636e\u7c7b\u578b":32,"\u5bf9\u4e8e\u5185\u5b58\u8f83\u5c0f\u7684\u673a\u5668":3,"\u5bf9\u4e8e\u5355\u5c42rnn":32,"\u5bf9\u4e8e\u5355\u5c42rnn\u7684\u6570\u636e\u4e00\u5171\u6709\u4e24\u4e2a\u6837\u672c":32,"\u5bf9\u4e8e\u53cc\u5c42rnn":32,"\u5bf9\u4e8e\u540c\u4e00\u6bb5c":20,"\u5bf9\u4e8e\u540c\u6837\u7684\u6570\u636e":32,"\u5bf9\u4e8e\u591a\u8bed\u8a00\u63a5\u53e3":20,"\u5bf9\u4e8e\u5927\u591a\u6570\u8bed\u8a00":20,"\u5bf9\u4e8e\u6211\u4eec\u652f\u6301\u7684\u5168\u90e8\u77e9\u9635\u64cd\u4f5c":37,"\u5bf9\u4e8e\u6811\u7684\u6bcf\u4e00\u5c42":62,"\u5bf9\u4e8e\u6bb5\u843d\u7684\u6587\u672c\u5206\u7c7b":32,"\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5355\u5c42rnn\u7684\u6570\u636e":32,"\u5bf9\u4e8e\u6bcf\u4f4d\u7528\u6237":59,"\u5bf9\u4e8e\u6bcf\u79cd\u7c7b\u578b":21,"\u5bf9\u4e8e\u6bcf\u79cdc":21,"\u5bf9\u4e8e\u7b80\u5355\u7684\u591a\u673a\u534f\u540c\u8bad\u7ec3\u4f7f\u7528\u4e0a\u8ff0\u65b9\u5f0f\u5373\u53ef":46,"\u5bf9\u4e8e\u7ed9\u5b9a\u7684\u4e00\u6761\u6587\u672c":57,"\u5bf9\u4e8e\u914d\u5907\u6709\u6ce8\u610f\u529b\u673a\u5236\u7684\u89e3\u7801\u5668":35,"\u5bf9\u4e8eamazon":57,"\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":40,"\u5bf9\u5168\u8fde\u63a5\u5c42\u6765\u8bf4":37,"\u5bf9\u56fe\u7247\u8fdb\u884c\u9884\u5904\u7406":54,"\u5bf9\u5e94\u4e00\u4e2a\u5b50\u53e5":34,"\u5bf9\u5e94\u4e00\u4e2a\u8bcd":34,"\u5bf9\u5e94\u4e8e\u5b57\u5178":53,"\u5bf9\u5e94\u7684":3,"\u5bf9\u6027\u80fd\u5c24\u5176\u662f\u5185\u5b58\u5360\u7528\u6709\u4e00\u5b9a\u7684\u5f00\u9500":46,"\u5bf9\u6570\u636e\u96c6\u8fdb\u884c\u9884\u5904\u7406\u7684\u57fa\u672c\u547d\u4ee4\u662f":62,"\u5bf9\u6574\u4e2a\u65b0\u5411\u91cf\u96c6\u5408\u7684\u6bcf\u4e00\u4e2a\u7ef4\u5ea6\u53d6\u6700\u5927\u503c\u6765\u8868\u793a\u6700\u540e\u7684\u53e5\u5b50":57,"\u5bf9\u6587\u6863\u5904\u7406\u540e\u5f62\u6210\u7684\u5355\u8bcd\u5411\u91cf":61,"\u5bf9\u673a\u5668\u7ffb\u8bd1\u7684\u4eba\u5de5\u8bc4\u4f30\u5de5\u4f5c\u5f88\u5e7f\u6cdb\u4f46\u4e5f\u5f88\u6602\u8d35":62,"\u5bf9\u6bcf\u4e2a\u8f93\u5165":37,"\u5bf9\u6bcf\u4e2a\u8f93\u5165\u4e58\u4e0a\u53d8\u6362\u77e9\u9635":37,"\u5bf9\u6bd4":20,"\u5bf9\u6fc0\u6d3b\u6c42\u5bfc":37,"\u5bf9\u7528\u6237\u6765\u8bf4":3,"\u5bf9\u8bad\u7ec3\u6570\u636e\u8fdb\u884cshuffl":3,"\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u975e\u7ebf\u6027\u53d8\u6362":57,"\u5bf9\u8c61":[24,46],"\u5bf9\u8c61\u5b58\u50a8\u4e3a\u6587\u4ef6":59,"\u5bf9\u8f93\u5165\u53c2\u6570\u7684\u5b89\u5168\u6027\u8fdb\u884c\u4e86\u5fc5\u8981\u7684\u5224\u65ad":21,"\u5bf9\u8f93\u51fa\u7684\u5408\u5e76":34,"\u5bf9\u8fd9\u4e2a\u7248\u672c\u7684\u63d0\u4ea4":23,"\u5bf9\u9762":32,"\u5bf9check":3,"\u5bf9sparse_binary_vector\u548csparse_float_vector":3,"\u5bfc\u51fa\u8fd9\u4e9b\u63a5\u53e3":21,"\u5bfc\u81f4\u4e86\u6d6e\u70b9\u6570\u6ea2\u51fa":24,"\u5bfc\u81f4\u53c2\u6570\u6536\u655b\u5230\u4e86\u4e00\u4e9b\u5947\u5f02\u7684\u60c5\u51b5":24,"\u5bfc\u81f4\u53c2\u6570\u7d2f\u52a0":24,"\u5bfc\u81f4\u7f16\u8bd1paddlepaddle\u5931\u8d25":24,"\u5bfc\u81f4\u8bad\u7ec3\u65f6\u95f4\u8fc7\u957f":24,"\u5c01\u88c5\u4e86":40,"\u5c01\u88c5\u8be5\u5c42\u7684python\u63a5\u53e3":37,"\u5c06":[3,23,24,40,46,57],"\u5c06\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u62c6\u5206\u6210\u591a\u4efd":18,"\u5c06\u4e0a\u4e00\u65f6\u95f4\u6b65\u6240\u751f\u6210\u7684\u8bcd\u7684\u5411\u91cf\u6765\u4f5c\u4e3a\u5f53\u524d\u65f6\u95f4\u6b65\u7684\u8f93\u5165":35,"\u5c06\u4ed6\u4eec\u79fb\u52a8\u5230\u76ee\u5f55":59,"\u5c06\u4f1a\u81ea\u52a8\u8ba1\u7b97\u51fa\u4e00\u4e2a\u5408\u9002\u7684\u503c":43,"\u5c06\u5176\u8bbe\u7f6e\u6210":24,"\u5c06\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e\u5148\u53d8\u6362\u6210\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e":32,"\u5c06\u542b\u6709\u5b50\u53e5":34,"\u5c06\u542b\u6709\u8bcd\u8bed\u7684\u53e5\u5b50\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":34,"\u5c06\u56fe\u7247\u6309\u7167\u4e0a\u8ff0\u7ed3\u6784\u5b58\u50a8\u597d\u4e4b\u540e":54,"\u5c06\u5728":54,"\u5c06\u5728\u8fd0\u884c\u65f6\u62a5\u9519":41,"\u5c06\u5916\u90e8\u7684\u5b58\u50a8\u670d\u52a1\u5728kubernetes\u4e2d\u63cf\u8ff0\u6210\u4e3a\u7edf\u4e00\u7684\u8d44\u6e90\u5f62\u5f0f":47,"\u5c06\u591a\u53e5\u8bdd\u770b\u6210\u4e00\u4e2a\u6574\u4f53\u540c\u65f6\u4f7f\u7528encoder\u538b\u7f29":32,"\u5c06\u591a\u53f0\u673a\u5668\u7684\u6d4b\u8bd5\u7ed3\u679c\u5408\u5e76":43,"\u5c06\u5927\u91cf\u7684":20,"\u5c06\u5b57\u5178\u7684\u5730\u5740\u4f5c\u4e3aargs\u4f20\u7ed9dataprovid":24,"\u5c06\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u524d\u5411\u548c\u53cd\u5411\u90e8\u5206\u6df7\u5408\u5728\u4e00\u8d77":35,"\u5c06\u6570\u636e\u5904\u7406\u6210\u89c4\u8303\u683c\u5f0f":53,"\u5c06\u6570\u636e\u7ec4\u5408\u6210batch\u8fdb\u884c\u8bad\u7ec3":3,"\u5c06\u6570\u636e\u8f6c\u6362\u4e3apaddle\u7684\u683c\u5f0f":54,"\u5c06\u65b0\u5206\u652f\u7684\u7248\u672c\u6253\u4e0atag":23,"\u5c06\u65b0\u5efa\u7684\u6743\u91cd\u52a0\u5165\u6743\u91cd\u8868":37,"\u5c06\u65e5\u5fd7\u5199\u5165\u6587\u4ef6":59,"\u5c06\u672c\u5730\u7684\u4fee\u6539\u63a8\u9001\u5230":36,"\u5c06\u6837\u672c\u7684\u5730\u5740\u653e\u5165\u53e6\u4e00\u4e2a\u6587\u672c\u6587\u4ef6":3,"\u5c06\u6b64\u76ee\u5f55\u6302\u8f7d\u4e3a\u5bb9\u5668\u7684":49,"\u5c06\u6bcf\u4e2a\u6e90\u8bed\u8a00\u5230\u76ee\u6807\u8bed\u8a00\u7684\u5e73\u884c\u8bed\u6599\u5e93\u6587\u4ef6\u5408\u5e76\u4e3a\u4e00\u4e2a\u6587\u4ef6":62,"\u5c06\u73af\u5883\u53d8\u91cf\u8f6c\u6362\u6210paddle\u7684\u547d\u4ee4\u884c\u53c2\u6570":49,"\u5c06\u7528\u6237\u7684\u539f\u59cb\u6570\u636e\u8f6c\u6362\u6210\u7cfb\u7edf\u53ef\u4ee5\u8bc6\u522b\u7684\u6570\u636e\u7c7b\u578b":46,"\u5c06\u7ed3\u679c\u4fdd\u5b58\u5230\u6b64\u76ee\u5f55\u91cc":49,"\u5c06\u884c\u4e2d\u7684\u6570\u636e\u8f6c\u6362\u6210\u4e0einput_types\u4e00\u81f4\u7684\u683c\u5f0f":3,"\u5c06\u88ab\u5206\u4e3a":53,"\u5c06\u8bad\u7ec3\u6587\u4ef6\u4e0e\u5207\u5206\u597d\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5171\u4eab\u5b58\u50a8":49,"\u5c06\u8be5\u53e5\u8bdd\u5305\u542b\u7684\u6240\u6709\u5355\u8bcd\u5411\u91cf\u6c42\u5e73\u5747":57,"\u5c06\u8df3\u8fc7\u5206\u53d1\u9636\u6bb5\u76f4\u63a5\u542f\u52a8\u6240\u6709\u8282\u70b9\u7684\u96c6\u7fa4\u4f5c\u4e1a":41,"\u5c06\u8fd9\u79cd\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684\u8fde\u63a5\u7528\u4e00\u4e2a\u7279\u6b8a\u7684\u795e\u7ecf\u7f51\u7edc\u5355\u5143\u5b9e\u73b0":32,"\u5c06\u8fdc\u7a0b\u4ed3\u5e93":36,"\u5c06\u900f\u660e":41,"\u5c06ip\u6392\u5e8f\u751f\u6210\u7684\u5e8f\u53f7\u4f5c\u4e3atrain":49,"\u5c06master\u5206\u652f\u7684\u5408\u5165commit\u6253\u4e0atag":23,"\u5c11\u4e8e5":41,"\u5c1a\u53ef":32,"\u5c31":32,"\u5c31\u4f1a\u751f\u6210\u975e\u5e38\u591a\u7684gener":3,"\u5c31\u53ef\u4ee5\u518d\u8fd0\u884c\u53e6\u4e00\u4e2anginx":27,"\u5c31\u53ef\u4ee5\u5c06\u6570\u636e\u4f20\u9001\u7ed9paddlepaddle\u4e86":3,"\u5c31\u53ef\u4ee5\u5c06\u8fd9\u4e9b\u6587\u4ef6\u6301\u4e45\u5316\u5b58\u50a8":47,"\u5c31\u5f88\u5bb9\u6613\u5bfc\u81f4\u5185\u5b58\u8d85\u9650":24,"\u5c31\u662f":32,"\u5c31\u662f\u6a21\u578b\u7684\u53c2\u6570":25,"\u5c31\u662f\u7528\u4e8e\u5c55\u793a\u4e0a\u8ff0\u5206\u6790\u5de5\u5177\u7684\u7528\u6cd5":40,"\u5c31\u80fd\u591f\u5f88\u65b9\u4fbf\u7684\u5b8c\u6210\u6570\u636e\u4e0b\u8f7d\u548c\u76f8\u5e94\u7684\u9884\u5904\u7406\u5de5\u4f5c":57,"\u5c31\u8fd9\u4e48\u7b80\u5355":27,"\u5c31\u901a\u5e38\u7684gpu\u6027\u80fd\u5206\u6790\u6765\u8bf4":40,"\u5c31\u9700\u8981\u5bf9\u8fd9\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00\u589e\u52a0\u4e00\u4e9b\u5b9a\u4e49":20,"\u5c31\u9700\u8981\u9009\u62e9\u4f7f\u7528no":27,"\u5c3a\u5bf8":55,"\u5c40\u90e8\u5173\u8054\u6027\u8d28\u548c\u7a7a\u95f4\u4e0d\u53d8\u6027\u8d28":54,"\u5c42\u540e\u5f97\u5230\u6df1\u5ea6":60,"\u5c42\u548c\u8f93\u5165\u7684\u914d\u7f6e":37,"\u5c42\u6743\u91cd":55,"\u5c42\u6b21\u5316\u7684rnn":34,"\u5c42\u7279\u5f81":55,"\u5c42\u7684\u540d\u79f0\u4e0e":35,"\u5c42\u7684\u5927\u5c0f":37,"\u5c42\u7684\u7279\u5f81":55,"\u5c42\u7684\u7c7b\u578b":37,"\u5c42\u7684\u8f93\u5165":60,"\u5c42\u7684\u8f93\u5165\u548c\u8f93\u51fa\u4f5c\u4e3a\u4e0b\u4e00\u4e2a":60,"\u5c42\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684":35,"\u5c42\u7ec4\u6210\u4e00\u5bf9":60,"\u5c45\u7136":32,"\u5c55\u793a\u4e86\u4e00\u79cd\u65b9\u6cd5":62,"\u5c55\u793a\u4e86\u4e0a\u8ff0\u7f51\u7edc\u6a21\u578b\u7684\u8bad\u7ec3\u6548\u679c":57,"\u5c55\u793a\u4e86\u5982\u4f55\u5c06\u6bcf\u4e2a\u7279\u5f81\u6620\u5c04\u5230\u4e00\u4e2a\u5411\u91cf":59,"\u5c5e\u6027":60,"\u5d4c\u5165\u5c42":59,"\u5d4c\u5165\u7279\u5f81\u5b57\u5178":59,"\u5d4c\u5165\u7f16\u53f7\u4f1a\u6839\u636e\u5355\u8bcd\u6392\u5e8f":59,"\u5de5\u4f5c\u6a21\u5f0f":43,"\u5de5\u4f5c\u7a7a\u95f4":41,"\u5de5\u4f5c\u7a7a\u95f4\u4e2d\u7684":41,"\u5de5\u4f5c\u7a7a\u95f4\u6839\u76ee\u5f55":41,"\u5de5\u4f5c\u7a7a\u95f4\u76ee\u5f55\u7684\u5de5\u4f5c\u7a7a\u95f4":41,"\u5de5\u4f5c\u7a7a\u95f4\u914d\u7f6e":41,"\u5de5\u5177":61,"\u5de5\u5177\u4e2d\u7684\u811a\u672c":61,"\u5de5\u5177\u6765\u7ba1\u7406":36,"\u5de5\u7a0b\u5e08":58,"\u5de6\u56fe\u6784\u9020\u7f51\u7edc\u6a21\u5757\u7684\u65b9\u5f0f\u88ab\u7528\u4e8e34\u5c42\u7684\u7f51\u7edc\u4e2d":55,"\u5de6\u8fb9\u662f":55,"\u5dee\u8bc4":57,"\u5df2\u6253\u5f00":36,"\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":43,"\u5df2\u7ecf\u63d0\u4f9b\u4e86\u811a\u672c\u6765\u5e2e\u52a9\u60a8\u521b\u5efa\u8fd9\u4e24\u4e2a\u6587\u4ef6":41,"\u5e02\u573a":58,"\u5e02\u9762\u4e0a\u5df2\u7ecf\u6709nvidia\u6216\u7b2c\u4e09\u65b9\u63d0\u4f9b\u7684\u4f17\u591a\u5de5\u5177":40,"\u5e0c\u671b\u52a0\u901f\u8bad\u7ec3":46,"\u5e0c\u671b\u80fd\u8ba9\u6211\u4eec\u77e5\u6653":59,"\u5e2e\u52a9\u6211\u4eec\u5b8c\u6210\u5bf9\u8f93\u5165\u5e8f\u5217\u7684\u62c6\u5206":34,"\u5e2e\u52a9\u6211\u4eec\u66f4\u597d\u5730\u63cf\u8ff0\u6bb5\u843d":34,"\u5e2e\u52a9\u6211\u4eec\u6784\u9020\u4e00\u4e9b\u590d\u6742\u7684\u8f93\u5165\u4fe1\u606f":31,"\u5e38\u5e38\u51fa\u73b0":24,"\u5e38\u7528\u4f18\u5316\u7b97\u6cd5\u5305\u62ecmomentum":57,"\u5e38\u89c1\u7684\u53ef\u9009\u5b58\u50a8\u670d\u52a1\u5305\u62ec":47,"\u5e72\u51c0":32,"\u5e73\u53f0\u4e3a\u60f3\u89c2\u6d4b\u8bcd\u5411\u91cf\u7684\u7528\u6237\u63d0\u4f9b\u4e86\u5c06\u4e8c\u8fdb\u5236\u8bcd\u5411\u91cf\u6a21\u578b\u8f6c\u6362\u4e3a\u6587\u672c\u6a21\u578b\u7684\u529f\u80fd":53,"\u5e73\u5747\u6545\u969c\u4fee\u590d\u65f6\u95f4":18,"\u5e73\u5747\u6545\u969c\u7387":18,"\u5e73\u5747\u7279\u5f81\u56fe\u7684\u9ad8\u5ea6\u53ca\u5bbd\u5ea6":54,"\u5e74\u9f84":58,"\u5e74\u9f84\u4ece\u4e0b\u5217\u5217\u8868\u8303\u56f4\u4e2d\u9009\u53d6":58,"\u5e74\u9f84\u548c\u804c\u4e1a":59,"\u5e76\u4e0d\u4fdd\u8bc1":37,"\u5e76\u4e0d\u662f\u4f7f\u7528\u53cc\u5c42rnn\u89e3\u51b3\u5b9e\u9645\u7684\u95ee\u9898":32,"\u5e76\u4e0d\u662fkubernetes\u4e2d\u7684node\u6982\u5ff5":49,"\u5e76\u4e0d\u771f\u6b63\u7684\u548c":32,"\u5e76\u4e14":[3,35],"\u5e76\u4e14\u4f7f\u7528":21,"\u5e76\u4e14\u5185\u5c42\u7684\u5e8f\u5217\u64cd\u4f5c\u4e4b\u95f4\u72ec\u7acb\u65e0\u4f9d\u8d56":32,"\u5e76\u4e14\u5206\u522b\u91cd\u547d\u540d\u6587\u4ef6\u540e\u7f00":62,"\u5e76\u4e14\u52a0\u4e0a\u4e0b\u9762\u7684\u547d\u4ee4\u884c\u53c2\u6570":45,"\u5e76\u4e14\u5305\u62ecunit":36,"\u5e76\u4e14\u53ea\u6709\u4e00\u4e2a\u6743\u91cd":55,"\u5e76\u4e14\u53ef\u4ee5\u88ab\u63d0\u4ea4\u7684job\u76f4\u63a5\u4f7f\u7528":19,"\u5e76\u4e14\u53ef\u80fd\u4f1a\u52a0\u901f\u8bad\u7ec3\u8fc7\u7a0b":24,"\u5e76\u4e14\u540e\u7eed\u4ecd\u5728\u4e0d\u65ad\u6539\u8fdb":25,"\u5e76\u4e14\u542f\u52a8\u8bad\u7ec3":49,"\u5e76\u4e14\u5728\u5185\u5b58\u8db3\u591f\u7684\u60c5\u51b5\u4e0b\u8d8a\u5927\u8d8a\u597d":3,"\u5e76\u4e14\u5728\u5e38\u89c1\u7684\u5e73\u53f0\u4e0a":20,"\u5e76\u4e14\u5728\u968f\u540e\u7684\u8bfb\u53d6\u6570\u636e\u8fc7\u7a0b\u4e2d\u586b\u5145\u8bcd\u8868":57,"\u5e76\u4e14\u5728dataprovider\u4e2d\u5b9e\u73b0\u5982\u4f55\u8bbf\u95ee\u8bad\u7ec3\u6587\u4ef6\u5217\u8868":2,"\u5e76\u4e14\u5b83\u4eec\u7684\u987a\u5e8f\u4e0e":55,"\u5e76\u4e14\u5bf9\u7528\u6237\u7684\u7279\u5f81\u505a\u540c\u6837\u7684\u64cd\u4f5c":59,"\u5e76\u4e14\u5c06\u9884\u5904\u7406\u597d\u7684\u6570\u636e\u96c6\u5b58\u653e\u5728":62,"\u5e76\u4e14\u67e5\u8be2paddlepaddle\u5355\u5143\u6d4b\u8bd5\u7684\u65e5\u5fd7":24,"\u5e76\u4e14\u7b2c\u4e8c\u4e2a\u662f\u53cd\u5411lstm":61,"\u5e76\u4e14\u7cfb\u7edf\u6bcf\u4e00\u8f6e\u8bad\u7ec3\u5f00\u59cb\u65f6\u4f1a\u91cd\u7f6edataprovid":46,"\u5e76\u4e14\u7f16\u8bd1\u80fd\u901a\u8fc7\u4ee3\u7801\u6837\u5f0f\u68c0\u67e5":36,"\u5e76\u4e14\u8ba9\u63a5\u53e3\u8131\u79bb\u5b9e\u73b0\u7ec6\u8282":20,"\u5e76\u4e14\u8f93\u51fa\u4e00\u4e2a":36,"\u5e76\u4e14\u901a\u8fc7\u7ed9\u51fa\u5f53\u524d\u76ee\u6807\u5355\u8bcd\u6765\u9884\u6d4b\u4e0b\u4e00\u4e2a\u76ee\u6807\u5355\u8bcd":62,"\u5e76\u4e14\u96c6\u7fa4\u4f5c\u4e1a\u4e2d\u7684\u6240\u6709\u8282\u70b9\u5c06\u5728\u6b63\u5e38\u60c5\u51b5\u4e0b\u5904\u7406\u5177\u6709\u76f8\u540c\u903b\u8f91\u4ee3\u7801\u7684\u6587\u4ef6":41,"\u5e76\u4e14\u9700\u8981\u91cd\u5199\u57fa\u7c7b\u4e2d\u7684\u4ee5\u4e0b\u51e0\u4e2a\u865a\u51fd\u6570":37,"\u5e76\u4e14softmax\u5c42\u7684\u4e24\u4e2a\u8f93\u5165\u4e5f\u4f7f\u7528\u4e86\u540c\u6837\u7684\u53c2\u6570":24,"\u5e76\u4f20\u5165\u76f8\u5e94\u7684\u547d\u4ee4\u884c\u53c2\u6570\u521d\u59cb\u5316paddlepaddl":5,"\u5e76\u4f7f\u7528":60,"\u5e76\u4f7f\u7528\u4e86dropout":57,"\u5e76\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u6765\u5bf9\u56fe\u7247\u8fdb\u884c\u5206\u7c7b":54,"\u5e76\u521b\u5efa\u4e86\u4e00\u4e2a\u65b0\u6587\u4ef6":36,"\u5e76\u5220\u9664":23,"\u5e76\u5220\u9664\u66f4\u65e9\u7684\u5feb\u7167":18,"\u5e76\u52a0\u8f7d\u5176\u4e2d\u7684\u53c2\u6570":18,"\u5e76\u5728\u4e58\u79ef\u7ed3\u679c\u4e0a\u518d\u52a0\u4e0a\u7ef4\u5ea6\u4e3a":37,"\u5e76\u5728\u6700\u5f00\u59cb\u521d\u59cb\u5316\u4e3a\u8d77\u59cb\u8bcd":35,"\u5e76\u5728\u7b14\u8bb0\u672c\u4e0a\u901a\u8fc7ssh\u4e0e\u5176\u8fde\u63a5":27,"\u5e76\u5728\u7c7b\u6784\u5efa\u51fd\u6570\u4e2d\u628a\u5b83\u653e\u5165\u4e00\u4e2a\u7c7b\u6210\u5458\u53d8\u91cf\u91cc":37,"\u5e76\u5728\u96c6\u7fa4\u4e2d\u8fd0\u884c\u591a\u4e2a\u5206\u5e03\u5f0f\u6570\u636e\u5904\u7406\u4efb\u52a1":19,"\u5e76\u5bf9\u76f8\u5e94\u7684\u53c2\u6570\u8c03\u7528":37,"\u5e76\u5c06\u5176\u6295\u5c04\u5230":35,"\u5e76\u5c06\u5b83\u4eec\u6309\u7167\u542f\u53d1\u4ee3\u4ef7":62,"\u5e76\u5c06\u5b83\u4eec\u653e\u5728":62,"\u5e76\u5c06\u6bcf\u8f6e\u7684\u6a21\u578b\u7ed3\u679c\u5b58\u653e\u5728":25,"\u5e76\u5c06c":21,"\u5e76\u5c06develop\u548ctest\u6570\u636e\u5206\u522b\u653e\u5165\u4e0d\u540c\u7684\u6587\u4ef6\u5939":62,"\u5e76\u60f3\u4f7f\u7528gpu\u6765\u8bad\u7ec3\u8bbe\u7f6e\u4e3atru":61,"\u5e76\u628a\u5feb\u7167\u4fdd\u5b58\u5230\u8fd9\u4e2a\u76ee\u5f55\u4e0b":18,"\u5e76\u6307\u5b9a\u7aef\u53e3\u53f7":46,"\u5e76\u6307\u5b9abatch":62,"\u5e76\u63d0\u4f9b\u4e86\u7b80\u5355\u7684cache\u529f\u80fd":3,"\u5e76\u6b22\u8fce\u60a8\u6765\u53c2\u4e0e\u8d21\u732e":61,"\u5e76\u6ca1\u6709paddle\u7279\u522b\u9700\u8981\u7684\u7279\u6027":20,"\u5e76\u7531":60,"\u5e76\u7ed9\u51fa\u5206\u7c7b\u7ed3\u679c":54,"\u5e76\u7ed9\u51fa\u7684\u76f8\u5173\u6a21\u578b\u683c\u5f0f\u7684\u5b9a\u4e49":53,"\u5e76\u88ab\u53cd\u5411\u5904\u7406":60,"\u5e76\u88ab\u5b58\u50a8\u5728\u8bf8\u5982hadoop":19,"\u5e76\u89c2\u5bdf\u7ed3\u679c":40,"\u5e76\u8bb0\u5f55\u5b83\u7684\u7f16\u53f7":36,"\u5e76\u8bbe\u7f6e":[29,41],"\u5e76\u9002\u5e94github\u7684\u7279\u6027\u505a\u4e86\u4e00\u4e9b\u533a\u522b":23,"\u5e76\u9010\u6e10\u5c55\u793a\u66f4\u52a0\u6df1\u5165\u7684\u529f\u80fd":57,"\u5e8a\u4e0a\u7528\u54c1":32,"\u5e8a\u57ab":32,"\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":31,"\u5e8f\u5217\u5230\u5e8f\u5217":62,"\u5e8f\u5217\u6570\u636e\u662f\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u9762\u5bf9\u7684\u4e00\u79cd\u4e3b\u8981\u8f93\u5165\u6570\u636e\u7c7b\u578b":34,"\u5e8f\u5217\u662f\u4e00\u79cd\u5e38\u89c1\u7684\u6570\u636e\u7c7b\u578b":31,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u5927\u591a\u9075\u5faaencod":34,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u7684\u8f93\u5165":34,"\u5e8f\u5217\u7684\u5f00\u59cb":62,"\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u539f\u6765\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u5143\u7d20\u7684\u5e73\u5747\u503c":31,"\u5e8f\u5217\u7684\u7ed3\u5c3e":62,"\u5e8f\u5217\u7684\u7ed3\u675f":62,"\u5e93":41,"\u5e93\u7684\u8def\u5f84":41,"\u5e94\u7528\u524d\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":35,"\u5e94\u7528\u53cd\u5411\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":35,"\u5e94\u7528\u6a21\u578b":57,"\u5e94\u8be5":32,"\u5e94\u8be5\u4e0e\u5b83\u7684memory\u540d\u5b57\u76f8\u540c":35,"\u5e94\u8be5\u964d\u4f4e\u5b66\u4e60\u7387":24,"\u5e95\u5c42\u8fdb\u7a0b":41,"\u5efa\u7acb\u4e00\u4e2a":36,"\u5efa\u7acb\u4e00\u4e2a\u6d3b\u8dc3\u7684\u5f00\u6e90\u793e\u533a":0,"\u5efa\u8bae":23,"\u5efa\u8bae\u5c06\u5176\u8bbe\u7f6e\u4e3a\u8f83\u5927":41,"\u5efa\u8bae\u5c06\u8be5\u53c2\u6570\u8bbe\u4e3atrue":43,"\u5f00\u53d1\u4e86\u6a21\u578b\u9884\u6d4b\u7684\u6837\u4f8b\u4ee3\u7801":21,"\u5f00\u53d1\u4eba\u5458\u4f7f\u7528":36,"\u5f00\u53d1\u4eba\u5458\u53ef\u4ee5\u5728docker\u5f00\u53d1\u955c\u50cf\u4e2d\u5f00\u53d1paddlepaddl":27,"\u5f00\u53d1\u8005\u4fee\u6539\u81ea\u5df1\u7684\u4ee3\u7801":23,"\u5f00\u53d1\u8005\u7ed9paddlepaddle\u4ee3\u7801\u589e\u52a0\u7684\u6ce8\u91ca\u4ee5pr\u7684\u5f62\u5f0f\u63d0\u4ea4\u5230github\u4e2d":38,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4e2d":23,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4f7f\u7528":23,"\u5f00\u53d1\u955c\u50cf":[27,36],"\u5f00\u53d1\u955c\u50cf\u5305\u542b\u4e86\u4ee5\u4e0b\u5de5\u5177":27,"\u5f00\u59cb":25,"\u5f00\u59cb\u63d0\u4f9b\u670d\u52a1":18,"\u5f00\u59cb\u6807\u8bb0":35,"\u5f00\u59cb\u8bad\u7ec3\u6a21\u578b":57,"\u5f00\u59cb\u9636\u6bb5":40,"\u5f02\u6b65\u8bfb\u53d6\u7b49\u95ee\u9898":46,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":42,"\u5f15\u5165\u4e86\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5f15\u5165lstm\u6a21\u578b\u4e3b\u8981\u662f\u4e3a\u4e86\u514b\u670d\u6d88\u5931\u68af\u5ea6\u7684\u95ee\u9898":61,"\u5f15\u5165paddlepaddle\u7684pydataprovider2\u5305":3,"\u5f15\u53d1":13,"\u5f15\u5bfc\u5c42":35,"\u5f15\u7528":41,"\u5f15\u7528memory\u5f97\u5230\u8fd9layer\u4e0a\u4e00\u65f6\u523b\u8f93\u51fa":34,"\u5f3a\u70c8\u63a8\u8350":32,"\u5f3a\u70c8\u63a8\u8350\u4f7f\u7528virtualenv\u6765\u521b\u9020\u4e00\u4e2a\u5e72\u51c0\u7684python\u73af\u5883":59,"\u5f52\u4e00\u5316\u6982\u7387\u5411\u91cf":35,"\u5f53":45,"\u5f53\u4f20\u9012\u76f8\u540c\u7684\u6837\u672c\u6570\u65f6":61,"\u5f53\u4f60\u6267\u884c\u547d\u4ee4":37,"\u5f53\u51fd\u6570\u8fd4\u56de\u7684\u65f6\u5019":3,"\u5f53\u524d\u5355\u8bcd\u5728\u76f8\u6bd4\u4e4b\u4e0b\u603b\u662f\u88ab\u5f53\u4f5c\u771f\u503c":62,"\u5f53\u524d\u5355\u8bcd\u662f\u89e3\u7801\u5668\u6700\u540e\u4e00\u6b65\u7684\u8f93\u51fa":62,"\u5f53\u524d\u65f6\u95f4\u6b65\u5904\u7684memory\u7684\u8f93\u51fa\u4f5c\u4e3a\u4e0b\u4e00\u65f6\u95f4\u6b65memory\u7684\u8f93\u5165":35,"\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":37,"\u5f53\u524d\u7684\u8f93\u5165y\u548c\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51farnn_state\u505a\u4e86\u4e00\u4e2a\u5168\u94fe\u63a5":32,"\u5f53\u524d\u8bc4\u4f30\u4e2d":62,"\u5f53\u524dbatch\u7684cost":62,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":57,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":57,"\u5f53\u529f\u80fd\u5206\u652f\u5f00\u53d1\u5b8c\u6bd5\u540e":23,"\u5f53\u5728\u7f51\u7edc\u5c42\u914d\u7f6e\u4e2d\u8bbe\u7f6e":43,"\u5f53\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":43,"\u5f53\u5bb9\u5668\u56e0\u4e3a\u5404\u79cd\u539f\u56e0\u88ab\u9500\u6bc1\u65f6":47,"\u5f53\u6240\u6709\u6570\u636e\u8bfb\u53d6\u5b8c\u4e00\u8f6e\u540e":46,"\u5f53\u6240\u6709pod\u90fd\u5904\u4e8erunning\u72b6\u6001":49,"\u5f53\u6a21\u578b\u53c2\u6570\u4e0d\u5b58\u5728\u65f6":43,"\u5f53\u6a21\u578b\u8bad\u7ec3\u597d\u4e86\u4e4b\u540e":57,"\u5f53\u6a21\u5f0f\u4e3a":43,"\u5f53\u7136":40,"\u5f53\u7528\u6237\u4f7f\u7528\u5b8c\u8fd9\u4e2a\u53c2\u6570\u540e":21,"\u5f53\u7f51\u7edc\u5c42\u7528\u4e00\u4e2a\u6279\u6b21\u505a\u8bad\u7ec3\u65f6":37,"\u5f53\u89e3\u8bfb\u6bcf\u4e00\u4e2a":35,"\u5f53\u8bad\u7ec3\u6570\u636e\u975e\u5e38\u591a\u65f6":3,"\u5f53\u8d85\u8fc7\u8be5\u9608\u503c\u65f6":43,"\u5f53\u8f93\u5165\u662f\u7ef4\u5ea6\u5f88\u9ad8\u7684\u7a00\u758f\u6570\u636e\u65f6":45,"\u5f53\u9700\u8981\u5feb\u901f\u6216\u8005\u9891\u7e41\u7684\u8bc4\u4f30\u65f6":62,"\u5f53classif":62,"\u5f62\u6210recurr":34,"\u5f62\u6210recurrent\u8fde\u63a5":34,"\u5f62\u72b6":55,"\u5f88":[32,57],"\u5f88\u591a":32,"\u5f88\u591a\u5f00\u53d1\u8005\u4f1a\u4f7f\u7528\u8fdc\u7a0b\u7684\u5b89\u88c5\u6709gpu\u7684\u670d\u52a1\u5668\u5de5\u4f5c":27,"\u5f88\u5b89\u9759":32,"\u5f88\u5bb9\u6613\u5bfc\u81f4\u67d0\u4e00\u4e2a\u53c2\u6570\u670d\u52a1\u5668\u6ca1\u6709\u5206\u914d\u5230\u4efb\u4f55\u53c2\u6570":46,"\u5f88\u5e72\u51c0":32,"\u5f88\u65b9\u4fbf":32,"\u5f88\u6709\u53ef\u80fd\u5b9e\u9645\u5e94\u7528\u5c31\u662f\u6ca1\u6709\u6309\u7167\u60a8\u7684\u9884\u671f\u60c5\u51b5\u8fd0\u884c":40,"\u5f88\u9002\u5408\u6784\u5efa\u7528\u4e8e\u7406\u89e3\u56fe\u7247\u5185\u5bb9\u7684\u6a21\u578b":54,"\u5f88\u96be\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":20,"\u5f88\u96be\u6574\u4f53\u4fee\u6b63":3,"\u5f8b\u5e08":58,"\u5f97":32,"\u5f97\u4f7f\u7528":20,"\u5f97\u5230\u53e5\u5b50\u7684\u8868\u793a":57,"\u5f97\u5230\u6700\u597d\u8f6e\u6b21\u4e0b\u7684\u6a21\u578b":59,"\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e2a\u65f6\u95f4\u6b65\u603b\u662f\u80fd\u591f\u5f15\u7528\u6240\u6709\u8f93\u5165":34,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4e2d":35,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4f5c\u4e3a\u4f7f\u7528":35,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u548c":35,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u9aa4\u987a\u5e8f\u5730\u5904\u7406\u5e8f\u5217":35,"\u5faa\u73af\u7f51\u7edc\u4ece":35,"\u5fc5\u8981":21,"\u5fc5\u987b":37,"\u5fc5\u987b\u4e00\u81f4":3,"\u5fc5\u987b\u4f7f\u7528python\u5173\u952e\u8bcd":3,"\u5fc5\u987b\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":32,"\u5fc5\u987b\u6307\u5411\u4e00\u4e2apaddlepaddle\u5b9a\u4e49\u7684lay":34,"\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":34,"\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":31,"\u5fc5\u987b\u7531\u53ea\u8bfbmemory\u7684":35,"\u5fd8\u8bb0\u95e8\u548c\u8f93\u51fa\u95e8":61,"\u5feb":[32,61],"\u5feb\u901f\u5165\u95e8":56,"\u5feb\u901f\u5728\u672c\u5730\u542f\u52a8\u4e00\u4e2a\u5355\u673a\u7684kubernetes\u670d\u52a1\u5668":47,"\u5feb\u901f\u90e8\u7f72\u96c6\u7fa4":47,"\u6027\u4ef7\u6bd4":32,"\u6027\u522b":[58,59],"\u6027\u80fd\u5206\u6790":40,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u662f\u7528\u4e8e\u7ed9\u5e94\u7528\u7a0b\u5e8f\u7684\u6027\u80fd\u505a\u5b9a\u91cf\u5206\u6790\u7684":40,"\u6027\u80fd\u5206\u6790\u662f\u6027\u80fd\u4f18\u5316\u7684\u5173\u952e\u4e00\u6b65":40,"\u6027\u80fd\u8c03\u4f18":42,"\u603b\u4f53\u6765\u8bf4":32,"\u603b\u8ba1\u7684\u53c2\u6570\u4e2a\u6570":53,"\u603b\u8bc4\u520610\u5206":61,"\u6050\u6016\u7247":58,"\u60a8\u4f1a\u5728\u63a5\u4e0b\u6765\u7684\u90e8\u5206\u4e2d\u83b7\u5f97\u66f4\u591a\u7684\u7ec6\u8282\u4ecb\u7ecd":40,"\u60a8\u53ef\u4ee5\u4efb\u610f\u4f7f\u7528\u4e00\u4e2a\u6216\u4e24\u4e2a\u6765\u5bf9\u611f\u5174\u8da3\u7684\u4ee3\u7801\u6bb5\u505a\u6027\u80fd\u5206\u6790":40,"\u60a8\u53ef\u4ee5\u5bfc\u5165":40,"\u60a8\u53ef\u4ee5\u91c7\u7528\u4e0b\u9762\u4e94\u4e2a\u6b65\u9aa4":40,"\u60a8\u5c06\u4e86\u89e3\u5982\u4f55":35,"\u60a8\u5c31\u80fd\u83b7\u5f97\u5982\u4e0b\u7684\u5206\u6790\u7ed3\u679c":40,"\u60a8\u6309\u5982\u4e0b\u6b65\u9aa4\u64cd\u4f5c\u5373\u53ef":40,"\u60a8\u6700\u597d\u5148\u786e\u8ba4":40,"\u60a8\u9700\u8981\u66f4\u6539":27,"\u60a8\u9996\u5148\u9700\u8981\u5728\u76f8\u5173\u4ee3\u7801\u6bb5\u4e2d\u52a0\u5165":40,"\u60ac\u7591\u7247":58,"\u60c5\u6001\u52a8\u8bcd":60,"\u60c5\u611f\u5206\u6790":[23,56],"\u60c5\u611f\u5206\u6790\u4e5f\u5e38\u7528\u4e8e\u57fa\u4e8e\u5927\u91cf\u8bc4\u8bba\u548c\u4e2a\u4eba\u535a\u5ba2\u6765\u76d1\u63a7\u793e\u4f1a\u5a92\u4f53":61,"\u60c5\u611f\u5206\u6790\u662f\u81ea\u7136\u8bed\u8a00\u7406\u89e3\u4e2d\u6700\u5178\u578b\u7684\u95ee\u9898\u4e4b\u4e00":61,"\u60c5\u611f\u5206\u6790\u6709\u8bb8\u591a\u5e94\u7528\u573a\u666f":61,"\u60ca\u9669\u7535\u5f71":58,"\u60f3\u4e86\u89e3\u66f4\u591a\u7ec6\u8282\u53ef\u4ee5\u53c2\u8003pydataprovider\u90e8\u5206\u7684\u6587\u6863":61,"\u610f\u5473\u7740\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u90fd\u662f\u76f8\u540c\u7684\u503c":35,"\u610f\u601d\u662f\u4e0d\u4f7f\u7528\u5e73\u5747\u53c2\u6570\u6267\u884c\u6d4b\u8bd5":43,"\u610f\u601d\u662f\u4e0d\u4fdd\u5b58\u7ed3\u679c":43,"\u610f\u601d\u662f\u4f7f\u7528\u7b2ctest":43,"\u610f\u601d\u662f\u5728gpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2agpu":43,"\u611f\u89c9":32,"\u620f\u5267":58,"\u6210\u529f\u8bad\u7ec3\u4e14\u9000\u51fa\u7684pod\u6570\u76ee\u4e3a3\u65f6":49,"\u6211\u4eec\u4e0d\u80fd\u901a\u8fc7\u5e38\u89c4\u7684\u68af\u5ea6\u68c0\u67e5\u7684\u65b9\u5f0f\u6765\u8ba1\u7b97\u68af\u5ea6":37,"\u6211\u4eec\u4e3a\u7528\u6237\u5b9a\u4ee5python\u63a5\u53e3\u6765\u914d\u7f6e\u7f51\u7edc":46,"\u6211\u4eec\u4e3b\u8981\u4f1a\u4ecb\u7ecdnvprof\u548cnvvp":40,"\u6211\u4eec\u4e5f\u53ef\u4ee5\u786e\u5b9a\u6bcf\u4e00\u4e2a\u53c2\u6570\u7684\u7c7b\u578b":21,"\u6211\u4eec\u4ec5\u4ec5\u662f\u5c06\u6bcf\u4e2a\u7279\u5f81\u79cd\u7c7b\u6620\u5c04\u5230\u4e00\u4e2a\u7279\u5f81\u5411\u91cf\u4e2d":59,"\u6211\u4eec\u4ec5\u6709\u4e00\u4e2a\u8f93\u5165":37,"\u6211\u4eec\u4ec5\u7528":59,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728":48,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884c\u5206\u5e03\u5f0fpaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":49,"\u6211\u4eec\u4ece\u63d0\u524d\u7ed9\u5b9a\u7684\u7c7b\u522b\u96c6\u5408\u4e2d\u9009\u62e9\u5176\u6240\u5c5e\u7c7b\u522b":57,"\u6211\u4eec\u4ee5mnist\u624b\u5199\u8bc6\u522b\u4e3a\u4f8b":3,"\u6211\u4eec\u4f1a\u53d1\u73b0\u6570\u636e\u96c6":62,"\u6211\u4eec\u4f1a\u5728":27,"\u6211\u4eec\u4f1a\u7ee7\u7eed\u4f7f\u7528\u73b0\u6709\u7684\u5185\u5b58\u5757":37,"\u6211\u4eec\u4f1a\u91cd\u65b0\u5206\u914d\u5185\u5b58":37,"\u6211\u4eec\u4f7f\u7528":[37,41,61],"\u6211\u4eec\u4f7f\u7528\u4e86":32,"\u6211\u4eec\u4f7f\u7528\u4e86\u4e00\u4e2a\u7f16\u89e3\u7801\u6a21\u578b\u7684\u6269\u5c55":62,"\u6211\u4eec\u4f7f\u7528\u4e86\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":61,"\u6211\u4eec\u4f7f\u7528\u5176\u4e2d\u7684":62,"\u6211\u4eec\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":20,"\u6211\u4eec\u4f7f\u7528\u6700\u5927\u6982\u7387\u7684\u6807\u7b7e\u4f5c\u4e3a\u7ed3\u679c":60,"\u6211\u4eec\u4f7f\u7528\u96c6\u675f\u641c\u7d22":62,"\u6211\u4eec\u4f7f\u7528paddlepaddle\u5728ilsvrc\u7684\u9a8c\u8bc1\u96c6\u517150":55,"\u6211\u4eec\u5047\u8bbe\u4e00\u53f0\u673a\u5668\u4e0a\u67094\u4e2agpu":45,"\u6211\u4eec\u5047\u8bbe\u623f\u4ea7\u7684\u4ef7\u683c":25,"\u6211\u4eec\u5148\u4ece\u4e00\u6761\u968f\u673a\u7684\u76f4\u7ebf":25,"\u6211\u4eec\u51c6\u5907\u4e86\u4e00\u4e2a\u811a\u672c":54,"\u6211\u4eec\u53ea\u4f7f\u7528\u5df2\u7ecf\u6807\u6ce8\u8fc7\u7684\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6":61,"\u6211\u4eec\u53ea\u6240\u4ee5\u4f7f\u7528lstm\u6765\u6267\u884c\u8fd9\u4e2a\u4efb\u52a1\u662f\u56e0\u4e3a\u5176\u6539\u8fdb\u7684\u8bbe\u8ba1\u5e76\u4e14\u5177\u6709\u95e8\u673a\u5236":61,"\u6211\u4eec\u53ea\u6f14\u793a\u4e00\u4e2a\u5355\u673a\u4f5c\u4e1a":48,"\u6211\u4eec\u53ea\u9700\u8981\u4f7f\u7528lstm":32,"\u6211\u4eec\u53ea\u9700\u8981\u8fd0\u884c":57,"\u6211\u4eec\u53ea\u9700\u8981\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u628a\u7f16\u8bd1\u597d\u7684paddlepaddle\u6253\u5305\u6210\u4e00\u4e2a":36,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528":54,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5b83\u6765\u751f\u6210\u5e8f\u5217":35,"\u6211\u4eec\u53ef\u4ee5\u5728":36,"\u6211\u4eec\u53ef\u4ee5\u5c06":41,"\u6211\u4eec\u53ef\u4ee5\u6309\u7167\u5982\u4e0b\u5c42\u6b21\u5b9a\u4e49\u975e\u5e8f\u5217":31,"\u6211\u4eec\u53ef\u4ee5\u751f\u6210":59,"\u6211\u4eec\u53ef\u4ee5\u76f4\u63a5\u89c2\u5bdf\u6a21\u578b\u7684\u53c2\u6570\u662f\u5426\u7b26\u5408\u9884\u671f\u6765\u8fdb\u884c\u68c0\u9a8c":25,"\u6211\u4eec\u53ef\u4ee5\u770b\u5230\u5728\u76ee\u5f55":61,"\u6211\u4eec\u53ef\u4ee5\u8bbe\u8ba1\u642d\u5efa\u4e00\u4e2a\u7075\u6d3b\u7684":34,"\u6211\u4eec\u53ef\u4ee5\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4\u6765\u505ableu\u8bc4\u4f30":62,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u4ee5\u4e0b\u547d\u4ee4\u6765\u8bad\u7ec3\u6a21\u578b":62,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u4ee5\u4e0b\u547d\u4ee4\u6765\u8fdb\u884c\u4ece\u6cd5\u8bed\u5230\u82f1\u8bed\u7684\u6587\u672c\u7ffb\u8bd1":62,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u5982\u4e0b\u547d\u4ee4\u8fdb\u884c\u9884\u5904\u7406\u5de5\u4f5c":54,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u65e5\u5fd7\u67e5\u770b\u5bb9\u5668\u8bad\u7ec3\u7684\u60c5\u51b5":49,"\u6211\u4eec\u57285\u5929\u91cc\u8bad\u7ec3\u4e8616\u4e2apass":62,"\u6211\u4eec\u5728\u51fd\u6570\u7684\u7ed3\u5c3e\u8fd4\u56de":35,"\u6211\u4eec\u5728\u62e5\u670950\u4e2a\u8282\u70b9\u7684\u96c6\u7fa4\u4e2d\u8bad\u7ec3\u6a21\u578b":62,"\u6211\u4eec\u5728\u8bad\u7ec3\u4e4b\u524d\u9700\u8981\u5e38\u89c1\u4e00\u4e2a\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":62,"\u6211\u4eec\u5728initialzier\u51fd\u6570\u91cc\u521d\u59cb\u5316\u8bcd\u8868":57,"\u6211\u4eec\u5bf9\u6a21\u578b\u8fdb\u884c\u4e86\u4ee5\u4e0b\u66f4\u6539":35,"\u6211\u4eec\u5c06":[49,59],"\u6211\u4eec\u5c06\u4e00\u6bb5\u8bdd\u770b\u6210\u53e5\u5b50\u7684\u6570\u7ec4":32,"\u6211\u4eec\u5c06\u4ecb\u7ecd\u5982\u4f55\u542f\u52a8\u5206\u5e03\u5f0f\u8bad\u7ec3\u4f5c\u4e1a":48,"\u6211\u4eec\u5c06\u4ee5":[41,57],"\u6211\u4eec\u5c06\u4ee5\u6700\u57fa\u672c\u7684\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u4f5c\u4e3a\u8d77\u70b9":57,"\u6211\u4eec\u5c06\u4f7f\u7528":35,"\u6211\u4eec\u5c06\u4f7f\u7528\u7b80\u5355\u7684":35,"\u6211\u4eec\u5c06\u4f7f\u7528cifar":54,"\u6211\u4eec\u5c06\u539f\u59cb\u6570\u636e\u7684\u6bcf\u4e00\u7ec4":32,"\u6211\u4eec\u5c06\u5728\u540e\u9762\u4ecb\u7ecd\u8bad\u7ec3\u548c\u9884\u6d4b\u6d41\u7a0b\u7684\u811a\u672c":57,"\u6211\u4eec\u5c06\u5b83\u4eec\u5212\u5206\u4e3a\u4e0d\u540c\u7684\u7c7b\u522b":42,"\u6211\u4eec\u5c06\u5bf9\u8fd9\u4e24\u4e2a\u6b65\u9aa4\u7ed9\u51fa\u4e86\u8be6\u7ec6\u7684\u89e3\u91ca":57,"\u6211\u4eec\u5c06\u653e\u7f6e\u4f9d\u8d56\u5e93":41,"\u6211\u4eec\u5c06\u8bc4\u5206\u5206\u6210\u4e24\u90e8\u5206":59,"\u6211\u4eec\u5c06\u9610\u91ca\u5982\u4f55\u5728\u96c6\u7fa4\u4e0a\u8fd0\u884c\u5206\u5e03\u5f0f":41,"\u6211\u4eec\u5c31\u53ef\u4ee5\u7740\u624b\u5bf9\u5206\u7c7b\u6a21\u578b\u8fdb\u884c\u8bad\u7ec3\u4e86":54,"\u6211\u4eec\u5c31\u53ef\u4ee5\u8bad\u7ec3\u6a21\u578b\u4e86":57,"\u6211\u4eec\u5c31\u53ef\u4ee5\u8fdb\u884c\u9884\u6d4b\u4e86":57,"\u6211\u4eec\u5c31\u5b8c\u6210\u4e86\u4e00\u6b21\u4ee3\u7801\u8d21\u732e\u7684\u8fc7\u7a0b":36,"\u6211\u4eec\u5c55\u793a\u5982\u4f55\u7528paddlepaddle\u89e3\u51b3":25,"\u6211\u4eec\u5df2\u7ecf\u5b9e\u73b0\u4e86\u5927\u591a\u6570\u5e38\u7528\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":35,"\u6211\u4eec\u5e0c\u671b\u80fd\u5728\u8fd9\u4e2a\u57fa\u7840\u4e0a\u4e0d\u65ad\u7684\u6539\u8fdb":0,"\u6211\u4eec\u5e0c\u671b\u80fd\u591f\u68c0\u9a8c\u6a21\u578b\u7684\u597d\u574f":25,"\u6211\u4eec\u5e94\u5f53\u4f1a\u5f97\u5230\u4e00\u4e2a\u540d\u4e3acifar":54,"\u6211\u4eec\u5efa\u8bae\u4f60\u4e3a\u4f60\u7684python\u5c01\u88c5\u5b9e\u73b0\u4e00\u4e2a":37,"\u6211\u4eec\u5efa\u8bae\u4f60\u5728\u5199\u65b0\u7f51\u7edc\u5c42\u65f6\u628a\u6d4b\u8bd5\u4ee3\u7801\u653e\u5165\u65b0\u7684\u6587\u4ef6\u4e2d":37,"\u6211\u4eec\u603b\u7ed3\u4e86\u5404\u4e2a\u7f51\u7edc\u7684\u590d\u6742\u5ea6\u548c\u6548\u679c":57,"\u6211\u4eec\u611f\u8c22":62,"\u6211\u4eec\u628apaddlepaddle\u7684\u7f16\u8bd1\u73af\u5883\u6253\u5305\u6210\u4e00\u4e2a\u955c\u50cf":27,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528":27,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":26,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528docker\u955c\u50cf\u6765\u90e8\u7f72\u73af\u5883":28,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u7f51\u7edc":61,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u8f6c\u6362\u65b9\u5f0f":19,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u6570\u636e\u9884\u5904\u7406\u811a\u672c":61,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u6784\u5efa\u811a\u672cbuild_doc":38,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u793a\u4f8b\u811a\u672c":55,"\u6211\u4eec\u63d0\u4f9b\u4e86\u56fd\u5185\u7684\u955c\u50cf\u670d\u52a1\u5668\u4f9b\u5927\u5bb6\u4f7f\u7528":27,"\u6211\u4eec\u63d0\u4f9b\u4e86\u811a\u672c\u6765\u6784\u5efa\u5b57\u5178\u548c\u9884\u5904\u7406\u6570\u6910":61,"\u6211\u4eec\u63d0\u4f9b\u4e86c":55,"\u6211\u4eec\u63d0\u4f9b\u4e86python\u5904\u7406\u6570\u636e\u7684\u63a5\u53e3":46,"\u6211\u4eec\u63d0\u4f9b\u53ef\u4ee5\u76f4\u63a5\u8fd0\u884cpaddlepaddl":27,"\u6211\u4eec\u662f\u5bf9\u6bcf\u4e00\u4e2a\u5b50\u5e8f\u5217\u53d6\u6700\u540e\u4e00\u4e2a\u5143\u7d20":32,"\u6211\u4eec\u6700\u7ec8\u7684\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165python\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u8bed\u8a00\u7684\u89e3\u91ca\u5668":20,"\u6211\u4eec\u6709\u4e00\u4e2a\u5e8f\u5217\u4f5c\u4e3a\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u72b6\u6001":35,"\u6211\u4eec\u7528":58,"\u6211\u4eec\u7528\u4ee5\u4e0b\u7684\u4e00\u4e9b":59,"\u6211\u4eec\u7528\u7f16\u53f7\u4f5c\u4e3akei":59,"\u6211\u4eec\u7528paddlepaddle\u89e3\u51b3\u4e86\u5355\u53d8\u91cf\u7ebf\u6027\u56de\u5f52\u95ee\u9898":25,"\u6211\u4eec\u7684\u5b57\u5178\u4f7f\u7528\u5185\u90e8\u7684\u5206\u8bcd\u5de5\u5177\u5bf9\u767e\u5ea6\u77e5\u9053\u548c\u767e\u5ea6\u767e\u79d1\u7684\u8bed\u6599\u8fdb\u884c\u5206\u8bcd\u540e\u4ea7\u751f":53,"\u6211\u4eec\u7684\u6807\u51c6\u5f00\u53d1\u6d41\u7a0b\u662f\u628a\u8fd9\u4e9b\u5de5\u5177\u90fd\u88c5\u8fdb\u4e00\u4e2adocker":36,"\u6211\u4eec\u7684\u8bad\u7ec3\u66f2\u7ebf\u5982\u4e0b":60,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u5c42rnn\u7684\u914d\u7f6e":32,"\u6211\u4eec\u770b\u4e00\u4e0b\u8bed\u4e49\u76f8\u540c\u7684\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":32,"\u6211\u4eec\u771f\u8bda\u5730\u611f\u8c22\u60a8\u7684\u5173\u6ce8":61,"\u6211\u4eec\u771f\u8bda\u5730\u611f\u8c22\u60a8\u7684\u8d21\u732e":36,"\u6211\u4eec\u79f0\u4e4b\u4e3a\u4e00\u4e2a0\u5c42\u7684\u5e8f\u5217":31,"\u6211\u4eec\u8fd8\u53ef\u4ee5\u767b\u5f55\u5230\u5bbf\u4e3b\u673a\u4e0a\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":48,"\u6211\u4eec\u8fd8\u5c06\u7f16\u7801\u5411\u91cf\u6295\u5c04\u5230":35,"\u6211\u4eec\u9009\u53d6\u5355\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e\u4e2d\u7684\u4e0d\u540c\u90e8\u5206":32,"\u6211\u4eec\u901a\u5e38\u5728\u6240\u6709\u8282\u70b9\u4e0a\u521b\u5efa\u4e00\u4e2a":41,"\u6211\u4eec\u901a\u5e38\u5c06\u4e00\u53e5\u8bdd\u7406\u89e3\u6210\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":32,"\u6211\u4eec\u901a\u8fc7\u8bfb\u53d6":49,"\u6211\u4eec\u9075\u5faa":62,"\u6211\u4eec\u90fd\u63d0\u4f9bpython\u7684\u8f6c\u6362\u5e93":19,"\u6211\u4eec\u91c7\u7528\u4e0a\u9762\u7684\u7279\u5f81\u6a21\u677f":60,"\u6211\u4eec\u91c7\u7528\u5355\u5c42lstm\u6a21\u578b":57,"\u6211\u4eec\u91c7\u7528\u7b2c\u4e8c\u4e2a\u65b9\u6cd5":27,"\u6211\u4eec\u9700\u8981\u5148\u521b\u5efa\u4e00\u4e2a\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":54,"\u6211\u4eec\u9700\u8981\u521b\u5efa\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":62,"\u6211\u4eec\u9700\u8981\u5236\u4f5c\u4e00\u4e2a\u5305\u542b\u8bad\u7ec3\u6570\u636e\u7684paddle\u955c\u50cf":48,"\u6211\u4eec\u9700\u8981\u5728\u96c6\u7fa4\u7684\u6240\u6709\u8282\u70b9\u4e0a\u5b89\u88c5":41,"\u6211\u4eec\u9700\u8981\u8ba1\u7b97":37,"\u6211\u4eec\u9700\u8981\u8bbe\u7f6e":59,"\u6211\u4eec\u9700\u8981\u9884\u5904\u7406\u6570\u6910\u5e76\u6784\u5efa\u4e00\u4e2a\u5b57\u5178":61,"\u6211\u4eec\u975e\u5e38\u6b22\u8fce\u60a8\u7528paddlepaddle\u6784\u5efa\u66f4\u597d\u7684\u793a\u4f8b":59,"\u6211\u4eec\u9884\u8bad\u7ec3\u5f97\u52304\u79cd\u4e0d\u540c\u7ef4\u5ea6\u7684\u8bcd\u5411\u91cf":53,"\u6211\u4eec\u9996\u5148\u5904\u7406\u7535\u5f71\u6216\u7528\u6237\u7684\u7279\u5f81\u6587\u4ef6":59,"\u6211\u4eec\u9ed8\u8ba4\u4f7f\u7528imdb\u7684\u6d4b\u8bd5\u6570\u636e\u96c6\u4f5c\u4e3a\u9a8c\u8bc1":61,"\u6211\u5220\u9664\u4e86":36,"\u6216":[3,40,46,54,60],"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"\u6216\u4e00\u4e2a\u5411\u91cf":34,"\u6216\u4e0d\u786e\u5b9a":58,"\u6216\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"\u6216\u662f\u624b\u52a8\u7f16\u8f91\u751f\u6210":59,"\u6216\u6700\u5927\u503c":31,"\u6216\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":2,"\u6216\u79f0pserver":46,"\u6216\u7b2c\u4e00\u4e2a":31,"\u6216\u7b2c\u4e00\u4e2a\u5143\u7d20":31,"\u6216\u8005":[20,21,24,27,29,31,32,40,46],"\u6216\u8005\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":34,"\u6216\u8005\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":[31,34],"\u6216\u8005\u4ece\u5de5\u5177\u7684\u754c\u9762\u91cc\u8fd0\u884c\u60a8\u7684\u5e94\u7528":40,"\u6216\u8005\u53cd\u5411\u5730\u4ece":35,"\u6216\u8005\u5728cpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2a\u7ebf\u7a0b":43,"\u6216\u8005\u5c06\u8fd9\u53f0\u8282\u70b9\u8fc1\u79fb\u5230\u53e6\u4e00\u4e2a\u8282\u70b9\u5e76\u542f\u52a8\u5373\u53ef\u6062\u590d\u8bad\u7ec3\u4efb\u52a1":18,"\u6216\u8005\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":42,"\u6216\u8005\u6570\u636e\u5e93\u8fde\u63a5\u8def\u5f84\u7b49":2,"\u6216\u8005\u6570\u7ec4\u7684\u6570\u7ec4\u8fd9\u4e2a\u6982\u5ff5":32,"\u6216\u8005\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":31,"\u6216\u8005\u662f\u4e00\u4e2a\u5c0f\u7684\u6587\u672c\u7247\u6bb5":61,"\u6216\u8005\u662f\u51fd\u6570\u8c03\u7528\u7684\u9891\u7387\u548c\u8017\u65f6\u7b49":40,"\u6216\u8005\u66f4\u65e9":24,"\u6216\u8005\u6bcf\u4e00\u4e2a\u7cfb\u5217\u91cc\u7684\u7279\u5f81\u6570\u636e":32,"\u6216\u8005\u7528tuple\u8868\u793a\u7684\u591a\u4e2a\u503c":19,"\u6216\u8005\u7531\u5b83\u4eec\u7ec4\u6210\u7684list":19,"\u6216\u8005\u76f4\u63a5\u4f7f\u7528\u4e0b\u9762\u7684shell\u547d\u4ee4":55,"\u6216\u8005\u76f4\u63a5\u6254\u6389\u975e\u5e38\u957f\u7684\u5e8f\u5217":24,"\u6216\u8005\u8f93\u5165\u6570\u636e\u5c3a\u5ea6\u8fc7\u5927":24,"\u6216\u8005\u91c7\u7528\u5e76\u884c\u8ba1\u7b97\u6765\u52a0\u901f\u67d0\u4e9b\u5c42\u7684\u66f4\u65b0":45,"\u6216\u8005\u9700\u8981\u53d1\u5e03\u60a8\u7684\u5e94\u7528\u7684\u955c\u50cf":27,"\u6216\u8005\u9700\u8981\u66f4\u9ad8\u7684\u6548\u7387":2,"\u6216\u8bbe\u7f6e\u4e3anone":2,"\u6216gpu":43,"\u6216gpu\u4e2a\u6570":61,"\u6218\u4e89\u7247":58,"\u623f":32,"\u623f\u95f4":32,"\u6240\u4ee5":[3,24,46],"\u6240\u4ee5\u4e00\u822c\u9700\u8981\u5bf9\u8bad\u7ec3\u7528\u7684\u6a21\u578b\u914d\u7f6e\u6587\u4ef6\u7a0d\u4f5c\u76f8\u5e94\u4fee\u6539\u624d\u80fd\u5728\u9884\u6d4b\u65f6\u4f7f\u7528":5,"\u6240\u4ee5\u4f60\u53ea\u7528\u6309\u4e0b\u9762\u7684\u7ed3\u6784\u6765\u7ec4\u7ec7\u6570\u6910\u5c31\u884c\u4e86":61,"\u6240\u4ee5\u505a\u6cd5\u53ef\u4ee5\u6709\u4e24\u79cd":24,"\u6240\u4ee5\u53ef\u4ee5\u5229\u7528\u5982\u4e0b\u65b9\u6cd5\u8bfb\u53d6\u6a21\u578b\u7684\u53c2\u6570":25,"\u6240\u4ee5\u53ef\u4ee5\u7b80\u5316\u5bf9\u73af\u5883\u7684\u8981\u6c42":48,"\u6240\u4ee5\u5728\u5199\u5165\u5feb\u7167\u7684\u8fc7\u7a0b\u4e2d":18,"\u6240\u4ee5\u5916\u5c42\u8f93\u51fa\u7684\u5e8f\u5217\u5f62\u72b6":32,"\u6240\u4ee5\u5982\u679c\u9700\u8981\u81ea\u884c\u914d\u7f6e\u5f00\u53d1\u73af\u5883\u9700\u8981\u8003\u8651\u7248\u672c\u7684\u56e0\u7d20":27,"\u6240\u4ee5\u5b83\u4eec\u4f7f\u7528\u540c\u4e00\u4e2aip\u5730\u5740":47,"\u6240\u4ee5\u5bf9":32,"\u6240\u4ee5\u5f88\u591a\u65f6\u5019\u4f60\u9700\u8981\u505a\u7684\u53ea\u662f\u5b9a\u4e49\u6b63\u786e\u7684\u7f51\u7edc\u5c42\u5e76\u628a\u5b83\u4eec\u8fde\u63a5\u8d77\u6765":25,"\u6240\u4ee5\u6027\u80fd\u4e5f\u5c31\u9010\u6b65\u53d8\u6210\u4e86\u6df1\u5ea6\u5b66\u4e60\u9886\u57df\u6700\u91cd\u8981\u7684\u6307\u6807":40,"\u6240\u4ee5\u6211\u4eec\u4f7f\u7528\u8fd9\u4e2a\u955c\u50cf\u6765\u4e0b\u8f7d\u8bad\u7ec3\u6570\u636e\u5230docker":48,"\u6240\u4ee5\u6211\u4eec\u53ef\u4ee5\u5728\u8fd9\u4e2a\u57fa\u7840\u4e0a":49,"\u6240\u4ee5\u6211\u4eec\u63a8\u8350\u4f7f\u7528\u57fa\u4e8edocker\u6765\u6784\u5efapaddlepaddle\u7684\u6587\u6863":38,"\u6240\u4ee5\u6211\u4eec\u9700\u8981\u5c06\u8f93\u5165\u6570\u636e\u6807\u8bb0\u6210":32,"\u6240\u4ee5\u63a8\u8350\u4f7f\u7528\u663e\u5f0f\u6307\u5b9a\u7684\u65b9\u5f0f\u6765\u8bbe\u7f6einput_typ":3,"\u6240\u4ee5\u653e\u4e00\u4e2a\u7a7a\u5217\u8868":25,"\u6240\u4ee5\u8bad\u7ec3":41,"\u6240\u4ee5\u8f93\u51fa\u7684value\u5305\u542b\u4e24\u4e2a\u5411\u91cf":5,"\u6240\u4ee5\u8fd9\u4e00\u6b65\u662f\u5fc5\u8981\u7684":37,"\u6240\u5bf9\u5e94\u7684\u8bcd\u8868index\u6570\u7ec4":32,"\u6240\u6709\u4e0e\u7c7b\u578b\u76f8\u5173\u7684\u51fd\u6570":21,"\u6240\u6709\u4ee3\u7801\u5fc5\u987b\u5177\u6709\u5355\u5143\u6d4b\u8bd5":36,"\u6240\u6709\u53c2\u6570\u7f6e\u4e3a\u96f6":43,"\u6240\u6709\u540c\u76ee\u5f55\u4e0b\u7684\u6587\u672c\u5b9e\u4f8b\u6587\u4ef6\u90fd\u662f\u540c\u7ea7\u522b\u7684":61,"\u6240\u6709\u547d\u4ee4\u884c\u9009\u9879\u53ef\u4ee5\u8bbe\u7f6e\u4e3a":41,"\u6240\u6709\u6587\u4ef6\u5217\u8868":3,"\u6240\u6709\u672c\u5730\u8bad\u7ec3":41,"\u6240\u6709\u6807\u8bb0\u7684\u6d4b\u8bd5\u96c6\u548c\u8bad\u7ec3\u96c6":61,"\u6240\u6709\u751f\u6210\u7684\u7684\u6587\u4ef6\u90fd\u5728build\u76ee\u5f55\u4e0b":27,"\u6240\u6709\u7684":[36,37],"\u6240\u6709\u7684\u4eba\u53e3\u7edf\u8ba1\u5b66\u4fe1\u606f\u7531\u7528\u6237\u81ea\u613f\u63d0\u4f9b":58,"\u6240\u6709\u7684\u5355\u6d4b\u90fd\u4f1a\u88ab\u6267\u884c\u4e00\u6b21":37,"\u6240\u6709\u7684\u63a5\u53e3\u5747\u4e3ac\u63a5\u53e3":21,"\u6240\u6709\u7684\u64cd\u4f5c\u90fd\u662f\u9488\u5bf9\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u8fdb\u884c\u7684":32,"\u6240\u6709\u7684\u7528\u6237\u4fe1\u606f\u90fd\u5305\u542b\u5728":58,"\u6240\u6709\u7684\u7535\u5f71\u4fe1\u606f\u90fd\u5305\u542b\u5728":58,"\u6240\u6709\u7684\u8bc4\u5206\u6570\u636e\u90fd\u5305\u542b\u5728":58,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u4f7f\u7528":37,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u5728":37,"\u6240\u6709\u7c7b\u578b\u540d\u4e3a":21,"\u6240\u6709\u7f51\u7edc\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u90fd\u4f4d\u4e8e":37,"\u6240\u6709\u8282\u70b9\u8fd0\u884c\u96c6\u7fa4\u4f5c\u4e1a\u7684\u4e3b\u673a\u540d\u6216":41,"\u6240\u6709\u8d21\u732e\u8005":0,"\u6240\u6709\u8f93\u5165\u5e8f\u5217\u5e94\u8be5\u6709\u76f8\u540c\u7684\u957f\u5ea6":35,"\u6240\u6709\u914d\u7f6e\u90fd\u80fd\u5728":57,"\u6240\u6784\u5efa\u7f51\u7edc\u7ed3\u6784\u7684\u7684\u6df1\u5ea6\u6bd4\u4e4b\u524d\u4f7f\u7528\u7684\u7f51\u7edc\u6709\u5927\u5e45\u5ea6\u7684\u63d0\u9ad8":55,"\u6240\u793a":60,"\u6240\u8c13\u65f6\u95f4\u6b65\u4fe1\u606f":3,"\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":20,"\u624b\u5de5\u827a\u8005":58,"\u624d\u4f1a\u91ca\u653e\u8be5\u6bb5\u5185\u5b58":3,"\u624d\u4f1astop":3,"\u624d\u80fd\u4fdd\u8bc1\u548c\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":32,"\u624d\u80fd\u5145\u5206\u5229\u7528mac":27,"\u624d\u80fd\u53d1\u6325\u5176\u5168\u90e8\u80fd\u529b":40,"\u6253\u5370\u5728\u5c4f\u5e55\u4e0a":59,"\u6253\u5370\u7684\u65e5\u5fd7\u53d8\u591a":26,"\u6253\u5f00":40,"\u6253\u5f00\u6587\u672c\u6587\u4ef6":3,"\u6253\u5f00\u6d4f\u89c8\u5668\u8bbf\u95ee\u5bf9\u5e94\u76ee\u5f55\u4e0b\u7684index":38,"\u6253\u5f00\u8fd9\u4e2a\u7f16\u8bd1\u9009\u9879":21,"\u6267\u884c":[29,60,61],"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5c31\u53ef\u4ee5\u9884\u5904\u7406\u6570\u6910":61,"\u6267\u884c\u4ee5\u4e0b\u64cd\u4f5c":35,"\u6267\u884c\u60a8\u7684\u4ee3\u7801":40,"\u6267\u884c\u7684\u547d\u4ee4\u5982\u4e0b":55,"\u6269\u5c55\u548c\u5ef6\u4f38":0,"\u6269\u5c55\u673a\u5236\u7b49\u529f\u80fd":47,"\u6279\u6b21\u540e\u6253\u5370\u65e5\u5fd7":59,"\u6279\u6b21\u5bf9\u5e73\u5747\u53c2\u6570\u8fdb\u884c\u6d4b\u8bd5":60,"\u6279\u6b21\u7684\u6570\u636e":59,"\u627e\u5230":35,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u539f\u56e0":40,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u90e8\u5206":40,"\u6280\u672f\u5458":58,"\u628a":[19,37],"\u628a\u4e4b\u524d\u793a\u4f8b\u4e2d\u8f6c\u6362\u5b8c\u6bd5\u7684random":19,"\u628a\u7528\u6237\u5728\u8d2d\u7269\u7f51\u7ad9":61,"\u628a\u7f16\u8bd1\u51fa\u6765\u7684paddlepaddle\u4e5f\u6253\u5305\u6210\u4e00\u4e2a\u955c":27,"\u628a\u8bad\u7ec3\u6570\u636e\u76f4\u63a5\u653e\u5728":48,"\u6293\u53d6\u4ea7\u54c1\u7684\u7528\u6237\u8bc4\u8bba\u5e76\u5206\u6790\u4ed6\u4eec\u7684\u60c5\u611f":61,"\u6295\u5c04\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b\u5230":35,"\u6295\u5c04\u7f16\u7801\u5411\u91cf\u5230":35,"\u62a5\u9519":29,"\u62bd\u53d6\u51fa\u7684\u65b0\u8bcd\u8868\u7684\u4fdd\u5b58\u8def\u5f84":53,"\u62bd\u53d6\u5bf9\u5e94\u7684\u8bcd\u5411\u91cf\u6784\u6210\u65b0\u7684\u8bcd\u8868":53,"\u62c6\u5206\u5230\u4e0d\u540c\u6587\u4ef6\u5939":62,"\u62c6\u89e3":34,"\u62c6\u89e3\u6210\u7684\u6bcf\u4e00\u53e5\u8bdd\u518d\u901a\u8fc7\u4e00\u4e2alstm\u7f51\u7edc":32,"\u62f7\u8d1d\u8bad\u7ec3\u6587\u4ef6\u5230\u5bb9\u5668\u5185":49,"\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5411\u91cf":57,"\u6301\u7eed\u96c6\u6210\u7cfb\u7edf\u4e5f\u53ef\u4ee5\u7b80\u5355\u5730\u590d\u73b0\u8fd9\u4e2a\u8fc7\u7a0b":27,"\u6307\u5411\u4e00\u4e2alayer":34,"\u6307\u5b9a":[24,34,35,46],"\u6307\u5b9a\u4e00\u53f0\u673a\u5668\u4e0a\u4f7f\u7528\u7684\u7ebf\u7a0b\u6570":43,"\u6307\u5b9a\u4e86dataprovider\u7684\u6587\u4ef6\u540d\u548c\u8fd4\u56de\u6570\u636e\u7684\u51fd\u6570\u540d":46,"\u6307\u5b9a\u4ee5\u592a\u7f51\u7c7b\u578b\u4e3atcp\u7f51\u7edc":46,"\u6307\u5b9a\u4f7f\u75282":24,"\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b\u8def\u5f84":57,"\u6307\u5b9a\u52a0\u8f7d\u7684\u65b9\u5f0f":43,"\u6307\u5b9a\u5de5\u4f5c\u6a21\u578b\u8fdb\u884c\u9884\u6d4b":55,"\u6307\u5b9a\u5de5\u4f5c\u6a21\u5f0f\u6765\u63d0\u53d6\u7279\u5f81":55,"\u6307\u5b9a\u63d0\u53d6\u7279\u5f81\u7684\u5c42":55,"\u6307\u5b9a\u662f\u5426\u4f7f\u7528gpu":55,"\u6307\u5b9a\u751f\u6210\u6570\u636e\u7684\u51fd\u6570":57,"\u6307\u5b9a\u7684\u5b57\u5178\u5355\u8bcd\u6570":62,"\u6307\u5b9a\u7684\u6570\u636e\u5c06\u4f1a\u88ab\u6d4b\u8bd5":57,"\u6307\u5b9a\u7684\u8f93\u5165\u4e0d\u4f1a\u88ab":34,"\u6307\u5b9a\u7f51\u7edc\u63a5\u53e3\u540d\u5b57\u4e3aeth0":46,"\u6307\u5b9a\u8bad\u7ec3\u6570\u636e\u548c\u6d4b\u8bd5\u6570\u636e":57,"\u6307\u5b9abatch":62,"\u6307\u5b9acudnn\u7684\u6700\u5927\u5de5\u4f5c\u7a7a\u95f4\u5bb9\u9650":43,"\u6307\u5bf9\u4e8e\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u8f93\u5165\u6570\u636e":32,"\u6307\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u4e4b\u540e\u5f97\u5230\u7684\u6240\u6709\u53c2\u6570":18,"\u6307\u793a\u4f7f\u7528\u54ea\u4e2agpu\u6838":43,"\u6307\u793a\u5728\u7b80\u5355\u7684recurrentlayer\u5c42\u7684\u8ba1\u7b97\u4e2d\u662f\u5426\u4f7f\u7528\u6279\u5904\u7406\u65b9\u6cd5":43,"\u6307\u793a\u5f53\u6307\u5b9a\u8f6e\u7684\u6d4b\u8bd5\u6a21\u578b\u4e0d\u5b58\u5728\u65f6":43,"\u6307\u793a\u662f\u5426\u4f7f\u7528\u591a\u7ebf\u7a0b\u6765\u8ba1\u7b97\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc":43,"\u6307\u793a\u662f\u5426\u5f00\u542f\u53c2\u6570\u670d\u52a1\u5668":43,"\u6307\u793a\u662f\u5426\u663e\u793a\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u65e5\u5fd7\u7ec6\u8282":43,"\u6307\u793a\u662f\u5426\u68c0\u67e5\u6240\u6709\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u7a00\u758f\u53c2\u6570\u7684\u5206\u5e03\u662f\u5747\u5300\u7684":43,"\u6307\u793a\u6d4b\u8bd5\u4efb\u52a1":60,"\u6307\u793a\u6d4b\u8bd5\u4efb\u52a1\u7684\u6807\u8bb0":60,"\u6309\u542f\u53d1\u5f0f\u635f\u5931\u7684\u5927\u5c0f\u9012\u589e\u6392\u5e8f":43,"\u6309\u7167\u4e0b\u9762\u6b65\u9aa4\u5373\u53ef":49,"\u6309\u94ae":36,"\u633a":32,"\u633a\u597d":32,"\u6355\u83b7":57,"\u635f\u5931\u51fd\u6570":46,"\u635f\u5931\u51fd\u6570\u5373\u4e3a\u7f51\u7edc\u7684\u4f18\u5316\u76ee\u6807":46,"\u635f\u5931\u51fd\u6570\u548c\u8bc4\u4f30\u5668":46,"\u6362":32,"\u6392\u6210\u4e00\u5217\u7684\u591a\u4e2a\u5143\u7d20":31,"\u63a5\u4e0b\u6765":[57,61],"\u63a5\u4e0b\u6765\u53ef\u4ee5\u8003\u8651\u4e0b\u65f6\u95f4\u7ebf\u7684\u5206\u6790":40,"\u63a5\u4e0b\u6765\u5c31\u53ef\u4ee5\u4f7f\u7528":40,"\u63a5\u4e0b\u6765\u6211\u4eec\u521b\u5efa\u4e00\u4e2a\u539f\u59cb":36,"\u63a5\u4e0b\u6765\u6211\u4eec\u53d6\u6d88\u5bf9":36,"\u63a5\u4e0b\u6765\u6211\u4eec\u5c06\u5c55\u793a\u5982\u4f55\u7528paddlepaddle\u8bad\u7ec3\u4e00\u4e2a\u6587\u672c\u5206\u7c7b\u6a21\u578b":57,"\u63a5\u4e0b\u6765\u7b49\u5f85":36,"\u63a5\u53d7":60,"\u63a5\u53d7\u7684\u4e1c\u897f":60,"\u63a5\u53d7\u8005":60,"\u63a5\u53e3":[20,21],"\u63a5\u53e3\u540d\u79f0":41,"\u63a5\u53e3\u5c42\u505a\u8fc7\u591a\u5c01\u88c5":21,"\u63a5\u53e3\u63d0\u53d6\u7684\u7ed3\u679c\u662f\u4e00\u81f4\u7684":55,"\u63a5\u53e3\u662f":19,"\u63a5\u53e3\u6709\u4e00\u4e2a":24,"\u63a5\u53e3\u6765\u52a0\u8f7d\u6570\u636e":57,"\u63a5\u53e3\u6765\u52a0\u8f7d\u8be5\u6587\u4ef6":55,"\u63a5\u53e3\u6765\u6253\u5f00\u8be5\u6587\u4ef6":55,"\u63a5\u53e3\u8bbe\u7f6e\u795e\u7ecf\u7f51\u7edc\u6240\u4f7f\u7528\u7684\u8bad\u7ec3\u53c2\u6570\u548c":46,"\u63a5\u7740\u6211\u4eec\u5c31\u80fd\u591f\u6253\u5f00\u6d4f\u89c8\u5668\u5728":27,"\u63a7\u5236":43,"\u63a7\u5236\u5982\u4f55\u6539\u53d8\u6a21\u578b\u53c2\u6570":25,"\u63a7\u5236\u662f\u5426\u751f\u6210\u5355\u5143\u6d4b\u8bd5":27,"\u63a7\u5236\u751f\u6210\u7684\u751f\u4ea7\u955c\u50cf\u662f\u5426\u652f\u6301avx":27,"\u63a7\u5236\u751f\u6210\u7684\u751f\u4ea7\u955c\u50cf\u662f\u5426\u652f\u6301gpu":27,"\u63a8\u5bfc\u8be5\u5c42\u524d\u5411\u548c\u540e\u5411\u4f20\u9012\u7684\u65b9\u7a0b":37,"\u63a8\u8350":32,"\u63a8\u8350\u4f7f\u7528":3,"\u63a8\u8350\u6e05\u7406\u6574\u4e2a\u7f16\u8bd1\u76ee\u5f55":26,"\u63a8\u8350\u76f4\u63a5\u5b58\u653e\u5230\u8bad\u7ec3\u76ee\u5f55":2,"\u63a8\u8350\u7cfb\u7edf":41,"\u63a8\u9001\u5230\u8fdc\u7a0b\u4ed3\u5e93":36,"\u63a8\u9500\u5458":58,"\u63cf\u8ff0":26,"\u63cf\u8ff0\u7f51\u7edc\u7ed3\u6784\u548c\u4f18\u5316\u7b97\u6cd5":57,"\u63cf\u8ff0\u95ee\u9898":36,"\u63cf\u8ff0kubernetes\u4e0a\u8fd0\u884c\u7684\u4f5c\u4e1a":47,"\u63d0\u4ea4\u65b9\u5f0f\u53ef\u53c2\u89c1":38,"\u63d0\u4f9b":41,"\u63d0\u4f9b\u4e86\u4e00\u4e2a\u542f\u52a8\u811a\u672c":49,"\u63d0\u4f9b\u4e86\u547d\u4ee4\u6837\u4f8b\u6765\u8fd0\u884c":41,"\u63d0\u4f9b\u4e86\u81ea\u52a8\u5316\u811a\u672c\u6765\u542f\u52a8\u4e0d\u540c\u8282\u70b9\u4e2d\u7684\u6240\u6709":41,"\u63d0\u4f9b\u51e0\u4e4e\u6240\u6709\u8bad\u7ec3\u7684\u5185\u90e8\u8f93\u51fa\u65e5\u5fd7":41,"\u63d0\u4f9b\u6269\u5c55\u7684\u957f\u5ea6\u4fe1\u606f":31,"\u63d0\u4f9b\u6700\u65b0":27,"\u63d0\u4f9b\u7ed9\u7528\u6237\u672c\u5730\u8f6c\u6362\u7684\u5e93":19,"\u63d0\u4f9b\u7ed9paddle\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e":19,"\u63d0\u4f9b\u8bad\u7ec3\u8fc7\u7a0b\u7684":41,"\u63d0\u51fa\u7684\u4ee3\u7801\u9700\u6c42":53,"\u63d0\u793a":24,"\u64cd\u4f5c":[32,46],"\u64cd\u6301\u5bb6\u52a1\u8005":58,"\u652f\u6301\u4e3b\u6d41x86\u5904\u7406\u5668\u5e73\u53f0":29,"\u652f\u6301\u5355\u673a\u6a21\u5f0f\u548c\u591a\u673a\u6a21\u5f0f":46,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":[33,34],"\u652f\u6301nvidia":29,"\u652f\u6301rbd":47,"\u653e\u5728\u8fd9\u4e2a\u76ee\u5f55\u91cc\u7684\u6587\u4ef6\u5176\u5b9e\u662f\u4fdd\u5b58\u5230\u4e86mfs\u4e0a":49,"\u653e\u5fc3":32,"\u6545\u800c\u662f\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":32,"\u6548\u679c\u603b\u7ed3":57,"\u6559\u7a0b\u6587\u6863\u5230":41,"\u6559\u80b2\u5de5\u4f5c\u8005":58,"\u6570":[34,60],"\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":34,"\u6570\u636e":62,"\u6570\u636e\u4e0b\u8f7d\u4e4b\u540e":54,"\u6570\u636e\u4e2d0":24,"\u6570\u636e\u5217\u8868":55,"\u6570\u636e\u5c06\u4fdd\u5b58\u5728":53,"\u6570\u636e\u5c42":[25,59],"\u6570\u636e\u5e94\u8be5\u5728\u542f\u52a8\u96c6\u7fa4\u4f5c\u4e1a\u4e4b\u524d\u51c6\u5907\u597d":41,"\u6570\u636e\u63d0\u4f9b\u5668":42,"\u6570\u636e\u63d0\u4f9b\u811a\u672c\u4ec5\u4ec5\u662f\u8bfb\u53d6meta":59,"\u6570\u636e\u63d0\u4f9b\u811a\u672c\u7684\u7ec6\u8282\u6587\u6863\u53ef\u4ee5\u53c2\u8003":59,"\u6570\u636e\u7684\u6574\u6570id":35,"\u6570\u636e\u76ee\u5f55\u4e2d\u7684\u6240\u6709\u6587\u4ef6\u88ab":41,"\u6570\u636e\u7c7b\u578b":5,"\u6570\u636e\u7f13\u5b58\u7684\u7b56\u7565":3,"\u6570\u636e\u8bbf\u95ee":1,"\u6570\u636e\u8bfb\u53d6\u5747\u4ea4\u7531\u5176\u4ed6\u8bed\u8a00\u5b8c\u6210":20,"\u6570\u636e\u8bfb\u53d6\u7a0b\u5e8f\u5f80\u5f80\u5b9a\u4e49\u5728\u4e00\u4e2a\u5355\u72ecpython\u811a\u672c\u6587\u4ef6\u91cc":46,"\u6570\u636e\u8f93\u5165":34,"\u6570\u636e\u8f93\u5165\u683c\u5f0f":3,"\u6570\u636e\u96c6":58,"\u6570\u636e\u96c6\u63cf\u8ff0":59,"\u6570\u636e\u96c6\u6587\u4ef6\u5939\u540d\u79f0":62,"\u6570\u636e\u9884\u5904\u7406\u4efb\u52a1":19,"\u6570\u636e\u9884\u5904\u7406\u5b8c\u6210\u4e4b\u540e":57,"\u6570\u636e\u9884\u6d4b":60,"\u6570\u6910\u5b9a\u4e49":61,"\u6570\u6910\u8bf4\u660e\u6587\u6863":61,"\u6570\u6910\u96c6\u548c":61,"\u6570\u76ee":45,"\u6574\u4f53":32,"\u6574\u4f53\u6570\u636e\u548c\u539f\u59cb\u6570\u636e\u5b8c\u5168\u4e00\u6837":32,"\u6574\u4f53\u7684\u7ed3\u6784\u56fe\u5982\u4e0b":49,"\u6574\u6570":37,"\u6574\u6570\u6807\u7b7e":3,"\u6574\u6d01":32,"\u6587\u4e66\u5de5\u4f5c":58,"\u6587\u4ef6":[20,36,48,60],"\u6587\u4ef6\u4e2d":[49,55,58,60],"\u6587\u4ef6\u4e2d\u6307\u5b9a\u6a21\u578b\u8def\u5f84\u548c\u8f93\u51fa\u7684\u76ee\u5f55":55,"\u6587\u4ef6\u4e2d\u6307\u5b9a\u8981\u63d0\u53d6\u7279\u5f81\u7684\u7f51\u7edc\u5c42\u7684\u540d\u5b57":55,"\u6587\u4ef6\u4e2d\u7684":55,"\u6587\u4ef6\u4e2d\u7684\u6bcf\u884c\u90fd\u5fc5\u987b\u662f\u4e00\u4e2a\u53e5\u5b50":62,"\u6587\u4ef6\u4e3a":[24,62],"\u6587\u4ef6\u4e4b\u5916":36,"\u6587\u4ef6\u4e5f\u53ef\u4ee5\u7528\u4e8e\u5bf9\u6837\u672c\u8fdb\u884c\u9884\u6d4b":55,"\u6587\u4ef6\u5185\u5bb9\u4e3a":20,"\u6587\u4ef6\u5206\u5272\u4e3a\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6587\u4ef6":59,"\u6587\u4ef6\u540d\u4e3a\u6b64uuid":18,"\u6587\u4ef6\u540d\u79f0\u4e3a":59,"\u6587\u4ef6\u59390":49,"\u6587\u4ef6\u5939\u4e2d\u7684\u6bcf\u4e2a\u6587\u4ef6\u7684\u6bcf\u4e00\u884c\u5305\u542b\u4e24\u90e8\u5206":62,"\u6587\u4ef6\u5f00\u5934":46,"\u6587\u4ef6\u7684\u5206\u9694\u7b26\u4e3a":59,"\u6587\u4ef6\u7684\u6539\u53d8":36,"\u6587\u4ef6\u7684\u683c\u5f0f\u53ef\u4ee5":59,"\u6587\u4ef6\u7a0d\u6709\u5dee\u522b":54,"\u6587\u4ef6\u7d22\u5f15":41,"\u6587\u4ef6\u7ed9\u51fa\u4e86\u5b8c\u6574\u4f8b\u5b50":57,"\u6587\u4ef6\u9700\u8981\u9884\u5148\u88ab\u8f6c\u6362\u6210paddlepaddle\u96c6\u7fa4\u5185\u90e8\u7684\u5b58\u50a8\u683c\u5f0f":19,"\u6587\u4ef6model":45,"\u6587\u5b57\u7684\u4ea4\u4e92\u5f0f\u6587\u6863":27,"\u6587\u672c\u4e2d\u7684\u5355\u8bcd\u7528\u7a7a\u683c\u5206\u9694":57,"\u6587\u672c\u4fe1\u606f\u5c31\u662f\u4e00\u4e2a\u5e8f\u5217\u6570\u636e":3,"\u6587\u672c\u5206\u7c7b\u95ee\u9898":57,"\u6587\u672c\u5377\u79ef\u5206\u53ef\u4e3a\u4e09\u4e2a\u6b65\u9aa4":57,"\u6587\u672c\u5377\u79ef\u91c7\u6837\u5c42":59,"\u6587\u672c\u6295\u5f71\u5c42":59,"\u6587\u672c\u683c\u5f0f\u7684\u5b9e\u4f8b\u6587\u4ef6":61,"\u6587\u6863":[24,46],"\u6587\u6863\u7f16\u5199\u7b49\u5de5\u4f5c":27,"\u6587\u6863\u81ea\u52a8\u5206\u7c7b\u548c\u95ee\u7b54":60,"\u6587\u6863\u90fd\u662f\u901a\u8fc7":38,"\u6587\u7ae0":49,"\u65b0":32,"\u65b0\u5efa\u4e00\u4e2a\u6743\u91cd":37,"\u65b0\u624b\u5165\u95e8":52,"\u65b0\u624b\u5165\u95e8\u7ae0\u8282":23,"\u65b9\u4fbf":32,"\u65b9\u4fbf\u4eca\u540e\u7684\u5d4c\u5165\u5f0f\u79fb\u690d\u5de5\u4f5c":26,"\u65b9\u4fbf\u5f00\u53d1\u8005\u76f4\u63a5\u767b\u5f55\u5230\u955c\u50cf\u4e2d\u8fdb\u884c\u5f00\u53d1":27,"\u65b9\u4fbf\u6d4b\u8bd5\u4eba\u5458\u6d4b\u8bd5paddle\u7684\u884c\u4e3a":23,"\u65b9\u5f0f1":24,"\u65b9\u5f0f2":24,"\u65b9\u6848\u6765\u6807\u8bb0\u6bcf\u4e2a\u53c2\u6570":60,"\u65b9\u6cd5\u4e00":45,"\u65b9\u6cd5\u4e09":45,"\u65b9\u6cd5\u4e8c":45,"\u65c1\u8fb9":32,"\u65c5\u6e38\u7f51\u7ad9":61,"\u65e0":32,"\u65e0\u4e1a\u4eba\u58eb":58,"\u65e0\u5ef6\u8fdf":43,"\u65e0\u6cd5\u505a\u5230\u5bf9\u4e8e\u5404\u79cd\u8bed\u8a00\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u7684\u9002\u914d":20,"\u65e0\u8bba\u5728\u672c\u5730\u8fd8\u662f\u5728\u4e91\u7aef":19,"\u65e0\u8bba\u662f\u5728\u672c\u5730\u6216\u662f\u4e91\u7aef\u8f6c\u6362":19,"\u65e5\u5fd7\u5c06\u4fdd\u5b58\u5728":61,"\u65e8\u5728\u5efa\u7acb\u4e00\u4e2a\u53ef\u4ee5\u88ab\u534f\u540c\u8c03\u81f3\u6700\u4f18\u7ffb\u8bd1\u6548\u679c\u7684\u5355\u795e\u7ecf\u5143\u7f51\u7edc":62,"\u65e9\u9910":32,"\u65f6":[18,24,31,35,37,43,49],"\u65f6\u5019":32,"\u65f6\u52a0\u4e0a":61,"\u65f6\u5e8f\u6a21\u578b\u5747\u4f7f\u7528\u8be5\u811a\u672c":57,"\u65f6\u5e8f\u6a21\u578b\u662f\u6307\u6570\u636e\u7684\u67d0\u4e00\u7ef4\u5ea6\u662f\u4e00\u4e2a\u5e8f\u5217\u5f62\u5f0f":3,"\u65f6\u76ee\u6807\u8bed\u8a00\u7684\u6587\u4ef6":62,"\u65f6\u88ab\u8bad\u7ec3\u7684":37,"\u65f6\u8bbe\u5907id\u53f7\u7684\u5206\u914d":45,"\u65f6\u95f4":32,"\u65f6\u95f4\u6233":58,"\u65f6\u95f4\u6233\u8868\u793a\u4e3a\u4ece1970":58,"\u65f6\u95f4\u6b65\u7684\u6982\u5ff5":32,"\u6620\u5c04\u5230\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":37,"\u662f":[26,32,46],"\u662f\u4e00\u4e2a\u51681\u7684\u5411\u91cf":37,"\u662f\u4e00\u4e2a\u5185\u7f6e\u7684\u5b9a\u65f6\u5668\u5c01\u88c5":40,"\u662f\u4e00\u4e2a\u52a8\u6001\u7a0b\u5e8f\u5206\u6790\u7684\u672f\u8bed":40,"\u662f\u4e00\u4e2a\u5305\u88c5\u6570\u636e\u7684":60,"\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u662f\u4e00\u4e2a\u53cc\u5c42\u7684\u5e8f\u5217":31,"\u662f\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3\u7684\u4ee3\u7801\u751f\u6210\u5668":20,"\u662f\u4e00\u4e2a\u5c01\u88c5\u5bf9\u8c61":40,"\u662f\u4e00\u4e2a\u5f88\u6709\u7528\u7684\u53c2\u6570":45,"\u662f\u4e00\u4e2a\u7b26\u5408\u9ad8\u65af\u5206\u5e03\u7684\u968f\u673a\u53d8\u91cf":25,"\u662f\u4e00\u4e2a\u7c7b\u578b\u7684\u6807\u5fd7":21,"\u662f\u4e00\u4e2a\u7ec4\u5408\u5c42":46,"\u662f\u4e00\u4e2a\u7edf\u8ba1\u5b66\u7684\u673a\u5668\u7ffb\u8bd1\u7cfb\u7edf":62,"\u662f\u4e00\u4e2a\u914d\u7f6e\u6587\u4ef6\u7684\u4f8b\u5b50":61,"\u662f\u4e00\u4e2a\u975e\u7ebf\u6027\u7684":37,"\u662f\u4e00\u4e2apython\u7684":3,"\u662f\u4e00\u4e2aunbound":34,"\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":3,"\u662f\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":34,"\u662f\u4e00\u7ec4":47,"\u662f\u4e0d\u5305\u62ec\u6e90\u7801\u7684":48,"\u662f\u4e0d\u5e38\u89c1\u7684\u505a\u6cd5":20,"\u662f\u4e0d\u662f\u5f88\u7b80\u5355\u5462":3,"\u662f\u4e0d\u662f\u8981\u5bf9\u6570\u636e\u505ashuffl":3,"\u662f\u4e3b\u8981\u7684\u53ef\u6267\u884cpython\u811a\u672c":60,"\u662f\u4ec0\u4e48\u4e5f\u6ca1\u5173\u7cfb":3,"\u662f\u4f17\u591a\u8bef\u5dee\u4ee3\u4ef7\u51fd\u6570\u5c42\u7684\u4e00\u79cd":25,"\u662f\u4f7f\u5f97\u8981\u5171\u4eab\u7684\u53c2\u6570\u4f7f\u7528\u540c\u6837\u7684":24,"\u662f\u504f\u5dee":35,"\u662f\u5176\u5927\u5c0f":25,"\u662f\u51e0\u4e4e\u4e0d\u5360\u5185\u5b58\u7684":3,"\u662f\u539f\u59cb\u6cd5\u8bed\u6587\u4ef6":62,"\u662f\u5404\u4e2a\u5b9e\u73b0\u4e2d\u5171\u4eab\u7684\u5934\u6587\u4ef6":21,"\u662f\u5411\u91cf":37,"\u662f\u5426\u4ee5\u9006\u5e8f\u5904\u7406\u8f93\u5165\u5e8f\u5217":34,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":26,"\u662f\u5426\u4f7f\u7528\u65e7\u7684remoteparameterupdat":43,"\u662f\u5426\u4f7f\u7528\u6743\u91cd":37,"\u662f\u5426\u4f7f\u7528gpu":59,"\u662f\u5426\u4f7f\u7528gpu\u8bad\u7ec3":62,"\u662f\u5426\u5141\u8bb8\u6682\u5b58\u7565\u5fae\u591a\u4f59pool_size\u7684\u6570\u636e":3,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":26,"\u662f\u5426\u5c06\u5168\u5c40\u79cd\u5b50\u5e94\u7528\u4e8e\u672c\u5730\u7ebf\u7a0b\u7684\u968f\u673a\u6570":43,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":26,"\u662f\u5426\u5f00\u542f\u8ba1\u65f6\u529f\u80fd":26,"\u662f\u5426\u5f00\u542frdma":26,"\u662f\u5426\u6253\u5370\u7248\u672c\u4fe1\u606f":43,"\u662f\u5426\u652f\u6301gpu":26,"\u662f\u5426\u663e\u793a":43,"\u662f\u5426\u7a00\u758f":37,"\u662f\u5426\u7f16\u8bd1\u4e2d\u82f1\u6587\u6587\u6863":26,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236\u6587\u4ef6":26,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":26,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":26,"\u662f\u5426\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":26,"\u662f\u5426\u9700\u8981\u7b49\u5f85\u8be5\u8f6e\u6a21\u578b\u53c2\u6570":43,"\u662f\u56e0\u4e3ac99\u652f\u6301":20,"\u662f\u56e0\u4e3apaddle\u7684\u7f51\u7edc\u901a\u4fe1\u4e2d":46,"\u662f\u56e0\u4e3apaddlepaddle\u914d\u7f6e\u6587\u4ef6\u4e0ec":46,"\u662f\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u52a0\u8f7d\u5b57\u5178\u5e76\u5b9a\u4e49\u6570\u636e\u63d0\u4f9b\u7a0b\u5e8f\u6a21\u5757\u548c\u7f51\u7edc\u67b6\u6784\u7684\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":60,"\u662f\u5728paddlepaddle\u4e2d\u6784\u9020\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u6700\u91cd\u8981\u7684\u6982\u5ff5":35,"\u662f\u57fa\u7840\u7684\u8ba1\u7b97\u5355\u5143":25,"\u662f\u5b58\u6709\u4e00\u7cfb\u5217\u53d8\u6362\u77e9\u9635\u7684\u6743\u91cd":37,"\u662f\u5b58\u6709\u504f\u7f6e\u5411\u91cf\u7684\u6743\u91cd":37,"\u662f\u5e8f\u5217":59,"\u662f\u5f85\u6269\u5c55\u7684\u6570\u636e":31,"\u662f\u6211\u4eec":36,"\u662f\u6307":21,"\u662f\u6307\u4e00\u4e2a\u6570\u636e\u5217\u8868\u6587\u4ef6":46,"\u662f\u6307\u4e00\u7cfb\u5217\u7684\u7279\u5f81\u6570\u636e":32,"\u662f\u6307recurrent_group\u7684\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":32,"\u662f\u6570\u636e\u8f93\u5165":35,"\u662f\u6574\u4e2a\u7684\u8bcd\u5d4c\u5165":59,"\u662f\u6709\u610f\u4e49\u7684":32,"\u662f\u6784\u5efa\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u6700\u91cd\u8981\u7684\u5de5\u5177":35,"\u662f\u6a21\u578b\u53c2\u6570\u4f18\u5316\u7684\u76ee\u6807\u51fd\u6570":25,"\u662f\u6d45\u5c42\u8bed\u4e49\u89e3\u6790\u7684\u4e00\u79cd\u5f62\u5f0f":60,"\u662f\u6e90\u8bed\u8a00\u7684\u6587\u4ef6":62,"\u662f\u7528\u6237\u4f7f\u7528c":21,"\u662f\u76ee\u6807\u82f1\u8bed\u6587\u4ef6":62,"\u662f\u77e9\u9635":37,"\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u5165\u53e3":25,"\u662f\u7f51\u7edc\u548c\u6570\u636e\u914d\u7f6e\u6587\u4ef6":54,"\u662f\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u6807\u8bc6\u7b26":37,"\u662f\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26":37,"\u662f\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":37,"\u662f\u7f51\u7edc\u5c42\u8f93\u51fa\u7684\u5927\u5c0f":37,"\u662f\u8be5\u5c42\u7684\u6807\u8bc6\u7b26":37,"\u662f\u8be5\u5c42\u7684\u7c7b\u540d":37,"\u662f\u8be5\u7f51\u7edc\u5c42\u7684":37,"\u662f\u8f93\u5165":35,"\u662f\u901a\u7528\u7269\u4f53\u5206\u7c7b\u9886\u57df\u4e00\u4e2a\u4f17\u6240\u5468\u77e5\u7684\u6570\u636e\u5e93":55,"\u662f\u9700\u8981\u4e86\u89e3\u54ea\u4e9b\u6b65\u9aa4\u62d6\u6162\u4e86\u6574\u4f53":40,"\u662fc":21,"\u662fdecoder\u7684\u6570\u636e\u8f93\u5165":34,"\u662fgoogle\u5f00\u6e90\u7684\u5bb9\u5668\u96c6\u7fa4\u7ba1\u7406\u7cfb\u7edf":47,"\u662fnvidia\u6027\u80fd\u5206\u6790\u5de5\u5177":40,"\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":34,"\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":57,"\u662fpod\u5185\u7684\u5bb9\u5668\u90fd\u53ef\u4ee5\u8bbf\u95ee\u7684\u5171\u4eab\u76ee\u5f55":47,"\u662fpython\u5c01\u88c5\u7684\u7c7b\u540d":37,"\u662frnn\u72b6\u6001":35,"\u663e":57,"\u663e\u5f0f\u6307\u5b9a\u8fd4\u56de\u7684\u662f\u4e00\u4e2a28":3,"\u665a":32,"\u666e\u901a\u7528\u6237\u8bf7\u8d70\u5b89\u88c5\u6d41\u7a0b":28,"\u6682\u4e0d\u8003\u8651\u5728\u5185":24,"\u6682\u65f6\u4e0d\u8003\u8651\u591a\u4e2aparamet":18,"\u66b4\u9732\u8fd9\u4e2a\u6982\u5ff5\u5fc5\u8981\u51fd\u6570":21,"\u66f4\u591a\u5173\u4e8edocker\u7684\u5b89\u88c5\u4e0e\u4f7f\u7528":24,"\u66f4\u591a\u5185\u5bb9\u53ef\u67e5\u770b\u53c2\u8003\u6587\u732e":61,"\u66f4\u591a\u7684\u7ec6\u8282\u53ef\u4ee5\u5728\u6587\u732e\u4e2d\u627e\u5230":61,"\u66f4\u597d\u5730\u5b8c\u6210\u4e00\u4e9b\u590d\u6742\u7684\u8bed\u8a00\u7406\u89e3\u4efb\u52a1":34,"\u66f4\u5feb":35,"\u66f4\u65b0":24,"\u66f4\u65b0\u6a21\u5f0f":24,"\u66f4\u65b9\u4fbf\u7684\u8bbe\u7f6e\u65b9\u5f0f":24,"\u66f4\u8be6\u7ec6\u6570\u636e\u683c\u5f0f\u548c\u7528\u4f8b\u8bf7\u53c2\u8003":57,"\u66f4\u8be6\u7ec6\u7684\u4f7f\u7528":46,"\u66f4\u8be6\u7ec6\u7684\u7f51\u7edc\u914d\u7f6e\u8fde\u63a5\u8bf7\u53c2\u8003":57,"\u66f4\u8be6\u7ec6\u7684\u8bf4\u660e":57,"\u66f4\u8fdb\u4e00\u6b65":34,"\u66f4\u9ad8":35,"\u66ff\u6211\u4eec\u5b8c\u6210\u4e86\u539f\u59cb\u8f93\u5165\u6570\u636e\u7684\u62c6\u5206":34,"\u6700":32,"\u6700\u4e0d\u540c\u7684\u7279\u8272\u662f\u5b83\u5e76\u6ca1\u6709\u5c06\u8f93\u5165\u8bed\u53e5\u7f16\u7801\u4e3a\u4e00\u4e2a\u5355\u72ec\u7684\u5b9a\u957f\u5411\u91cf":62,"\u6700\u4e3b\u8981\u7684\u5de5\u4f5c\u5c31\u662f\u89e3\u6790\u51fa":49,"\u6700\u4f73\u63a8\u8350":3,"\u6700\u540e":[3,36,37,41,57,61],"\u6700\u540e\u4e00\u4e2a":31,"\u6700\u540e\u4e00\u90e8\u5206\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u914d\u7f6e":25,"\u6700\u540e\u5220\u9664":23,"\u6700\u540e\u6211\u4eec\u4f7f\u7528\u94fe\u5f0f\u6cd5\u5219\u8ba1\u7b97":37,"\u6700\u597d\u7684\u6a21\u578b\u662f":61,"\u6700\u5c11\u663e\u793a\u591a\u5c11\u4e2a\u8282\u70b9":43,"\u6700\u5e38\u89c1\u7684\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662fexcept":20,"\u6700\u65b0\u7684\u4ee3\u7801":36,"\u6700\u65b0log":61,"\u6700\u7ec8":37,"\u6700\u7ec8\u5b9e\u73b0\u4e00\u4e2a\u5c42\u6b21\u5316\u7684\u590d\u6742rnn":34,"\u6700\u7ec8\u751f\u6210\u540d\u4e3apaddl":27,"\u6700\u7ec8\u7684\u8f93\u51fa\u7ed3\u679c":34,"\u6700\u7ec8\u8d8b\u4e8e\u63a5\u8fd1":25,"\u6708\u6e56":32,"\u6709":32,"\u6709\u4e00\u4e2a\u57fa\u672c\u7684\u8ba4\u8bc6":47,"\u6709\u4e24\u4e2a\u7279\u6b8a\u6807\u8bc6":62,"\u6709\u4e86meta\u914d\u7f6e\u6587\u4ef6\u4e4b\u540e":59,"\u6709\u4e9b\u5c42\u53ef\u80fd\u9700\u8981\u9ad8\u7cbe\u5ea6\u6765\u4fdd\u8bc1\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u6b63\u786e\u6267\u884c":37,"\u6709\u4e9b\u5c42\u6216\u8005\u6fc0\u6d3b\u9700\u8981\u505a\u5f52\u4e00\u5316\u4ee5\u4fdd\u8bc1\u5b83\u4eec\u7684\u8f93\u51fa\u7684\u548c\u662f\u4e00\u4e2a\u5e38\u6570":37,"\u6709\u4e9b\u7279\u5f81\u7684\u53d6\u503c\u8fbe\u5230\u6570\u767e\u4e07":24,"\u6709\u4e9b\u7535\u5f71id\u53ef\u80fd\u4e0e\u5b9e\u9645\u7535\u5f71\u4e0d\u76f8\u7b26\u5408":58,"\u6709\u5173":32,"\u6709\u5173\u5982\u4f55\u7f16\u5199\u6570\u636e\u63d0\u4f9b\u7a0b\u5e8f\u7684\u66f4\u591a\u7ec6\u8282\u63cf\u8ff0":35,"\u6709\u5173kubernetes\u76f8\u5173\u6982\u5ff5\u4ee5\u53ca\u5982\u4f55\u642d\u5efa\u548c\u914d\u7f6ekubernetes\u96c6\u7fa4":49,"\u6709\u52a9\u4e8e\u7406\u89e3\u7528\u6237\u5bf9\u4e0d\u540c\u516c\u53f8":61,"\u6709\u52a9\u4e8e\u8bca\u65ad\u5206\u5e03\u5f0f\u9519\u8bef":41,"\u6709\u56db\u4e2a\u8bad\u7ec3\u8fdb\u7a0b":46,"\u6709\u65f6\u79f0\u4e3a":61,"\u6709\u6807\u51c6\u7684":20,"\u6709\u7684\u65f6\u5019":20,"\u6709\u7684\u65f6\u5019\u7b80\u7b80\u5355\u5355\u7684\u6539\u53d8\u5c31\u80fd\u5728\u6027\u80fd\u4e0a\u4ea7\u751f\u660e\u663e\u7684\u4f18\u5316\u6548\u679c":40,"\u670d\u52a1":32,"\u670d\u52a1\u5458":32,"\u672a\u5305\u542b\u5728\u5b57\u5178\u4e2d\u7684\u5355\u8bcd":62,"\u672a\u6807\u8bb0\u7684\u8bc4\u4ef7\u6837\u672c":61,"\u672a\u77e5\u8bcd":53,"\u672c\u4f8b\u4e2d\u4e3a0":53,"\u672c\u4f8b\u4e2d\u4e3a32":53,"\u672c\u4f8b\u4e2d\u4e3a4":53,"\u672c\u4f8b\u4e2d\u4f7f\u7528for\u5faa\u73af\u8fdb\u884c\u591a\u6b21\u8c03\u7528":3,"\u672c\u4f8b\u4e2d\u7684\u539f\u59cb\u6570\u636e\u4e00\u5171\u670910\u4e2a\u6837\u672c":32,"\u672c\u4f8b\u4e2d\u7684\u8f93\u5165\u7279\u5f81\u662f\u8bcdid\u7684\u5e8f\u5217":3,"\u672c\u4f8b\u6839\u636e\u7f51\u7edc\u914d\u7f6e\u4e2d":3,"\u672c\u4f8b\u6bcf\u884c\u4fdd\u5b58\u4e00\u6761\u6837\u672c":57,"\u672c\u4f8b\u7531\u6613\u5230\u96be\u5c55\u793a4\u79cd\u4e0d\u540c\u7684\u6587\u672c\u5206\u7c7b\u7f51\u7edc\u914d\u7f6e":57,"\u672c\u4f8b\u7684":3,"\u672c\u4f8b\u7684\u6240\u6709\u5b57\u7b26\u90fd\u5c06\u8f6c\u6362\u4e3a\u8fde\u7eed\u6574\u6570\u8868\u793a\u7684id\u4f20\u7ed9\u6a21\u578b":57,"\u672c\u4f8b\u91c7\u7528\u82f1\u6587\u60c5\u611f\u5206\u7c7b\u7684\u6570\u636e":3,"\u672c\u4f8b\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":57,"\u672c\u5217\u8868\u8bf4\u660epaddle\u53d1\u7248\u4e4b\u524d\u9700\u8981\u6d4b\u8bd5\u7684\u529f\u80fd\u70b9":23,"\u672c\u5730\u6d4b\u8bd5":42,"\u672c\u5730\u8bad\u7ec3":42,"\u672c\u5730\u8bad\u7ec3\u7684\u5b9e\u9a8c":45,"\u672c\u5b9e\u4f8b\u4e2d":53,"\u672c\u5c0f\u8282\u6211\u4eec\u5c06\u4ecb\u7ecd\u6a21\u578b\u7f51\u7edc\u7ed3\u6784":57,"\u672c\u5c42\u5c3a\u5bf8":55,"\u672c\u5c42\u6709\u56db\u4e2a\u53c2\u6570":55,"\u672c\u6559\u7a0b\u4e2d\u6211\u4eec\u7ed9\u51fa\u4e86\u4e09\u4e2aresnet\u6a21\u578b":55,"\u672c\u6559\u7a0b\u5c06\u4ecb\u7ecd\u4f7f\u7528\u6df1\u5ea6\u53cc\u5411\u957f\u77ed\u671f\u8bb0\u5fc6":60,"\u672c\u6559\u7a0b\u5c06\u6307\u5bfc\u4f60\u5982\u4f55\u5728":35,"\u672c\u6559\u7a0b\u5c06\u6307\u5bfc\u60a8\u5b8c\u6210\u957f\u671f\u77ed\u671f\u8bb0\u5fc6":61,"\u672c\u6559\u7a0b\u63d0\u4f9b\u4e86\u4e00\u4e2a\u7528\u4e8eimagenet\u4e0a\u7684\u5377\u79ef\u5206\u7c7b\u7f51\u7edc\u6a21\u578b":55,"\u672c\u6587\u4e2d\u6240\u6709\u7684\u4f8b\u5b50":32,"\u672c\u6587\u4e2d\u7531\u4e8e\u8f93\u5165\u6570\u636e\u662f\u968f\u673a\u751f\u6210\u7684\u4e0d\u9700\u8981\u8bfb\u8f93\u5165\u6587\u4ef6":25,"\u672c\u6587\u4e2d\u793a\u4f8b\u6240\u4f7f\u7528\u7684\u5355\u5143\u6d4b\u8bd5\u6587\u4ef6\u662f":32,"\u672c\u6587\u4ee5paddlepaddle\u7684\u53cc\u5c42rnn\u5355\u5143\u6d4b\u8bd5\u4e3a\u793a\u4f8b":32,"\u672c\u6587\u53ea\u4f7f\u7528\u4e86\u9ed8\u8ba4\u547d\u540d\u7a7a\u95f4":47,"\u672c\u6587\u5c06\u4ecb\u7ecd\u5728kubernetes\u5bb9\u5668\u7ba1\u7406\u5e73\u53f0\u4e0a\u5feb\u901f\u6784\u5efapaddlepaddle\u5bb9\u5668\u96c6\u7fa4":49,"\u672c\u6587\u6863\u4ecb\u7ecd\u5982\u4f55\u5728paddlepaddle\u5e73\u53f0\u4e0a":53,"\u672c\u6587\u6863\u5185\u4e0d\u91cd\u590d\u4ecb\u7ecd":47,"\u672c\u6587\u6863\u63cf\u8ff0paddl":21,"\u672c\u6587\u9996\u5148\u4ecb\u7ecdtrainer\u8fdb\u7a0b\u4e2d\u7684\u4e00\u4e9b\u4f7f\u7528\u6982\u5ff5":46,"\u672c\u6765":32,"\u672c\u6b21\u8bad\u7ec3\u6587\u4ef6\u6240\u5728\u76ee\u5f55":49,"\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6\u53ef\u4ee5\u5199\u6210":49,"\u672c\u6b21\u8bad\u7ec3\u8981\u6c42\u67093\u4e2apaddlepaddle\u8282\u70b9":49,"\u672c\u6b21\u8bd5\u9a8c":57,"\u672c\u793a\u4f8b\u4e2d\u4f7f\u7528\u7684\u539f\u59cb\u6570\u636e\u5982\u4e0b":32,"\u672c\u793a\u4f8b\u610f\u56fe\u4f7f\u7528\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":32,"\u672c\u793a\u4f8b\u7684\u9884\u6d4b\u7ed3\u679c":61,"\u672c\u7bc7\u6559\u7a0b\u5728paddlepaddle\u4e2d\u91cd\u73b0\u4e86\u8fd9\u4e00\u826f\u597d\u7684\u8bad\u7ec3\u7ed3\u679c":62,"\u672c\u7bc7\u6559\u7a0b\u5c06\u4f1a\u6307\u5bfc\u4f60\u901a\u8fc7\u8bad\u7ec3\u4e00\u4e2a":62,"\u672c\u8d28\u4e0a\u4e0e\u8bad\u7ec3\u6a21\u578b\u4e00\u6837":62,"\u673a\u5668\u7684\u8bbe\u5907":45,"\u673a\u5668\u7ffb\u8bd1":[23,56],"\u6743\u91cd\u66f4\u65b0\u7684\u68af\u5ea6":43,"\u6761\u4ef6\u4e0b":47,"\u6765":32,"\u6765\u4ee3\u66ff":36,"\u6765\u4fdd\u8bc1\u8bad\u7ec3\u8fc7\u7a0b\u53ef\u4ee5\u4ece\u4e2d\u95f4\u72b6\u6001\u91cd\u65b0\u542f\u52a8":18,"\u6765\u505a\u68af\u5ea6\u68c0\u67e5":37,"\u6765\u505ableu\u8bc4\u4f30":62,"\u6765\u505c\u6b62\u8bad\u7ec3":59,"\u6765\u5206\u6790\u6267\u884c\u6587\u4ef6":40,"\u6765\u5206\u79bb\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6587\u4ef6":59,"\u6765\u5206\u9694\u6bcf\u4e00\u884c":59,"\u6765\u521d\u59cb\u5316\u53c2\u6570":24,"\u6765\u5b89\u88c5":41,"\u6765\u5b9a\u4e49\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":35,"\u6765\u5bf9\u6bd4\u5206\u6790\u4e24\u8005\u8bed\u4e49\u76f8\u540c\u7684\u539f\u56e0":32,"\u6765\u5e2e\u52a9\u4f60\u7406\u89e3paddlepaddle\u7684\u5185\u90e8\u8fd0\u884c\u673a\u5236":57,"\u6765\u5f00\u542f\u672c\u5730\u7684\u8bad\u7ec3":61,"\u6765\u5f97\u5230\u67d0\u4e2a\u7279\u5b9a\u53c2\u6570\u7684\u68af\u5ea6\u77e9\u9635":37,"\u6765\u6307\u5b9a\u7f51\u7edc\u5c42\u7684\u6570\u76ee":55,"\u6765\u63a5\u53d7\u4e0d\u4f7f\u7528\u7684\u51fd\u6570\u4ee5\u4fdd\u8bc1\u517c\u5bb9\u6027":3,"\u6765\u6ce8\u518c\u8be5\u5c42":37,"\u6765\u6df7\u5408\u4f7f\u7528gpu\u548ccpu\u8ba1\u7b97\u7f51\u7edc\u5c42\u7684\u53c2\u6570":45,"\u6765\u751f\u6210\u5e8f\u5217":62,"\u6765\u7684\u79d2\u6570":58,"\u6765\u786e\u4fdd\u628a":20,"\u6765\u786e\u5b9a\u5bf9\u5e94\u5173\u7cfb":3,"\u6765\u81ea\u5b9a\u4e49\u4f20\u6570\u636e\u7684\u8fc7\u7a0b":2,"\u6765\u83b7\u5f97\u8f93\u51fa\u7684\u68af\u5ea6":37,"\u6765\u8868\u793a":35,"\u6765\u8868\u793a\u53c2\u6570\u4f4d\u7f6e":60,"\u6765\u8868\u793a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":35,"\u6765\u8868\u793apaddle\u5185\u90e8\u7c7b":20,"\u6765\u8ba1\u7b97\u68af\u5ea6":37,"\u6765\u8bb2\u89e3\u5982\u4f55\u4f7f\u7528\u53cc\u5c42rnn":32,"\u6765\u8bbe\u7f6e":24,"\u6765\u8bf4\u660epydataprovider2\u7684\u7b80\u5355\u4f7f\u7528\u573a\u666f":3,"\u6765\u8c03\u6574":36,"\u6765\u8fd0\u884c":41,"\u6765\u8fd0\u884c\u6027\u80fd\u5206\u6790\u548c\u8c03\u4f18":40,"\u6765\u8fd0\u884c\u955c\u50cf":27,"\u6765\u8fdb\u884c\u8ba8\u8bba":21,"\u6765\u9884\u6d4b\u8fd9\u4e2a\u4e2d\u95f4\u7684\u8bcd":24,"\u676f\u5b50":32,"\u6784\u6210\u4e86\u8f93\u51fa\u53cc\u5c42\u5e8f\u5217\u7684\u7b2ci\u4e2a":31,"\u6784\u9020":49,"\u6784\u9020paddl":5,"\u67b6\u6784\u5bf9\u591a\u4e2a\u8282\u70b9\u7684":46,"\u67b6\u6784\u6765\u8bad\u7ec3\u60c5\u611f\u5206\u6790\u6a21\u578b":61,"\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u4e00\u4e2a\u8f93\u5165\u4e3a\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7f51\u7edc\u4e2d\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u8f93\u51fa":32,"\u67d0\u4e9b\u53c2\u6570\u53ea\u53ef\u7528\u4e8e\u7279\u5b9a\u7684\u5c42\u4e2d":42,"\u67e5\u770b":[36,57],"\u67e5\u770b\u5b89\u88c5\u540e\u7684paddl":29,"\u67e5\u770b\u5f53\u524d\u72b6\u6001":36,"\u67e5\u770b\u5f53\u524d\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":36,"\u67e5\u770b\u6587\u4ef6\u5177\u4f53\u88ab\u4fee\u6539\u7684\u5185\u5bb9":36,"\u67e5\u770bjob\u7684\u8be6\u7ec6\u60c5\u51b5":48,"\u6807\u51c6\u5dee\u4e3a":24,"\u6807\u51c6\u8868\u793apaddle\u7248\u672c\u53f7":23,"\u6807\u51c6lstm\u4ee5\u6b63\u5411\u5904\u7406\u8be5\u5e8f\u5217":60,"\u6807\u793a\u56fe\u7247\u662f\u5f69\u8272\u56fe\u6216\u7070\u5ea6\u56fe":54,"\u6807\u793a\u662f\u5426\u4e3a\u5f69\u8272\u56fe\u7247":54,"\u6807\u7b7e0\u8868\u793a\u8d1f\u9762\u7684\u8bc4\u8bba":61,"\u6807\u7b7e1\u8868\u793a\u6b63\u9762\u7684\u8bc4\u8bba":61,"\u6807\u7b7e\u4e0b\u627e\u5230\u6700\u65b0\u7684paddle\u955c\u50cf\u7248\u672c":27,"\u6807\u7b7e\u6587\u4ef6":60,"\u6807\u7b7e\u65b9\u6848\u6765\u81ea":60,"\u6807\u8bb0":46,"\u6807\u8bb0\u7f51\u7edc\u8f93\u51fa\u7684\u51fd\u6570\u4e3a":46,"\u6807\u8bc6\u662f\u5426\u4e3a\u8fde\u7eed\u7684batch\u8ba1\u7b97":43,"\u6839\u636e\u4f60\u7684\u4efb\u52a1":45,"\u6839\u636e\u524d\u6587\u7684\u63cf\u8ff0":49,"\u6839\u636e\u5728\u6a21\u578b\u914d\u7f6e\u6587\u4ef6\u4e2d\u4f7f\u7528\u7684\u540d\u4e3a":41,"\u6839\u636e\u6570\u636e\u91cf\u89c4\u6a21":58,"\u6839\u636e\u7528\u6237\u6307\u5b9a\u7684\u5b57\u5178":53,"\u6839\u636e\u7d22\u5f15\u77e9\u9635\u548c\u5b57\u5178\u6253\u5370\u6587\u672c":35,"\u6839\u636e\u7f51\u7edc\u914d\u7f6e\u4e2d\u7684":43,"\u6839\u636e\u8fd9\u4e9b\u53c2\u6570\u7684\u4f7f\u7528\u573a\u5408":42,"\u6839\u636e\u9ed8\u8ba4\u503c\u9012\u589e":43,"\u6839\u636e\u9ed8\u8ba4\u7aef\u53e3\u53f7\u9012\u589e":43,"\u6839\u636ecpu":27,"\u6839\u636ejob\u5bf9\u5e94\u7684pod\u4fe1\u606f":48,"\u683c\u5f0f":43,"\u683c\u5f0f\u5982\u4e0b":57,"\u683c\u5f0f\u8bf4\u660e":53,"\u68af\u5ea6\u4f1a\u5c31\u5730":37,"\u68af\u5ea6\u53c2\u6570\u7684\u5206\u5757\u6570\u76ee":43,"\u68af\u5ea6\u5c31\u53ef\u4ee5\u901a\u8fc7\u8fd9\u4e2a\u65b9\u7a0b\u8ba1\u7b97\u5f97\u5230":37,"\u68af\u5ea6\u670d\u52a1\u5668\u7684\u6570\u91cf":43,"\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5\u901a\u8fc7\u6709\u9650\u5dee\u5206\u6cd5\u6765\u9a8c\u8bc1\u4e00\u4e2a\u5c42\u7684\u68af\u5ea6":37,"\u68af\u5ea6\u68c0\u67e5\u7684\u8f93\u5165\u6570\u636e\u7684\u6279\u6b21\u5927\u5c0f":37,"\u68c0\u67e5\u70b9\u4fdd\u5b58\u7a0b\u5e8f\u6d41\u7a0b":18,"\u68d2":57,"\u697c\u5c42":32,"\u6a21\u5757":54,"\u6a21\u5757\u4e2d\u7684":3,"\u6a21\u5757\u63a5\u7ba1\u4e86shuffl":46,"\u6a21\u5757\u901a\u4fe1\u7684\u6700\u57fa\u7840\u534f\u8bae\u662fprotobuf":46,"\u6a21\u578b":60,"\u6a21\u578b\u4e00\u76f4\u4e0d\u6536\u655b":24,"\u6a21\u578b\u4fdd\u5b58\u5728\u76ee\u5f55":61,"\u6a21\u578b\u5171\u5305\u542b1":53,"\u6a21\u578b\u5217\u8868\u6587\u4ef6":60,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\u901a\u8fc7\u5b9a\u671f\u5411\u78c1\u76d8\u4e0a\u4fdd\u5b58\u4e00\u4efd\u5b58\u50a8\u5728paramet":18,"\u6a21\u578b\u53ca\u53c2\u6570\u4f1a\u88ab\u4fdd\u5b58\u5728\u8def\u5f84":54,"\u6a21\u578b\u5b58\u50a8\u8def\u5f84":57,"\u6a21\u578b\u5c06\u4fdd\u5b58\u5728\u76ee\u5f55":60,"\u6a21\u578b\u5c31\u8bad\u7ec3\u6210\u529f\u4e86":62,"\u6a21\u578b\u6570\u636e\u68c0\u67e5\u70b9\u7684\u5b9e\u73b0":18,"\u6a21\u578b\u6587\u4ef6\u5c06\u88ab\u5199\u5165\u8282\u70b9":41,"\u6a21\u578b\u6765\u5c06\u6cd5\u8bed\u7ffb\u8bd1\u6210\u82f1\u8bed":62,"\u6a21\u578b\u6765\u6307\u5bfc\u4f60\u5b8c\u6210\u8fd9\u4e9b\u6b65\u9aa4":35,"\u6a21\u578b\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":35,"\u6a21\u578b\u7684\u4ee3\u7801\u53ef\u4ee5\u5728":35,"\u6a21\u578b\u7684\u7ed3\u6784\u548c\u8bad\u7ec3\u8fc7\u7a0b":53,"\u6a21\u578b\u7684\u7f16\u7801\u5668\u90e8\u5206\u5982\u4e0b\u6240\u793a":35,"\u6a21\u578b\u88ab\u4fdd\u5b58\u5728":59,"\u6a21\u578b\u8bad\u7ec3\u4f1a\u770b\u5230\u7c7b\u4f3c\u4e0a\u9762\u8fd9\u6837\u7684\u65e5\u5fd7\u4fe1\u606f":57,"\u6a21\u578b\u8bad\u7ec3\u548c\u6700\u540e\u7684\u7ed3\u679c\u9a8c\u8bc1":25,"\u6a21\u578b\u8def\u5f84":[55,60],"\u6a21\u578b\u8f93\u51fa\u8def\u5f84":60,"\u6a21\u578b\u914d\u7f6e":[1,46],"\u6a21\u578b\u914d\u7f6e\u89e3\u6790":20,"\u6a21\u578b\u91c7\u7528":53,"\u6a21\u578b\u9884\u6d4b":5,"\u6b21":32,"\u6b22\u8fce\u901a\u8fc7":36,"\u6b63\u5219\u65b9\u6cd5\u7b49":46,"\u6b63\u6837\u672c":57,"\u6b63\u786e\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":24,"\u6b63\u8d1f\u5bf9\u9a8c\u8bc1":42,"\u6b63\u9762\u7684\u8bc4\u8bba\u7684\u5f97\u5927\u4e8e\u7b49\u4e8e7":61,"\u6b63\u9762\u8bc4\u4ef7\u6837\u672c":61,"\u6b64\u5904":53,"\u6b64\u5904\u90fd\u4e3a2":32,"\u6b64\u6559\u7a0b\u5c06\u5411\u60a8\u5206\u6b65\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528\u5185\u7f6e\u7684\u5b9a\u65f6\u5de5\u5177":40,"\u6b64\u6570\u636e\u96c6\u5305\u542b\u7535\u5f71\u8bc4\u8bba\u53ca\u5176\u76f8\u5173\u8054\u7684\u7c7b\u522b\u6807\u7b7e":61,"\u6b64\u65f6\u60a8\u53ef\u4ee5\u8fd0\u884c\u8fd9\u4e2a\u547d\u4ee4\u5728\u5f00\u53d1\u673a\u4e0a\u8fdb\u884c\u6d4b\u8bd5\u8fd0\u884c":27,"\u6b64\u65f6master\u5c06\u8d1f\u8d23\u542f\u52a8\u4e00\u4e2a\u65b0\u7684train":18,"\u6bb5\u843d\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u53cc\u5c42\u7684\u5e8f\u5217":34,"\u6bcf100\u4e2abatch\u6253\u5370\u4e00\u6b21\u7edf\u8ba1\u4fe1\u606f":61,"\u6bcf100\u4e2abatch\u663e\u793a\u53c2\u6570\u7edf\u8ba1":60,"\u6bcf20\u4e2abatch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":61,"\u6bcf20\u4e2abatch\u8f93\u51fa\u65e5\u5fd7":60,"\u6bcf\u4e00\u4e2a":23,"\u6bcf\u4e00\u4e2a\u4efb\u52a1\u6d41\u7a0b\u90fd\u53ef\u4ee5\u88ab\u5212\u5206\u4e3a\u5982\u4e0b\u4e94\u4e2a\u6b65\u9aa4":57,"\u6bcf\u4e00\u4e2a\u6587\u4ef6\u662f\u6570\u636e\u96c6\u7684\u4e00\u4e2ashard":19,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65":32,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u95f4\u7684\u795e\u7ecf\u7f51\u7edc\u5177\u6709\u4e00\u5b9a\u7684\u76f8\u5173\u6027":32,"\u6bcf\u4e00\u4e2a\u6d4b\u8bd5\u5468\u671f\u6d4b\u8bd5\u4e00\u6b21\u6240\u6709\u6570\u636e":59,"\u6bcf\u4e00\u4e2a\u8282\u70b9\u90fd\u6709\u76f8\u540c\u7684\u65e5\u5fd7\u7ed3\u6784":41,"\u6bcf\u4e00\u4e2akey\u7531":59,"\u6bcf\u4e00\u7ec4\u5185\u7684\u6240\u6709\u53e5\u5b50\u548clabel":32,"\u6bcf\u4e00\u884c\u8868\u793a\u4e00\u4e2a\u5b9e\u4f8b":61,"\u6bcf\u4e2a":[35,41,60],"\u6bcf\u4e2a\u503c\u7684\u7c7b\u578b\u53ef\u4ee5\u662f\u6574\u5f62":19,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u6bcf\u4e2a\u5355\u5c42rnn":34,"\u6bcf\u4e2a\u5355\u8bcd\u7684\u9884\u6d4b\u9519\u8bef\u7387":62,"\u6bcf\u4e2a\u53e5\u5b50\u53c8\u662f\u5355\u8bcd\u7684\u6570\u7ec4":32,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u5f00\u59cb\u6807\u8bb0\u5f00\u5934":35,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u7ed3\u675f\u6807\u8bb0\u7ed3\u5c3e":35,"\u6bcf\u4e2a\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u4e00\u81f4":32,"\u6bcf\u4e2a\u5b50\u6587\u4ef6\u5939\u4e0b\u5b58\u50a8\u76f8\u5e94\u5206\u7c7b\u7684\u56fe\u7247":54,"\u6bcf\u4e2a\u5b57\u5178\u5305\u542b\u603b\u517130000\u4e2a\u5355\u8bcd":62,"\u6bcf\u4e2a\u5b57\u5178\u90fd\u6709dictsize\u4e2a\u5355\u8bcd":62,"\u6bcf\u4e2a\u5c42\u5728\u5176":37,"\u6bcf\u4e2a\u5c42\u90fd\u6709\u4e00\u4e2a\u6216\u591a\u4e2ainput":57,"\u6bcf\u4e2a\u6279\u6b21\u6570\u636e":43,"\u6bcf\u4e2a\u6574\u6570\u5217\u8868\u88ab\u89c6\u4e3a\u4e00\u4e2a\u6574\u6570\u5e8f\u5217":35,"\u6bcf\u4e2a\u6587\u4ef6\u5939\u90fd\u5305\u542b\u6cd5\u8bed\u5230\u82f1\u8bed\u7684\u5e73\u884c\u8bed\u6599\u5e93":62,"\u6bcf\u4e2a\u6587\u4ef6\u662f\u4e00\u4e2a\u7535\u5f71\u8bc4\u8bba":61,"\u6bcf\u4e2a\u6587\u672c\u6587\u4ef6\u5305\u542b\u4e00\u4e2a\u6216\u8005\u591a\u4e2a\u5b9e\u4f8b":61,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":34,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u90fd\u7528\u4e86\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u7ed3\u679c":32,"\u6bcf\u4e2a\u6743\u91cd\u5bf9\u5e94\u4e00\u4e2a\u8f93\u5165":37,"\u6bcf\u4e2a\u6837\u672c\u7531\u4e24\u90e8\u5206\u7ec4\u6210":32,"\u6bcf\u4e2a\u6837\u672c\u95f4\u7528\u7a7a\u884c\u5206\u5f00":32,"\u6bcf\u4e2a\u6d4b\u8bd5\u5468\u671f\u6d4b\u8bd5":59,"\u6bcf\u4e2a\u7279\u5f81\u7684meta\u914d\u7f6e":59,"\u6bcf\u4e2a\u72b6\u6001":34,"\u6bcf\u4e2a\u7c7b\u522b\u4e2d\u968f\u673a\u62bd\u53d6\u4e8610\u5f20\u56fe\u7247":54,"\u6bcf\u4e2a\u7c7b\u5305\u542b6000\u5f20":54,"\u6bcf\u4e2a\u7ebf\u7a0b":43,"\u6bcf\u4e2a\u7ebf\u7a0b\u5206\u914d\u5230128\u4e2a\u6837\u672c\u7528\u4e8e\u8bad\u7ec3":43,"\u6bcf\u4e2a\u8282\u70b9\u6709\u4e24\u4e2a6\u6838cpu":62,"\u6bcf\u4e2a\u8bad\u7ec3\u8282\u70b9\u5fc5\u987b\u6307\u5b9a\u4e00\u4e2a\u552f\u4e00\u7684id\u53f7":43,"\u6bcf\u4e2a\u8bb0\u5fc6\u5355\u5143\u5305\u542b\u56db\u4e2a\u4e3b\u8981\u7684\u5143\u7d20":61,"\u6bcf\u4e2a\u8bc4\u8bba\u7684\u7f51\u5740":61,"\u6bcf\u4e2a\u8f93\u5165\u90fd\u662f\u4e00\u4e2a":37,"\u6bcf\u4e2a\u8f93\u51fa\u8282\u70b9\u90fd\u8fde\u63a5\u5230\u6240\u6709\u7684\u8f93\u5165\u8282\u70b9\u4e0a":37,"\u6bcf\u4e2a\u91cc\u9762\u90fd\u5305\u542b202mb\u7684\u5168\u90e8\u7684\u6a21\u578b\u53c2\u6570":62,"\u6bcf\u4e2adata":19,"\u6bcf\u4e2alayer\u8fd4\u56de\u7684\u90fd\u662f\u4e00\u4e2a":46,"\u6bcf\u4e2aparamet":18,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":57,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":57,"\u6bcf\u4e2apod\u5305\u542b\u4e00\u4e2apaddlepaddle\u5bb9\u5668":49,"\u6bcf\u4e2ashard\u5206\u522b\u5b58\u50a8\u5728\u5176\u4e2d\u4e00\u53f0paramet":18,"\u6bcf\u4f4d\u7528\u6237\u81f3\u5c11\u670920\u6761\u8bc4\u5206":58,"\u6bcf\u5c42\u4e0a\u53ea\u80fd\u4fdd\u5b58\u56fa\u5b9a\u6570\u76ee\u4e2a\u6700\u597d\u7684\u72b6\u6001":43,"\u6bcf\u5c42\u4f7f\u7528\u7684gpu\u53f7\u4f9d\u8d56\u4e8e\u53c2\u6570train":45,"\u6bcf\u5f53\u6a21\u578b\u5728\u7ffb\u8bd1\u8fc7\u7a0b\u4e2d\u751f\u6210\u4e86\u4e00\u4e2a\u5355\u8bcd":62,"\u6bcf\u5f53\u7cfb\u7edf\u9700\u8981\u65b0\u7684\u6570\u636e\u8bad\u7ec3\u65f6":46,"\u6bcf\u6279\u6b21":43,"\u6bcf\u6b21":27,"\u6bcf\u6b21\u63d0\u4ea4\u4ee3\u7801":36,"\u6bcf\u6b21\u6d4b\u8bd5\u90fd\u6d4b\u8bd5\u6240\u6709\u6570\u636e":61,"\u6bcf\u6b21\u751f\u62101\u4e2a\u5e8f\u5217":62,"\u6bcf\u6b21\u8bfb\u53d6\u4e00\u6761\u6570\u636e\u540e":57,"\u6bcf\u6b21\u8f93\u51fa\u4e00\u4e2adata":19,"\u6bcf\u6b21\u90fd\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":3,"\u6bcf\u884c\u5b58\u50a8\u4e00\u4e2a\u8bcd":53,"\u6bcf\u884c\u5b58\u50a8\u7684\u662f\u4e00\u4e2a\u6837\u672c\u7684\u7279\u5f81":55,"\u6bcf\u884c\u6253\u537032\u4e2a\u53c2\u6570\u4ee5":53,"\u6bcf\u884c\u8868\u793a\u4e00\u4e2a\u6279\u6b21\u4e2d\u7684\u5355\u4e2a\u8f93\u5165":37,"\u6bcf\u884c\u90fd\u662f\u4e00\u4e2a\u6cd5\u8bed\u6216\u8005\u82f1\u8bed\u7684\u53e5\u5b50":62,"\u6bcf\u8f6e\u4f1a\u5c06\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u8bad\u7ec3\u6837\u672c\u4f7f\u7528\u4e00\u6b21":43,"\u6bcf\u8f6e\u7ed3\u675f\u65f6\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":43,"\u6bcf\u8f6e\u90fd\u4f1a\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c":43,"\u6bcf\u8fd0\u884c\u591a\u5c11\u4e2a\u6279\u6b21\u6267\u884c\u4e00\u6b21\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u68c0\u67e5":43,"\u6bcf\u969410\u5206\u949f":18,"\u6bcf\u9694\u591a\u5c11batch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":57,"\u6bcfdot":43,"\u6bcflog":43,"\u6bcfsave":43,"\u6bcftest":43,"\u6bd4\u5982":[19,24,27,36,57],"\u6bd4\u5982\u4e00\u53e5\u8bdd\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5355\u8bcd":32,"\u6bd4\u5982\u5982\u679c\u8981build\u4e00\u4e2a\u4e0d\u4f9d\u8d56gpu":36,"\u6bd4\u5982\u6bcf\u969410\u5206\u949f\u6700\u65b0\u7684\u5feb\u7167":18,"\u6bd4\u5982\u6d41\u5f0f\u6570\u636e\u5904\u7406":19,"\u6bd4\u5982\u8bbe\u7f6e\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u53c2\u6570\u521d\u59cb\u5316\u65b9\u5f0f\u548cbias\u521d\u59cb\u5316\u65b9\u5f0f":24,"\u6bd4\u5982\u901a\u8fc78080\u7aef\u53e3":47,"\u6bd4\u5982ide\u914d\u7f6e\u91cc":36,"\u6bd4\u5982imagenet\u8fd9\u4e2a\u6570\u636e\u96c6\u53ef\u80fd\u88ab\u5206\u62101000\u4e2ashard":19,"\u6bd4\u5982mnist":19,"\u6bd4\u8f83\u5bb9\u6613\u5728\u5927\u6a21\u578b\u4e0b\u6ea2\u51fa":46,"\u6c34\u6e29":32,"\u6c49\u5ead":32,"\u6c60\u5316\u5c42":54,"\u6ca1":32,"\u6ca1\u6709\u4f5c\u7528":3,"\u6ca1\u6709\u4f7f\u7528avx\u6307\u4ee4\u96c6":29,"\u6ca1\u6709\u5b9e\u9645\u610f\u4e49":53,"\u6ca1\u6709\u6d4b\u8bd5\u6570\u636e":3,"\u6ca1\u6709\u8fdb\u884c\u6b63\u786e\u6027\u7684\u68c0\u67e5":58,"\u6ca1\u6709\u8fdb\u884c\u7ed3\u6784\u7684\u5fae\u8c03":59,"\u6cd5\u8bed":62,"\u6ce8":18,"\u6ce8\u610f":[3,26,27,35,37,38,49,54],"\u6ce8\u610f\u4e0a\u8ff0\u547d\u4ee4\u4e2d":49,"\u6ce8\u610f\u5230\u6211\u4eec\u5df2\u7ecf\u5047\u8bbe\u673a\u5668\u4e0a\u67094\u4e2agpu":45,"\u6ce8\u610f\u5e94\u8be5\u786e\u4fdd\u9ed8\u8ba4\u6a21\u578b\u8def\u5f84":61,"\u6ce8\u610f\u9884\u6d4b\u6570\u636e\u901a\u5e38\u4e0d\u5305\u542blabel":5,"\u6ce8\u610fnode":49,"\u6ce8\u91ca\u6389":61,"\u6cf3\u6c60":32,"\u6d41":32,"\u6d41\u7a0b\u6765\u63d0\u4ea4\u4ee3\u7801":36,"\u6d44":32,"\u6d4b\u8bd5":36,"\u6d4b\u8bd5\u6570\u636e":41,"\u6d4b\u8bd5\u6570\u636e\u4e5f\u5305\u542b":41,"\u6d4b\u8bd5\u6570\u636e\u548c\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":41,"\u6d4b\u8bd5\u6570\u636e\u548c\u751f\u6210\u6570\u636e":62,"\u6d4b\u8bd5\u6570\u636e\u653e\u7f6e\u5728\u5de5\u4f5c\u7a7a\u95f4\u4e2d\u4e0d\u540c\u76ee\u5f55\u7684\u8981\u6c42":41,"\u6d4b\u8bd5\u6570\u636e\u7684\u6240\u6709\u76f8\u5bf9\u6216\u7edd\u5bf9\u6587\u4ef6\u8def\u5f84":41,"\u6d4b\u8bd5\u6570\u6910\u96c6":61,"\u6d4b\u8bd5\u65f6\u6307\u5b9a\u7684\u5b58\u50a8\u6a21\u578b\u5217\u8868\u7684\u6587\u4ef6":43,"\u6d4b\u8bd5\u65f6\u9ed8\u8ba4\u4e0dshuffl":3,"\u6d4b\u8bd5\u662f":36,"\u6d4b\u8bd5\u6837\u672c":41,"\u6d4b\u8bd5\u6a21\u578b\u662f\u6307\u4f7f\u7528\u8bad\u7ec3\u51fa\u7684\u6a21\u578b\u8bc4\u4f30\u5df2\u6807\u8bb0\u7684\u9a8c\u8bc1\u96c6":61,"\u6d4b\u8bd5\u7684\u6a21\u578b\u5305\u62ec\u4ece\u7b2cm\u8f6e\u5230\u7b2cn":45,"\u6d4b\u8bd5\u811a\u672c\u662f":60,"\u6d4b\u8bd5\u96c6\u548c\u8bad\u7ec3\u96c6\u76ee\u5f55\u5305\u542b\u4e0b\u9762\u7684\u6587\u4ef6":61,"\u6d4b\u8bd5docker\u955c\u50cf":23,"\u6d4b\u8bd5model_list":42,"\u6d4b\u8bd5save_dir":42,"\u6d4f\u89c8\u4ee3\u7801":27,"\u6d6a\u6f2b\u7247":58,"\u6d6e\u70b9\u578b\u6570\u636e":19,"\u6d6e\u70b9\u5f02\u5e38\u901a\u5e38\u7684\u539f\u56e0\u662f\u6d6e\u70b9\u6570\u6ea2\u51fa":24,"\u6d6e\u70b9\u6570\u5360\u7528\u7684\u5b57\u8282\u6570":53,"\u6d6e\u70b9\u7a00\u758f\u6570\u636e":37,"\u6dd8\u5b9d\u7b49":61,"\u6df1\u5ea6\u53cc\u5411lstm\u5c42\u63d0\u53d6softmax\u5c42\u7684\u7279\u5f81":60,"\u6df1\u5ea6\u5b66\u4e60\u5165\u95e8\u8bfe\u7a0b":30,"\u6df7\u5408":60,"\u6df7\u5408\u5f53\u524d\u8bcd\u5411\u91cf\u548cattention\u52a0\u6743\u7f16\u7801\u5411\u91cf":35,"\u6dfb\u52a0\u542f\u52a8\u811a\u672c":49,"\u6e05\u7406\u6389\u8001\u65e7\u7684paddlepaddle\u5b89\u88c5\u5305":24,"\u6e29\u99a8":32,"\u6e90":62,"\u6e90\u4ee3\u7801":[27,57],"\u6e90\u4ee3\u7801\u683c\u5f0f":36,"\u6e90\u5b57\u5178":62,"\u6e90\u5e8f\u5217":35,"\u6e90\u7801":27,"\u6e90\u7801\u4e0edemo":48,"\u6e90\u7801\u7f16\u8bd1\u6587\u6863":38,"\u6e90\u8bed\u8a00\u5230\u76ee\u6807\u8bed\u8a00\u7684\u5e73\u884c\u8bed\u6599\u5e93\u6587\u4ef6":62,"\u6e90\u8bed\u8a00\u548c\u76ee\u6807\u8bed\u8a00\u5171\u4eab\u76f8\u540c\u7684\u7f16\u7801\u5b57\u5178":53,"\u6e90\u8bed\u8a00\u548c\u76ee\u6807\u8bed\u8a00\u90fd\u662f\u76f8\u540c\u7684\u8bed\u8a00":53,"\u6e90\u8bed\u8a00\u77ed\u8bed\u548c\u76ee\u6807\u8bed\u8a00\u77ed\u8bed\u7684\u5b57\u5178\u5c06\u88ab\u5408\u5e76":53,"\u6ee4\u6ce2\u5668\u6838\u5728\u5782\u76f4\u65b9\u5411\u4e0a\u7684\u5c3a\u5bf8":55,"\u6ee4\u6ce2\u5668\u6838\u5728\u6c34\u5e73\u65b9\u5411\u4e0a\u7684\u5c3a\u5bf8":55,"\u6f14\u793a\u4e2d\u4f7f\u7528\u7684":60,"\u6f14\u793a\u91c7\u7528":60,"\u6fc0\u6d3b":37,"\u6fc0\u6d3b\u51fd\u6570":46,"\u6fc0\u6d3b\u51fd\u6570\u4e3asoftmax":46,"\u6fc0\u6d3b\u51fd\u6570\u7c7b\u578b":57,"\u6fc0\u6d3b\u65b9\u7a0b":37,"\u6fc0\u6d3b\u7684\u7c7b\u578b":37,"\u6fc0\u6d3b\u7c7b\u578b\u7b49":46,"\u7075\u6d3b\u6027\u548c\u53ef\u6269\u5c55\u6027":0,"\u70ed\u60c5":32,"\u7136\u540e":[40,41,53,59],"\u7136\u540e\u4ea4\u7ed9\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u51fd\u6570":25,"\u7136\u540e\u4ea4\u7ed9step\u51fd\u6570":34,"\u7136\u540e\u4ecb\u7ecdpserver\u8fdb\u7a0b\u4e2d\u6982\u5ff5":46,"\u7136\u540e\u4f60\u53ea\u9700\u8981\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4":62,"\u7136\u540e\u4f7f\u7528\u4e0b\u9762\u7684\u811a\u672c":61,"\u7136\u540e\u518d\u505a\u4e00\u6b21\u6587\u672c\u5377\u79ef\u7f51\u7edc\u64cd\u4f5c":59,"\u7136\u540e\u5229\u7528\u89c2\u6d4b\u6570\u636e\u8c03\u6574":25,"\u7136\u540e\u52a0":46,"\u7136\u540e\u5355\u51fb":36,"\u7136\u540e\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":49,"\u7136\u540e\u53ef\u4ee5\u8f6c\u6362\u4e3a\u56fe\u7247":55,"\u7136\u540e\u5728":62,"\u7136\u540e\u5728\u4e0b\u4e00\u4e2a\u65f6\u95f4\u6b65\u8f93\u5165\u7ed9\u53e6\u4e00\u4e2a\u795e\u7ecf\u5143":32,"\u7136\u540e\u5728\u6d4f\u89c8\u5668\u4e2d\u8f93\u5165\u4ee5\u4e0b\u7f51\u5740":27,"\u7136\u540e\u5728\u89e3\u7801\u88ab\u7ffb\u8bd1\u7684\u8bed\u53e5\u65f6":62,"\u7136\u540e\u5728dataprovider\u91cc\u9762\u6839\u636e\u8be5\u5730\u5740\u52a0\u8f7d\u5b57\u5178":24,"\u7136\u540e\u5728etcd\u7684":18,"\u7136\u540e\u5b9a\u4e49":35,"\u7136\u540e\u5c06\u6784\u5efa\u6210\u529f\u7684\u955c\u50cf\u4e0a\u4f20\u5230\u955c\u50cf\u4ed3\u5e93":49,"\u7136\u540e\u5f97\u5230\u5e73\u5747\u91c7\u6837\u7684\u7ed3\u679c":59,"\u7136\u540e\u6211\u4eec\u5229\u7528\u591a\u8f93\u5165\u7684":59,"\u7136\u540e\u6211\u4eec\u53d1\u73b0pass":62,"\u7136\u540e\u6211\u4eec\u5806\u53e0\u4e00\u5bf9\u5bf9\u7684":60,"\u7136\u540e\u6211\u4eec\u6c42\u8fd9\u4e24\u4e2a\u7279\u5f81\u7684\u4f59\u5f26\u76f8\u4f3c\u5ea6":59,"\u7136\u540e\u6240\u6709\u7528":36,"\u7136\u540e\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4":55,"\u7136\u540e\u628a\u8fd9\u4e2a\u5305\u542b\u4e86\u8bad\u7ec3\u6570\u636e\u7684container\u4fdd\u5b58\u4e3a\u4e00\u4e2a\u65b0\u7684\u955c\u50cf":48,"\u7136\u540e\u63d0\u4ea4\u65b0\u6dfb\u52a0\u7684":36,"\u7136\u540e\u63d0\u53d6\u9690\u85cflstm\u5c42\u7684\u6240\u6709\u65f6\u95f4\u6b65\u957f\u7684\u6700\u5927\u8bcd\u5411\u91cf\u4f5c\u4e3a\u6574\u4e2a\u5e8f\u5217\u7684\u8868\u793a":61,"\u7136\u540e\u662f\u5bf9\u5e94\u7684\u82f1\u8bed\u5e8f\u5217":62,"\u7136\u540e\u6dfb\u52a0\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42":59,"\u7136\u540e\u70b9\u51fb":36,"\u7136\u540e\u7528\u5bc6\u7801":27,"\u7136\u540e\u7528pickle\u547d\u4ee4\u5c06\u7279\u5f81":59,"\u7136\u540e\u7533\u660e\u4e00\u4e2a\u5b58\u50a8\u5377":49,"\u7136\u540e\u89c2\u5bdf\u5230\u8f93\u51fa\u7684\u53d8\u5316\u4e3a":37,"\u7136\u540e\u89e3\u538b":62,"\u7136\u540e\u89e3\u7801\u5668\u901a\u8fc7\u8fd9\u4e2a\u5411\u91cf\u751f\u6210\u4e00\u4e2a\u76ee\u6807\u8bed\u53e5":62,"\u7136\u540e\u8f93\u51fa\u9884\u6d4b\u5206\u6570":59,"\u7136\u540e\u8fd4\u56de\u7ed9paddlepaddle\u8fdb\u7a0b":3,"\u7136\u540e\u8fdb\u884c\u968f\u673a\u6253\u4e71":59,"\u7136\u540e\u901a\u8fc7\u51fd\u6570":49,"\u7136\u540e\u901a\u8fc7\u81ea\u8eab\u7684ip\u5730\u5740\u5728":49,"\u7136\u800c":[35,43],"\u7136\u800c\u6709\u4e9b\u8bc4\u8bba\u4e0a\u4e0b\u6587\u975e\u5e38\u957f":61,"\u7248\u672c":29,"\u7248\u672c\u5206\u652f":23,"\u7248\u672c\u53f7":23,"\u7248\u672c\u53f7rc":23,"\u7248\u672c\u5728":36,"\u7248\u672cfork\u51fa\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f":23,"\u7279\u522b\u611f\u8c22paddlepaddle\u7684":0,"\u7279\u522b\u662f\u5728lstm\u7b49rnn\u4e2d":24,"\u7279\u522b\u662f\u5f53\u76f8\u540c\u7684\u8bcd\u5728\u53e5\u5b50\u4e2d\u51fa\u73b0\u591a\u4e8e\u4e00\u6b21\u65f6":60,"\u7279\u5f81":59,"\u7279\u5f81\u56fe\u5747\u503c":55,"\u7279\u5f81\u56fe\u65b9\u5dee":55,"\u7279\u5f81\u5c06\u4f1a\u5b58\u5230":55,"\u7279\u5f81\u6587\u4ef6":60,"\u7279\u5f81\u7684\u7c7b\u578b\u548c\u7ef4\u5ea6":59,"\u72af\u7f6a\u7247":58,"\u73af\u5883\u53d8\u91cf":49,"\u73af\u5883\u53d8\u91cf\u6765\u6307\u5b9a\u7279\u5b9a\u7684gpu":24,"\u73b0\u5728\u6211\u4eec\u53ef\u4ee5\u5f00\u59cbpaddle\u8bad\u7ec3\u4e86":59,"\u73b0\u9636\u6bb5paddle\u6709\u4e00\u4e2a\u95ee\u9898\u662f":20,"\u751a\u81f3\u4e0d\u540c\u7ade\u4e89\u5bf9\u624b\u4ea7\u54c1\u7684\u504f\u597d":61,"\u751a\u81f3\u53ef\u4ee5\u76f4\u63a5\u914d\u7f6e\u4e00\u4e2a\u5b8c\u6574\u7684lstm":46,"\u751a\u81f3\u80fd\u89e3\u91ca\u4e3a\u4ec0\u4e48\u67d0\u4e2a\u64cd\u4f5c\u82b1\u4e86\u5f88\u957f\u65f6\u95f4":40,"\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u8bad\u7ec3\u6570\u636e\u96c6\u901a\u5e38\u4f53\u79ef\u5f88\u5927":19,"\u751f\u4ea7\u73af\u5883\u7684\u65e5\u5fd7\u6570\u636e\u4f1a\u901a\u8fc7\u5b9e\u65f6\u6d41\u7684\u65b9\u5f0f":19,"\u751f\u4ea7\u955c\u50cf":[27,36],"\u751f\u4ea7\u955c\u50cf\u7684\u751f\u6210\u5206\u4e3a\u4e24\u6b65":27,"\u751f\u6210":49,"\u751f\u6210\u5404\u79cd\u8bed\u8a00\u7684\u7ed1\u5b9a\u4ee3\u7801":20,"\u751f\u6210\u540e\u7684\u6587\u6863\u5206\u522b\u5b58\u50a8\u5728\u7f16\u8bd1\u76ee\u5f55\u7684":38,"\u751f\u6210\u5e8f\u5217\u7684\u6700\u5927\u957f\u5ea6":35,"\u751f\u6210\u5f53\u524d\u5c42\u7684\u6240\u6709\u540e\u7ee7\u72b6\u6001":62,"\u751f\u6210\u6570\u636e\u51fd\u6570\u63a5\u53e3":46,"\u751f\u6210\u6570\u636e\u7684\u76ee\u5f55":62,"\u751f\u6210\u6587\u6863":20,"\u751f\u6210\u751f\u4ea7\u955c\u50cf\u53ea\u9700\u8981\u8fd0\u884c":27,"\u751f\u6210\u7684":19,"\u751f\u6210\u7684\u6570\u636e\u5c06\u4f1a\u5b58\u50a8\u5728\u8fd9\u4e2avolume\u4e0b":49,"\u751f\u6210\u7684\u6570\u636e\u7f13\u5b58\u5728\u5185\u5b58\u91cc":24,"\u751f\u6210\u7684\u7ed3\u679c\u6587\u4ef6":62,"\u751f\u6210\u7684html\u7248\u672c\u7684c":27,"\u751f\u6210\u7684meta\u914d\u7f6e\u6587\u4ef6\u5982\u4e0b\u6240\u793a":59,"\u751f\u6210\u7ed3\u679c\u6587\u4ef6\u7684\u8def\u5f84":35,"\u751f\u6210\u7ed9\u5b9asstable\u6587\u4ef6\u5bf9\u5e94\u7684data":19,"\u751f\u6210\u7f51\u7edc\u5c42\u914d\u7f6e":37,"\u751f\u6210\u81ea\u5df1\u76ee\u5f55\u4e0b\u7684\u4ed3\u5e93":36,"\u751f\u6210\u8bad\u7ec3\u9700\u8981\u7684\u6837\u672c":59,"\u751f\u6210\u8fd0\u884c\u7a0b\u5e8f":27,"\u751f\u6210api\u6587\u6863":20,"\u751f\u6210docker\u955c\u50cf\u7684\u65b9\u5f0f\u6709\u4e24\u4e2a":27,"\u7528":[58,59,60],"\u75280\u548c1\u8868\u793a":3,"\u7528\u4e86\u4e24\u4e2a\u6708\u4e4b\u540e\u8fd9\u4e2a\u663e\u793a\u5668\u5c4f\u5e55\u788e\u4e86":57,"\u7528\u4e8e":41,"\u7528\u4e8e\u4e5f\u53ef\u4ee5\u4e0a\u4f20label":19,"\u7528\u4e8e\u5207\u5206\u5355\u5355\u8bcd\u548c\u6807\u70b9\u7b26\u53f7":61,"\u7528\u4e8e\u521d\u59cb\u5316\u53c2\u6570\u548c\u8bbe\u7f6e":37,"\u7528\u4e8e\u5c06\u4e0b\u4e00\u884c\u7684\u6570\u636e\u8f93\u5165\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovider2":3,"\u7528\u4e8e\u5c06\u53c2\u6570\u4f20\u9012\u7ed9\u7f51\u7edc\u914d\u7f6e":45,"\u7528\u4e8e\u5c06\u8bcdid\u8f6c\u6362\u4e3a\u8bcd\u7684\u5b57\u5178\u6587\u4ef6":35,"\u7528\u4e8e\u6307\u5b9a\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":43,"\u7528\u4e8e\u653e\u7f6e":41,"\u7528\u4e8e\u6784\u6210\u65b0\u7684\u8bcd\u8868":53,"\u7528\u4e8e\u6807\u8bc6\u751f\u6210\u7684\u6587\u4ef6\u4e2d\u7684\u76f8\u5e94\u8f93\u51fa":35,"\u7528\u4e8e\u7a00\u758f\u8bad\u7ec3\u4e2d":43,"\u7528\u4e8e\u7edf\u8ba1\u8bcd\u9891\u7684bow\u6a21\u578b\u7279\u5f81":61,"\u7528\u4e8e\u81ea\u5b9a\u4e49\u6bcf\u6761\u6570\u636e\u7684batch":3,"\u7528\u4e8e\u8ba1\u7b97\u7f16\u7801\u5411\u91cf\u7684\u52a0\u6743\u548c":35,"\u7528\u4e8e\u8bbe\u7f6e\u8bad\u7ec3\u7b97\u6cd5":54,"\u7528\u4e8e\u8bfb\u53d6\u8bad\u7ec3":41,"\u7528\u4e8e\u96c6\u7fa4\u901a\u4fe1\u901a\u9053\u7684\u7aef\u53e3\u6570":41,"\u7528\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7f16\u7801":35,"\u7528\u547d\u4ee4":46,"\u7528\u591a\u5bf9\u6548\u679c\u5b8c\u5168\u76f8\u540c\u7684":32,"\u7528\u6237":41,"\u7528\u62371\u7684\u7279\u5f81":59,"\u7528\u6237\u4e5f\u53ef\u4ee5\u5728c":2,"\u7528\u6237\u53ea\u9700\u5b9a\u4e49rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":34,"\u7528\u6237\u53ea\u9700\u6267\u884c":60,"\u7528\u6237\u53ea\u9700\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u5c31\u53ef\u4ee5\u4e0b\u8f7d\u5e76\u5904\u7406\u539f\u59cb\u6570\u636e":60,"\u7528\u6237\u53ef\u4ee5\u4e0a\u4f20\u81ea\u5df1\u7684\u6570\u636e\u96c6":19,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528\u5f00\u53d1\u955c\u50cf\u4ee3\u66ff\u914d\u7f6e\u672c\u5730\u73af\u5883":27,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528ssh\u767b\u5f55\u5230\u8fd9\u53f0\u670d\u52a1\u5668\u4e0a\u5e76\u6267\u884c":27,"\u7528\u6237\u53ef\u4ee5\u5206\u522b\u67e5\u770b\u6700\u65b0\u7684":38,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003":46,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003sphinx\u6559\u7a0b\u8fdb\u884c\u4e66\u5199":38,"\u7528\u6237\u53ef\u4ee5\u5728\u8f93\u51fa\u7684\u6587\u672c\u6a21\u578b\u4e2d\u770b\u5230":53,"\u7528\u6237\u53ef\u4ee5\u5b89\u5168\u7684\u91ca\u653e\u67d0\u4e2ac":21,"\u7528\u6237\u53ef\u4ee5\u6839\u636e\u8bad\u7ec3\u65e5\u5fd7":57,"\u7528\u6237\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u8fd9\u4e2a\u52a8\u6001\u5e93\u6765\u5f15\u5165paddl":21,"\u7528\u6237\u53ef\u4ee5\u7f16\u5199\u7a0b\u5e8f\u5b8c\u6210\u8f6c\u6362":19,"\u7528\u6237\u53ef\u4ee5\u81ea\u5b9a\u4e49beam":43,"\u7528\u6237\u53ef\u4ee5\u8bbe\u7f6e":45,"\u7528\u6237\u53ef\u4ee5\u901a\u8fc7\u7b80\u5355\u4f7f\u7528python\u63a5\u53e3":2,"\u7528\u6237\u53ef\u4ee5\u901a\u8fc7\u7f51\u9875\u6d4f\u89c8\u6587\u6863":27,"\u7528\u6237\u53ef\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e\u5b83\u4eec":26,"\u7528\u6237\u53ef\u5728cmake\u7684\u547d\u4ee4\u884c\u4e2d":26,"\u7528\u6237\u5728\u4f7f\u7528paddlepaddl":24,"\u7528\u6237\u5728\u9996\u6b21\u7f16\u8bd1\u5b89\u88c5paddlepaddle\u65f6":38,"\u7528\u6237\u5b9a\u4e49\u7684\u53c2\u6570":3,"\u7528\u6237\u5c06\u914d\u7f6e\u4e0e\u8bad\u7ec3\u6570\u636e\u5207\u5206\u597d\u653e\u5728\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u9884\u5148\u5206\u914d\u597d\u7684\u76ee\u5f55\u4e2d":49,"\u7528\u6237\u5e94\u8be5\u63d0\u4f9b\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":60,"\u7528\u6237\u5f3a\u5236\u6307\u5b9a\u7279\u5b9a\u7684python\u7248\u672c":24,"\u7528\u6237\u6307\u5b9a\u65b0\u7684\u5b57\u5178\u7684\u8def\u5f84":53,"\u7528\u6237\u6587\u4ef6\u4e2d\u6709\u56db\u79cd\u7c7b\u578b\u7684\u5b57\u6bb5":59,"\u7528\u6237\u7279\u5f81":59,"\u7528\u6237\u8fd8\u53ef\u4ee5\u6839\u636e\u6982\u7387\u5206\u5e03\u77e9\u9635\u5b9e\u73b0\u67f1\u641c\u7d22\u6216\u7ef4\u7279\u6bd4\u89e3\u7801":60,"\u7528\u6237\u901a\u8fc7c":21,"\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":45,"\u7528\u6237\u9700\u8981\u6307\u5b9a\u672c\u673a\u4e0apython\u7684\u8def\u5f84":24,"\u7528\u6237\u9700\u8981\u9996\u5148\u786e\u8ba4py_paddle\u5305\u5df2\u7ecf\u5b89\u88c5":38,"\u7528\u6237\u9884\u6d4b\u7684\u547d\u4ee4\u884c\u754c\u9762\u5982\u4e0b":59,"\u7528\u6237id":58,"\u7528\u6237id\u8303\u56f4\u4ece1\u52306040":58,"\u7528\u6765\u4ece\u53c2\u6570\u670d\u52a1\u5668\u9884\u53d6\u53c2\u6570\u77e9\u9635\u76f8\u5e94\u7684\u884c":37,"\u7528\u6765\u4f30\u8ba1\u7ebf\u6027\u51fd\u6570\u7684\u53c2\u6570w":25,"\u7528\u6765\u5177\u4f53\u63cf\u8ff0":59,"\u7528\u6765\u5177\u4f53\u8bf4\u660e\u6570\u636e\u96c6\u7684\u5b57\u6bb5\u548c\u6587\u4ef6\u683c\u5f0f":59,"\u7528\u6765\u8ba1\u7b97\u6a21\u578b\u7684\u8bef\u5dee":25,"\u7528\u8fd9\u4e2a\u955c\u50cf\u521b\u5efa\u7684\u5bb9\u5668\u9700\u8981\u6709\u4ee5\u4e0b\u4e24\u4e2a\u529f\u80fd":49,"\u7531":34,"\u7531\u4e8e\u4e0d\u540c\u7684paddle\u7684\u7248\u672c\u53ef\u80fd\u9700\u8981\u4e0d\u540c\u7684\u4f9d\u8d56\u548c\u5de5\u5177":27,"\u7531\u4e8e\u5b83\u5185\u90e8\u5305\u542b\u4e86\u6bcf\u7ec4\u6570\u636e\u4e2d\u7684\u6240\u6709\u53e5\u5b50":32,"\u7531\u4e8e\u5bb9\u5668\u4e4b\u95f4\u5171\u4eabnet":47,"\u7531\u4e8e\u5bf9parameters\u7684\u66f4\u65b0\u9700\u8981\u83b7\u53d6parameters\u5185\u5b58\u7684":18,"\u7531\u4e8e\u5df2\u7ecf\u77e5\u9053\u4e86\u771f\u5b9e\u7b54\u6848":25,"\u7531\u4e8e\u610f\u5916\u7684\u526f\u672c\u8bb0\u5f55\u548c\u6d4b\u8bd5\u8bb0\u5f55":58,"\u7531\u4e8e\u6211\u4eec\u60f3\u8981\u7684\u53d8\u6362\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":32,"\u7531\u4e8e\u6211\u4eec\u652f\u6301\u8bad\u7ec3\u6570\u636e\u6709\u4e0d\u540c\u7684\u6279\u6b21\u5927\u5c0f":37,"\u7531\u4e8e\u6570\u636e\u8bb8\u53ef\u7684\u539f\u56e0":60,"\u7531\u4e8e\u6807\u51c6\u7684\u7ffb\u8bd1\u7ed3\u679c\u5df2\u7ecf\u4e0b\u8f7d\u5230\u8fd9\u91cc":62,"\u7531\u4e8e\u6bcf\u4e2a\u5377\u79ef\u5c42\u540e\u9762\u8fde\u63a5\u7684\u662fbatch":55,"\u7531\u4e8e\u8fd9\u4e2a\u5730\u5740\u4f1a\u88abdataprovider\u4f7f\u7528":2,"\u7531\u4e8e\u8fd9\u6837\u505a\u53ef\u4ee5\u907f\u514d\u5f88\u591a\u6b7b\u9501\u95ee\u9898":3,"\u7531\u4e8e\u96c6\u7fa4\u4e2d\u540c\u65f6\u5b58\u5728\u4e24\u53f0\u673a\u5668\u6545\u969c\u7684\u6982\u7387\u6781\u4f4e":18,"\u7531\u4e8e\u987a\u5e8f\u8c03\u7528\u8fd9\u4e9bgenerator\u4e0d\u4f1a\u51fa\u73b0\u4e0a\u8ff0\u95ee\u9898":3,"\u7531\u4e8ec":20,"\u7531\u4e8epaddlepaddle\u5df2\u7ecf\u5b9e\u73b0\u4e86\u4e30\u5bcc\u7684\u7f51\u7edc\u5c42":25,"\u7531\u4e8estep":34,"\u7531\u4e8etest_data\u5305\u542b\u4e24\u6761\u9884\u6d4b\u6570\u636e":5,"\u7531\u8bcd\u8bed\u6784\u6210\u7684\u53e5\u5b50":31,"\u7531grouplen":58,"\u7535\u5f711\u7684\u7279\u5f81":59,"\u7535\u5f71\u4fe1\u606f\u4ee5\u53ca\u7535\u5f71\u8bc4\u5206":58,"\u7535\u5f71\u540d\u5b57\u6bb5":59,"\u7535\u5f71\u540d\u79f0":58,"\u7535\u5f71\u548c\u7528\u6237":59,"\u7535\u5f71\u548c\u7528\u6237\u6709\u8bb8\u591a\u7684\u7279\u5f81":59,"\u7535\u5f71\u5927\u90e8\u5206\u662f\u624b\u5de5\u8f93\u5165\u6570\u636e":58,"\u7535\u5f71\u7279\u5f81":59,"\u7535\u5f71\u7c7b\u578b":58,"\u7535\u5f71\u7c7b\u578b\u5982\u7b26\u5408\u591a\u79cd\u7528\u7ba1\u9053\u7b26\u53f7":58,"\u7535\u5f71id":58,"\u7535\u5f71id\u8303\u56f4\u4ece1\u52303952":58,"\u7535\u8111":32,"\u767e\u4e07\u6570\u636e\u96c6":58,"\u7684":[32,36,41,48,49,57,61],"\u768410\u7ef4\u6574\u6570\u6807\u7b7e":3,"\u7684\u4e00\u4e2a\u7b80\u5355\u8c03\u7528\u5982\u4e0b":34,"\u7684\u4e00\u4e2a\u7ebf\u6027\u51fd\u6570":25,"\u7684\u4e00\u79cd":62,"\u7684\u4e3a0":43,"\u7684\u4e3b\u8981\u90e8\u5206":60,"\u7684\u4efb\u4e00\u4e00\u79cd":24,"\u7684\u4efb\u52a1":62,"\u7684\u4f5c\u7528":46,"\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":31,"\u7684\u504f\u7f6e\u5411\u91cf":37,"\u7684\u5185\u5b58":24,"\u7684\u5185\u5bb9\u5982\u4e0b\u6240\u793a":62,"\u7684\u5185\u5bb9\u6765\u5b9a\u5236imag":49,"\u7684\u5185\u6838block\u4f7f\u7528\u60c5\u51b5":40,"\u7684\u51fd\u6570":46,"\u7684\u5206\u7c7b\u4efb\u52a1\u4e2d\u8d62\u5f97\u4e86\u7b2c\u4e00\u540d":55,"\u7684\u522b\u540d":[6,7],"\u7684\u5355\u8bcd\u7ea7\u522b\u7684\u5377\u79ef\u795e\u7ecf\u7f51\u7edc":59,"\u7684\u53cd\u5411\u4f20\u64ad\u5c06\u4f1a\u6253\u5370\u65e5\u5fd7\u4fe1\u606f":43,"\u7684\u53d8\u6362\u77e9\u9635":37,"\u7684\u53e5\u5b50\u7684\u60c5\u611f":61,"\u7684\u540d\u5b57":3,"\u7684\u540d\u79f0\u76f8\u540c":35,"\u7684\u540e\u7f00":58,"\u7684\u5411\u91cf":37,"\u7684\u542f\u52a8\u53c2\u6570":49,"\u7684\u542f\u52a8\u53c2\u6570\u5e76\u6267\u884c\u8fdb\u7a0b":49,"\u7684\u547d\u540d\u98ce\u683c\u5e76\u4e0d\u80fd\u9002\u5e94\u5176\u4ed6\u7b2c\u4e09\u65b9\u8bed\u8a00":20,"\u7684\u5730\u5740":47,"\u7684\u5730\u65b9":36,"\u7684\u5747\u5300\u5206\u5e03":24,"\u7684\u5934\u6587\u4ef6":20,"\u7684\u5dee\u8ddd\u4e0d\u65ad\u51cf\u5c0f":25,"\u7684\u5e73\u5747\u503c":31,"\u7684\u5e8f\u5217\u5f62\u72b6\u4e00\u81f4":32,"\u7684\u5f00\u53d1\u5de5\u4f5c\u90fd\u5e94\u8be5\u5728\u4e00\u4e2a\u65b0\u7684\u5206\u652f\u4e0a\u5b8c\u6210":36,"\u7684\u603b":41,"\u7684\u63a5\u53e3\u6837\u5f0f":20,"\u7684\u63cf\u8ff0\u8bf4\u660e\u4e2d":36,"\u7684\u6570\u636e":46,"\u7684\u6570\u636e\u6d41\u56fe":19,"\u7684\u6570\u636e\u8bfb\u53d6\u811a\u672c\u548c\u7c7b\u4f3c\u4e8e":57,"\u7684\u6570\u76ee\u4e00\u81f4":31,"\u7684\u6587\u4ef6\u4e5f\u5e26\u5230\u65b0\u5206\u652f\u4e0a":36,"\u7684\u65b9\u6cd5\u5df2\u88ab\u8bc1\u660e\u662f\u4e00\u79cd\u5f3a\u5927\u7684\u6a21\u578b":62,"\u7684\u65b9\u7a0b":37,"\u7684\u65f6\u5019\u5982\u679c\u62a5\u4e00\u4e9b\u4f9d\u8d56\u672a\u627e\u5230\u7684\u9519\u8bef\u662f\u6b63\u5e38\u7684":29,"\u7684\u65f6\u95f4\u6b65\u4fe1\u606f\u6210\u6b63\u6bd4":24,"\u7684\u66f4\u8be6\u7ec6\u51c6\u786e\u7684\u5b9a\u4e49":32,"\u7684\u6700\u5c0f\u503c":43,"\u7684\u6700\u65b0\u4ee3\u7801\u5e76\u66f4\u65b0\u5f53\u524d\u5206\u652f":36,"\u7684\u67b6\u6784\u7684\u793a\u4f8b":35,"\u7684\u6837\u5f0f":36,"\u7684\u6838\u5fc3\u662f\u8bbe\u8ba1step\u51fd\u6570\u7684\u8ba1\u7b97\u903b\u8f91":34,"\u7684\u6bb5\u843d\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":34,"\u7684\u6d4b\u8bd5\u6570\u636e\u96c6":60,"\u7684\u6e90\u7801\u4ee5\u53ca\u751f\u6210\u6587\u6863\u9700\u8981\u591a\u79cd\u5f00\u53d1\u5de5\u5177":36,"\u7684\u6e90\u7801\u91cc\u4f7f\u7528\u4e86":20,"\u7684\u7248\u672c\u53f7":53,"\u7684\u7279\u5f81":55,"\u7684\u72b6\u6001":34,"\u7684\u7528\u6237\u53c2\u8003":41,"\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":59,"\u7684\u76f8\u5173\u6587\u6863\u8fdb\u884c\u914d\u7f6e":46,"\u7684\u771f\u5b9e\u5173\u7cfb\u4e3a":25,"\u7684\u77e9\u9635":37,"\u7684\u795e\u7ecf\u7f51\u7edc\u673a\u5668\u7ffb\u8bd1":62,"\u7684\u7a20\u5bc6\u5411\u91cf\u4f5c\u4e3a\u8f93\u5165":37,"\u7684\u7aef\u5230\u7aef\u7cfb\u7edf\u6765\u89e3\u51b3srl\u4efb\u52a1":60,"\u7684\u7b2ci\u4e2a\u503c":37,"\u7684\u7b2cj\u4e2a\u503c":37,"\u7684\u7d22\u5f15\u6587\u4ef6\u5f15\u7528\u8bad\u7ec3":41,"\u7684\u7ed3\u6784\u5982\u4e0b":59,"\u7684\u7ef4\u5ea6":53,"\u7684\u884c\u6570\u5e94\u8be5\u4e00\u81f4":62,"\u7684\u89c4\u8303":20,"\u7684\u8bad\u7ec3\u6a21\u578b\u811a\u672c":57,"\u7684\u8bdd":24,"\u7684\u8def\u5f84\u4e2d":61,"\u7684\u8f93\u5165":34,"\u7684\u8f93\u51fa":40,"\u7684\u8f93\u51fa\u4fe1\u606f\u5165\u624b\u662f\u4e2a\u4e0d\u9519\u7684\u9009\u62e9":40,"\u7684\u8f93\u51fa\u51fd\u6570\u8fd4\u56de\u7684\u662f\u4e0b\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa\u8bcd\u7684":35,"\u7684\u8f93\u51fa\u683c\u5f0f":32,"\u7684\u8f93\u51fa\u88ab\u7528\u4f5c":35,"\u7684\u8fd4\u56de\u503c\u4e00\u81f4":58,"\u7684\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":36,"\u7684\u90e8\u5206":41,"\u7684\u914d\u7f6e":[46,53],"\u7684\u9875\u9762\u5220\u9664\u8fdc\u7a0b\u4ed3\u5e93\u7684\u5206\u652f":36,"\u7684docker\u955c\u50cf":27,"\u76ee\u524d":[34,60],"\u76ee\u524d\u4f7f\u7528":36,"\u76ee\u524d\u53ea\u8003\u8651\u52a8\u6001\u6269\u5bb9trainer\u6570\u91cf":18,"\u76ee\u524d\u5d4c\u5165python\u89e3\u91ca\u5668":20,"\u76ee\u524d\u5df2\u88ab\u767e\u5ea6\u5185\u90e8\u591a\u4e2a\u4ea7\u54c1\u7ebf\u5e7f\u6cdb\u4f7f\u7528":0,"\u76ee\u524d\u652f\u6301\u4e24\u79cd":31,"\u76ee\u524d\u652f\u6301fail":43,"\u76ee\u524d\u8be5\u53c2\u6570\u4ec5\u7528\u4e8eaucvalidationlayer\u548cpnpairvalidationlayer\u5c42":43,"\u76ee\u524d\u8fd8\u672a\u652f\u6301":34,"\u76ee\u524dpaddle\u7684\u8fdb\u7a0b\u6a21\u578b\u662fc":20,"\u76ee\u524dpaddlepaddle\u7684develop\u5206\u652f\u7684\u6587\u6863\u662f\u81ea\u52a8\u89e6\u53d1\u66f4\u65b0\u7684":38,"\u76ee\u5f55":[41,48,49,61],"\u76ee\u5f55\u4e0b":[21,37,57,62],"\u76ee\u5f55\u4e0b\u627e\u5230":57,"\u76ee\u5f55\u4e0b\u7684demo\u8bad\u7ec3\u51fa\u6765":5,"\u76ee\u5f55\u4e0b\u7684python\u5305":24,"\u76ee\u5f55\u4e2d":[41,59],"\u76ee\u5f55\u4e2d\u7684":[40,41],"\u76ee\u5f55\u4e2dpaddl":49,"\u76ee\u5f55\u4f1a\u51fa\u73b0\u5982\u4e0b\u51e0\u4e2a\u65b0\u7684\u6587\u4ef6":60,"\u76ee\u5f55\u5c31\u6210\u4e3a\u4e86\u5171\u4eab\u5b58\u50a8":49,"\u76ee\u5f55\u7ed3\u6784\u5982\u4e0b":62,"\u76ee\u5f55\u91cc\u63d0\u4f9b\u4e86\u8be5\u6570\u636e\u7684\u4e0b\u8f7d\u811a\u672c\u548c\u9884\u5904\u7406\u811a\u672c":57,"\u76ee\u6807":62,"\u76ee\u6807\u51fd\u6570\u662f\u6807\u7b7e\u7684\u4ea4\u53c9\u71b5":60,"\u76ee\u6807\u5411\u91cf":35,"\u76ee\u6807\u5b57\u5178":62,"\u76f4\u5230\u8bad\u7ec3\u6536\u655b\u4e3a\u6b62":24,"\u76f4\u5230\u903c\u8fd1\u771f\u5b9e\u89e3":25,"\u76f4\u63a5\u4f7f\u7528c\u8bed\u8a00\u7684":20,"\u76f4\u63a5\u5220\u9664\u8fd9\u4e2a\u53c2\u6570\u5373\u53ef":21,"\u76f4\u63a5\u5bfc\u51fa\u5230c\u7684\u63a5\u53e3\u6bd4\u8f83\u56f0\u96be":20,"\u76f4\u63a5\u8fd0\u884c":27,"\u76f4\u63a5\u8fd4\u56de\u8ba1\u7b97\u7ed3\u679c":5,"\u76f4\u63a5\u8fdb\u5165\u8bad\u7ec3\u6a21\u578b\u7ae0\u8282":57,"\u76f8\u5173\u6982\u5ff5\u662f":3,"\u76f8\u5173\u7684\u9e1f\u7c7b\u6570\u636e\u96c6\u53ef\u4ee5\u4ece\u5982\u4e0b\u5730\u5740\u4e0b\u8f7d":54,"\u76f8\u5173\u8bba\u6587":60,"\u76f8\u53cd":62,"\u76f8\u540c\u540d\u5b57\u7684\u53c2\u6570":24,"\u76f8\u5bf9":32,"\u76f8\u5bf9\u4e8epaddlepaddle\u7a0b\u5e8f\u8fd0\u884c\u65f6\u7684\u8def\u5f84":2,"\u76f8\u5bf9mnist\u800c\u8a00":3,"\u76f8\u5e94\u7684\u6570\u636e\u8bfb\u53d6\u811a\u672c\u548c\u8bad\u7ec3\u6a21\u578b\u811a\u672c":57,"\u76f8\u5e94\u7684\u6570\u636e\u8fed\u4ee3\u5668\u5982\u4e0b":60,"\u76f8\u5e94\u7684\u6807\u8bb0\u53e5\u5b50\u662f":60,"\u76f8\u5f53":32,"\u77e9\u9635":42,"\u7814\u7a76\u4eba\u5458\u5206\u6790\u4e86\u51e0\u4e2a\u5173\u4e8e\u6d88\u8d39\u8005\u4fe1\u5fc3\u548c\u653f\u6cbb\u89c2\u70b9\u7684\u8c03\u67e5":61,"\u7814\u7a76\u751f":58,"\u786e\u4fdd\u7f16\u8bd1\u5668\u9009\u9879":36,"\u793a":57,"\u793a\u4f8b":[24,55],"\u793a\u4f8b3\u5bf9\u4e8e\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u6570\u636e\u5b8c\u5168\u76f8\u540c":32,"\u793a\u4f8b3\u7684\u914d\u7f6e\u4f7f\u7528\u4e86\u5355\u5c42rnn\u548c\u53cc\u5c42rnn":32,"\u793a\u4f8b3\u7684\u914d\u7f6e\u5206\u522b\u4e3a":32,"\u793e\u533a\u53c2\u4e0e\u56f0\u96be":20,"\u793e\u533a\u8d21\u732e\u4ee3\u7801\u5b66\u4e60\u6210\u672c\u9ad8":20,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u7684\u53c2\u6570":18,"\u795e\u7ecf\u7f51\u7edc\u5728\u8bad\u7ec3\u7684\u65f6\u5019":24,"\u795e\u7ecf\u7f51\u7edc\u673a\u5668\u7ffb\u8bd1":62,"\u795e\u7ecf\u7f51\u7edc\u7684\u67d0\u4e00\u5c42":46,"\u795e\u7ecf\u7f51\u7edc\u7684\u7f51\u7edc\u7ed3\u6784\u4e2d\u5177\u6709\u6709\u5411\u73af\u7ed3\u6784":32,"\u795e\u7ecf\u7f51\u7edc\u7684\u8bad\u7ec3\u672c\u8eab\u662f\u4e00\u4e2a\u975e\u5e38\u6d88\u8017\u5185\u5b58\u548c\u663e\u5b58\u7684\u5de5\u4f5c":24,"\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":25,"\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e\u4e3b\u8981\u5305\u62ec\u7f51\u7edc\u8fde\u63a5":46,"\u79bb":32,"\u79bb\u7ebf\u6279\u5904\u7406":19,"\u79d1\u5b66\u5bb6":58,"\u79d1\u5e7b\u7247":58,"\u79f0\u4e3a":[35,36,46],"\u79f0\u4e3a\u5f00\u53d1\u955c\u50cf":27,"\u79f0\u4e3a\u751f\u4ea7\u955c\u50cf":27,"\u79f0\u4e3a\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6":46,"\u79f0\u4e4b\u4e3a\u53cc\u5c42\u5e8f\u5217\u7684\u4e00\u4e2a\u5b50\u5e8f\u5217":31,"\u79f0\u4e4b\u4e3a\u96c6\u675f\u5927\u5c0f":43,"\u7a00\u758f\u6570\u636e\u7684\u683c\u5f0f":37,"\u7a00\u758f\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":49,"\u7a00\u758f\u768401\u5411\u91cf":3,"\u7a00\u758f\u7684\u5411\u91cf":3,"\u7a00\u758f\u77e9\u9635\u7684\u4e58\u79ef\u5e94\u7528\u4e8e\u524d\u5411\u4f20\u64ad\u8fc7\u7a0b":45,"\u7a0b\u5e8f\u4ece\u6b64\u76ee\u5f55\u62f7\u8d1d\u6587\u4ef6\u5230\u5bb9\u5668\u5185\u8fdb\u884c\u8bad\u7ec3":49,"\u7a0b\u5e8f\u505c\u6b62":43,"\u7a0b\u5e8f\u5458":58,"\u7a0b\u5e8f\u76f4\u63a5\u9000\u51fa":43,"\u7a0d\u505a\u8be6\u7ec6\u8bf4\u660e":46,"\u7a20\u5bc6\u5411\u91cf":37,"\u7a20\u5bc6\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":49,"\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":3,"\u7a97\u6237":32,"\u7acb\u523b\u9000\u51fa":24,"\u7aef\u53e3":41,"\u7aef\u53e3\u6570":41,"\u7aef\u53e3\u9644\u52a0\u5230\u4e3b\u673a\u540d\u4e0a":41,"\u7aef\u81ea\u5b9a\u4e49\u4e00\u4e2a":2,"\u7aef\u8bfb\u53d6\u6570\u636e":24,"\u7b2c":32,"\u7b2c\u4e00\u4e2a\u53c2\u6570\u662fsettings\u5bf9\u8c61":3,"\u7b2c\u4e00\u4e2a\u65b9\u6cd5\u7684\u597d\u5904\u662f\u7b80\u5355\u5feb\u6377":27,"\u7b2c\u4e00\u4e2a\u6837\u672c\u540c\u65f6encode\u4e24\u6761\u6570\u636e\u6210\u4e24\u4e2a\u5411\u91cf":32,"\u7b2c\u4e00\u4e2apass\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":3,"\u7b2c\u4e00\u4e2atag\u4e3a":23,"\u7b2c\u4e00\u5929":32,"\u7b2c\u4e00\u6b65\u662f\u8fd0\u884c":27,"\u7b2c\u4e00\u884c\u4ece":61,"\u7b2c\u4e00\u884c\u5b58\u7684\u662f\u56fe\u50cf":55,"\u7b2c\u4e00\u884c\u662f":53,"\u7b2c\u4e00\u884c\u7684":62,"\u7b2c\u4e00\u90e8\u5206\u5b9a\u4e49\u4e86\u6570\u636e\u8f93\u5165":25,"\u7b2c\u4e00\u90e8\u5206\u662f\u56fe\u7247\u7684\u6807\u7b7e":3,"\u7b2c\u4e09":61,"\u7b2c\u4e09\u5217\u662f\u751f\u6210\u7684\u82f1\u8bed\u5e8f\u5217":62,"\u7b2c\u4e09\u6b65":55,"\u7b2c\u4e09\u6b65\u5b8c\u6210\u540e":23,"\u7b2c\u4e8c":61,"\u7b2c\u4e8c\u4e2a\u4e3a":23,"\u7b2c\u4e8c\u4e2a\u65b9\u6cd5\u7684\u597d\u5904\u662fdockerfile\u53ef\u4ee5\u628a\u6574\u4e2a\u751f\u6210\u6d41\u7a0b\u63cf\u8ff0\u5f88\u6e05\u695a":27,"\u7b2c\u4e8c\u5217\u662f\u96c6\u675f\u641c\u7d22\u7684\u5f97\u5206":62,"\u7b2c\u4e8c\u6b65":[53,55],"\u7b2c\u4e8c\u6b65\u662f\u8fd0\u884c":27,"\u7b2c\u4e8c\u884c\u5b58\u7684\u662f\u56fe\u50cf":55,"\u7b2c\u4e8c\u90e8\u5206\u4e3b\u8981\u662f\u9009\u62e9\u5b66\u4e60\u7b97\u6cd5":25,"\u7b2c\u4e8c\u90e8\u5206\u662f28":3,"\u7b2ci\u884c\u7b2cj\u5217\u7684\u6570\u503c":37,"\u7b49":21,"\u7b49\u5168\u90e8\u9759\u6001\u5e93\u4e2d\u7684\u76ee\u6807\u6587\u4ef6\u5168\u90e8\u6253\u5305\u540e\u4ea7\u751f\u7684\u6587\u4ef6":21,"\u7b49\u5176\u4ed6":62,"\u7b49\u53c2\u6570":49,"\u7b49\u591a\u79cd\u516c\u6709\u4e91\u73af\u5883":47,"\u7b49\u5f85\u8fd9\u4e2a\u7a0b\u5e8f\u6267\u884c\u6210\u529f\u5e76\u8fd4\u56de0\u5219\u6210\u529f\u9000\u51fa":47,"\u7b49\u6587\u4ef6":21,"\u7b49\u7b49":[57,62],"\u7b49\u90fd\u5c5e\u4e8e\u4e00\u4e2a\u547d\u540d\u7a7a\u95f4":47,"\u7b80\u4ecb":39,"\u7b80\u5355\u6765\u8bf4":40,"\u7b80\u5355\u7684\u5168\u8fde\u63a5\u7f51\u7edc":24,"\u7b80\u5355\u7684\u57fa\u4e8e\u5b57\u6bcd\u7684\u8bcd\u5d4c\u5165":59,"\u7b80\u5355\u7684\u6027\u80fd\u5206\u6790":40,"\u7b80\u5355\u7684\u6574\u4e2a\u8bcd\u5d4c\u5165":59,"\u7b80\u5355\u7684pydataprovider2\u6837\u4f8b\u5c31\u8bf4\u660e\u5b8c\u6bd5\u4e86":3,"\u7b80\u5355\u7684yaml\u6587\u4ef6\u5982\u4e0b":48,"\u7b80\u76f4":32,"\u7b97\u6cd5":[24,25,35,61],"\u7b97\u6cd5\u4e2d\u7684beam\u5927\u5c0f":35,"\u7b97\u6cd5\u914d\u7f6e":61,"\u7ba1\u7406\u4eba\u5458":58,"\u7ba1\u7406\u5458":58,"\u7c7b\u4f3c":[21,31],"\u7c7b\u4f3c\u5730":60,"\u7c7b\u4f5c\u4e3a\u53c2\u6570\u7684\u62bd\u8c61":37,"\u7c7b\u522b\u4e2a\u6570":54,"\u7c7b\u522b\u4e2d\u7684\u53c2\u6570\u53ef\u7528\u4e8e\u6240\u6709\u573a\u5408":42,"\u7c7b\u522bid":57,"\u7c7b\u522bid\u548c\u6587\u672c\u4fe1\u606f\u7528":57,"\u7c7b\u540d\u548cc":20,"\u7c7b\u578b":[20,43,59],"\u7c7b\u578b\u53ef\u4ee5\u662fpaddlepaddle\u652f\u6301\u7684\u4efb\u610f\u8f93\u5165\u6570\u636e\u7c7b\u578b":31,"\u7c7b\u578b\u662fsparse_binary_vector":3,"\u7c7b\u578b\u662fsparse_float_vector":3,"\u7c7b\u578b\u6765\u8bbe\u7f6e":3,"\u7c7b\u578b\u7684":32,"\u7c7b\u7684\u6784\u9020\u51fd\u6570\u548c\u6790\u6784\u51fd\u6570":37,"\u7c7b\u9700\u8981\u5b9e\u73b0\u521d\u59cb\u5316":37,"\u7cfb\u7edf\u7f16\u8bd1wheel\u5305\u7684\u65f6\u5019":24,"\u7cfb\u7edfc":46,"\u7d2f\u52a0\u6c42\u548c":46,"\u7ea2\u697c\u68a6":53,"\u7eaa\u5f55\u7247":58,"\u7eafcpu\u955c\u50cf\u4ee5\u53cagpu\u955c\u50cf\u90fd\u4f1a\u7528\u5230avx\u6307\u4ee4\u96c6":27,"\u7ebf\u6027\u56de\u5f52\u7684\u8f93\u5165\u662f\u4e00\u6279\u70b9":25,"\u7ebf\u6027\u56de\u5f52\u7684\u8f93\u51fa\u662f\u4ece\u8fd9\u6279\u70b9\u4f30\u8ba1\u51fa\u6765\u7684\u53c2\u6570":25,"\u7ebf\u6027\u8ba1\u7b97\u7f51\u7edc\u5c42":25,"\u7ebf\u7a0bid\u53f7":45,"\u7ec4\u6210":46,"\u7ec6\u8282\u63cf\u8ff0":44,"\u7ecf\u5e38\u4f1a\u6d88\u8017\u657010gb\u7684\u5185\u5b58\u548c\u6570gb\u7684\u663e\u5b58":24,"\u7ed3\u5408":47,"\u7ed3\u675f\u6807\u8bb0":35,"\u7ed3\u6784\u5982\u4e0b":61,"\u7ed3\u6784\u5982\u4e0b\u56fe":53,"\u7ed3\u679c\u4fdd\u5b58\u5728":60,"\u7ed3\u679c\u53d1\u73b0\u5b83\u4eec\u4e0e\u540c\u65f6\u671f\u7684twitter\u6d88\u606f\u4e2d\u7684\u60c5\u7eea\u8bcd\u9891\u7387\u76f8\u5173":61,"\u7ed3\u8bba":20,"\u7ed9":32,"\u7ed9\u51fa\u56fe\u7247\u5c3a\u5bf8":54,"\u7ed9\u51fa\u8f93\u5165\u6570\u636e\u6240\u5728\u8def\u5f84":54,"\u7ed9\u5b9a\u52a8\u8bcd":60,"\u7ed9\u5b9a\u7684\u6587\u672c\u53ef\u4ee5\u662f\u4e00\u4e2a\u6587\u6863":61,"\u7ed9\u5b9aencoder\u8f93\u51fa\u548c\u5f53\u524d\u8bcd":34,"\u7ee7\u7eed\u6df1\u5165\u4e86\u89e3":46,"\u7ee7\u7eed\u8bad\u7ec3\u6216\u9884\u6d4b":3,"\u7ef4\u57fa\u767e\u79d1\u4e2d\u6587\u9875\u9762":32,"\u7ef4\u57fa\u767e\u79d1\u9875\u9762":32,"\u7ef4\u5ea6\u4e3aword_dim":57,"\u7ef4\u5ea6\u662f\u7c7b\u522b\u4e2a\u6570":57,"\u7ef4\u5ea6\u662f\u8bcd\u5178\u5927\u5c0f":57,"\u7ef4\u62a4":47,"\u7ef4\u7a7a\u95f4":35,"\u7ef4\u7a7a\u95f4\u5b8c\u6210":35,"\u7f13\u5b58\u6c60\u7684\u51cf\u5c0f":24,"\u7f13\u5b58\u8bad\u7ec3\u6570\u636e\u5230\u5185\u5b58":3,"\u7f16\u5199\u597d\u6570\u636e\u63d0\u4f9b\u811a\u672c\u540e":59,"\u7f16\u5199\u5b8cyaml\u6587\u4ef6\u540e":49,"\u7f16\u5199\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6":49,"\u7f16\u53f7":59,"\u7f16\u53f7\u4ece0\u5f00\u59cb":24,"\u7f16\u53f7\u5b57\u6bb5":59,"\u7f16\u7801\u5411\u91cf":35,"\u7f16\u7801\u5668\u8f93\u51fa":35,"\u7f16\u7801\u6e90\u5e8f\u5217":35,"\u7f16\u89e3\u7801\u6a21\u578b\u5c06\u4e00\u4e2a\u6e90\u8bed\u53e5\u7f16\u7801\u4e3a\u4e00\u4e2a\u5b9a\u957f\u7684\u5411\u91cf":62,"\u7f16\u8bd1":[27,36],"\u7f16\u8bd1\u5668\u6ca1\u6709":20,"\u7f16\u8bd1\u578b\u8bed\u8a00":20,"\u7f16\u8bd1\u5b8c\u6210\u4e4b\u540e":38,"\u7f16\u8bd1\u5b8c\u6210\u540e":38,"\u7f16\u8bd1\u6210\u52a8\u6001\u5e93":43,"\u7f16\u8bd1\u6d41\u7a0b\u4e3b\u8981\u63a8\u8350\u9ad8\u7ea7\u7528\u6237\u67e5\u770b":28,"\u7f16\u8bd1\u751f\u6210":38,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684docker\u53d1\u884c\u955c\u50cf":23,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684ubuntu":23,"\u7f16\u8bd1\u9009\u9879":26,"\u7f16\u8bd1c":21,"\u7f16\u8bd1master\u5206\u652f\u7684docker\u53d1\u884c\u955c\u50cf":23,"\u7f16\u8bd1ubuntu\u7684deb\u5305":23,"\u7f16\u8f91":47,"\u7f29\u653e\u53c2\u6570":55,"\u7f51\u7edc":[60,61],"\u7f51\u7edc\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":47,"\u7f51\u7edc\u540d\u79f0":57,"\u7f51\u7edc\u5c42\u53ef\u4ee5\u6709\u591a\u4e2a\u8f93\u5165":37,"\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u4e3a":37,"\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":37,"\u7f51\u7edc\u5c42\u7684\u7ec6\u8282\u53ef\u4ee5\u901a\u8fc7\u4e0b\u9762\u8fd9\u4e9b\u4ee3\u7801\u7247\u6bb5\u6765\u6307\u5b9a":37,"\u7f51\u7edc\u5c42\u7684\u8f93\u51fa\u662f\u7ecf\u8fc7\u6fc0\u6d3b\u51fd\u6570\u4e4b\u540e\u7684\u503c":43,"\u7f51\u7edc\u5c42\u914d\u7f6e\u5305\u542b\u4ee5\u4e0b\u51e0\u9879":37,"\u7f51\u7edc\u63a5\u53e3\u5361":41,"\u7f51\u7edc\u6a21\u5757":55,"\u7f51\u7edc\u6a21\u578b\u5c06\u8f93\u51fa\u6807\u7b7e\u7684\u6982\u7387\u5206\u5e03":60,"\u7f51\u7edc\u7684\u8bad\u7ec3\u8fc7\u7a0b":61,"\u7f51\u7edc\u7684\u8f93\u51fa\u4e3a\u795e\u7ecf\u7f51\u7edc\u7684\u4f18\u5316\u76ee\u6807":46,"\u7f51\u7edc\u7684\u8f93\u51fa\u4e5f\u53ef\u901a\u8fc7":46,"\u7f51\u7edc\u7ed3\u6784\u5982\u4e0b\u56fe\u6240\u793a":59,"\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e\u4e09\u90e8\u5206":46,"\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e\u8fd9\u4e09\u90e8\u5206\u8be5\u6982\u5ff5":46,"\u7f51\u7edc\u8fde\u63a5":46,"\u7f51\u7edc\u901a\u4fe1":37,"\u7f51\u7edc\u914d\u7f6e":[41,57,61],"\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":[55,60],"\u800c":[25,35,48,59],"\u800c\u4e0d\u4f7f\u7528imdb\u6570\u6910\u96c6\u4e2d\u7684imdb":61,"\u800c\u4e0d\u5fc5\u5728\u610fpaddl":21,"\u800c\u4e0d\u652f\u6301pypy\u89e3\u91ca\u5668":20,"\u800c\u4e0d\u662f\u4f7f\u7528\u540c\u6b65":41,"\u800c\u4e0d\u662f\u65b0\u6570\u636e\u9a71\u52a8\u7cfb\u7edf":46,"\u800c\u4e0d\u662f\u6e90\u7801\u76ee\u5f55\u91cc":24,"\u800c\u4e0d\u662f\u7279\u5f81\u7684\u96c6\u5408":32,"\u800c\u4e0d\u662f\u7ec4\u5408\u4e0a\u4e0b\u6587\u7ea7\u522b\u4fe1\u606f":61,"\u800c\u4e0d\u66b4\u9732\u6982\u5ff5\u7684\u5b9e\u73b0":21,"\u800c\u4e0d\u7528\u5173\u5fc3\u6570\u636e\u5982\u4f55\u4f20\u8f93":3,"\u800c\u4e14":62,"\u800c\u4e4b\u524d\u7684\u53c2\u6570\u5c06\u4f1a\u88ab\u5220\u9664":43,"\u800c\u4ece\u5e94\u7528\u7684\u89d2\u5ea6":40,"\u800c\u4f18\u5316\u6027\u80fd\u7684\u9996\u8981\u4efb\u52a1":40,"\u800c\u5176\u4ed6\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":45,"\u800c\u51fa\u73b0\u9636\u6bb5\u6027\u7684\u8fd0\u884c\u505c\u6ede":18,"\u800c\u53cc\u5c42rnn\u662f\u53ef\u4ee5\u5904\u7406\u8fd9\u79cd\u8f93\u5165\u6570\u636e\u7684\u7f51\u7edc\u7ed3\u6784":32,"\u800c\u53f3\u56fe\u7684\u74f6\u9888\u8fde\u63a5\u6a21\u5757\u7528\u4e8e50\u5c42":55,"\u800c\u5728cpp\u91cc\u9762\u5b9e\u73b0\u8fd9\u4e2ac\u7684\u63a5\u53e3":20,"\u800c\u591a\u8bed\u8a00\u63a5\u53e3\u9700\u8981\u76f4\u63a5\u8bfb\u53d6\u751f\u6210\u7684\u4e8c\u8fdb\u5236":20,"\u800c\u5927\u591a\u6570\u65b9\u6cd5\u53ea\u662f\u5229\u7528n":61,"\u800c\u5bf9\u4e8e\u53cc\u5c42\u5e8f\u5217":32,"\u800c\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5185\u5c42\u7279\u5f81\u6570\u636e\u800c\u8a00":32,"\u800c\u5bf9\u4e8egolang":20,"\u800c\u5bf9\u4e8egolang\u9519\u8bef\u5904\u7406\u5e94\u8be5\u4f7f\u7528\u8fd4\u56de\u503c":20,"\u800c\u5c06\u8fd9\u4e2a\u6bb5\u843d\u7684\u6bcf\u4e00\u53e5\u8bdd\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u5411\u91cf":32,"\u800c\u5f53\u524d\u5df2\u7ecf\u67095":40,"\u800c\u662f\u76f4\u63a5\u4ece\u5185\u5b58\u7684\u7f13\u5b58\u91cc\u8bfb\u53d6\u6570\u636e":24,"\u800c\u662f\u76f4\u63a5\u4fee\u6539paddl":21,"\u800c\u66f4\u6df1\u5165\u7684\u5206\u6790":40,"\u800c\u6709\u4e9b\u53c2\u6570\u9700\u8981\u5728\u96c6\u7fa4\u591a\u673a\u8bad\u7ec3\u4e2d\u4f7f\u7528\u7b49":42,"\u800c\u6ca1\u6709\u77ed\u65f6\u8bb0\u5fc6\u7684\u635f\u5931":61,"\u800c\u6e90\u5e8f\u5217\u7684\u7f16\u7801\u5411\u91cf\u53ef\u4ee5\u88ab\u65e0\u8fb9\u754c\u7684memory\u8bbf\u95ee":35,"\u800c\u7a00\u758f\u66f4\u65b0\u5728\u53cd\u5411\u4f20\u64ad\u4e4b\u540e\u7684\u6743\u91cd\u66f4\u65b0\u65f6\u8fdb\u884c":45,"\u800c\u7cfb\u7edf\u4e2d\u7684":24,"\u800c\u8fd9\u4e00\u53e5\u8bdd\u5c31\u53ef\u4ee5\u8868\u793a\u6210\u8fd9\u4e9b\u4f4d\u7f6e\u7684\u6570\u7ec4":32,"\u800c\u8fd9\u4e2acontext\u53ef\u80fd\u4f1a\u975e\u5e38\u5927":3,"\u800c\u8fd9\u6bcf\u4e00\u4e2a\u6570\u7ec4\u5143\u7d20":32,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":26,"\u800cpaddlepaddle\u5219\u4f1a\u5e2e\u7528\u6237\u505a\u4ee5\u4e0b\u5de5\u4f5c":3,"\u800crnn\u662f\u6700\u6d41\u884c\u7684\u9009\u62e9":34,"\u800cswig\u53ea\u80fd\u7b80\u5355\u7684\u66b4\u9732c":20,"\u800cweight":54,"\u804c\u4e1a":58,"\u804c\u4e1a\u4ece\u4e0b\u9762\u6240\u5217\u4e2d\u9009\u62e9":58,"\u80fd\u591f\u5904\u7406\u53cc\u5c42\u5e8f\u5217":34,"\u80fd\u591f\u5bf9\u53cc\u5411\u5e8f\u5217\u8fdb\u884c\u5904\u7406\u7684\u6709":34,"\u80fd\u591f\u627e\u5230\u8fd9\u91cc\u4f7f\u7528\u7684\u6240\u6709\u6570\u636e":57,"\u80fd\u591f\u8bb0\u5f55\u4e0a\u4e00\u4e2asubseq":34,"\u80fd\u83b7\u53d6":41,"\u811a\u672c":[41,54,59],"\u811a\u672c\u4fdd\u5b58\u5728":54,"\u811a\u672c\u5f00\u59cb\u65f6":49,"\u811a\u672c\u63d0\u4f9b\u4e86\u4e00\u4e2a\u9884\u6d4b\u63a5\u53e3":61,"\u811a\u672c\u65f6\u9700\u8981\u52a0\u4e0a":61,"\u811a\u672c\u8fd0\u884c\u5b8c\u6210\u540e":54,"\u81ea\u52a8\u5173\u95ed\u5bf9\u5e94\u7684":36,"\u81ea\u52a8\u5730\u5c06\u8fd9\u4e9b\u9009\u9879\u5e94\u7528\u5230":41,"\u81ea\u52a8\u5b8c\u6210\u8fd9\u4e00\u8fc7\u7a0b":34,"\u81ea\u52a8\u6302\u8f7d\u5206\u5e03\u5f0f\u5b58\u50a8\u76ee\u5f55":18,"\u81ea\u52a8\u751f\u6210":38,"\u81ea\u52a8\u83b7\u53d6\u4e0a\u4e00\u4e2a\u751f\u6210\u7684\u8bcd":35,"\u81ea\u5e95\u5411\u4e0a\u6cd5":57,"\u81ea\u7136\u8bed\u8a00\u5904\u7406\u7b49":45,"\u81ea\u7531\u804c\u4e1a\u8005":58,"\u81f3\u4e8e\u4e3a\u4ec0\u4e48\u9700\u8981c":21,"\u81f3\u6b64":[3,32,36],"\u8212\u9002":32,"\u826f\u597d\u7684\u6587\u6863":20,"\u827a\u672f\u5bb6":58,"\u8282\u70b9\u4e2d\u7684":41,"\u82e5":37,"\u82e5\u5e72\u4e2a\u53e5\u5b50\u6784\u6210\u4e00\u4e2a\u6bb5\u843d":31,"\u82e5\u6709\u4e0d\u4e00\u81f4\u4e4b\u5904":40,"\u82e5\u6709\u5fc5\u8981":37,"\u82e5\u8f93\u51fa\u662f\u5355\u5c42\u5e8f\u5217":31,"\u82e5\u8f93\u51fa\u662f\u53cc\u5c42\u5e8f\u5217":31,"\u82f1\u6587\u6587\u6863":38,"\u82f1\u6587\u6587\u6863\u76ee\u5f55":38,"\u82f1\u8bed":62,"\u8303\u56f4":45,"\u83b7\u53d6":36,"\u83b7\u53d6\u5229\u7528":57,"\u83b7\u53d6\u5b57\u5178\u7ef4\u5ea6":61,"\u83b7\u53d6\u6700\u65b0\u7684\u68c0\u67e5\u70b9\u7684\u6587\u4ef6uuid":18,"\u83b7\u53d6\u8be5\u6761\u6837\u672c\u7c7b\u522bid":57,"\u83b7\u53d6\u901a\u8fc7":61,"\u83b7\u53d6trainer":49,"\u83b7\u5f97\u53c2\u6570\u5c3a\u5bf8":37,"\u867d\u7136":25,"\u867d\u7136\u4e0d\u9f13\u52b1\u8fd9\u6837":21,"\u867d\u7136\u6bcf\u4e2agenerator\u5728\u6ca1\u6709\u8c03\u7528\u7684\u65f6\u5019":3,"\u867d\u7136\u8fd9\u4e9b\u6587\u4ef6\u5e76\u975e\u90fd\u9700\u8981\u96c6\u7fa4\u8bad\u7ec3":41,"\u867d\u7136paddle\u770b\u8d77\u6765\u5305\u542b\u4e86\u4f17\u591a\u53c2\u6570":42,"\u884c":53,"\u884c\u4f18\u5148\u6b21\u5e8f\u5b58\u50a8":55,"\u884c\u5185\u4f7f\u7528":3,"\u884c\u653f\u5de5\u4f5c":58,"\u884c\u955c\u50cf\u5305\u62ec\u7eafcpu\u7248\u672c\u548cgpu\u7248\u672c\u4ee5\u53ca\u5176\u5bf9\u5e94\u7684\u975eavx\u7248\u672c":27,"\u8868\u660e\u4e86\u8fd9\u4e9b\u884c\u7684\u6807\u53f7":37,"\u8868\u660e\u8fd9\u4e2a\u5c42\u7684\u4e00\u4e2a\u5b9e\u4f8b\u662f\u5426\u9700\u8981\u504f\u7f6e":37,"\u8868\u793a\u4e00\u4e2akubernetes\u96c6\u7fa4\u4e2d\u7684\u4e00\u4e2a\u5de5\u4f5c\u8282\u70b9":47,"\u8868\u793a\u4e3adeviceid":45,"\u8868\u793a\u5728\u96c6\u7fa4\u4f5c\u4e1a":41,"\u8868\u793a\u5973\u6027":58,"\u8868\u793a\u5c06\u5916\u5c42\u7684outer_mem\u4f5c\u4e3a\u5185\u5c42memory\u7684\u521d\u59cb\u72b6\u6001":32,"\u8868\u793a\u5f53\u524d\u96c6\u7fa4\u4f5c\u4e1a\u7684\u8282\u70b9":41,"\u8868\u793a\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":41,"\u8868\u793a\u751f\u6210\u6570\u636e\u7684\u5e8f\u5217id":62,"\u8868\u793a\u7528\u4e8e\u8bad\u7ec3\u6216\u9884\u6d4b":3,"\u8868\u793a\u7537\u6027":58,"\u8868\u793a\u7684\u6bcf\u4e2a\u5355\u8bcd":57,"\u8868\u793a\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u7684\u5206\u7c7b\u9519\u8bef":61,"\u8868\u793a\u8bad\u7ec3\u4e86xx\u4e2a\u6837\u672c":61,"\u8868\u793a\u8bad\u7ec3\u4e86xx\u4e2abatch":61,"\u8868\u793a\u8bfb\u8005\u6240\u4f7f\u7528\u7684docker\u955c\u50cf\u4ed3\u5e93\u5730\u5740":49,"\u8868\u793a\u8fc7\u4e8620\u4e2abatch":57,"\u8868\u793a\u8fc7\u4e862560\u4e2a\u6837\u672c":57,"\u8868\u793a\u8fd9\u4e2ajob\u7684\u540d\u5b57":49,"\u88ab":36,"\u88ab\u6269\u5c55\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"\u88ab\u653e\u5728":37,"\u88ab\u79f0\u4e3a":35,"\u88ab\u79f0\u4e3a\u6570\u636e\u63d0\u4f9b\u5668":46,"\u897f\u90e8\u7247":58,"\u8981\u4e0b\u8f7d\u548c\u89e3\u538b\u6570\u636e\u96c6":59,"\u8981\u4e0b\u8f7d\u89e3\u538b\u8fd9\u4e2a\u6a21\u578b":62,"\u8981\u4f7f\u7528\u547d\u4ee4\u884c\u5206\u6790\u5de5\u5177":40,"\u8981\u5728\u5df2\u6709\u7684kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884cpaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":49,"\u8981\u5728\u6240\u6709\u8282\u70b9\u4e0a\u5b58\u5728":41,"\u8981\u5bf9\u4e00\u4e2a\u56fe\u7247\u7684\u8fdb\u884c\u5206\u7c7b\u9884\u6d4b":54,"\u8981\u5c06\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6\u8f6c\u5316\u4e3ameta\u914d\u7f6e\u6587\u4ef6":59,"\u8981\u6c42\u5355\u5c42\u5e8f\u5217\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee":31,"\u8981\u751f\u6210\u7684\u76ee\u6807\u5e8f\u5217":34,"\u8981\u8c03\u7528":37,"\u89e3\u51b3\u529e\u6cd5\u662f":24,"\u89e3\u51b3\u65b9\u6848\u662f":24,"\u89e3\u538b":62,"\u89e3\u6790\u5668\u80fd\u901a\u8fc7\u6587\u4ef6\u7684\u6269\u5c55\u540d\u81ea\u52a8\u8bc6\u522b\u6587\u4ef6\u7684\u683c\u5f0f":59,"\u89e3\u6790\u6570\u636e\u96c6\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5b57\u6bb5":59,"\u89e3\u6790\u6a21\u578b\u914d\u7f6e\u6587\u4ef6":5,"\u89e3\u6790\u73af\u5883\u53d8\u91cf\u5f97\u5230":49,"\u89e3\u6790\u8bad\u7ec3\u6a21\u578b\u65f6\u7528\u7684\u914d\u7f6e\u6587\u4ef6":5,"\u89e3\u7801\u5668\u4f7f\u7528":35,"\u89e3\u7801\u5668\u57fa\u4e8e\u7f16\u7801\u6e90\u5e8f\u5217\u548c\u6700\u540e\u751f\u6210\u7684\u76ee\u6807\u8bcd\u9884\u6d4b\u4e0b\u4e00\u76ee\u6807\u8bcd":35,"\u89e3\u7801\u5668\u662f\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":35,"\u89e3\u7801\u5668\u6839\u636e\u4e0a\u4e0b\u6587\u5411\u91cf\u9884\u6d4b\u51fa\u4e00\u4e2a\u76ee\u6807\u5355\u8bcd":62,"\u89e3\u91ca":57,"\u89e3\u91ca\u578b\u8bed\u8a00\u53ea\u80fd\u8c03\u7528\u52a8\u6001\u5e93":20,"\u89e3\u91ca\u6027\u8bed\u8a00\u5b9e\u9645\u8fd0\u884c\u7684\u4e8c\u8fdb\u5236\u662f\u89e3\u91ca\u5668\u672c\u8eab":20,"\u8ba1\u7b97":35,"\u8ba1\u7b97\u504f\u7f6e\u7684\u68af\u5ea6":37,"\u8ba1\u7b97\u5355\u5143\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u6838\u5fc3":25,"\u8ba1\u7b97\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":35,"\u8ba1\u7b97\u53d8\u6362\u77e9\u9635\u7684\u5927\u5c0f\u548c\u683c\u5f0f":37,"\u8ba1\u7b97\u5f53\u524d\u5c42\u6743\u91cd\u7684\u68af\u5ea6":37,"\u8ba1\u7b97\u6bcf\u4e2a\u8bcd\u7684\u8bcd\u5411\u91cf":35,"\u8ba1\u7b97\u6fc0\u6d3b\u51fd\u6570\u7684\u68af\u5ea6":37,"\u8ba1\u7b97\u7684\u7ec6\u8282\u5c06\u5728\u4e0b\u9762\u7684\u5c0f\u8282\u7ed9\u51fa":37,"\u8ba1\u7b97\u8bef\u5dee\u51fd\u6570":25,"\u8ba1\u7b97\u8f6c\u6362\u77e9\u9635\u548c\u8f93\u5165\u7684\u68af\u5ea6":37,"\u8ba1\u7b97\u8f93\u5165\u548c\u53c2\u6570\u7684\u68af\u5ea6":37,"\u8ba1\u7b97\u8f93\u5165\u5c42\u7684\u504f\u5dee":37,"\u8ba1\u7b97\u8f93\u51fa":37,"\u8ba1\u7b97\u8fd9\u4e2a\u6587\u4ef6\u7684md5":18,"\u8ba9\u6a21\u578b\u80fd\u591f\u5f97\u5230\u8bad\u7ec3\u66f4\u65b0":57,"\u8ba9\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fdb\u884c\u8bad\u7ec3\u6216\u9884\u6d4b":2,"\u8ba9\u8fd9\u4e2a\u793a\u4f8b\u53d8\u5f97\u66f4\u597d":59,"\u8ba9paddle\u6838\u5fc3\u4e2d":21,"\u8bad\u7ec3":[42,61],"\u8bad\u7ec3\u4e0e\u5e94\u7528":1,"\u8bad\u7ec3\u4efb\u52a1\u7684\u8fd0\u884c\u53ef\u80fd\u4f1a\u5360\u6ee1trainer\u548cparamet":18,"\u8bad\u7ec3\u4f5c\u4e1a":41,"\u8bad\u7ec3\u53ca\u6d4b\u8bd5\u8bef\u5dee\u66f2\u7ebf\u56fe\u4f1a\u88ab":54,"\u8bad\u7ec3\u53ef\u4ee5\u8bbe\u7f6e\u4e3atrue":60,"\u8bad\u7ec3\u540e":60,"\u8bad\u7ec3\u548c\u7eaf\u4f7f\u7528":23,"\u8bad\u7ec3\u5931\u8d25\u65f6\u53ef\u4ee5\u68c0\u67e5\u9519\u8bef\u65e5\u5fd7":41,"\u8bad\u7ec3\u597d\u4e00\u4e2a\u6df1\u5c42\u795e\u7ecf\u7f51\u7edc\u901a\u5e38\u8981\u8017\u8d39\u975e\u5e38\u957f\u7684\u65f6\u95f4":40,"\u8bad\u7ec3\u5b8c\u6210\u540e":25,"\u8bad\u7ec3\u6570\u636e\u548c\u6d4b\u8bd5\u6570\u636e\u7684\u76ee\u5f55":62,"\u8bad\u7ec3\u6570\u636e\u662f":3,"\u8bad\u7ec3\u6570\u636e\u6709\u95ee\u9898":24,"\u8bad\u7ec3\u6570\u636e\u7684\u683c\u5f0f\u5f80\u5f80\u5404\u4e0d\u76f8\u540c":46,"\u8bad\u7ec3\u6570\u6910\u96c6":61,"\u8bad\u7ec3\u65f6":49,"\u8bad\u7ec3\u65f6\u6240\u9700\u8bbe\u7f6e\u7684\u4e3b\u8981\u53c2\u6570\u5982\u4e0b":57,"\u8bad\u7ec3\u65f6\u9ed8\u8ba4shuffl":3,"\u8bad\u7ec3\u6a21\u578b\u4e4b\u524d":61,"\u8bad\u7ec3\u6a21\u578b\u540e":35,"\u8bad\u7ec3\u6a21\u578b\u6b63\u786e\u6027":23,"\u8bad\u7ec3\u7684\u635f\u5931\u51fd\u6570\u9ed8\u8ba4\u6bcf\u969410\u4e2abatch\u6253\u5370\u4e00\u6b21":62,"\u8bad\u7ec3\u7684\u811a\u672c\u662f":60,"\u8bad\u7ec3\u7b97\u6cd5":46,"\u8bad\u7ec3\u7b97\u6cd5\u901a\u5e38\u5b9a\u4e49\u5728\u53e6\u4e00\u5355\u72ecpython\u6587\u4ef6\u4e2d":46,"\u8bad\u7ec3\u7ed3\u675f\u540e\u67e5\u770b\u8f93\u51fa\u7ed3\u679c":49,"\u8bad\u7ec3\u811a\u672c":57,"\u8bad\u7ec3\u811a\u672c\u7b49\u7b49":57,"\u8bad\u7ec3\u81f3\u591a":59,"\u8bad\u7ec3\u8282\u70b9\u6570\u91cf":49,"\u8bad\u7ec3\u8bed\u8a00\u6a21\u578b\u8ddd\u79bb":24,"\u8bad\u7ec3\u8f6e\u6b21":57,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u53c2\u6570\u6216\u8005\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u7684\u68af\u5ea6\u5c3a\u5ea6\u8fc7\u5927":24,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6d4b\u8bd5test_period":42,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4e3a\u672c\u5730\u6a21\u5f0f":43,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4f7f\u7528gpu":43,"\u8bad\u7ec3\u8fdb\u7a0b":46,"\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u8bbe\u5907\u5c5e\u6027\u5c06\u4f1a\u65e0\u6548":43,"\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6\u4e3b\u8981\u5305\u62ec\u6570\u636e\u6e90":46,"\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6\u7684\u6570\u636e\u6e90\u914d\u7f6e\u4e2d\u6307\u5b9adataprovider\u6587\u4ef6\u540d\u5b57":46,"\u8bad\u7ec3\u9636\u6bb5":46,"\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\u6587\u4ef6\u5217\u8868":61,"\u8bad\u7ec3\u96c6\u5df2\u7ecf\u505a\u4e86\u968f\u673a\u6253\u4e71\u6392\u5e8f\u800c\u6d4b\u8bd5\u96c6\u6ca1\u6709":61,"\u8bad\u7ec3\u96c6\u5df2\u7ecf\u968f\u673a\u6253\u4e71":61,"\u8bad\u7ec3\u96c6\u5e73\u5747\u503c":54,"\u8bad\u7ec3dot_period":42,"\u8bb0\u5f55\u4e0b\u6240\u6709\u5931\u8d25\u7684\u4f8b\u5b50":23,"\u8bb0\u5fc6\u6a21\u5757":35,"\u8bba\u6587":55,"\u8bbe\u4e3a\u5df2\u90e8\u7f72\u7684\u5de5\u4f5c\u7a7a\u95f4\u76ee\u5f55":41,"\u8bbe\u4e3a\u672c\u5730":41,"\u8bbe\u7f6e":21,"\u8bbe\u7f6e\u4e3a":37,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528\u672c\u5730\u8bad\u7ec3\u6216\u8005\u4f7f\u7528\u96c6\u7fa4\u4e0a\u7684\u4e00\u4e2a\u8282\u70b9":43,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528gpu\u6a21\u5f0f":43,"\u8bbe\u7f6e\u4efb\u52a1\u7684\u6a21\u5f0f\u4e3a\u6d4b\u8bd5":62,"\u8bbe\u7f6e\u4fdd\u5b58\u6a21\u578b\u7684\u8f93\u51fa\u8def\u5f84":62,"\u8bbe\u7f6e\u5168\u5c40\u5b66\u4e60\u7387":61,"\u8bbe\u7f6e\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":3,"\u8bbe\u7f6e\u5185\u5b58\u4e2d\u6700\u5c0f\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":3,"\u8bbe\u7f6e\u53c2\u6570\u7684\u540d\u5b57":24,"\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570":[24,39,59],"\u8bbe\u7f6e\u5b57\u5178\u6587\u4ef6":61,"\u8bbe\u7f6e\u5de5\u4f5c\u6a21\u5f0f\u4e3a\u8bad\u7ec3":61,"\u8bbe\u7f6e\u5e73\u5747sgd\u7a97\u53e3":61,"\u8bbe\u7f6e\u6210":24,"\u8bbe\u7f6e\u6210\u4e00\u4e2a\u5c0f\u4e00\u4e9b\u7684\u503c":24,"\u8bbe\u7f6e\u6570\u636e\u904d\u5386\u6b21\u6570":60,"\u8bbe\u7f6e\u6807\u7b7e\u7c7b\u522b\u5b57\u5178":61,"\u8bbe\u7f6e\u6a21\u578b\u8def\u5f84":61,"\u8bbe\u7f6e\u7684\u547d\u4ee4\u884c\u53c2\u6570":61,"\u8bbe\u7f6e\u795e\u7ecf\u7f51\u7edc\u7684\u914d\u7f6e\u6587\u4ef6":62,"\u8bbe\u7f6e\u7c7b\u522b\u6570":61,"\u8bbe\u7f6e\u7ebf\u7a0b\u6570":[60,61],"\u8bbe\u7f6e\u7f51\u7edc\u914d\u7f6e":61,"\u8bbe\u7f6e\u8f93\u51fa\u7684\u5c3a\u5bf8":37,"\u8bbe\u7f6e\u8f93\u51fa\u8def\u5f84\u4ee5\u4fdd\u5b58\u8bad\u7ec3\u5b8c\u6210\u7684\u6a21\u578b":61,"\u8bbe\u7f6e\u8fd9\u4e2apydataprovider2\u8fd4\u56de\u4ec0\u4e48\u6837\u7684\u6570\u636e":3,"\u8bbe\u7f6e\u9ed8\u8ba4\u8bbe\u5907\u53f7\u4e3a0":45,"\u8bbe\u7f6ebatch":61,"\u8bbe\u7f6ecpu\u7ebf\u7a0b\u6570\u6216\u8005gpu\u8bbe\u5907\u6570":62,"\u8bbe\u7f6egpu":43,"\u8bbe\u7f6epass":61,"\u8bbe\u7f6epasses\u7684\u6570\u91cf":62,"\u8bbf\u95eekubernetes\u7684\u63a5\u53e3\u6765\u67e5\u8be2\u6b64job\u5bf9\u5e94\u7684\u6240\u6709pod\u4fe1\u606f":49,"\u8bc4\u4ef7\u9884\u6d4b\u7684\u6548\u679c":25,"\u8bc4\u4f30\u5668":46,"\u8bc4\u4f30\u5668\u53ef\u4ee5\u8bc4\u4ef7\u6a21\u578b\u7ed3\u679c":46,"\u8bc4\u4f30\u8be5\u4ea7\u54c1\u7684\u8d28\u91cf":57,"\u8bc4\u5206":[58,59],"\u8bc4\u5206\u6587\u4ef6\u7684\u6bcf\u4e00\u884c\u4ec5\u4ec5\u63d0\u4f9b\u7535\u5f71\u6216\u7528\u6237\u7684\u7f16\u53f7\u6765\u4ee3\u8868\u76f8\u5e94\u7684\u7535\u5f71\u6216\u7528\u6237":59,"\u8bc4\u5206\u88ab\u8c03\u6574\u4e3a5\u661f\u7684\u89c4\u6a21":58,"\u8bc6\u522b\u6570\u5b57":23,"\u8bcd\u5411\u91cf":[23,53],"\u8bcd\u5411\u91cf\u6a21\u578b":56,"\u8bcd\u5411\u91cf\u6a21\u578b\u540d\u79f0":53,"\u8bcd\u672c\u8eab\u548c\u8bcd\u9891":53,"\u8bcd\u9891\u6700\u9ad8\u7684":62,"\u8bd5\u7740\u8ba9\u8f93\u51fa\u7684\u5206\u6790\u6570\u636e\u548c\u7406\u8bba\u503c\u5bf9\u5e94":40,"\u8be5":[41,60],"\u8be5\u51fd\u6570\u5177\u6709\u4e24\u4e2a\u53c2\u6570":3,"\u8be5\u51fd\u6570\u5728\u521d\u59cb\u5316\u7684\u65f6\u5019\u4f1a\u88ab\u8c03\u7528":3,"\u8be5\u51fd\u6570\u7684\u529f\u80fd\u662f":3,"\u8be5\u53c2\u6570\u5728\u7f51\u7edc\u914d\u7f6e\u7684output":43,"\u8be5\u53c2\u6570\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":43,"\u8be5\u53c2\u6570\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":43,"\u8be5\u53c2\u6570\u5fc5\u987b\u80fd\u88abflag":43,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u65e5\u5fd7\u622a\u65ad\u4fe1\u606f":43,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u9519\u8bef\u622a\u65ad\u65e5\u5fd7":43,"\u8be5\u53c2\u6570\u7528\u4e8e\u6307\u5b9a\u52a8\u6001\u5e93\u8def\u5f84":43,"\u8be5\u53c2\u6570\u7684\u610f\u601d\u662f\u8bad\u7ec3num":43,"\u8be5\u53c2\u6570\u9ed8\u8ba4\u4e3anull":43,"\u8be5\u5bf9\u8c61\u5177\u6709\u4ee5\u4e0b\u4e24\u4e2a\u5c5e\u6027":3,"\u8be5\u5c42\u4ec5\u9700\u8981\u8fd9\u4e9b\u975e\u96f6\u6837\u672c\u4f4d\u7f6e\u6240\u5bf9\u5e94\u7684\u53d8\u6362\u77e9\u9635\u7684\u90a3\u4e9b\u884c":37,"\u8be5\u5c42\u795e\u7ecf\u5143\u4e2a\u6570":57,"\u8be5\u622a\u65ad\u4f1a\u5f71\u54cd":43,"\u8be5\u6279\u6b21\u7684\u8f93\u5165\u4e2d\u4ec5\u6709\u4e00\u4e2a\u5b50\u96c6\u662f\u975e\u96f6\u7684":37,"\u8be5\u63a5\u53e3\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bfb\u53d6\u6570\u636e":3,"\u8be5\u63a5\u53e3\u53ef\u7528\u4e8e\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":26,"\u8be5\u6570\u636e\u53ca\u6709\u5f88\u591a\u4e0d\u540c\u7684\u7248\u672c":58,"\u8be5\u6570\u636e\u96c6":53,"\u8be5\u6570\u636e\u96c6\u4e8e2003\u5e742\u6708\u53d1\u5e03":58,"\u8be5\u6570\u636e\u96c6\u5305\u542b\u4e00\u4e9b\u7528\u6237\u4fe1\u606f":58,"\u8be5\u6570\u76ee\u662f\u63d0\u524d\u5b9a\u4e49\u597d\u7684":43,"\u8be5\u6587\u4ef6\u53ef\u4ee5\u4ece\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6\u751f\u6210":59,"\u8be5\u6587\u4ef6\u662f\u7531cpickle\u4ea7\u751f\u7684":55,"\u8be5\u6587\u4ef6\u662fpython\u7684pickle\u5bf9\u8c61":59,"\u8be5\u6587\u4ef6\u8d1f\u8d23\u4ea7\u751f\u56fe\u7247\u6570\u636e\u5e76\u4f20\u9012\u7ed9paddle\u7cfb\u7edf":54,"\u8be5\u6a21\u578b\u4f9d\u7136\u4f7f\u7528\u903b\u8f91\u56de\u5f52\u5206\u7c7b\u7f51\u7edc\u7684\u6846\u67b6":57,"\u8be5\u6a21\u578b\u5728\u957f\u8bed\u53e5\u7ffb\u8bd1\u7684\u573a\u666f\u4e0b\u6548\u679c\u63d0\u5347\u66f4\u52a0\u660e\u663e":62,"\u8be5\u6a21\u578b\u7684\u7f51\u7edc\u914d\u7f6e\u5982\u4e0b":25,"\u8be5\u6a21\u578b\u7684\u8bf4\u660e\u5982\u4e0b\u56fe\u6240\u793a":35,"\u8be5\u6a21\u578b\u7f51\u7edc\u53ea\u662f\u7528\u4e8e\u8fdb\u884cdemo\u5c55\u793apaddle\u5982\u4f55\u5de5\u4f5c":59,"\u8be5\u793a\u4f8b\u5c06\u5c55\u793apaddle\u5982\u4f55\u8fdb\u884c\u8bcd\u5411\u91cf\u5d4c\u5165":59,"\u8be5\u793a\u4f8b\u7684\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":59,"\u8be5\u7b97\u6cd5\u6bcf\u6279\u91cf":25,"\u8be5\u7c7b\u7684\u5b9e\u73b0\u7ec6\u8282\u5728":37,"\u8be5\u811a\u672c\u4ec5\u4ec5\u662f\u5f00\u59cb\u4e00\u4e2apaddle\u8bad\u7ec3\u8fc7\u7a0b":59,"\u8be5\u811a\u672c\u4f1a\u751f\u6210\u4e00\u4e2adot\u6587\u4ef6":55,"\u8be5\u811a\u672c\u5c06\u8f93\u51fa\u9884\u6d4b\u5206\u7c7b\u7684\u6807\u7b7e":54,"\u8be5\u8bed\u53e5\u4f1a\u4e3a\u6bcf\u4e2a\u5c42\u521d\u59cb\u5316\u5176\u6240\u9700\u8981\u7684\u53d8\u91cf\u548c\u8fde\u63a5":37,"\u8be5layer\u5c06\u591a\u4e2a\u8f93\u5165":46,"\u8be5python\u4ee3\u7801\u53ef\u4ee5\u751f\u6210protobuf\u5305":46,"\u8be6\u7ec6\u4ecb\u7ecd\u53ef\u4ee5\u53c2\u8003":32,"\u8be6\u7ec6\u4fe1\u606f\u8bf7\u68c0\u67e5":41,"\u8be6\u7ec6\u5185\u5bb9\u8bf7\u53c2\u89c1":57,"\u8be6\u7ec6\u53ef\u4ee5\u53c2\u8003":46,"\u8be6\u7ec6\u5730\u5c55\u793a\u4e86\u6574\u4e2a\u7279\u5f81\u63d0\u53d6\u7684\u8fc7\u7a0b":55,"\u8be6\u7ec6\u6587\u6863\u53c2\u8003":24,"\u8be6\u7ec6\u7684\u53c2\u6570\u89e3\u91ca":57,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":26,"\u8be6\u7ec6\u89c1":31,"\u8bed\u4e49\u89d2\u8272\u6807\u6ce8":[56,60],"\u8bed\u610f\u89d2\u8272\u6807\u6ce8":23,"\u8bed\u8a00\u6a21\u578b":53,"\u8bf4\u660e":[18,26,49],"\u8bf4\u660e\u6bcf\u4e2a\u7279\u5f81\u6587\u4ef6\u5177\u4f53\u5b57\u6bb5\u662f":59,"\u8bf4\u660e\u8fd9\u4e2a\u5c42\u7684\u8f93\u5165":37,"\u8bf7\u4e0d\u8981\u6df7\u6dc6":46,"\u8bf7\u518d\u6b21\u786e\u8ba4py_paddle\u5305\u5df2\u7ecf\u5b89\u88c5":38,"\u8bf7\u53c2\u7167\u7f51\u7edc\u914d\u7f6e\u7684\u6587\u6863\u4e86\u89e3\u66f4\u8be6\u7ec6\u7684\u4fe1\u606f":45,"\u8bf7\u53c2\u8003":[3,21,24,32,35,37,38,46,57],"\u8bf7\u53c2\u8003\u5982\u4e0b\u8868\u683c":57,"\u8bf7\u53c2\u8003\u9875\u9762":59,"\u8bf7\u53c2\u8003layer\u6587\u6863":54,"\u8bf7\u53c2\u89c1":36,"\u8bf7\u53c2\u9605":35,"\u8bf7\u53c2\u9605\u60c5\u611f\u5206\u6790\u7684\u6f14\u793a\u4ee5\u4e86\u89e3\u6709\u5173\u957f\u671f\u77ed\u671f\u8bb0\u5fc6\u5355\u5143\u7684\u66f4\u591a\u4fe1\u606f":60,"\u8bf7\u5b89\u88c5cuda":29,"\u8bf7\u5c06with_doc\u9009\u9879\u5173\u95ed":38,"\u8bf7\u628a\u6587\u6863\u91cc\u547d\u4ee4\u4e2d\u7684paddlepaddl":27,"\u8bf7\u6307\u5b9a\u8be5\u76ee\u5f55":43,"\u8bf7\u67e5\u770b":53,"\u8bf7\u6ce8\u610f":[27,35,48,53],"\u8bf7\u770b\u4e0b\u9762\u7684\u4f8b\u5b50":45,"\u8bf7\u786e\u4fdd":36,"\u8bf7\u8bb0\u4f4f":41,"\u8bf7\u8fd0\u884c":27,"\u8bf7\u9009\u62e9\u6b63\u786e\u7684\u7248\u672c":24,"\u8bf8\u5982\u56fe\u50cf\u5206\u7c7b":45,"\u8bfb\u53d612\u4e2a\u91c7\u6837\u6570\u636e\u8fdb\u884c\u968f\u673a\u68af\u5ea6\u8ba1\u7b97\u6765\u66f4\u65b0\u66f4\u65b0":25,"\u8bfb\u53d6\u6570\u636e":3,"\u8bfb\u53d6\u6bcf\u4e00\u884c":3,"\u8bfb\u53d6volume\u4e2d\u7684\u6570\u636e\u8fdb\u884c\u8fd9\u6b21\u5206\u5e03\u5f0f\u8bad\u7ec3":49,"\u8bfb\u8005\u53ef\u4ee5\u67e5\u770b":49,"\u8bfb\u8005\u9700\u8981\u66ff\u6362\u6210\u81ea\u5df1\u4f7f\u7528\u7684\u4ed3\u5e93\u5730\u5740":49,"\u8c03\u7528":[37,54],"\u8c03\u7528\u4e00\u6b21":3,"\u8c03\u7528\u4e0a\u9762\u7684process\u51fd\u6570\u83b7\u5f97\u89c2\u6d4b\u6570\u636e":25,"\u8c03\u7528\u7684pydataprovider2\u662f":3,"\u8c03\u7528\u7b2c\u4e8c\u6b21\u7684\u65f6\u5019":3,"\u8c03\u7528\u8be5\u51fd\u6570\u540e":37,"\u8c03\u7528\u8fd9\u4e2apydataprovider2\u7684\u65b9\u6cd5":3,"\u8c13\u8bcd\u4e0a\u4e0b\u6587":60,"\u8d1f\u6837\u672c":57,"\u8d1f\u9762\u7684\u8bc4\u8bba\u7684\u5f97\u5206\u5c0f\u4e8e\u7b49\u4e8e4":61,"\u8d1f\u9762\u8bc4\u4ef7\u6837\u672c":61,"\u8d21\u732e\u6587\u6863":38,"\u8d44\u6e90\u5bf9\u8c61\u7684\u540d\u5b57\u662f\u552f\u4e00\u7684":47,"\u8d77":32,"\u8def\u5f84\u4e0b":[25,55],"\u8df3\u8f6c\u5230":36,"\u8f6c\u4e3ajpeg\u6587\u4ef6\u5e76\u5b58\u5165\u7279\u5b9a\u7684\u76ee\u5f55":54,"\u8f6c\u6362\u751f\u6210\u7684\u6587\u4ef6\u540d\u4f1a\u662f\u4ee5\u4e0b\u683c\u5f0f":19,"\u8f6c\u6362\u8fc7\u6765\u7684":55,"\u8f6e":59,"\u8f83":32,"\u8f93\u5165":[31,35],"\u8f93\u5165\u5168\u662f\u5176\u4ed6layer\u7684\u8f93\u51fa":46,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u5355\u5c42\u5e8f\u5217":34,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u53cc\u5c42\u5e8f\u5217":34,"\u8f93\u5165\u56fe\u7247\u7684\u9ad8\u5ea6\u53ca\u5bbd\u5ea6":54,"\u8f93\u5165\u5c42\u5c3a\u5bf8":55,"\u8f93\u5165\u6570\u636e\u4e3a\u4e00\u4e2a\u5b8c\u6574\u7684\u65f6\u95f4\u5e8f\u5217":32,"\u8f93\u5165\u6570\u636e\u4e3a\u5728\u5355\u5c42rnn\u6570\u636e\u91cc\u9762":32,"\u8f93\u5165\u6570\u636e\u6574\u4f53\u4e0a\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":32,"\u8f93\u5165\u6570\u636e\u7684\u5b57\u5178\u7ef4\u6570\u662f1\u767e\u4e07":45,"\u8f93\u5165\u6570\u6910\u6240\u5728\u76ee\u5f55":61,"\u8f93\u5165\u6587\u672c":53,"\u8f93\u5165\u6587\u672c\u4e2d\u6ca1\u6709\u5934\u90e8":53,"\u8f93\u5165\u662f\u5426\u662f\u8f6c\u7f6e\u7684":37,"\u8f93\u5165\u662f\u7531\u4e00\u4e2alist\u4e2d\u7684\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u7ec4\u6210\u7684":37,"\u8f93\u5165\u7279\u5f81\u56fe\u7684\u901a\u9053\u6570\u76ee":55,"\u8f93\u5165\u7684":53,"\u8f93\u5165\u7684\u539f\u59cb\u6570\u636e\u96c6\u8def\u5f84":62,"\u8f93\u5165\u7684\u540d\u5b57":37,"\u8f93\u5165\u7684\u5927\u5c0f":37,"\u8f93\u5165\u7684\u6587\u672c\u683c\u5f0f\u5982\u4e0b":53,"\u8f93\u5165\u7684\u6587\u672c\u8bcd\u5411\u91cf\u6a21\u578b\u540d\u79f0":53,"\u8f93\u5165\u7684\u7c7b\u578b":37,"\u8f93\u5165\u95e8":61,"\u8f93\u5165\u9884\u6d4b\u6837\u672c":61,"\u8f93\u5165n\u4e2a\u5355\u8bcd":57,"\u8f93\u51fa":[31,35],"\u8f93\u51fa\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":34,"\u8f93\u51fa\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":34,"\u8f93\u51fa\u4e3an\u4e2aword_dim\u7ef4\u5ea6\u5411\u91cf":57,"\u8f93\u51fa\u51fd\u6570":35,"\u8f93\u51fa\u5e8f\u5217\u7684\u7c7b\u578b":31,"\u8f93\u51fa\u5e8f\u5217\u7684\u8bcd\u8bed\u6570\u548c\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":34,"\u8f93\u51fa\u5e94\u8be5\u7c7b\u4f3c\u5982\u4e0b":59,"\u8f93\u51fa\u6587\u4ef6\u7684\u683c\u5f0f\u8bf4\u660e":53,"\u8f93\u51fa\u65e5\u5fd7\u4fdd\u5b58\u5728\u8def\u5f84":61,"\u8f93\u51fa\u65e5\u5fd7\u8bf4\u660e\u5982\u4e0b":61,"\u8f93\u51fa\u67092\u5217":53,"\u8f93\u51fa\u7279\u5f81\u56fe\u7684\u901a\u9053\u6570\u76ee":55,"\u8f93\u51fa\u7684\u4e8c\u8fdb\u5236\u8bcd\u5411\u91cf\u6a21\u578b\u540d\u79f0":53,"\u8f93\u51fa\u7684\u6587\u672c\u6a21\u578b\u540d\u79f0":53,"\u8f93\u51fa\u7684\u68af\u5ea6":43,"\u8f93\u51fa\u76ee\u5f55":55,"\u8f93\u51fa\u7ed3\u679c\u53ef\u80fd\u4f1a\u968f\u7740\u5bb9\u5668\u7684\u6d88\u8017\u800c\u88ab\u5220\u9664":48,"\u8fbe\u5230\u5bb9\u707e\u7684\u76ee\u7684":18,"\u8fc7\u4e86\u4e00\u4e2a\u5f88\u7b80\u5355\u7684recurrent_group":32,"\u8fc7\u5b8c\u6240\u6709\u8bad\u7ec3\u6570\u636e\u5373\u4e3a\u4e00\u4e2apass":24,"\u8fd0":27,"\u8fd0\u884c":29,"\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u5373\u53ef":59,"\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4\u6765\u4e0b\u8f7d\u8fd9\u4e2a\u811a\u672c":62,"\u8fd0\u884c\u4ee5\u4e0b\u6307\u4ee4":27,"\u8fd0\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4\u4e0b\u8f7d\u548c\u83b7\u53d6\u6211\u4eec\u7684\u5b57\u5178\u548c\u9884\u8bad\u7ec3\u6a21\u578b":53,"\u8fd0\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4\u4e0b\u8f7d\u6570\u636e\u96c6":53,"\u8fd0\u884c\u4ee5\u4e0b\u8bad\u7ec3\u547d\u4ee4":25,"\u8fd0\u884c\u4ee5\u53ca\u53d1\u5e03\u60a8\u7684ai\u7a0b\u5e8f":27,"\u8fd0\u884c\u5206\u5e03\u5f0f\u4f5c\u4e1a":41,"\u8fd0\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":[24,39,57],"\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5":27,"\u8fd0\u884c\u5931\u8d25":45,"\u8fd0\u884c\u5b8c\u4ee5\u4e0a\u547d\u4ee4":53,"\u8fd0\u884c\u5b8c\u6210\u540e":41,"\u8fd0\u884c\u5b8c\u811a\u672c":61,"\u8fd0\u884c\u6210\u529f\u4ee5\u540e":53,"\u8fd0\u884c\u6210\u529f\u540e\u76ee\u5f55":61,"\u8fd0\u884c\u65e5\u5fd7":41,"\u8fd0\u884c\u7684\u4e00\u4e9b\u53c2\u6570\u901a\u8fc7\u8fd9\u79cd\u65b9\u5f0f\u4f20\u9012\u5230\u5bb9\u5668\u5185":49,"\u8fd1":32,"\u8fd1\u671f\u63d0\u51fa\u7684nmt\u6a21\u578b\u901a\u5e38\u90fd\u5c5e\u4e8e\u7f16\u89e3\u7801\u6a21\u578b":62,"\u8fd4\u56de":[8,9,10,13,15],"\u8fd4\u56de0":3,"\u8fd4\u56de8\u4e2a\u7279\u5f81list\u548c1\u4e2a\u6807\u7b7elist":60,"\u8fd4\u56de\u4e00\u6761\u5b8c\u6574\u7684\u6837\u672c":3,"\u8fd4\u56de\u6570\u636e\u7684\u6bcf\u4e00\u6761\u6837\u672c\u7ed9":59,"\u8fd4\u56de\u65f6":3,"\u8fd4\u56de\u7684\u662f":3,"\u8fd4\u56de\u7684\u987a\u5e8f\u9700\u8981\u548cinput_types\u4e2d\u5b9a\u4e49\u7684\u987a\u5e8f\u4e00\u81f4":3,"\u8fd4\u56de\u7b2c\u4e8c\u6b65":23,"\u8fd4\u56de\u7b2ci\u4e2a\u8f93\u5165\u77e9\u9635":37,"\u8fd4\u56de\u7c7b\u578b":[8,9,10,13,15],"\u8fd8\u4f1a":32,"\u8fd8\u4f1a\u8f93\u51fa\u4e00\u4e2a":36,"\u8fd8\u662f":32,"\u8fd8\u6709":32,"\u8fd8\u80fd\u5904\u7406\u5176\u4ed6\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e":61,"\u8fd8\u91c7\u7528\u4e86\u4e24\u4e2a\u5176\u4ed6\u7279\u5f81":60,"\u8fd8\u9700\u8981\u8bbe\u7f6e\u4e0b\u9762\u4e24\u4e2a\u53c2\u6570":46,"\u8fd8\u9700\u8981\u8fdb\u884c\u9884\u5904\u7406":54,"\u8fd9":[24,32,57],"\u8fd9\u4e00\u5757\u7684\u8017\u65f6\u6bd4\u4f8b\u771f\u7684\u592a\u9ad8":40,"\u8fd9\u4e00\u5c42\u8fdb\u884c\u5c01\u88c5":21,"\u8fd9\u4e00\u6982\u5ff5\u4e0d\u518d\u7410\u788e":21,"\u8fd9\u4e00\u8fc7\u7a0b\u5bf9\u7528\u6237\u662f\u5b8c\u5168\u900f\u660e\u7684":34,"\u8fd9\u4e09\u4e2a\u5206\u652f":23,"\u8fd9\u4e09\u4e2a\u6b65\u9aa4\u53ef\u914d\u7f6e\u4e3a":57,"\u8fd9\u4e0e\u672c\u5730\u8bad\u7ec3\u76f8\u540c":41,"\u8fd9\u4e24\u4e2a\u6587\u4ef6\u5939\u4e0b\u5404\u81ea\u670910\u4e2a\u5b50\u6587\u4ef6\u5939":54,"\u8fd9\u4e24\u4e2a\u6807\u51c6":60,"\u8fd9\u4e24\u4e2a\u9700\u8981\u4e0e":46,"\u8fd9\u4e2a":[32,47],"\u8fd9\u4e2a\u4efb\u52a1\u7684\u914d\u7f6e\u4e3a":24,"\u8fd9\u4e2a\u4efb\u52a1\u7684dataprovider\u4e3a":24,"\u8fd9\u4e2a\u51fd\u6570\u7684":35,"\u8fd9\u4e2a\u51fd\u6570\u8fdb\u884c\u53d8\u6362":32,"\u8fd9\u4e2a\u51fd\u6570\u9700\u8981\u8bbe\u7f6e":35,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u7684\u8fde\u63a5\u53c2\u6570\u4e0epaddle\u7684\u5176\u4ed6\u4e8c\u8fdb\u5236":21,"\u8fd9\u4e2a\u5305\u91cc\u9762\u5305\u542b\u4e86\u6a21\u578b\u914d\u7f6e\u9700\u8981\u7684\u5404\u4e2a\u6a21\u5757":46,"\u8fd9\u4e2a\u53c2\u6570\u4e5f\u4e0d\u4f1a\u4e00\u5e76\u5220\u9664":21,"\u8fd9\u4e2a\u5411\u91cf\u4e0e\u6e90\u4e2d\u641c\u7d22\u51fa\u7684\u4f4d\u7f6e\u548c\u6240\u6709\u4e4b\u524d\u751f\u6210\u7684\u76ee\u6807\u5355\u8bcd\u6709\u5173":62,"\u8fd9\u4e2a\u5730\u5740\u5219\u4e3a\u5b83\u7684\u7edd\u5bf9\u8def\u5f84\u6216\u76f8\u5bf9\u8def\u5f84":2,"\u8fd9\u4e2a\u5730\u5740\u6765\u8868\u793a\u6b64\u6b65\u9aa4\u6240\u6784\u5efa\u51fa\u7684\u955c\u50cf":49,"\u8fd9\u4e2a\u57fa\u7c7b":37,"\u8fd9\u4e2a\u5934\u6587\u4ef6\u4e0d\u5047\u8bbe\u5176\u4ed6\u6587\u4ef6\u7684\u5f15\u7528\u987a\u5e8f":21,"\u8fd9\u4e2a\u5b57\u5178\u662f\u6574\u6570\u6807\u7b7e\u548c\u5b57\u7b26\u4e32\u6807\u7b7e\u7684\u4e00\u4e2a\u5bf9\u5e94":61,"\u8fd9\u4e2a\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":34,"\u8fd9\u4e2a\u63a5\u53e3\u9700\u8981\u505a\u5230":20,"\u8fd9\u4e2a\u6570\u636e\u4e5f\u88ab\u5355\u5c42rnn\u7f51\u7edc\u76f4\u63a5\u4f7f\u7528":32,"\u8fd9\u4e2a\u6570\u636e\u5217\u8868\u6587\u4ef6\u4e2d\u5305\u542b\u7684\u662f\u6bcf\u4e00\u4e2a\u8bad\u7ec3\u6216\u8005\u6d4b\u8bd5\u6587\u4ef6\u7684\u8def\u5f84":46,"\u8fd9\u4e2a\u6570\u91cf\u79f0\u4e3abeam":62,"\u8fd9\u4e2a\u6587\u4ef6\u5177\u6709\u72ec\u7279\u7684\u8bed\u6cd5":20,"\u8fd9\u4e2a\u663e\u793a\u5668\u5f88\u68d2":57,"\u8fd9\u4e2a\u6a21\u578b\u5bf9\u4e8e\u7f16\u89e3\u7801\u6a21\u578b\u6765\u8bf4":62,"\u8fd9\u4e2a\u76ee\u5f55\u4e2d\u9664\u4e86":21,"\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u5355\u5143\u5c31\u53ebmemori":32,"\u8fd9\u4e2a\u7a0b\u5e8f\u662f\u60a8\u5728\u5f00\u53d1\u673a\u4e0a\u4f7f\u7528\u5f00\u53d1\u955c\u50cf\u5b8c\u6210\u5f00\u53d1":27,"\u8fd9\u4e2a\u7c7b\u7684\u53c2\u6570\u5305\u62ec":37,"\u8fd9\u4e2a\u7c7b\u9700\u8981\u7ee7\u627f":37,"\u8fd9\u4e2a\u7cfb\u7edf\u5c06srl\u4efb\u52a1\u89c6\u4e3a\u5e8f\u5217\u6807\u6ce8\u95ee\u9898":60,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u4e2d\u7684\u53e6\u4e00\u4e2a\u9879\u76ee\u662f":21,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u5305\u542b\u4e24\u4e2a\u9879\u76ee":21,"\u8fd9\u4e2a\u8282\u70b9\u53ef\u4ee5\u662f\u7269\u7406\u673a\u6216\u8005\u865a\u62df\u673a":47,"\u8fd9\u4e2a\u8868\u683c":47,"\u8fd9\u4e2a\u8fc7\u7a0b\u5bf9\u7528\u6237\u4e5f\u662f\u900f\u660e\u7684":34,"\u8fd9\u4e2a\u8fc7\u7a0b\u5c31\u662f\u6a21\u578b\u7684\u8bad\u7ec3\u8fc7\u7a0b":25,"\u8fd9\u4e2a\u8fc7\u7a0b\u9664\u4e86\u7f16\u8bd1paddlepaddle\u4e3a":36,"\u8fd9\u4e2a\u914d\u7f6e\u4e0e":53,"\u8fd9\u4e2a\u914d\u7f6e\u6587\u4ef6":47,"\u8fd9\u4e2a\u914d\u7f6e\u6587\u4ef6\u7f51\u7edc\u7531":46,"\u8fd9\u4e2a\u914d\u7f6e\u662f\u5426\u7528\u6765\u751f\u6210":62,"\u8fd9\u4e2a\u955c\u50cf\u5305\u542b\u4e86paddle\u76f8\u5173\u7684\u5f00\u53d1\u5de5\u5177\u4ee5\u53ca\u7f16\u8bd1\u548c\u8fd0\u884c\u73af\u5883":27,"\u8fd9\u4e2a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u751f\u6210\u4e00\u7cfb\u5217\u6743\u91cd":35,"\u8fd9\u4e2a\u95ee\u9898\u662fpydataprovider\u8bfb\u6570\u636e\u65f6\u5019\u7684\u903b\u8f91\u95ee\u9898":3,"\u8fd9\u4e2a\u9759\u6001\u5e93\u5305\u542b\u4e86paddle\u7684\u5168\u90e8\u7b26\u53f7":21,"\u8fd9\u4e2adataprovider\u8f83\u590d\u6742":3,"\u8fd9\u4e2ainstance\u53ef\u4ee5\u662f\u5355\u4e2a\u503c":19,"\u8fd9\u4e2ajob\u624d\u7b97\u6210\u529f\u7ed3\u675f":49,"\u8fd9\u4e2alayer\u7684\u8f93\u51fa\u4f1a\u4f5c\u4e3a\u6574\u4e2a":34,"\u8fd9\u4e5f\u4f1a\u6781\u5927\u51cf\u5c11\u6570\u636e\u8bfb\u5165\u7684\u8017\u65f6":24,"\u8fd9\u4e9b":41,"\u8fd9\u4e9b\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1\u901a\u5e38\u4f1a\u628a\u6570\u636e\u5207\u5272\u6210\u591a\u4e2a\u5206\u7247\u5206\u5e03\u5f0f\u7684\u5b58\u50a8\u5728\u591a\u4e2a\u8282\u70b9\u4e4b\u4e0a":19,"\u8fd9\u4e9b\u53c2\u6570\u7684\u5177\u4f53\u63cf\u8ff0":49,"\u8fd9\u4e9b\u53c2\u6570\u7684\u7b80\u77ed\u4ecb\u7ecd\u5982\u4e0b":59,"\u8fd9\u4e9b\u540d\u5b57\u5fc5\u987b\u8981\u5199\u5bf9":37,"\u8fd9\u4e9b\u6570\u636e\u4f1a\u88ab\u7528\u6765\u66f4\u65b0\u53c2\u6570":24,"\u8fd9\u4e9b\u6570\u636e\u4f7f\u7528\u7684\u5185\u5b58\u4e3b\u8981\u548c\u4e24\u4e2a\u53c2\u6570\u6709\u5173\u7cfb":24,"\u8fd9\u4e9b\u6587\u4ef6\u5c06\u4f1a\u88ab\u4fdd\u5b58\u5728":55,"\u8fd9\u4e9b\u6a21\u578b\u90fd\u662f\u7531\u539f\u4f5c\u8005\u63d0\u4f9b\u7684\u6a21\u578b":55,"\u8fd9\u4e9b\u7279\u5f81\u503c\u4e0e\u4e0a\u8ff0\u4f7f\u7528c":55,"\u8fd9\u4e9b\u7279\u5f81\u548c\u6807\u7b7e\u5b58\u50a8\u5728":60,"\u8fd9\u4e9b\u7279\u5f81\u6570\u636e\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u6709\u610f\u4e49\u7684":32,"\u8fd9\u4efd\u6559\u7a0b\u5c55\u793a\u4e86\u5982\u4f55\u5728paddlepaddle\u4e2d\u5b9e\u73b0\u4e00\u4e2a\u81ea\u5b9a\u4e49\u7684\u7f51\u7edc\u5c42":37,"\u8fd9\u4efd\u7b80\u77ed\u7684\u4ecb\u7ecd\u5c06\u5411\u4f60\u5c55\u793a\u5982\u4f55\u5229\u7528paddlepaddle\u6765\u89e3\u51b3\u4e00\u4e2a\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u95ee\u9898":25,"\u8fd9\u4f1a\u63d0\u793a\u5f53\u524d\u76ee\u5f55\u7684\u4e00\u4e9b\u53d8\u5316":36,"\u8fd9\u4f1a\u81ea\u52a8\u8fdb\u884c\u7f51\u7edc\u914d\u7f6e\u4e2d\u58f0\u660e\u7684\u6fc0\u6d3b\u64cd\u4f5c":37,"\u8fd9\u4f7f\u5f97nmt\u6a21\u578b\u5f97\u4ee5\u89e3\u653e\u51fa\u6765":62,"\u8fd9\u4fbf\u662f\u4e00\u79cd\u53cc\u5c42rnn\u7684\u8f93\u5165\u6570\u636e":32,"\u8fd9\u51e0\u4e2a\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":26,"\u8fd9\u53ef\u4ee5\u8ba9\u5176\u4ed6\u4eba\u77e5\u9053\u8fd9\u6b21\u63d0\u4ea4\u505a\u4e86\u54ea\u4e9b\u6539\u53d8":36,"\u8fd9\u53ef\u4ee5\u901a\u8fc7":36,"\u8fd9\u548c\u5355\u5c42rnn\u7684\u914d\u7f6e\u662f\u7b49\u4ef7\u7684":32,"\u8fd9\u56db\u4e2a\u7b80\u5355\u7684\u7279\u5f81\u662f\u6211\u4eec\u7684srl\u7cfb\u7edf\u6240\u9700\u8981\u7684":60,"\u8fd9\u56db\u6761\u6570\u636e\u540c\u65f6\u5904\u7406\u7684\u53e5\u5b50\u6570\u91cf\u4e3a":32,"\u8fd9\u5728\u5f88\u5927\u7a0b\u5ea6\u4e0a\u4f18\u4e8e\u5148\u524d\u7684\u6700\u5148\u8fdb\u7684\u7cfb\u7edf":60,"\u8fd9\u5728\u6784\u9020\u975e\u5e38\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u662f\u6709\u7528\u7684":35,"\u8fd9\u5bf9\u4e8e\u901a\u5e38\u7684java\u7684\u5f00\u53d1\u8005\u6765\u8bf4":20,"\u8fd9\u5c06\u82b1\u8d39\u6570\u5206\u949f\u7684\u65f6\u95f4":62,"\u8fd9\u5c31\u662f":46,"\u8fd9\u5df2\u7ecf\u5728":61,"\u8fd9\u610f\u5473\u7740":35,"\u8fd9\u610f\u5473\u7740\u6a21\u578b\u5728\u8bad\u7ec3\u6570\u636e\u4e0a\u4e0d\u65ad\u7684\u6539\u8fdb":25,"\u8fd9\u610f\u5473\u7740\u9664\u4e86\u6307\u5b9adevic":45,"\u8fd9\u65f6\u8fdb\u884c\u77e9\u9635\u4e58\u6cd5\u8fd0\u7b97\u5c31\u53ef\u80fd\u5bfc\u81f4\u6d6e\u70b9\u6570\u6ea2\u51fa":24,"\u8fd9\u662f\u4e00\u4e2a\u57fa\u4e8e\u7edf\u8ba1\u7684\u673a\u5668\u7ffb\u8bd1\u7cfb\u7edf":61,"\u8fd9\u662f\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":31,"\u8fd9\u662f\u56e0\u4e3a":20,"\u8fd9\u662f\u56e0\u4e3a\u5b83\u53d1\u6398\u51fa\u4e86\u56fe\u7247\u7684\u4e24\u7c7b\u91cd\u8981\u4fe1\u606f":54,"\u8fd9\u662f\u666e\u901a\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u4ee3\u7801":32,"\u8fd9\u662f\u76ee\u524dcmake\u5bfb\u627epython\u7684\u903b\u8f91\u5b58\u5728\u7f3a\u9677":24,"\u8fd9\u662f\u96c6\u675f\u641c\u7d22\u7684\u7ed3\u679c":62,"\u8fd9\u6765\u81ea\u4e8epaddlepaddle\u7684\u5185\u5b58\u4e2d":62,"\u8fd9\u6837":[21,25,41],"\u8fd9\u6837\u4fdd\u8bc1":23,"\u8fd9\u6837\u505a\u53ef\u4ee5\u6781\u5927\u7684\u51cf\u5c11\u5185\u5b58\u5360\u7528":24,"\u8fd9\u6837\u5355\u4e2a\u5b50\u7ebf\u7a0b\u7684\u957f\u5ea6\u5c31\u4e0d\u4f1a\u6ea2\u51fa\u4e86":46,"\u8fd9\u6837\u53ef\u4ee5\u51cf\u5c0fgpu\u5185\u5b58":45,"\u8fd9\u6837\u5bb9\u5668\u7684":49,"\u8fd9\u6837\u5c31\u4f1a\u751f\u6210\u4e24\u4e2a\u6587\u4ef6":59,"\u8fd9\u6837\u5c31\u53ef\u4ee5\u5728\u4e91\u7aef\u6267\u884c\u591a\u79cd\u6570\u636e\u7c7b\u8ba1\u7b97\u4efb\u52a1":19,"\u8fd9\u6837\u5f00\u53d1\u4eba\u5458\u53ef\u4ee5\u4ee5\u4e00\u81f4\u7684\u65b9\u5f0f\u5728\u4e0d\u540c\u7684\u5e73\u53f0\u4e0a\u5de5\u4f5c":27,"\u8fd9\u6837\u7684\u88c5\u9970\u5668":37,"\u8fd9\u6837\u7684\u8bdd":48,"\u8fd9\u6837\u7684\u8bdd\u6bcf\u4f4d\u7528\u6237\u5728\u6d4b\u8bd5\u6587\u4ef6\u4e2d\u5c06\u4e0e\u8bad\u7ec3\u6587\u4ef6\u542b\u6709\u540c\u6837\u7684\u4fe1\u606f":59,"\u8fd9\u6b63\u662f\u5b83\u4eec\u901f\u5ea6\u5feb\u7684\u539f\u56e0":40,"\u8fd9\u6bb5\u7b80\u77ed\u7684\u914d\u7f6e\u5c55\u793a\u4e86paddlepaddle\u7684\u57fa\u672c\u7528\u6cd5":25,"\u8fd9\u7528\u4e8e\u5728\u591a\u7ebf\u7a0b\u548c\u591a\u673a\u4e0a\u66f4\u65b0\u53c2\u6570":37,"\u8fd9\u79cd\u521d\u59cb\u5316\u65b9\u5f0f\u5728\u4e00\u822c\u60c5\u51b5\u4e0b\u4e0d\u4f1a\u4ea7\u751f\u5f88\u5dee\u7684\u7ed3\u679c":24,"\u8fd9\u79cd\u60c5\u51b5\u4e0b\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":37,"\u8fd9\u79cd\u65b9\u5f0f\u5fc5\u987b\u4f7f\u7528paddle\u5b58\u50a8\u7684\u6a21\u578b\u8def\u5f84\u683c\u5f0f":45,"\u8fd9\u79cd\u751f\u6210\u6280\u672f\u53ea\u7528\u4e8e\u7c7b\u4f3c\u89e3\u7801\u5668\u7684\u751f\u6210\u8fc7\u7a0b":35,"\u8fd9\u79cd\u7c7b\u578b\u7684\u8f93\u5165\u5fc5\u987b\u901a\u8fc7":34,"\u8fd9\u79cd\u96c6\u7fa4\u8282\u70b9\u7ba1\u7406\u65b9\u5f0f\u4f1a\u5728\u5c06\u6765\u4f7f\u7528":49,"\u8fd9\u7bc7\u6587\u6863":36,"\u8fd9\u7bc7\u6587\u7ae0":62,"\u8fd9\u7ec4\u8bed\u4e49\u76f8\u540c\u7684\u793a\u4f8b\u914d\u7f6e\u5982\u4e0b":32,"\u8fd9\u901a\u8fc7\u83b7\u5f97\u53cd\u5411\u5faa\u73af\u7f51\u7edc\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":35,"\u8fd9\u90fd\u9700\u8981\u8fd9\u4e2a\u63a5\u53e3\u6309\u7167\u7ea6\u5b9a\u4fd7\u6210\u7684\u89c4\u5219\u6765\u6ce8\u91ca\u5b8c\u5907":20,"\u8fd9\u91cc":[24,27,35,36,46,47,49,55,60],"\u8fd9\u91cc\u4e5f\u53ef\u53eb\u5206\u7c7b\u5c42":46,"\u8fd9\u91cc\u4ecb\u7ecd\u5982\u751f\u6210\u9020\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf":27,"\u8fd9\u91cc\u4ee5":57,"\u8fd9\u91cc\u4f7f\u7528\u4e00\u4e2a\u57fa\u4e8emomentum\u7684\u968f\u673a\u68af\u5ea6\u4e0b\u964d":25,"\u8fd9\u91cc\u4f7f\u7528\u4e86\u4e09\u79cd\u7f51\u7edc\u5355\u5143":25,"\u8fd9\u91cc\u4f7f\u7528\u4e86paddlepaddle\u7684python\u63a5\u53e3\u6765\u52a0\u8f7d\u6570\u6910":61,"\u8fd9\u91cc\u4f7f\u7528\u4e86paddlepaddle\u9884\u5b9a\u4e49\u597d\u7684rnn\u5904\u7406\u51fd\u6570":32,"\u8fd9\u91cc\u4f7f\u7528\u7b80\u5355\u7684":24,"\u8fd9\u91cc\u5229\u7528\u5b83\u5efa\u6a21\u53d8\u91cf\u4e4b\u95f4\u7684\u7ebf\u6027\u5173\u7cfb":25,"\u8fd9\u91cc\u53ea\u52a0\u8f7d":62,"\u8fd9\u91cc\u53ea\u7b80\u5355\u4ecb\u7ecd\u4e86\u5355\u673a\u8bad\u7ec3":57,"\u8fd9\u91cc\u5bf9":46,"\u8fd9\u91cc\u5c55\u793a\u5982\u4f55\u4f7f\u7528\u89c2\u6d4b\u6570\u636e\u6765\u62df\u5408\u8fd9\u4e00\u7ebf\u6027\u5173\u7cfb":25,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528":59,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7684\u662f\u4e00\u4e2a\u5c0f\u7684vgg\u7f51\u7edc":54,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7684\u662fgpu\u6a21\u5f0f\u8fdb\u884c\u8bad\u7ec3":54,"\u8fd9\u91cc\u6211\u4eec\u5728movielens\u6570\u636e\u96c6\u63cf\u8ff0\u4e00\u79cd":59,"\u8fd9\u91cc\u6211\u4eec\u5c55\u793a\u4e00\u4efd\u7b80\u5316\u8fc7\u7684\u4ee3\u7801":37,"\u8fd9\u91cc\u6211\u4eec\u7528paddl":27,"\u8fd9\u91cc\u6211\u4eec\u901a\u8fc7\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2ajob\u6765\u4e0b\u8f7d\u5e76\u5207\u5272\u6570\u636e":49,"\u8fd9\u91cc\u6307\u5b9a\u8bcd\u5178":57,"\u8fd9\u91cc\u6570\u636e\u5c42\u6709\u4e24\u4e2a":25,"\u8fd9\u91cc\u662f\u4e00\u4e2a\u4f8b\u5b50":62,"\u8fd9\u91cc\u6709\u4e00\u4e9b\u4e0d\u540c\u7684\u53c2\u6570\u9700\u8981\u6307\u5b9a":62,"\u8fd9\u91cc\u68c0\u9a8c\u8fd0\u884c\u65f6\u95f4\u6a21\u578b\u7684\u6536\u655b":41,"\u8fd9\u91cc\u6bcf\u4e2a5\u4e2abatch\u6253\u5370\u4e00\u4e2a\u70b9":62,"\u8fd9\u91cc\u6bcf\u9694100\u4e2abatch\u663e\u793a\u4e00\u6b21\u53c2\u6570\u7edf\u8ba1\u4fe1\u606f":62,"\u8fd9\u91cc\u6bcf\u969410\u4e2abatch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":62,"\u8fd9\u91cc\u7684\u5217\u51fa\u7684\u548c":54,"\u8fd9\u91cc\u76f4\u63a5\u901a\u8fc7\u9884\u6d4b\u811a\u672c":57,"\u8fd9\u91cc\u7b80\u5355\u4ecb\u7ecdlayer":46,"\u8fd9\u91cc\u7ed9\u51fa\u96c6\u4e2d\u5e38\u89c1\u7684\u90e8\u7f72\u65b9\u6cd5":47,"\u8fd9\u91cc\u8bbe\u7f6e\u4e3a\u4f7f\u7528cpu":62,"\u8fd9\u91cc\u8bbe\u7f6e\u4e3afals":62,"\u8fd9\u91cc\u8bbe\u7f6e\u4e3atrue":62,"\u8fd9\u91cc\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":57,"\u8fd9\u91cc\u9700\u8981\u7528\u6237\u989d\u5916\u6ce8\u610f":18,"\u8fdb\u4e3b\u4ed3\u5e93\u540e":36,"\u8fdb\u5165":61,"\u8fdb\u5165\u5bb9\u5668":48,"\u8fdb\u5165\u5f00\u53d1\u955c\u50cf\u5e76\u5f00\u59cb\u5de5\u4f5c":27,"\u8fdb\u7a0b":46,"\u8fdb\u7a0b\u4e2d\u53ef\u4ee5\u542f\u52a8\u591a\u4e2a\u5b50\u7ebf\u7a0b\u53bb\u63a5\u53d7":46,"\u8fdb\u7a0b\u4e4b\u540e":46,"\u8fdb\u7a0b\u5171\u7ed1\u5b9a\u591a\u5c11\u4e2a\u7aef\u53e3\u7528\u6765\u505a\u7a20\u5bc6\u66f4\u65b0":46,"\u8fdb\u7a0b\u542f\u52a8\u7684\u5fc5\u8981\u53c2\u6570":49,"\u8fdb\u7a0b\u7684":41,"\u8fdb\u7a0b\u7684\u542f\u52a8\u53c2\u6570":49,"\u8fdb\u7a0b\u7684\u8fd0\u884c\u73af\u5883":49,"\u8fdb\u7a0b\u7aef\u53e3\u662f":46,"\u8fdb\u7a0b\u9700\u8981\u7684":49,"\u8fdb\u800c\u8fdb\u884c\u4ee3\u7801\u8bc4\u5ba1":23,"\u8fdb\u884c\u4e86":32,"\u8fdb\u884c\u4f7f\u7528":54,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6848":49,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6cd5":49,"\u8fdb\u884c\u540c\u6b65":46,"\u8fdb\u884c\u5f00\u53d1":36,"\u8fdb\u884c\u62c6\u89e3":32,"\u8fdb\u884c\u6fc0\u6d3b\u64cd\u4f5c":37,"\u8fdb\u884c\u8bfb\u5165\u548c\u9884\u5904\u7406\u4ece\u800c\u5f97\u5230\u771f\u5b9e\u8f93\u5165":25,"\u8fdb\u884c\u9884\u6d4b":57,"\u8fdb\u9636\u6307\u5357":52,"\u8fde\u63a5":34,"\u8fde\u63a5\u4e09\u4e2alstm\u9690\u85cf\u5c42":61,"\u9000\u4f11\u4eba\u5458":58,"\u9000\u51fa\u5bb9\u5668":48,"\u9002\u4e2d":32,"\u9002\u5408\u81ea\u5df1\u5b9e\u9a8c":27,"\u9009":32,"\u9009\u62e9":32,"\u9009\u62e9\u6d4b\u8bd5\u7ed3\u679c\u6700\u597d\u7684\u6a21\u578b\u6765\u9884\u6d4b":57,"\u9009\u62e9\u76ee\u6807\u5206\u652f":36,"\u9009\u62e9\u8def\u5f84\u6765\u52a8\u6001\u52a0\u8f7dnvidia":43,"\u9009\u62e9\u8fc7\u540e\u7684":62,"\u9009\u62e9\u9002\u5408\u60a8\u7684\u573a\u666f\u7684\u5408\u9002\u65b9\u6848":47,"\u9009\u62e9cephfs\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u670d\u52a1":19,"\u9009\u81ea\u4e0b\u5217\u7c7b\u578b":58,"\u9009\u9879":[26,53],"\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":42,"\u901a\u5e38":[21,41,61],"\u901a\u5e38\u4f1a\u4f7f\u7528\u73af\u5883\u53d8\u91cf\u914d\u7f6ejob\u7684\u914d\u7f6e\u4fe1\u606f":49,"\u901a\u5e38\u4f7f\u7528\u7a00\u758f\u8bad\u7ec3\u6765\u52a0\u901f\u8ba1\u7b97\u8fc7\u7a0b":45,"\u901a\u5e38\u505a\u6cd5\u662f\u4ece\u4e00\u4e2a\u6bd4\u8f83\u5927\u7684learning_rate\u5f00\u59cb\u8bd5":24,"\u901a\u5e38\u540d\u5b57\u662f":36,"\u901a\u5e38\u5728\u9ad8\u7ea7\u60c5\u51b5\u4e0b":46,"\u901a\u5e38\u60c5\u51b5\u4e0b":40,"\u901a\u5e38\u6211\u4eec\u4f1a\u5b89\u88c5ceph\u7b49\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u6765\u5b58\u50a8\u8bad\u7ec3\u6570\u636e":48,"\u901a\u5e38\u6307\u5c06\u4e00\u4e2a\u6574\u4f53\u62c6\u5206\u6210\u591a\u4efd\u7684\u5176\u4e2d\u7684\u4e00\u4efd":18,"\u901a\u5e38\u662f\u4e00\u4e2apython\u51fd\u6570":46,"\u901a\u5e38\u6bcf\u4e2a\u914d\u7f6e\u6587\u4ef6\u90fd\u4f1a\u5305\u62ec":46,"\u901a\u5e38\u6bcf\u4e2ajob\u5305\u62ec\u4e00\u4e2a\u6216\u8005\u591a\u4e2apod":47,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u4f7f\u7528":35,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u5c06\u914d\u7f6e\u5b58\u4e8e":37,"\u901a\u5e38\u8981\u6c42\u65f6\u95f4\u6b65\u4e4b\u95f4\u5177\u6709\u4e00\u4e9b\u4f9d\u8d56\u6027":32,"\u901a\u5e38\u90fd\u4f1a\u4f7f\u7528\u4e0b\u9762\u8fd9\u4e9b\u547d\u4ee4\u884c\u53c2\u6570":45,"\u901a\u7528":42,"\u901a\u77e5":32,"\u901a\u77e5\u7cfb\u7edf\u4e00\u8f6e\u6570\u636e\u8bfb\u53d6\u7ed3\u675f":46,"\u901a\u8fc7":[24,32,36,37,41,46,57],"\u901a\u8fc7\u4e24\u4e2a\u5d4c\u5957\u7684":34,"\u901a\u8fc7\u4ea4\u66ff\u4f7f\u7528\u5377\u79ef\u548c\u6c60\u5316\u5904\u7406":54,"\u901a\u8fc7\u4f7f\u7528":26,"\u901a\u8fc7\u51fd\u6570":49,"\u901a\u8fc7\u5377\u79ef\u64cd\u4f5c\u4ece\u56fe\u7247\u6216\u7279\u5f81\u56fe\u4e2d\u63d0\u53d6\u7279\u5f81":54,"\u901a\u8fc7\u547d\u4ee4\u884c\u53c2\u6570":24,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u7684\u8f93\u51fa":34,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa":34,"\u901a\u8fc7\u6240\u6709\u5355\u5143\u6d4b\u8bd5":36,"\u901a\u8fc7\u6240\u6709\u8bad\u7ec3\u96c6\u4e00\u6b21\u79f0\u4e3a\u4e00\u904d":61,"\u901a\u8fc7\u67e5\u770b\u4e70\u5bb6\u5bf9\u67d0\u4e2a\u4ea7\u54c1\u7684\u8bc4\u4ef7\u53cd\u9988":57,"\u901a\u8fc7\u6a21\u578b\u63a8\u65adapi\u7684\u5b9e\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u6837\u4f8b":21,"\u901a\u8fc7\u7f16\u8bd1\u4f1a\u751f\u6210py_paddle\u8f6f\u4ef6\u5305":5,"\u901a\u8fc7\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u6765\u6307\u5b9a":37,"\u901a\u8fc7\u8c03\u7528":5,"\u901a\u8fc7\u914d\u7f6e\u7c7b\u4f3c\u4e8e":57,"\u901a\u8fc7data":34,"\u901a\u8fc7volum":47,"\u903b\u8f91\u56de\u5f52":57,"\u9053\u6b49":32,"\u9069":32,"\u9075\u5faa\u4ee5\u4e0b\u6d41\u7a0b":23,"\u9075\u5faa\u5982\u4e0b\u7684\u683c\u5f0f":58,"\u9075\u5faa\u6587\u7ae0":53,"\u90a3\u4e48":[21,34,37],"\u90a3\u4e480\u5c42\u5e8f\u5217\u5373\u4e3a\u4e00\u4e2a\u8bcd\u8bed":34,"\u90a3\u4e48\u53ef\u4ee5\u8ba4\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":24,"\u90a3\u4e48\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4e0d\u4f1a\u6267\u884c\u6d4b\u8bd5\u64cd\u4f5c":2,"\u90a3\u4e48\u5982\u4f55\u5224\u65ad\u8bad\u7ec3\u4e0d\u6536\u655b\u5462":24,"\u90a3\u4e48\u5e38\u6570\u8f93\u51fa\u6240\u80fd\u8fbe\u5230\u7684\u6700\u5c0fcost\u662f":24,"\u90a3\u4e48\u5f53check\u51fa\u6570\u636e\u4e0d\u5408\u6cd5\u65f6":3,"\u90a3\u4e48\u6211\u4eec\u53ef\u4ee5\u5224\u65ad\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":24,"\u90a3\u4e48\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u6536\u96c6\u5e02\u573a\u4e0a\u623f\u5b50\u7684\u5927\u5c0f\u548c\u4ef7\u683c":25,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528":35,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u65b9\u6cd5":35,"\u90a3\u4e48\u6536\u655b\u53ef\u80fd\u5f88\u6162":24,"\u90a3\u4e48\u6700\u597d\u5c06\u6570\u636e\u6587\u4ef6\u5728\u6bcf\u6b21\u8bfb\u53d6\u4e4b\u524d\u505a\u4e00\u6b21shuffl":24,"\u90a3\u4e48\u7528\u6237\u9700\u8981\u5728\u672c\u5730\u7f16\u8bd1\u5b89\u88c5paddlepaddl":38,"\u90a3\u4e48\u8bad\u7ec3\u6709\u53ef\u80fd\u4e0d\u6536\u655b":24,"\u90a3\u4e48\u8be5\u4f18\u5316\u7b97\u6cd5\u81f3\u5c11\u9700\u8981":24,"\u90a3\u4e48fc1\u548cfc2\u5c42\u5c06\u4f1a\u4f7f\u7528\u7b2c1\u4e2agpu\u6765\u8ba1\u7b97":45,"\u90a3\u4e48paddlepaddle\u4f1a\u6839\u636elayer\u7684\u58f0\u660e\u987a\u5e8f":3,"\u90a3\u4e5f\u5c31\u4e0d\u9700\u8981\u6025\u7740\u4f18\u5316\u6027\u80fd\u5566":40,"\u90a3\u4f30\u8ba1\u8fd9\u91cc\u7684\u6f5c\u529b\u5c31\u6ca1\u5565\u597d\u6316\u7684\u4e86":40,"\u90a3\u51cf\u5c11\u5b66\u4e60\u738710\u500d\u7ee7\u7eed\u8bd5\u9a8c":24,"\u90a3\u6211\u4f1a\u671f\u671b\u5206\u6790\u5de5\u5177\u7edf\u8ba1\u5230\u901f\u5ea6\u662f100gb":40,"\u90a3\u7a0b\u5e8f\u5206\u6790\u5de5\u5177\u662f\u5fc5\u4e0d\u53ef\u5c11\u7684\u5229\u5668":40,"\u90ae\u7f16":58,"\u90e8\u7f72\u548c\u914d\u7f6e\u6bd4\u8f83\u7b80\u5355":47,"\u90fd":32,"\u90fd\u4f1a\u4ea7\u751f\u5f53\u524d\u5c42\u72b6\u6001\u7684\u6240\u6709\u7ee7\u627f\u7ed3\u679c":43,"\u90fd\u4f7f\u7528\u968f\u673a\u503c\u521d\u59cb\u5316":25,"\u90fd\u53ea\u662f\u4ecb\u7ecd\u53cc\u5c42rnn\u7684api\u63a5\u53e3":32,"\u90fd\u662f\u4e94\u4f4d\u7684\u6570\u5b57":19,"\u90fd\u662f\u5bf9layer1\u5143\u7d20\u7684\u62f7\u8d1d":31,"\u90fd\u662f\u5c06\u6bcf\u4e00\u53e5\u5206\u597d\u8bcd\u540e\u7684\u53e5\u5b50":32,"\u90fd\u662fabi\u8c03\u7528\u6807\u51c6\u7684":20,"\u90fd\u7528":36,"\u90fd\u9700\u8981\u5199\u63d0\u4ea4\u8bf4\u660e":36,"\u90fd\u9700\u8981\u8c03\u7528\u4e00\u6b21":37,"\u914d\u5408\u4f7f\u7528":46,"\u914d\u7f6e":61,"\u914d\u7f6e\u4e86\u7f51\u7edc":59,"\u914d\u7f6e\u51fa\u975e\u5e38\u590d\u6742\u7684\u7f51\u7edc":46,"\u914d\u7f6e\u521b\u5efa\u5b8c\u6bd5\u540e":54,"\u914d\u7f6e\u5982\u4e0b":53,"\u914d\u7f6e\u6253\u5f00":40,"\u914d\u7f6e\u6570\u636e\u6e90":46,"\u914d\u7f6e\u6587\u4ef6":57,"\u914d\u7f6e\u6587\u4ef6\u63a5\u53e3\u662ffc_layer":37,"\u914d\u7f6e\u6a21\u578b\u6587\u4ef6":53,"\u914d\u7f6e\u7b49\u6587\u4ef6\u7684\u76ee\u5f55\u89c6\u4e3a":41,"\u914d\u7f6e\u7b80\u5355\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u4f8b\u5b50":35,"\u914d\u7f6e\u7f51\u7edc\u5c42\u7684\u8f93\u5165":37,"\u914d\u7f6eapi":31,"\u9152\u5e97":32,"\u91c7\u6837\u5c42":59,"\u91c7\u7528":60,"\u91c7\u7528\u53e6\u4e00\u79cd\u65b9\u6cd5\u6765\u5806\u53e0lstm\u5c42":60,"\u91c7\u7528\u5747\u5300\u5206\u5e03\u6216\u8005\u9ad8\u65af\u5206\u5e03\u521d\u59cb\u5316":43,"\u91c7\u7528multi":24,"\u91ca\u653e\u5bf9paramters\u5185\u5b58\u7684\u9501\u5b9a":18,"\u91cc\u4ecb\u7ecd\u4e86\u7528paddle\u6e90\u7801\u4e2d\u7684\u811a\u672c\u4e0b\u8f7d\u8bad\u7ec3\u6570\u636e\u7684\u8fc7\u7a0b":48,"\u91cc\u4f1a\u7ee7\u7eed\u5b89\u88c5":29,"\u91cc\u6240\u6709\u7684\u7b26\u53f7\u90fd\u5199\u5165\u81ea\u5df1\u7684\u7a0b\u5e8f\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cc":20,"\u91cc\u6307\u5b9a\u56fe\u50cf\u6570\u636e\u5217\u8868":55,"\u91cc\u7684\u65e5\u5fd7":41,"\u91cc\u901a\u8fc7train_list\u548ctest_list\u6307\u5b9a\u662f\u8bad\u7ec3\u6587\u4ef6\u5217\u8868\u548c\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":46,"\u91cc\u9762\u6db5\u76d6\u4e86":27,"\u91cc\u9762\u6db5\u76d6\u4e86paddlepaddle\u8fd0\u884c\u6240\u9700\u7684\u6240\u6709\u73af\u5883":27,"\u91cd\u547d\u540d\u6210":20,"\u91cd\u65b0\u7f16\u8bd1paddlepaddl":40,"\u9488\u5bf9\u4efb\u52a1\u8fd0\u884c\u5b8c\u6210\u540e\u5bb9\u5668\u81ea\u52a8\u9000\u51fa\u7684\u573a\u666f":48,"\u9488\u5bf9\u5185\u5b58\u548c\u663e\u5b58":24,"\u9488\u5bf9\u6587\u672c":59,"\u94fe\u63a5\u4f55\u79cdblas\u5e93\u7b49":26,"\u94fe\u63a5\u5230\u81ea\u5df1\u7684\u7a0b\u5e8f\u91cc":20,"\u94fe\u63a5\u5f85\u8865\u5145":57,"\u9500\u552e":58,"\u9519\u8bef\u5904\u7406":20,"\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662f\u8fd4\u56de\u503c":20,"\u9519\u8bef\u5904\u7406\u7684\u65b9\u5f0f\u4e5f\u4e0d\u5c3d\u76f8\u540c":20,"\u9519\u8bef\u7387":57,"\u9519\u8bef\u7684define_py_data_sources2\u7c7b\u4f3c":24,"\u955c\u50cf\u91cc\u6709":48,"\u957f\u5ea6":24,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570":35,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u8f93\u51famemori":35,"\u95ee\u9898":25,"\u95f4\u9694":57,"\u9650\u5236\u5957\u63a5\u5b57\u53d1\u9001\u7f13\u51b2\u533a\u7684\u5927\u5c0f":43,"\u9650\u5236\u5957\u63a5\u5b57\u63a5\u6536\u7f13\u51b2\u533a\u7684\u5927\u5c0f":43,"\u9664\u4e86":3,"\u9664\u4e86boot_lay":32,"\u9664\u53bbdata\u5c42":57,"\u9664\u6784\u9020\u67d0\u79cd\u7c7b\u578b\u7684\u51fd\u6570":21,"\u9664\u8bcd\u5411\u91cf\u6a21\u578b\u5916\u7684\u53c2\u6570\u5c06\u4f7f\u7528\u6b63\u6001\u5206\u5e03\u968f\u673a\u521d\u59cb\u5316":53,"\u9664\u96f6\u7b49\u95ee\u9898":24,"\u968f\u540e\u53ef\u4ee5\u7528\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf\u5f00build":36,"\u968f\u673a\u521d\u59cb\u4e0d\u5b58\u5728\u7684\u53c2\u6570":60,"\u968f\u673a\u6570\u7684\u79cd\u5b50":43,"\u968f\u673a\u6570seed":42,"\u968f\u7740\u8f6e\u6570\u589e\u52a0\u8bef\u5dee\u4ee3\u4ef7\u51fd\u6570\u7684\u8f93\u51fa\u5728\u4e0d\u65ad\u7684\u51cf\u5c0f":25,"\u9694\u5f00":55,"\u96c6":58,"\u96c6\u675f\u641c\u7d22\u4e2d\u7684\u6269\u5c55\u5e7f\u5ea6":62,"\u96c6\u675f\u641c\u7d22\u4f7f\u7528\u5e7f\u5ea6\u4f18\u5148\u641c\u7d22\u6765\u6784\u5efa\u641c\u7d22\u6811":62,"\u96c6\u675f\u641c\u7d22\u4f7f\u7528\u5e7f\u5ea6\u4f18\u5148\u641c\u7d22\u7684\u65b9\u5f0f\u6784\u5efa\u67e5\u627e\u6811":43,"\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673a\u4f7f\u7528cpu\u7684paddle\u8bad\u7ec3\u4f5c\u4e1a":48,"\u96c6\u7fa4\u4f5c\u4e1a\u4e2d\u6240\u6709\u8fdb\u7a0b\u7684\u73af\u5883\u8bbe\u7f6e":41,"\u96c6\u7fa4\u4f5c\u4e1a\u5c06\u4f1a\u5728\u51e0\u79d2\u540e\u542f\u52a8":41,"\u96c6\u7fa4\u5de5\u4f5c":41,"\u96c6\u7fa4\u6d4b\u8bd5":42,"\u96c6\u7fa4\u8bad\u7ec3":42,"\u96c6\u7fa4\u8fdb\u7a0b":41,"\u96c6\u7fa4\u901a\u4fe1\u4fe1\u9053\u7684\u7aef\u53e3\u6570":41,"\u96c6\u7fa4\u901a\u4fe1\u901a\u9053\u7684":41,"\u96c6\u7fa4\u901a\u4fe1\u901a\u9053\u7684\u7aef\u53e3\u53f7":41,"\u9700\u5728nvvp\u754c\u9762\u4e2d\u9009\u4e0a\u624d\u80fd\u5f00\u542f":40,"\u9700\u8981\u4e0e":46,"\u9700\u8981\u4f7f\u7528\u5176\u5236\u5b9a\u7684\u65b9\u5f0f\u6302\u8f7d\u540e\u5e76\u5bfc\u5165\u6570\u636e":49,"\u9700\u8981\u4fdd\u5b58\u4e00\u4e9b\u9884\u7f6e\u7684\u516c\u5f00\u6570\u636e\u96c6":19,"\u9700\u8981\u4fdd\u6301\u5f53\u524d\u5206\u652f\u76ee\u5f55":36,"\u9700\u8981\u5148\u6302\u8f7d\u5230\u670d\u52a1\u5668node\u4e0a\u518d\u901a\u8fc7kubernet":47,"\u9700\u8981\u540c\u6b65\u539f\u4ed3\u5e93":36,"\u9700\u8981\u542f\u52a8":46,"\u9700\u8981\u542f\u52a8\u7684\u8282\u70b9\u4e2a\u6570\u4ee5\u53ca":49,"\u9700\u8981\u5728":41,"\u9700\u8981\u5728\u521b\u5efa\u5bb9\u5668\u524d\u6302\u8f7d\u5377\u4ee5\u4fbf\u6211\u4eec\u4fdd\u5b58\u8bad\u7ec3\u7ed3\u679c":48,"\u9700\u8981\u5728\u7cfb\u7edf\u91cc\u5148\u5b89\u88c5\u597ddocker\u5de5\u5177\u5305":38,"\u9700\u8981\u5728cmake\u7684\u65f6\u5019":21,"\u9700\u8981\u5b89\u88c5graphviz\u6765\u8f6c\u6362dot\u6587\u4ef6\u4e3a\u56fe\u7247":55,"\u9700\u8981\u5bf9":47,"\u9700\u8981\u5c06\u5176parameter\u8bbe\u7f6e\u6210":24,"\u9700\u8981\u5c06\u6807\u8bb0\u7684\u8bad\u7ec3\u548c\u6d4b\u8bd5\u6837\u672c\u79fb\u52a8\u5230\u53e6\u4e00\u4e2a\u8def\u5f84":61,"\u9700\u8981\u5c06bugfix\u7684\u5206\u652f\u540c\u65f6merge\u5230":23,"\u9700\u8981\u5f15\u7528":21,"\u9700\u8981\u6307\u5b9a\u4e0e\u67d0\u4e00\u4e2a\u8f93\u5165\u7684\u5e8f\u5217\u4fe1\u606f\u662f\u4e00\u81f4\u7684":32,"\u9700\u8981\u660e\u786e\u6307\u5b9a":43,"\u9700\u8981\u6709\u4e00\u4e2a\u5916\u90e8\u7684\u5b58\u50a8\u670d\u52a1\u6765\u4fdd\u5b58\u8bad\u7ec3\u6240\u9700\u6570\u636e\u548c\u8bad\u7ec3\u8f93\u51fa":47,"\u9700\u8981\u6709\u7a33\u5b9a\u7684\u5bfc\u51fa\u7b26\u53f7":20,"\u9700\u8981\u6839\u636e\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u6765\u7ed1\u5b9a\u4e00\u4e2a":49,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[23,43,46,59],"\u9700\u8981\u6ce8\u610f\u7684\u662f\u68af\u5ea6\u68c0\u67e5\u4ec5\u4ec5\u9a8c\u8bc1\u4e86\u68af\u5ea6\u7684\u8ba1\u7b97":37,"\u9700\u8981\u6ce8\u610f\u7684\u662fpaddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5b50\u5e8f\u5217\u6570\u76ee\u4e00\u6837\u7684\u591a\u8f93\u5165\u53cc\u5c42rnn":32,"\u9700\u8981\u88ab\u66b4\u9732\u5230\u5176\u4ed6\u8bed\u8a00":21,"\u9700\u8981\u9075\u5faa\u4ee5\u4e0b\u7ea6\u5b9a":34,"\u9700\u8981import\u8fd9\u4e9b\u51fd\u6570":46,"\u9700\u8981python\u63a5\u53e3\u91cc\u5904\u7406shuffl":46,"\u975e\u5e38\u6570":37,"\u975e\u96f6\u6570\u5b57\u7684\u4e2a\u6570":37,"\u97f3\u4e50\u5267":58,"\u987a\u5e8f":32,"\u9884\u5904\u7406\u6570\u636e\u4e00\u822c\u7684\u547d\u4ee4\u4e3a":59,"\u9884\u5904\u7406\u811a\u672c":61,"\u9884\u5b9a\u4e49\u7f51\u7edc":61,"\u9884\u5b9a\u4e49\u7f51\u7edc\u5982\u56fe3\u6240\u793a":61,"\u9884\u63d0\u4ea4\u94a9\u5b50":36,"\u9884\u6d4b\u540e":60,"\u9884\u6d4b\u63a5\u53e3\u811a\u672c":61,"\u9884\u6d4b\u6982\u7387\u53d6\u5e73\u5747":55,"\u9884\u6d4b\u7a0b\u5e8f\u5c06\u8bfb\u53d6\u7528\u6237\u7684\u8f93\u5165":59,"\u9884\u6d4b\u7ed3\u679c\u4ee5\u6587\u672c\u7684\u5f62\u5f0f\u4fdd\u5b58\u5728":57,"\u9884\u6d4b\u811a\u672c\u662f":60,"\u9884\u6d4b\u9636\u6bb5":46,"\u9884\u6d4bid":57,"\u9884\u6d4bimdb\u7684\u672a\u6807\u8bb0\u8bc4\u8bba\u7684\u4e00\u4e2a\u5b9e\u4f8b\u5982\u4e0b":61,"\u9884\u8bad\u7ec3\u6a21\u578b\u4f7f\u7528\u7684\u5b57\u5178\u7684\u8def\u5f84":53,"\u9884\u8bad\u7ec3\u8bcd\u5411\u91cf\u5b57\u5178\u6a21\u578b\u7684\u8def\u5f84":53,"\u989c\u8272\u901a\u9053\u987a\u5e8f\u4e3a":55,"\u989d\u5916\u7684\u53c2\u6570":57,"\u9996\u5148":[3,25,32,35,37,53,55,57,60,61],"\u9996\u5148\u4e0b\u8f7dcifar":54,"\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":26,"\u9996\u5148\u5b89\u88c5\u5e76\u5728\u5f53\u524d\u76ee\u5f55\u8fd0\u884c\u5b83":36,"\u9996\u5148\u5b89\u88c5paddlepaddl":61,"\u9996\u5148\u5bf9\u8f93\u5165\u505a\u4e00\u4e2a\u5c0f\u7684\u6270\u52a8":37,"\u9996\u5148\u6211\u4eec\u9700\u8981\u63a8\u5bfc\u8be5\u7f51\u7edc\u5c42\u7684":37,"\u9996\u5148\u662f\u6cd5\u8bed\u5e8f\u5217":62,"\u9996\u5148\u901a\u8fc7":36,"\u9a71\u52a8":38,"\u9ad8\u4e2d\u6bd5\u4e1a\u751f":58,"\u9ad8\u4eae\u90e8\u5206":32,"\u9ad8\u53ef\u7528":47,"\u9ad8\u5ea6\u652f\u6301\u7075\u6d3b\u548c\u9ad8\u6548\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":35,"\u9ad8\u6548\u6027":0,"\u9ad8\u65af\u5206\u5e03":24,"\u9ed1\u8272\u7535\u5f71":58,"\u9ed8\u8ba4":[3,43,62],"\u9ed8\u8ba4\u4e00\u4e2apass\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":57,"\u9ed8\u8ba4\u4e0d\u663e\u793a":43,"\u9ed8\u8ba4\u4e0d\u8bbe\u7f6e":34,"\u9ed8\u8ba4\u4e3a0":[43,45],"\u9ed8\u8ba4\u4e3a1":[3,45],"\u9ed8\u8ba4\u4e3a100":45,"\u9ed8\u8ba4\u4e3a4096mb":43,"\u9ed8\u8ba4\u4e3a\u4e0d\u4f7f\u7528":59,"\u9ed8\u8ba4\u4e3a\u7b2c\u4e00\u4e2a\u8f93\u5165":34,"\u9ed8\u8ba4\u4e3anull":43,"\u9ed8\u8ba4\u4f7f\u7528\u591a\u7c7b\u4ea4\u53c9\u71b5\u635f\u5931\u51fd\u6570\u548c\u5206\u7c7b\u9519\u8bef\u7387\u7edf\u8ba1\u8bc4\u4f30\u5668":46,"\u9ed8\u8ba4\u4f7f\u7528concurrentremoteparameterupdat":43,"\u9ed8\u8ba4\u503c":[26,31,45],"\u9ed8\u8ba4\u521d\u59cb\u72b6\u4e3a0":34,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b":[24,41,61],"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u4f7f\u7528\u6b64\u7f51\u7edc":61,"\u9ed8\u8ba4\u6307\u5b9a\u7b2c\u4e00\u4e2a\u8f93\u5165":32,"\u9ed8\u8ba4\u662f0":46,"\u9ed8\u8ba4\u662f1":46,"\u9ed8\u8ba4\u7528\u6765\u5207\u5206\u5355\u8bb0\u548c\u6807\u70b9\u7b26\u53f7":61,"\u9ed8\u8ba4\u7684":48,"\u9ed8\u8ba4\u8bbe\u7f6e\u4e3a\u771f":45,"\u9ed8\u8ba4\u914d\u7f6e\u5982\u4e0b":41,"adamax\u7b49":57,"amazon\u7535\u5b50\u4ea7\u54c1\u8bc4\u8bba\u6570\u636e":57,"api\u4e2d\u4f7f\u7528":20,"api\u5bf9\u6bd4\u4ecb\u7ecd":33,"api\u5bfc\u51fa\u7684\u52a8\u6001\u5e93":21,"api\u5bfc\u51fa\u7684\u9759\u6001\u5e93":21,"api\u63a5\u53d7\u7684\u7c7b\u578b\u5168\u662f":21,"api\u63a5\u53e3":47,"api\u63a5\u53e3\u7684\u53c2\u6570\u8f6c\u53d1\u7ed9":21,"api\u6587\u6863\u751f\u6210\u8fc7\u7a0b\u4f9d\u8d56\u4e8epy_paddl":38,"api\u65f6":21,"api\u65f6\u6240\u552f\u4e00\u9700\u8981\u5f15\u5165\u7684\u5934\u6587\u4ef6":21,"api\u662f\u591a\u8bed\u8a00api\u7684\u57fa\u7840\u90e8\u5206":21,"api\u66b4\u9732\u7684\u7c7b\u578b":21,"api\u751f\u6210\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u4f1a\u88ab\u5b89\u88c5\u5230":21,"api\u7684\u5b9e\u4f8b":21,"api\u7684\u5b9e\u73b0\u7ec6\u8282":21,"api\u7684\u63a5\u53e3":21,"api\u7684\u65f6\u5019\u63a8\u8350paddle\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":21,"api\u7684\u7f16\u8bd1\u9009\u9879\u9ed8\u8ba4\u5173\u95ed":21,"api\u76ee\u5f55\u7ed3\u6784\u5982\u4e0a\u56fe\u8868\u6240\u793a":21,"api\u83b7\u5f97\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u5b9e\u4f8b":21,"async_sgd\u8fdb\u884c\u8bad\u7ec3\u65f6":24,"atlas\u7684\u8def\u5f84":26,"avx\u7684\u955c\u50cf":27,"awselasticblockstore\u7b49":47,"batch\u4e2d\u5305\u542b":24,"batches\u4e2a\u6279\u6b21\u4fdd\u5b58\u4e00\u6b21\u53c2\u6570":43,"batches\u6b21":43,"bin\u548c\u8bc4\u5206\u6587\u4ef6":59,"blas\u7684\u8def\u5f84":26,"block\u6784\u6210\u4e00\u4e2amodel":18,"book\u4e00\u5b9a\u662f\u60a8\u6700\u597d\u7684\u9009\u62e9":27,"book\u4e2d\u6240\u6709\u7ae0\u8282\u529f\u80fd\u7684\u6b63\u786e\u6027":23,"book\u662f\u4e3a\u7528\u6237\u548c\u5f00\u53d1\u8005\u5236\u4f5c\u7684\u4e00\u4e2a\u4ea4\u4e92\u5f0f\u7684jupyt":27,"book\u7684docker\u955c\u50cf":27,"bool\u578b\u53c2\u6570":3,"boolean":[9,20],"bugfix\u5206\u652f\u4e5f\u662f\u5728\u5f00\u53d1\u8005\u81ea\u5df1\u7684fork\u7248\u672c\u5e93\u7ef4\u62a4":23,"bugfix\u5206\u652f\u9700\u8981\u5206\u522b\u7ed9\u4e3b\u7248\u672c\u5e93\u7684":23,"build\u6307\u4ee4\u6309\u7167dockerfile\u751f\u6210\u955c\u50cf":27,"build\u8fd9\u4e2a\u547d\u4ee4\u7684":27,"byte":24,"c99\u662f\u76ee\u524dc\u6700\u5e7f\u6cdb\u7684\u4f7f\u7528\u6807\u51c6":20,"c\u6709\u6807\u51c6\u7684abi":20,"c\u8bed\u8a00\u662f\u6709\u5bfc\u51fa\u7b26\u53f7\u7684\u6807\u51c6\u7684":20,"caoying\u7684pul":62,"case":[9,21,22,40],"class":[6,7,8,9,10,11,12,13,15,16,20,24,37,61],"cmake\u4e2d\u5c06":40,"cmake\u627e\u5230\u7684python\u5e93\u548cpython\u89e3\u91ca\u5668\u7248\u672c\u53ef\u80fd\u6709\u4e0d\u4e00\u81f4\u73b0\u8c61":24,"cmake\u7f16\u8bd1\u65f6":26,"cmake\u914d\u7f6e\u4e2d\u5c06":40,"conf\u4f5c\u4e3a\u914d\u7f6e":62,"const":37,"container\u4e2d":48,"core\u4e2d\u7684\u6a21\u578b\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u53c2\u6570":21,"core\u4e2d\u8fd9\u4e00\u7c7b\u578b\u63a5\u53e3\u7684\u667a\u80fd\u6307\u9488":21,"core\u662f\u5426\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u5b9e\u4f8b":21,"core\u6982\u5ff5":21,"cost\u8fd8\u5927\u4e8e\u8fd9\u4e2a\u6570":24,"count\u4e2agpu\u4e0a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97\u67d0\u4e00\u5c42":45,"count\u548cgpu":45,"cpu\u7248\u672c":29,"cuda\u5e73\u53f0":29,"cuda\u5e93":43,"cudnn\u5e93":[26,43],"dat\u4e2d":59,"data\u5230\u5206\u5e03\u5f0f\u5b58\u50a8\u8865\u5145\u8bad\u7ec3\u6570\u636e":19,"data\u76ee\u5f55\u4e2d\u5b58\u653e\u5207\u5206\u597d\u7684\u6570\u636e":49,"dataprovider\u5171\u8fd4\u56de\u4e24\u4e2a\u6570\u636e":32,"dataprovider\u5171\u8fd4\u56de\u4e24\u7ec4\u6570\u636e":32,"dataprovider\u662f\u88ab\u7cfb\u7edf\u8c03\u7528":46,"dataprovider\u662fpaddlepaddle\u7cfb\u7edf\u7684\u6570\u636e\u63d0\u4f9b\u5668":46,"dataprovider\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":2,"dataprovider\u7684\u4ecb\u7ecd":[4,57],"dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":24,"dataprovider\u8fd4\u56de\u7a7a\u6570\u636e":46,"dataprovider\u91cc\u5b9a\u4e49\u6570\u636e\u8bfb\u53d6\u51fd\u6570":46,"deb\u5305":23,"deb\u5305\u7f16\u8bd1\u95ee\u9898":23,"deb\u5b89\u88c5\u5305":29,"decay\u5219\u4e3a0":54,"decoder\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4f1a\u5f15\u7528\u5168\u90e8\u7ed3\u679c":34,"decoder\u63a5\u53d7\u4e24\u4e2a\u8f93\u5165":34,"decoder\u6bcf\u6b21\u9884\u6d4b\u4ea7\u751f\u4e0b\u4e00\u4e2a\u6700\u53ef\u80fd\u7684\u8bcd\u8bed":34,"decoer\u67b6\u6784":34,"default":[7,8,9,10,12,13,15,16,45,48,49,61],"demo\u9884\u6d4b\u8f93\u51fa\u5982\u4e0b":5,"dictionary\u662f\u4ece\u7f51\u7edc\u914d\u7f6e\u4e2d\u4f20\u5165\u7684dict\u5bf9\u8c61":3,"dictionary\u7531\u89e3\u6790\u81ea\u52a8\u751f\u6210":59,"dir\u4e2d\u670916\u4e2a\u5b50\u76ee\u5f55":62,"docker\u5b58\u5728\u95ee\u9898":27,"docker\u5b89\u88c5\u8bf7\u53c2\u8003":38,"docker\u7684\u5b98\u7f51":38,"docker\u955c\u50cf\u6765\u670d\u52a1html\u4ee3\u7801":27,"dockerfile\u4f4d\u4e8epaddlepaddl":27,"dockers\u8bbe\u7f6e":27,"dropout\u7684\u6bd4\u4f8b":37,"elec\u6d4b\u8bd5\u96c6":57,"embedding\u6a21\u578b\u9700\u8981\u7a0d\u5fae\u6539\u53d8\u63d0\u4f9b\u6570\u636e\u7684python\u811a\u672c":57,"encode\u6210\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":32,"encoder\u548cdecoder\u53ef\u4ee5\u662f\u80fd\u591f\u5904\u7406\u5e8f\u5217\u7684\u4efb\u610f\u795e\u7ecf\u7f51\u7edc\u5355\u5143":34,"encoder\u8f93\u51fa":34,"entropy\u4f5c\u4e3acost":24,"evaluator\u7684\u503c\u4f4e\u4e8e0":62,"export":[24,27,29,54],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":3,"false\u7684\u60c5\u51b5":3,"fc1\u548cfc2\u5c42\u5728gpu\u4e0a\u8ba1\u7b97":45,"fc3\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":45,"final":[10,59],"float":[3,7,8,9,11,13,25,40,55,59],"float\u7b49":45,"function":[9,10,11,13,16,17,22,35,61],"gen\u6587\u4ef6\u5939\u4e2d\u7684\u6587\u4ef6\u5217\u8868":62,"generator\u4fbf\u4f1a\u5b58\u4e0b\u5f53\u524d\u7684\u4e0a\u4e0b\u6587":3,"generator\u81f3\u5c11\u9700\u8981\u8c03\u7528\u4e24\u6b21\u624d\u4f1a\u77e5\u9053\u662f\u5426\u505c\u6b62":3,"git\u6d41\u5206\u652f\u6a21\u578b":36,"github\u9996\u9875":36,"golang\u53ef\u4ee5\u4f7f\u7528":20,"golang\u7684":20,"gpu\u4e8c\u8fdb\u5236\u6587\u4ef6":26,"gpu\u5219\u8fd8\u9700\u8981\u9ad8\u5e76\u884c\u6027":40,"gpu\u53cc\u7f13\u5b58":3,"gpu\u548c\u975eavx\u533a\u5206\u4e86\u5982\u4e0b4\u4e2a\u955c\u50cf":27,"gpu\u548ccpu\u901a\u4fe1":46,"gpu\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18":39,"gpu\u6838\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u6307\u5b9a":43,"gpu\u7248\u672c":29,"gpu\u7248\u672c\u5e76\u60f3\u4f7f\u7528":60,"gpu\u7684docker\u955c\u50cf\u7684\u65f6\u5019":24,"gram\u7ea7\u522b\u7684\u77e5\u8bc6":61,"group\u6559\u7a0b":33,"gru\u6216lstm":35,"gru\u6a21\u578b":57,"gru\u6a21\u578b\u914d\u7f6e":57,"h\u5e76\u4e0d\u56f0\u96be":20,"hot\u7a20\u5bc6\u5411\u91cf":59,"html\u5373\u53ef\u8bbf\u95ee\u672c\u5730\u6587\u6863":38,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":3,"id\u4e3a0\u7684\u6982\u7387":57,"id\u4e3a1\u7684\u6982\u7387":57,"id\u6307\u5b9a\u4f7f\u7528\u54ea\u4e2agpu\u6838":43,"id\u6307\u5b9a\u7684gpu":45,"id\u65e0\u6548":43,"image\u91cc":48,"imagenet\u6570\u636e\u96c6\u7b49":19,"images\u6570\u636e\u96c6\u4e0a\u4f20\u5230\u4e91\u7aef\u7684":19,"imdb\u6570\u636e\u96c6\u5305\u542b25":61,"imdb\u6709\u4e24\u4e2a\u6807\u7b7e":61,"imdb\u7684\u6570\u6910\u96c6":61,"import":[3,5,8,9,16,24,25,38,46,53,54,55,59,61,62],"include\u4e0b\u9700\u8981\u5305\u542bcbla":26,"include\u4e0b\u9700\u8981\u5305\u542bmkl":26,"init_hook\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":3,"instance\u4e0e\u751f\u6210\u6570\u636e\u96c6\u65f6":19,"instance\u5305\u6db5\u4e24\u4e2a\u503c":19,"instance\u662f\u4e00\u6a21\u4e00\u6837\u7684":19,"int":[3,7,8,9,10,13,20,21,22,32,37,45,57,59,60],"interface\u6587\u4ef6\u7684\u5199\u6cd5\u975e\u5e38":20,"issue\u7f16\u53f7":36,"job\u542f\u52a8\u540e\u4f1a\u521b\u5efa\u8fd9\u4e9bpod\u5e76\u5f00\u59cb\u6267\u884c\u4e00\u4e2a\u7a0b\u5e8f":47,"job\u5b8c\u6210\u8f6c\u6362":19,"job\u6216\u8005\u5e94\u7528\u7a0b\u5e8f\u5728\u5bb9\u5668\u4e2d\u8fd0\u884c\u65f6\u751f\u6210\u7684\u6570\u636e\u4f1a\u5728\u5bb9\u5668\u9500\u6bc1\u65f6\u6d88\u5931":47,"job\u662f\u672c\u6b21\u8bad\u7ec3\u5bf9\u5e94\u7684job":49,"job\u7684\u540d\u5b57":49,"kubernetes\u4e3a\u8fd9\u6b21\u8bad\u7ec3\u521b\u5efa\u4e863\u4e2apod\u5e76\u4e14\u8c03\u5ea6\u5230\u4e863\u4e2anode\u4e0a\u8fd0\u884c":49,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":39,"kubernetes\u5355\u673a\u8bad\u7ec3":39,"kubernetes\u53ef\u4ee5\u5728\u7269\u7406\u673a\u6216\u865a\u62df\u673a\u4e0a\u8fd0\u884c":47,"kubernetes\u53ef\u4ee5\u901a\u8fc7yaml\u6587\u4ef6\u6765\u521b\u5efa\u76f8\u5173\u5bf9\u8c61":49,"kubernetes\u5c31\u4f1a\u521b\u5efa3\u4e2apod\u4f5c\u4e3apaddlepaddle\u8282\u70b9\u7136\u540e\u62c9\u53d6\u955c\u50cf":49,"kubernetes\u63d0\u4f9b\u4e86\u591a\u79cd\u96c6\u7fa4\u90e8\u7f72\u7684\u65b9\u6848":47,"kubernetes\u652f\u6301\u591a\u79cdvolum":47,"kubernetes\u6709job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u652f\u6301":48,"kubernetes\u96c6\u7fa4\u5c31\u662f\u7531node\u8282\u70b9\u4e0emaster\u8282\u70b9\u7ec4\u6210\u7684":47,"label\u662f\u539f\u59cb\u6570\u636e\u4e2d\u5bf9\u4e8e\u6bcf\u4e00\u53e5\u8bdd\u7684\u5206\u7c7b\u6807\u7b7e":32,"labels\u662f\u6bcf\u7ec4\u5185\u6bcf\u4e2a\u53e5\u5b50\u7684\u6807\u7b7e":32,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":31,"layer\u62ff\u5230\u7684\u7528\u6237\u8f93\u5165":34,"layer\u7c7b\u53ef\u4ee5\u81ea\u52a8\u8ba1\u7b97\u4e0a\u9762\u7684\u5bfc\u6570":37,"layer\u91cc\u9762\u53ef\u4ee5\u5b9a\u4e49\u53c2\u6570\u5c5e\u6027":46,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u548catlas\u4e24\u4e2a\u5e93":26,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u5e93":26,"lib\u4e0b\u9700\u8981\u5305\u542bopenblas\u5e93":26,"lib\u76ee\u5f55\u4e0b\u9700\u8981\u5305\u542bmkl_cor":26,"list\u4e2d\u7684\u6bcf\u4e00\u884c\u90fd\u4f20\u9012\u7ed9process\u51fd\u6570":3,"list\u4f5c\u4e3a\u68c0\u67e5\u5217\u8868":23,"list\u5199\u5165\u90a3\u4e2a\u6587\u672c\u6587\u4ef6\u7684\u5730\u5740":3,"list\u548ctest":2,"list\u5982\u4e0b\u6240\u793a":45,"list\u5b58\u653e\u5728\u672c\u5730":2,"list\u6216\u8005test":46,"list\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b\u5217\u8868":45,"long":[9,10,13],"lstm\u67b6\u6784\u7684\u6700\u5927\u4f18\u70b9\u662f\u5b83\u53ef\u4ee5\u5728\u957f\u65f6\u95f4\u95f4\u9694\u5185\u8bb0\u5fc6\u4fe1\u606f":61,"lstm\u6a21\u578b":57,"lstm\u6a21\u578b\u914d\u7f6e":57,"lstm\u7f51\u7edc\u7c7b\u4f3c\u4e8e\u5177\u6709\u9690\u85cf\u5c42\u7684\u6807\u51c6\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":61,"memory\u4e0d\u80fd\u72ec\u7acb\u5b58\u5728":34,"memory\u4e5f\u53ef\u4ee5\u5177\u6709":35,"memory\u4e5f\u53ef\u4ee5\u662f\u5e8f\u5217":35,"memory\u53ea\u80fd\u5728":34,"memory\u53ef\u4ee5\u7f13\u5b58\u4e0a\u4e00\u4e2a\u65f6\u523b\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u8f93\u51fa":32,"memory\u6307\u5411\u4e00\u4e2alay":34,"memory\u662f\u5728\u5355\u6b65\u51fd\u6570\u4e2d\u5faa\u73af\u4f7f\u7528\u7684\u72b6\u6001":35,"memory\u662fpaddlepaddle\u5b9e\u73b0rnn\u65f6\u5019\u4f7f\u7528\u7684\u4e00\u4e2a\u6982\u5ff5":32,"memory\u7684":35,"memory\u7684\u521d\u59cb\u72b6\u6001":34,"memory\u7684\u65f6\u95f4\u5e8f\u5217\u957f\u5ea6\u4e00\u81f4\u7684\u60c5\u51b5":32,"memory\u7684\u66f4\u591a\u8ba8\u8bba\u8bf7\u53c2\u8003\u8bba\u6587":34,"memory\u7684\u8f93\u51fa\u5b9a\u4e49\u5728":35,"memory\u7684i":34,"memory\u9ed8\u8ba4\u521d\u59cb\u5316\u4e3a0":34,"mkl\u7684\u8def\u5f84":26,"mkl_sequential\u548cmkl_intel_lp64\u4e09\u4e2a\u5e93":26,"mnist\u662f\u4e00\u4e2a\u5305\u542b\u670970":3,"mode\u548cattent":62,"mode\u7684python\u51fd\u6570":62,"model\u505a\u5206\u652f\u7ba1\u7406":23,"model\u53ef\u4ee5\u901a\u8fc7":5,"model\u6765\u5b9e\u73b0\u624b\u5199\u8bc6\u522b\u7684\u9884\u6d4b\u4ee3\u7801":5,"movielens\u6570\u636e\u96c6":59,"name\u662f\u4f53\u88c1":59,"name\u662f\u5e74\u9f84":59,"name\u662f\u6027\u522b":59,"name\u662f\u7535\u5f71\u540d":59,"name\u662f\u804c\u4e1a":59,"name\u7ec4\u5408\u53ef\u4ee5\u627e\u5230\u672c\u6b21\u8bad\u7ec3\u9700\u8981\u7684\u6587\u4ef6\u8def\u5f84":49,"ndarray\u7c7b\u578b\u7684\u503c\u548c\u6574\u578b\u7684\u503c":19,"new":[9,13,17,22,36,37],"nfs\u7684\u90e8\u7f72\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":47,"nmt\u6a21\u578b\u53d7\u5236\u4e8e\u6e90\u8bed\u53e5\u7684\u7f16\u7801":62,"noavx\u7248\u672c":29,"normalization\u5c42":55,"normalization\u5c42\u7684\u53c2\u6570":55,"note\u7684\u4e66\u5199":23,"notebook\u662f\u4e00\u4e2a\u5f00\u6e90\u7684web\u7a0b\u5e8f":27,"null":[9,37,43,59],"num_gradient_servers\u53c2\u6570":49,"openblas\u7684\u8def\u5f84":26,"operator\u7684\u6982\u5ff5":46,"out\u4e0b\u5305\u542b":54,"out\u7684\u6587\u4ef6\u5939":54,"outer_mem\u662f\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":32,"output\u6587\u4ef6\u5939\u5b58\u653e\u8bad\u7ec3\u7ed3\u679c\u4e0e\u65e5\u5fd7":49,"packages\u91cc\u9762":24,"packages\u91cc\u9762\u7684python\u5305":24,"paddepaddle\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":26,"paddle\u4e00\u4e2a\u52a8\u6001\u5e93\u53ef\u4ee5\u5728\u4efb\u4f55linux\u7cfb\u7edf\u4e0a\u8fd0\u884c":20,"paddle\u4e2d\u7684\u4e00\u6761pass\u8868\u793a\u8bad\u7ec3\u6570\u636e\u96c6\u4e2d\u6240\u6709\u7684\u6837\u672c\u4e00\u6b21":62,"paddle\u4e2d\u7ecf\u5e38\u4f1a\u5c06\u65f6\u95f4\u5e8f\u5217\u6210\u4e3a":32,"paddle\u4e8c\u8fdb\u5236\u5728\u8fd0\u884c\u65f6\u6355\u83b7\u4e86\u6d6e\u70b9\u6570\u5f02\u5e38":24,"paddle\u4f7f\u7528git":23,"paddle\u5185\u5d4c\u7684python\u89e3\u91ca\u5668\u548c\u5916\u90e8\u4f7f\u7528\u7684python\u5982\u679c\u7248\u672c\u4e0d\u540c":20,"paddle\u5185\u90e8\u7684\u7c7b\u4e3ac":20,"paddle\u5f00\u53d1\u8fc7\u7a0b\u4f7f\u7528":23,"paddle\u66ff\u6362\u6210":27,"paddle\u6bcf\u6b21\u53d1\u65b0\u7684\u7248\u672c":23,"paddle\u6bcf\u6b21\u53d1\u7248\u672c\u9996\u5148\u8981\u4fdd\u8bc1paddl":23,"paddle\u7684\u4e3b\u7248\u672c\u5e93\u9075\u5faa":23,"paddle\u7684\u5404\u7248\u672c\u955c\u50cf\u53ef\u4ee5\u53c2\u8003":48,"paddle\u7684\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0\u5305\u62ec\u4e00\u4e0b\u51e0\u4e2a\u65b9\u9762":20,"paddle\u7684\u7c7b\u578b\u5168\u90e8\u9000\u5316\u6210":21,"paddle\u7684\u94fe\u63a5\u65b9\u5f0f\u6bd4\u8f83\u590d\u6742":20,"paddle\u7684c":21,"paddle\u7684dock":48,"paddle\u7684docker\u5f00\u53d1\u955c\u50cf\u5e26\u6709\u4e00\u4e2a\u901a\u8fc7":27,"paddle\u8bad\u7ec3\u4efb\u52a1":19,"paddle\u8def\u5f84\u4e0b":21,"paddle\u955c\u50cf":48,"paddle\u9700\u8981\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3":20,"paddle\u9700\u8981\u66b4\u9732\u7684api\u5f88\u591a":21,"paddle\u9759\u6001\u5e93\u94fe\u63a5\u590d\u6742":20,"paddle_\u7c7b\u578b\u540d":21,"paddle_\u7c7b\u578b\u540d_\u51fd\u6570\u540d":21,"paddlepaddle\u4e2d":[31,34],"paddlepaddle\u4e2d\u7684\u4e00\u4e2apass\u610f\u5473\u7740\u5bf9\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u6837\u672c\u8fdb\u884c\u4e00\u6b21\u8bad\u7ec3":61,"paddlepaddle\u4e2d\u7684\u8bb8\u591alayer\u5e76\u4e0d\u5728\u610f\u8f93\u5165\u662f\u5426\u662f\u65f6\u95f4\u5e8f\u5217":32,"paddlepaddle\u4f1a\u5728\u8c03\u7528\u8bfb\u53d6\u6570\u636e\u7684python\u811a\u672c\u4e4b\u524d":57,"paddlepaddle\u4f7f\u7528\u5747\u503c0":24,"paddlepaddle\u4f7f\u7528avx":24,"paddlepaddle\u4f7f\u7528swig\u5bf9\u5e38\u7528\u7684\u9884\u6d4b\u63a5\u53e3\u8fdb\u884c\u4e86\u5c01\u88c5":5,"paddlepaddle\u4fdd\u7559\u6dfb\u52a0\u53c2\u6570\u7684\u6743\u529b":3,"paddlepaddle\u5148\u4ece\u4e00\u4e2a\u6587\u4ef6\u5217\u8868\u91cc\u83b7\u5f97\u6570\u636e\u6587\u4ef6\u5730\u5740":25,"paddlepaddle\u5305\u62ec\u5f88\u591a\u635f\u5931\u51fd\u6570\u548c\u8bc4\u4f30\u8d77":46,"paddlepaddle\u53d1\u5e03\u65b0\u7248\u672c\u7684\u65f6\u5019\u90fd\u4f1a\u53d1\u5e03\u5bf9\u5e94\u7248\u672c\u7684\u751f\u4ea7\u955c\u50cf\u4ee5\u53ca\u5f00\u53d1\u955c\u50cf":27,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528cudnn":26,"paddlepaddle\u53ef\u4ee5\u6267\u884c\u7528\u6237\u7684python\u811a\u672c\u7a0b\u5e8f\u6765\u8bfb\u53d6\u5404\u79cd\u683c\u5f0f\u7684\u6570\u636e\u6587\u4ef6":57,"paddlepaddle\u53ef\u4ee5\u6bd4\u8f83\u7b80\u5355\u7684\u5224\u65ad\u54ea\u4e9b\u8f93\u51fa\u662f\u5e94\u8be5\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684":32,"paddlepaddle\u53ef\u4ee5\u901a\u8fc7\u8be5\u673a\u5236\u5224\u65ad\u662f\u5426\u5df2\u7ecf\u6536\u96c6\u9f50\u6240\u6709\u7684\u68af\u5ea6":37,"paddlepaddle\u5728\u5b9e\u73b0rnn\u7684\u65f6\u5019":32,"paddlepaddle\u591a\u673a\u91c7\u7528\u7ecf\u5178\u7684":46,"paddlepaddle\u5b58\u7684\u662f\u6709\u503c\u4f4d\u7f6e\u7684\u7d22\u5f15":3,"paddlepaddle\u5b9a\u4e49\u7684\u53c2\u6570":3,"paddlepaddle\u5c06\u4ee5\u8bbe\u7f6e\u53c2\u6570\u7684\u65b9\u5f0f\u6765\u8bbe\u7f6e":57,"paddlepaddle\u5c06\u5728\u89c2\u6d4b\u6570\u636e\u96c6\u4e0a\u8fed\u4ee3\u8bad\u7ec330\u8f6e":25,"paddlepaddle\u5c06\u6bcf\u4e2a\u6a21\u578b\u53c2\u6570\u4f5c\u4e3a\u4e00\u4e2anumpy\u6570\u7ec4\u5355\u72ec\u5b58\u4e3a\u4e00\u4e2a\u6587\u4ef6":25,"paddlepaddle\u5c06train":3,"paddlepaddle\u5f00\u53d1\u955c\u50cf\u5c31\u88ab\u6784\u5efa\u5b8c\u6bd5\u4e86":27,"paddlepaddle\u63d0\u4f9b\u4e13\u7528\u7684":19,"paddlepaddle\u63d0\u4f9b\u4e86\u57fa\u4e8e":46,"paddlepaddle\u63d0\u4f9b\u4e86\u5f88\u591a\u4f18\u79c0\u7684\u5b66\u4e60\u7b97\u6cd5":25,"paddlepaddle\u63d0\u4f9b\u4e86ubuntu":29,"paddlepaddle\u63d0\u4f9b\u6570\u4e2a\u9884\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6765\u8fdb\u884c\u5b89\u88c5":28,"paddlepaddle\u652f\u6301\u4ee5\u4e0b\u4efb\u610f\u4e00\u79cdblas\u5e93":26,"paddlepaddle\u652f\u6301\u5927\u91cf\u7684\u8ba1\u7b97\u5355\u5143\u548c\u4efb\u610f\u6df1\u5ea6\u7684\u7f51\u7edc\u8fde\u63a5":25,"paddlepaddle\u652f\u6301\u975e\u5e38\u591a\u7684\u4f18\u5316\u7b97\u6cd5":24,"paddlepaddle\u652f\u6301sparse\u7684\u8bad\u7ec3":24,"paddlepaddle\u6587\u6863\u4f7f\u7528":38,"paddlepaddle\u6587\u6863\u9700\u8981\u51c6\u5907\u7684\u73af\u5883\u76f8\u5bf9\u8f83\u590d\u6742":38,"paddlepaddle\u662f\u4e00\u4e2a\u6700\u65e9\u7531\u767e\u5ea6\u79d1\u5b66\u5bb6\u548c\u5de5\u7a0b\u5e08\u5171\u540c\u7814\u53d1\u7684\u5e76\u884c\u5206\u5e03\u5f0f\u6df1\u5ea6\u5b66\u4e60\u5e73\u53f0":0,"paddlepaddle\u662f\u4e00\u4e2a\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6":46,"paddlepaddle\u662f\u6e90\u4e8e\u767e\u5ea6\u7684\u4e00\u4e2a\u6df1\u5ea6\u5b66\u4e60\u5e73\u53f0":25,"paddlepaddle\u6bcf\u6b21\u53d1\u5e03\u65b0\u7248\u672c\u90fd\u4f1a\u53d1\u5e03\u5bf9\u5e94\u7684\u5f00\u53d1\u955c\u50cf\u4f9b\u5f00\u53d1\u8005\u76f4\u63a5\u4f7f\u7528":27,"paddlepaddle\u7684\u5185\u5b58\u5360\u7528\u4e3b\u8981\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u65b9\u9762":24,"paddlepaddle\u7684\u53c2\u6570\u4f7f\u7528\u540d\u5b57":24,"paddlepaddle\u7684\u6570\u636e\u5305\u62ec\u56db\u79cd\u4e3b\u8981\u7c7b\u578b":3,"paddlepaddle\u7684\u6587\u6863\u5305\u62ec\u82f1\u6587\u6587\u6863":38,"paddlepaddle\u7684\u6587\u6863\u6784\u5efa\u6709\u76f4\u63a5\u6784\u5efa\u548c\u57fa\u4e8edocker\u6784\u5efa\u4e24\u79cd\u65b9\u5f0f":38,"paddlepaddle\u7684\u6e90\u7801":36,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":28,"paddlepaddle\u7684bas":37,"paddlepaddle\u7684docker\u5bb9\u5668\u4f7f\u7528\u65b9\u5f0f":28,"paddlepaddle\u7684trainer\u8fdb\u7a0b\u91cc\u5185\u5d4c\u4e86python\u89e3\u91ca\u5668":46,"paddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":32,"paddlepaddle\u76ee\u524d\u552f\u4e00\u5b98\u65b9\u652f\u6301\u7684\u8fd0\u884c\u7684\u65b9\u5f0f\u662fdocker\u5bb9\u5668":27,"paddlepaddle\u76ee\u524d\u5df2\u7ecf\u5f00\u653e\u6e90\u7801":0,"paddlepaddle\u76ee\u524d\u63d0\u4f9b\u4e24\u79cd\u53c2\u6570\u521d\u59cb\u5316\u7684\u65b9\u5f0f":24,"paddlepaddle\u8c03\u7528process\u51fd\u6570\u6765\u8bfb\u53d6\u6570\u636e":57,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u68af\u5ea6\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":34,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u8bef\u5dee\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":34,"paddlepaddle\u955c\u50cf\u9700\u8981\u63d0\u4f9b":49,"paddlepaddle\u9700\u8981\u7528\u6237\u5728\u7f51\u7edc\u914d\u7f6e":2,"paddlepaddle\u9700\u8981\u7684\u6240\u6709\u7f16\u8bd1\u5de5\u5177":27,"pass\u4e2a\u6a21\u578b\u5230\u7b2c":43,"pass\u5230":62,"pass\u5c06\u4e0d\u8d77\u4f5c\u7528":43,"pass\u8f6e\u5f00\u59cb\u8bad\u7ec3":43,"pass\u8f6e\u7684\u6a21\u578b\u7528\u4e8e\u6d4b\u8bd5":43,"passes\u8f6e":43,"patch\u53f7":23,"patch\u53f7\u52a0\u4e00":23,"path\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b":45,"period\u4e2a\u6279\u6b21\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":43,"period\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":43,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u53c2\u6570\u7edf\u8ba1":43,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u7b26\u53f7":43,"period\u4e2abatch\u5904\u7406\u7684\u5f53\u524d\u635f\u5931":61,"period\u4e2abatch\u7684\u5206\u7c7b\u9519\u8bef":61,"period\u6574\u9664":43,"period\u8f6e\u4fdd\u5b58\u8bad\u7ec3\u53c2\u6570":43,"pod\u4e2d\u7684\u5bb9\u5668\u5171\u4eabnet":47,"pod\u662fkubernetes\u7684\u6700\u5c0f\u8c03\u5ea6\u5355\u5143":47,"pooling\u5bf9\u7279\u5f81\u56fe\u4e0b\u91c7\u6837":54,"process\u51fd\u6570\u4f1a\u7528yield\u8bed\u53e5\u8f93\u51fa\u8fd9\u6761\u6570\u636e":57,"prod\u7684\u751f\u4ea7\u955c\u50cf":27,"pserver\u8fdb\u7a0b\u7528\u4e8e\u534f\u8c03\u591a\u4e2atrainer\u8fdb\u7a0b\u4e4b\u95f4\u7684\u901a\u4fe1":46,"public":[13,37,48,61],"py_paddle\u91cc\u9762\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5de5\u5177\u7c7b":5,"pydataprovider2\u4f1a\u5c3d\u53ef\u80fd\u591a\u7684\u4f7f\u7528\u5185\u5b58":3,"pydataprovider2\u63d0\u4f9b\u4e86\u4e24\u79cd\u7b80\u5355\u7684cache\u7b56\u7565":3,"pydataprovider2\u662fpaddlepaddle\u4f7f\u7528python\u63d0\u4f9b\u6570\u636e\u7684\u63a8\u8350\u63a5\u53e3":3,"pydataprovider2\u7684\u4f7f\u7528":[2,4,24,35,46,57,59],"pydataprovider\u4f7f\u7528\u7684\u662f\u5f02\u6b65\u52a0\u8f7d":24,"python\u4ee3\u7801\u5c06\u968f\u673a\u4ea7\u751f2000\u4e2a\u89c2\u6d4b\u70b9":25,"python\u5305":38,"python\u53ef\u4ee5\u89e3\u9664\u6389\u5185\u90e8\u53d8\u91cf\u7684\u5f15\u7528":3,"python\u5c01\u88c5\u7684\u5b9e\u73b0\u4f7f\u5f97\u6211\u4eec\u53ef\u4ee5\u5728\u914d\u7f6e\u6587\u4ef6\u4e2d\u4f7f\u7528\u65b0\u5b9e\u73b0\u7684\u7f51\u7edc\u5c42":37,"python\u811a\u672c\u91cc\u5b9a\u4e49\u4e86\u6a21\u578b\u914d\u7f6e":46,"query\u6539\u5199":62,"rate\u4e3a0":62,"rate\u4e3a5":62,"rate\u88ab\u8bbe\u7f6e\u4e3a0":54,"reader\u7684\u4f7f\u7528\u65b9\u5f0f\u90fd\u662f\u4e00\u81f4\u7684":19,"reader\u8f93\u51fa\u7684data":19,"recommendation\u6587\u4ef6\u5939\u5185\u5b58\u653e\u8bad\u7ec3\u6587\u4ef6":49,"release\u9875\u9762":23,"repo\u7684\u6839\u76ee\u5f55":27,"research\u5b9e\u9a8c\u5ba4\u641c\u96c6\u6574\u7406":58,"resnet\u6a21\u578b":56,"return":[3,9,10,12,13,15,16,19,25,32,35,37,49,55,57,59],"rnn\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":32,"rnn\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u901a\u8fc7\u4e86\u4e00\u4e2alstm\u7f51\u7edc":32,"rnn\u603b\u662f\u5f15\u7528\u4e0a\u4e00\u65f6\u523b\u9884\u6d4b\u51fa\u7684\u8bcd\u7684\u8bcd\u5411\u91cf":34,"rnn\u6a21\u578b":57,"rnn\u76f8\u5173\u6a21\u578b":39,"s3\u4e4b\u7c7b\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u4e4b\u4e0a":19,"search\u7684\u65b9\u6cd5":43,"sentences\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684\u6570\u636e":32,"seq\u53c2\u6570\u5fc5\u987b\u4e3afals":34,"server\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":43,"server\u4e4b\u4e0a":18,"server\u4e4b\u95f4\u7684\u7f51\u7edc\u5e26\u5bbd":18,"server\u4f1a\u6682\u505c\u53c2\u6570\u66f4\u65b0\u5e76\u7b49\u5f85":18,"server\u4f1a\u83b7\u53d6parameters\u5185\u5b58\u7684":18,"server\u5185\u5b58\u4e2d\u7684\u6a21\u578b\u6570\u636e\u7684\u5b8c\u6574\u955c\u50cf":18,"server\u540c\u6b65\u7684\u4fdd\u5b58\u4e00\u4e2a\u7279\u5b9a\u65f6\u95f4\u70b9\u7684\u5168\u5c40\u68c0\u67e5\u70b9":18,"server\u5728\u96c6\u7fa4\u4e2d\u542f\u52a8\u540e":18,"server\u6545\u969c\u540e\u88abkubernetes\u91cd\u65b0\u542f\u52a8":18,"server\u6b64\u65f6\u8fd8\u9700\u8981\u901a\u8fc7\u7f51\u7edc\u8bbf\u95ee\u5206\u5e03\u5f0f\u5b58\u50a8\u4ee5\u4fdd\u5b58\u5feb\u7167":18,"server\u751f\u6210\u4e00\u4e2auuid":18,"server\u7684\u5355\u70b9\u6216\u591a\u70b9\u540c\u65f6\u6545\u969c":18,"server\u7684\u6570\u636e\u5feb\u7167":18,"server\u7684\u68c0\u67e5\u70b9\u5404\u81ea\u72ec\u7acb\u4fdd\u5b58":18,"server\u7b2c\u4e00\u6b21\u542f\u52a8\u6216\u4efb\u610f\u65f6\u95f4paramet":18,"sh\u6765\u8bad\u7ec3\u6a21\u578b":54,"sh\u6765\u8fdb\u884c\u6784\u5efa":38,"sh\u8c03\u7528\u4e86":55,"short":[9,10],"simd\u6307\u4ee4\u63d0\u9ad8cpu\u6267\u884c\u6548\u7387":24,"size\u4e3a1":62,"size\u4e3a50":62,"size\u4e3a512":43,"size\u53ef\u80fd\u4f1a\u5bf9\u8bad\u7ec3\u7ed3\u679c\u4ea7\u751f\u5f71\u54cd":24,"size\u5927\u5c0f\u4e3a128":61,"size\u662f3":62,"size\u672c\u8eab\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u8d85\u53c2\u6570":24,"size\u7684\u503c":3,"softmax\u5c42":53,"softmax\u6fc0\u6d3b\u7684\u8f93\u51fa\u7684\u548c\u603b\u662f1":37,"sparse\u8bad\u7ec3\u9700\u8981\u8bad\u7ec3\u7279\u5f81\u662f":24,"srl\u4f5c\u4e3a\u5f88\u591a\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d\u7684\u4e2d\u95f4\u6b65\u9aa4\u662f\u5f88\u6709\u7528\u7684":60,"ssh\u65b9\u5f0f\u7684\u4e00\u4e2a\u4f18\u70b9\u662f\u6211\u4eec\u53ef\u4ee5\u4ece\u591a\u4e2a\u7ec8\u7aef\u8fdb\u5165\u5bb9\u5668":27,"ssh\u8fdb\u5165\u5bb9\u5668":27,"static":21,"step\u51fd\u6570\u4e2d\u7684memori":34,"step\u51fd\u6570\u5185\u90e8\u53ef\u4ee5\u81ea\u7531\u7ec4\u5408paddlepaddle\u652f\u6301\u7684\u5404\u79cdlay":34,"subseq\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":31,"super":37,"swig\u652f\u6301\u7684\u8bed\u8a00\u6216\u8005\u89e3\u91ca\u5668\u6709\u5c40\u9650":20,"swig\u66b4\u9732\u7684\u63a5\u53e3\u4fdd\u7559\u4e86c":20,"swig\u751f\u6210\u7684\u4ee3\u7801\u4e0d\u80fd\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":20,"swig\u76f4\u63a5\u8bfb\u53d6c":20,"swig\u9700\u8981\u5199\u4e00\u4e2ainterface\u6587\u4ef6":20,"swig_paddle\u4e2d\u7684\u9884\u6d4b\u63a5\u53e3\u7684\u53c2\u6570\u662f\u81ea\u5b9a\u4e49\u7684c":5,"switch":21,"t\u6307\u5b9a\u4e86\u751f\u6210\u7684\u955c\u50cf\u7684\u540d\u5b57":27,"tag\u4e3a":23,"test\u548cgen\u8fd9\u4e09\u4e2a\u6587\u4ef6\u5939\u662f\u56fa\u5b9a\u7684":62,"tests\u7684paddlepaddl":36,"tflops\u4e86":40,"trainer\u8fdb\u7a0b\u4f1a\u8c03\u7528dataprovider\u51fd\u6570\u8fd4\u56de\u6570\u636e":46,"trainer\u8fdb\u7a0b\u53ef\u4ee5\u5229\u7528\u8fd9\u4e2a\u89e3\u91ca\u5668\u6267\u884cpython\u811a\u672c":46,"true":[7,8,9,10,12,13,15,16,22,24,32,35,37,45,49,55,59,60,61,62],"true\u8868\u793a\u53cd\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":35,"try":[11,17,22,24,59],"type\u5b57\u6bb5\u5747\u4e0d\u5c3d\u76f8\u540c":21,"type\u662fon":59,"ubuntu\u5b89\u88c5\u5305\u7684\u529f\u80fd\u6b63\u786e\u6027":23,"ubuntu\u7684deb\u5b89\u88c5\u5305\u7b49":28,"ubuntu\u90e8\u7f72paddlepaddl":28,"update\u53c2\u6570\u65f6\u624d\u6709\u6548":43,"utf8\u7f16\u7801":53,"uts\u7b49linux":47,"v2\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2a\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":26,"vocab\u4e2d\u6bcf\u4e2a\u5207\u5206\u5355\u8bcd\u7684\u9884\u671f\u8bc4\u7ea7":61,"vocab\u505a\u4e3a\u5b57\u5178":61,"void":[20,21,37],"volume\u6302\u8f7d\u5230\u5bb9\u5668\u4e2d":47,"w0\u548c":55,"wbias\u662f\u9700\u8981\u5b66\u4e60\u7684\u53c2\u6570":55,"while":[7,13,22,49,62],"words\u5373\u4e3a\u8fd9\u4e2a\u6570\u636e\u4e2d\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":32,"words\u662f\u539f\u59cb\u6570\u636e\u4e2d\u7684\u6bcf\u4e00\u53e5\u8bdd":32,"x\u548cwindow":27,"x\u548cwindows\u4e0a\u7684\u786c\u4ef6\u8d44\u6e90":27,"yaml\u6587\u4ef6\u4e2d\u5404\u4e2a\u5b57\u6bb5\u7684\u5177\u4f53\u542b\u4e49":49,"yaml\u6587\u4ef6\u63cf\u8ff0\u4e86\u8fd9\u6b21\u8bad\u7ec3\u4f7f\u7528\u7684docker\u955c\u50cf":49,"zero\u4e09\u79cd\u64cd\u4f5c":43,AGE:48,AWS:[19,47,50,51],And:[8,9,11,13,22],But:[9,10,24],EOS:9,For:[5,8,9,11,13,16,22,29,40],IDs:13,NFS:47,Not:[16,17],One:[8,10,15],QoS:48,TLS:16,That:[9,13,22],The:[3,6,7,8,9,10,11,13,15,16,17,21,22,37,49,57,59,60,62],Their:[9,17],Then:[9,59],There:[8,9,13,15,16,17],Use:[16,22],Used:10,Using:[17,61],Will:[13,15],With:[9,10],Yes:27,___embedding_0__:49,___embedding_1__:49,__init__:37,__list_to_map__:59,__main__:[5,55],__meta__:59,__mse_cost_0__:49,__name__:[5,55],__rnn_step__:35,_link:10,_proj:9,_recurrent_group:35,_res2_1_branch1_bn:55,_source_language_embed:[35,53],_target_language_embed:[35,53],aaaaa:19,abc:9,abl:[9,16],about:[9,10,60,62],abov:[3,9,16,17,40],abs:10,acceler:18,accept:[13,16,22,60],access:[9,10,16],accord:[8,9],accrod:10,accumul:17,accuraci:8,acl:61,aclimdb:61,across:9,act:[9,10,24,25,32,35,46],act_typ:57,activ:[9,10,14,46,57],activi:10,actual:9,adadelta:[24,57],adagrad:57,adam:[16,57,61],adamoptim:[46,53,57,61,62],adapt:[8,11],add:[9,10,13,36,59],add_input:37,add_test:37,add_to:9,add_unittest_without_exec:37,addbia:37,added:[8,36],addit:[9,10],address:17,addrow:37,addtolay:9,adversari:22,affect:9,afi:3,after:[9,13],again:[16,17],age:[13,49,59],agg_level:[9,31,32],aggregatelevel:[9,31,32],aircraft:62,airplan:54,aistat:9,alex:[9,61],alexnet_pass1:45,alexnet_pass2:45,algo_hrnn_demo:32,algorithm:[9,11,53,61,62],align:[9,10,13,62],all:[3,7,8,9,11,16,17,21,34,49,59,60,61],alloc:7,allow:[16,57],allow_only_one_model_on_one_gpu:[42,43,45],almost:10,alreadi:[17,24,29],also:[9,10,13,16,22,40,57],alwai:[9,10,15,22,49],amazon:48,ambigu:22,amd64:47,analysi:[60,61],ani:[9,10,13,16,17,22],annot:60,annual:60,anoth:[9,16],anyth:[13,22,60],api:[13,15,16,23,40,46,49,52,61],api_trainer_config:59,api_trainer_config_helpers_data_sourc:59,api_trainer_config_helpers_evalu:46,api_trainer_config_helpers_lay:[35,57],api_trainer_config_helpers_layers_context_project:59,api_trainer_config_helpers_layers_cos_sim:59,api_trainer_config_helpers_layers_cost_lay:46,api_trainer_config_helpers_layers_data_lay:[46,59],api_trainer_config_helpers_layers_embedding_lay:59,api_trainer_config_helpers_layers_fc_lay:[46,59],api_trainer_config_helpers_layers_mixed_lay:46,api_trainer_config_helpers_layers_pooling_lay:59,api_trainer_config_helpers_optim:46,api_trainer_config_helpers_optimizers_set:46,apiserv:47,apivers:[47,48,49],apo:62,append:[3,15,22,32,35,49,59],append_gradient_machin:15,appleyard:40,appli:[9,10],applic:[40,48],approach:[9,57],apt:[29,54],arbitrari:9,architectur:62,archiv:[13,20,21],arg:[3,8,9,10,13,24,25,49,54,55,57,59,60,61],arg_nam:9,argpars:49,args_ext:49,argument:[3,9,13,49,57,59,60],argumentpars:49,argv:55,around:[3,9],arrai:[5,9,13,15,22,55],articl:36,arxiv:[9,10,61],ask:17,assert:5,assign:9,associ:60,assum:9,astyp:22,asyc:17,async:[17,42],async_count:[42,43],async_lagged_grad_discard_ratio:43,async_lagged_ratio_default:[42,43],async_lagged_ratio_min:[42,43],asynchron:17,atla:26,atlas_root:26,attenion:10,attent:[9,10,62],attr:[7,9,10],attribut:[9,10,14],auc:42,auc_evalu:8,author:47,authorized_kei:41,auto:[20,37,40],automat:[9,16,62],automaticli:9,automobil:54,avail:17,averag:[8,9,12,60],average_test_period:[42,43,60],averagepool:9,avg:[31,40,57],avgcost:[57,59,61,62],avgpool:[9,57],avoid:[17,40],avx:27,await:48,awar:[16,17],azur:47,b2t:53,b363:48,b8561f5c79193550d64fa47418a9e67ebdd71546186e840f88de5026b8097465:48,back:17,backward:[6,9,10,37],backward_first:35,backwardactiv:37,bag:57,baidu:[9,48],balasubramanyan:61,bank:60,bardward:10,bare:[47,48],barrierstatset:40,base:[9,10,12,13,16],baseactiv:9,baseev:15,basematrix:37,basenam:8,basepoolingtyp:[9,10],basestr:[7,8,9,10,12,15,59],bash:[27,36,38,48,49],basic:[9,15],batch:[9,10,11,13,15,16,17,19,25,41,48,49,54,55,57,59,61,62],batch_0:55,batch_id:15,batch_norm:10,batch_norm_typ:9,batch_read:[19,22],batch_siz:[13,24,25,41,46,53,54,57,59,61,62],batchsiz:[9,37],bbbbb:19,bcd:9,bcebo:13,beam:[9,35,62],beam_gen:[9,35],beam_search:[15,34,35],beam_siz:[9,35,42,43,45],beamsiz:62,becaus:[9,13,16,17,22,32],been:60,befor:[9,10,17,22,24,59],begin:[8,9],beginiter:[15,16],beginn:35,beginpass:[15,16],begintrain:16,being:[17,22],belong:9,below:[9,13,17,22],benefit:10,bengio:9,bertolami:61,besid:[9,13],best:[9,59],best_model_path:60,besteffort:48,beta1:11,beta2:11,beta:55,better:[9,10],between:[9,11,17,21,62],bgr:55,bi_gru:10,bi_lstm:10,bia:[9,10,11,37,46,55],bias:9,bias_attr:[9,10,24,25,32,35],bias_param:24,bias_param_attr:10,biases_:37,biasparameter_:37,biassiz:37,bidi:48,bidirect:[10,60,61],bidirectional_lstm_net:61,bigger:17,bilinear:9,bilinear_interpol:9,bilinearfwdbwd:40,bin:[27,29,41,47,48,49,59],binari:[8,9,13,57],bird:54,bitext:62,blank:9,bleu:62,block:[9,18],block_i:9,block_x:9,bn_attr:10,bn_bias_attr:10,bn_param_attr:10,bollen:61,book:13,bool:[7,8,9,10,12,13,37,43,45,57,59,61],boot:[9,34,35],boot_bia:9,boot_bias_active_typ:9,boot_lay:[9,32,35],boot_with_const_id:9,bos_id:[9,35],both:[6,7,9,10,16,17],bottom:57,bow:[19,57],branch:[9,16,23,36],brendan:61,broadcast:17,broken:36,browser:27,bryan:61,buf_siz:13,buffer:[13,22],buffered_read:22,bug:36,build:[13,27,36,49,50,51,62],build_dict:13,build_doc:38,built:40,bunk:61,c11:20,c703c041:36,c99:21,cach:[24,57,59,60],cache_pass_in_mem:[3,24,57,59,60],cachetyp:[3,24,57,59,60],calc_batch_s:[3,60],calcul:[8,9,10,11,17],call:[9,10,16,40,49,57],callabl:[9,13],callback:37,calrnn:32,caltech:54,can:[7,8,9,10,13,16,17,22,40,57],can_over_batch_s:[3,60],candid:9,capi:20,capi_prvi:21,caption:62,card:41,care:[10,22],cat:[27,49,54,55,61],categori:[9,13,17,57],categorig:13,categoryfil:48,caus:17,ccb2_pc30:62,cde:9,cdn:13,ceil:9,ceil_mod:9,cell:[9,10],ceph:[19,47],certif:[16,47],cffi:20,cfg:48,cgo:20,chain:13,challeng:17,chanc:16,chang:[9,13,22,36,61],channel:[9,40],char_bas:59,check:[3,13,24,29,36,37,43],check_align:13,check_eq:37,check_fail_continu:3,check_l:37,check_sparse_distribution_batch:[42,43],check_sparse_distribution_in_pserv:[42,43],check_sparse_distribution_ratio:[42,43],check_sparse_distribution_unbalance_degre:[42,43],checkgrad:43,checkgrad_ep:43,checkout:36,chunk_evalu:8,chunk_schem:8,chunktyp:8,cifar:54,cifar_vgg_model:54,claimnam:49,clang:[20,27,36],class1:61,class2:61,class_dim:61,classic:9,classif:[9,57,61,62],classifi:55,classification_cost:[24,32,46,54,57],classification_error_evalu:[8,57,61,62],classification_error_printer_evalu:8,clean:[24,36],client:47,clip:[7,43,57],clock:9,clone:27,close:[3,22,36],cloud:17,cluster:[16,17,41,47,49],cluster_train:41,cmake:[21,24,26,36,38,40],cmakelist:37,cmatrix:[20,21],cna:9,cnn:[9,48,57],code:[3,5,13,16,22,27,37,48,59],coded_stream:24,codedinputstream:24,coeff:9,coeffici:9,collabor:17,collect:[9,13,15],collectbia:37,color:[54,55],colour:13,column:[8,9,22],column_sum_evalu:8,com:[9,10,13,27,29,36,47,48,55],combin:[9,10,13,15,59],command:[37,45,48,49,50,51],commandlin:[40,49],comment:[10,32,49,57],commit:48,common_util:[41,59],commun:17,compil:29,complet:[9,10,13,15,17,48,49],complex:[10,22],complic:9,compos:[13,16],composenotalign:13,comput:[9,10,16,17,60,61],conat:9,concat:62,concat_lay:35,concaten:10,concept:16,concern:16,concurr:17,condit:[9,48],conf:[5,9,24,32,41,49,53,55,62],conf_paddle_gradient_num:49,conf_paddle_n:49,conf_paddle_port:49,conf_paddle_ports_num:49,conf_paddle_ports_num_spars:49,config:[7,9,10,25,37,41,42,43,45,46,47,48,49,54,57,59,60,61,62],config_:43,config_arg:[42,43,45,55,57,60,61],config_bas:[8,9,10,15],config_fil:60,config_gener:[41,59],config_lay:37,config_pars:[5,37],configprotostr:24,configur:[9,37,53,55,62],conflict:36,confront:62,conll05st:60,conll:[13,60],connect:[10,48,57],connectionist:[9,61],connor:61,consequ:[9,10],consid:[8,9,11],consider:10,consist:[9,13,22],construct:[16,59],construct_featur:59,consum:17,contact:17,contain:[3,8,9,10,12,13,15,16,48,49,57],context:[3,9,10,13,35,47],context_attr:10,context_len:[9,10,57,59],context_proj_nam:10,context_proj_param_attr:10,context_project:[10,59],context_start:[9,10,57],continu:17,contrast:9,control:[7,48,62],conv:10,conv_act:10,conv_attr:10,conv_batchnorm_drop_r:10,conv_bias_attr:10,conv_filter_s:10,conv_num_filt:10,conv_op:9,conv_pad:10,conv_param_attr:10,conv_strid:10,conv_with_batchnorm:10,conveni:16,convert:[3,5,13,19,22,57,59],convlay:9,convolut:[9,10,46],convoper:9,convtran:9,convtranslay:9,cool:36,coordin:17,copi:[15,16,59],core:[7,21],corpora:62,corpu:[13,60],correct:[8,9],correctli:[8,13],correspoind:16,correspond:16,corss_entropi:16,cos:9,cos_sim:59,cosin:9,cost:[11,15,16,25,59,61,62],cost_id:9,could:[9,13,15,16,22],couldn:29,count:[17,22,40,43,45,48,59,60,61,62],counter:17,cpickl:59,cpp:[20,21,24,32,37,40,49,57,59,62],cpu:[3,7,9,23,27,29,40,45,48,49,60],cpuinfo:27,cpusparsematrix:21,crash:[17,40],creat:[7,9,13,15,16,17,36,37,48,49],create_bias_paramet:37,create_input_paramet:37,createfromconfigproto:5,creator:[13,19],crf:60,critic:61,crlf:36,crop:55,crop_siz:55,cross:[9,24,57],cross_entropi:[9,16],cross_entropy_with_selfnorm:9,crt:47,csc:37,cslm:62,csr:37,ctc:8,ctc_error_evalu:8,ctest:[27,36],ctrl:[41,59],ctx:60,ctx_0:60,ctx_0_slot:60,ctx_n1:60,ctx_n1_slot:60,ctx_n2:60,ctx_n2_slot:60,ctx_p1:60,ctx_p1_slot:60,ctx_p2:60,ctx_p2_slot:60,cub:54,cuda:[29,40,41,43],cuda_dir:[42,43],cuda_so:[24,27],cuda_visible_devic:24,cudaconfigurecal:40,cudadevicegetattribut:40,cudaeventcr:40,cudaeventcreatewithflag:40,cudafre:40,cudagetdevic:40,cudagetdevicecount:40,cudagetdeviceproperti:40,cudagetlasterror:40,cudahostalloc:40,cudalaunch:40,cudamalloc:40,cudamemcpi:40,cudaprofilerstart:40,cudaprofilerstop:40,cudaprofilestop:40,cudaruntimegetvers:40,cudasetdevic:40,cudasetupargu:40,cudastreamcr:40,cudastreamcreatewithflag:40,cudastreamsynchron:40,cudeviceget:40,cudevicegetattribut:40,cudevicegetcount:40,cudevicegetnam:40,cudevicetotalmem:40,cudnn:[9,12],cudnn_batch_norm:9,cudnn_conv:9,cudnn_conv_workspace_limit_in_mb:[42,43],cudnn_convt:9,cudnn_dir:[42,43],cudnnv5:26,cudrivergetvers:40,cuinit:40,cumul:9,curl:47,current:[3,9,17,47,57],current_word:35,currentcost:[57,59,61,62],currentev:[57,59,61,62],curv:16,custom:16,custom_batch_read:22,cutoff:13,cycl:17,cyclic:9,cython:20,dalla:3,dan:60,darwin:47,dat:[13,41,59],data:[3,8,10,11,15,16,19,24,29,32,41,42,46,48,49,50,53,54,55,57,59,60,61,62],data_config:5,data_dir:[53,54,61,62],data_feed:13,data_fil:25,data_initialz:57,data_lay:[3,24,25,32,35,46,54,57,59,60],data_nam:13,data_read:[13,22],data_reader_creator_random_imag:22,data_typ:[9,13],databas:13,datadim:9,datalay:9,dataprovid:[2,24,25,35,41,46,49,57,59,60],dataprovider_:57,dataprovider_bow:57,dataprovider_emb:57,dataproviderconvert:5,datasci:9,dataset:[19,22,55,57,58,61,62],date:60,db_lstm:60,dcudnn_root:26,dead:17,deb:[29,36],decai:11,decid:[16,22],declar:9,decod:[9,10,34,35,62],decoder_boot:35,decoder_group_nam:35,decoder_input:35,decoder_mem:35,decoder_prev:10,decoder_s:35,decoder_st:[10,35],deconv:9,deconvolut:9,decor:[3,13],deep:[9,40,54,55],deer:54,def:[3,5,9,13,16,19,22,24,25,32,35,37,49,55,57,59,60],defalut:9,default_decor:49,default_devic:45,default_valu:45,defin:[3,9,10,13,16,17,22,24,57,59],define_py_data_sources2:[3,24,25,46,54,55,57,59],defini:62,definit:[13,17,53],degre:9,del:59,delar:57,delet:[17,36],delimit:[8,59],demand:17,demo:[5,9,13,35,41,48,50,53,54,55,57,59,61,62],dens:[9,13,59],dense_vector:[3,5,9,13,25,59],dense_vector_sequ:13,depend:17,deriv:[6,16],descent:[9,17],describ:[16,48,57],descript:49,deseri:15,design:[9,13,20],desir:[17,48],detail:[7,9,10,11],detect:[8,36],determin:[9,13],dev:[24,27,36,54,59,62],develop:[23,36,62],deviat:7,devic:[7,24,27,45],deviceid:45,devid:9,dez:61,dict:[3,13,15,24,32,49,57,59,61,62],dict_dim:[24,32,61],dict_fil:[8,32,35,57,60],dict_siz:13,dictionai:57,dictionari:[3,8,9,13,15,16,24,57,62],dictrionari:57,dictsiz:62,diff:36,differ:[8,9,17],digit:9,dim:[13,37,53,61],dimens:[6,9,12,13,24,57],dimes:9,din:59,dir:[41,49,55,59,60,61,62],direct:[9,10],directli:10,directori:[19,36,40,48],disabl:24,discard:[13,17,36,43],discount:9,discov:17,discuss:16,disput:62,dist_train:16,distanc:8,distribut:[9,43,50,51],distribute_test:[42,43],disucss:16,divid:11,diy_beam_search_prob_so:[42,43],dmkl_root:26,do_forward_backward:22,doc:[5,10,13,27,38,49],doc_cn:38,docker:[23,24,27,36,48,49,50,51],docker_build:16,docker_push:16,dockerfil:[27,36,49],dockerhub:27,document:10,documentari:3,doe:[10,17,22],doesn:[7,9,13,16,22],dog:[54,55],don:[10,16,22],done:[9,10,17,40,49],dot:62,dot_period:[43,45,49,54,59,61,62],dotmuloper:9,dotmulproject:9,doubl:43,down:40,download:[13,17,48],download_cifar:54,doxygen:36,dpkg:29,dpython_execut:24,dpython_include_dir:24,dpython_librari:24,drop_rat:[7,46],dropout:[7,9],dropout_r:10,drwxr:48,dso_handl:29,dtoh:40,dtype:[5,25,55],dubai:62,due:59,dure:[3,9,17,57,62],dwith_c_api:21,dwith_gpu:26,dwith_profil:40,dwith_python:21,dwith_swig_pi:21,dwith_tim:40,dynam:[3,21,22],dynamic_cast:37,each:[3,8,9,12,13,15,17,22,57,59],each_feature_vector:6,each_meta:59,each_pixel_str:3,each_sequ:[9,31,32],each_time_step_output:6,each_timestep:[9,31],each_word:3,eaqual:9,eas:[13,22],easi:22,easier:[16,22],easili:[16,22],ec2:47,echo:[24,27,59,61],edit:8,edu:[13,48,54],efg:9,either:[9,13,15,16,57],electron:48,elem_dim:9,element:[8,9,10,13,15,22],elif:[16,59],els:[9,16,27,32,37,55,57,59],emb1:32,emb2:32,emb:[24,32,48,57],emb_group:32,emb_sum:24,embed:[16,53,59,61],embedding_lay:[24,32,35,57,59],embedding_nam:[9,35],embedding_s:[9,35],empir:9,emplace_back:37,empti:[8,13,17,25],enabl:[7,40],enable_grad_shar:[42,43],enable_parallel_vector:43,enc_proj:[10,35],enc_seq:10,enc_vec:35,encod:[10,32,62],encoded_proj:[10,35],encoded_sequ:[10,35],encoded_vector:35,encoder1:32,encoder1_expand:32,encoder1_rep:32,encoder2:32,encoder2_rep:32,encoder_last:9,encoder_proj:35,encoder_s:35,end:[8,9,22,35,36,60,61],end_pass:16,enditer:[15,16],endpass:[15,16],endtrain:16,english:[9,62],ensembl:10,ensur:17,entir:[9,10],entri:13,entropi:[9,57],enumer:[6,24,57,59],enumerate_data_types_of_data_lay:13,env:[24,49],environ:[16,24,40,48],eol:36,eos_id:[9,35],epsilon:11,equal:[9,10,17,32],equat:[9,10,11],equival:[9,16],error:[7,8,9,11,16,24,43,57,59,61,62],error_clipping_threshold:7,especi:10,essenc:16,essenti:[9,16],estim:[9,16],eta:48,etc:[13,22,62],etcd:17,eth0:[41,46,49],eval:[8,57,59,61,62],eval_bleu:62,evalu:[9,14,15,40,41,59,61,62],evaluate_pass:61,even:[16,22],event:48,event_handl:[15,16],everi:[8,9,10,13,16,17],exactli:[8,9,10],exampl:[9,10,11,13,15,22,55,57],exc_path:24,exceed:9,except:[13,59],excluded_chunk_typ:8,exconv:9,exconvt:9,exdb:13,exe:47,exec:27,execut:17,exist:[16,17,22,61],exit:48,expand:32,expand_a:[9,31,32],expand_lay:32,expand_level:[9,31],expandconvlay:9,expandlevel:[9,31],expect:9,expir:17,explain:[8,17],explan:9,explicit:37,explicitli:16,explor:9,exponenti:6,express:16,extend:59,extent:21,extern:[20,21],extra:[7,9,10],extra_input:9,extra_lay:15,extraattr:[7,45,46],extraattribut:[9,10],extraattributenon:9,extract:[9,53,55,60],extract_fea_c:55,extract_fea_pi:55,extralayerattribut:[7,9],extrapaddl:10,extrem:9,f1205:24,f120da72:48,fa0wx:48,fabric:41,facotr:9,factor:[7,9,11],factori:20,fail:[17,24,29,43,48],failur:17,fake_imag:22,fals:[7,8,9,10,11,13,22,24,25,32,35,37,45,48,53,57,59,60,61,62],false_label:22,false_read:22,faq:52,fast:[9,40],faster:[9,10,17],fbd1f2bb71f4:48,fc1:[37,45],fc2:45,fc3:45,fc4:45,fc_act:10,fc_attr:10,fc_bias_attr:10,fc_layer:[9,24,25,32,45,46,57,59],fc_mat:15,fc_name:10,fc_param:24,fc_param_attr:10,fclayer:37,fdata:[32,60],fea:55,fea_output:55,feat:61,featur:[3,6,9,13,36,55,57,59,60],feature_a:24,feature_b:24,feature_map:59,feed:[10,13,15,16],feeder:13,fernan:61,festiv:3,fetch:[13,17,36],few:[17,22],fewer:9,fg0:9,field:[9,15,59],figur:[16,53,62],file1:62,file2:62,file:[3,8,9,13,15,16,17,19,21,22,36,55,57,58,59,61,62],file_list:3,file_nam:[25,32,55,57,60],filenam:[3,19,24,59],filer:9,fill:[9,17,57],filter:9,filter_s:[9,10,46],filter_size_i:9,find:[9,11,17,29],fine:7,finish:[17,48],first:[9,13,16,17,57,59],first_seq:35,firstn:13,firstseen:48,fit:13,fix:[7,20,36],flag:13,flexiabl:22,flexibl:[9,10,16],flight:62,float32:[5,13,22,25,55],floor:9,flow:23,fly:57,folder:[19,62],follow:[8,9,10,11,13,16,17,22,50,51,59],forbid:16,force_load:20,forget:[11,16],form:[10,11],format:[8,36,37],former:16,formula:[9,10],formular:9,forward:[6,10,37],forwardactiv:37,forwardtest:5,found:9,frame:8,framework:[16,57],free:13,french:62,frequenc:13,frequent:22,frog:54,from:[3,5,9,10,13,15,17,19,22,24,25,27,34,40,46,48,53,54,57,59,60,61,62],from_sequ:31,from_timestep:[9,31],fromfil:[22,25,55],fulfil:40,full:[9,17],full_matrix_project:[10,32,35,46],fulli:57,fullmatrixproject:9,fully_matrix_project:10,fullyconnectedlay:37,func:[3,13],further:9,fusion:59,gain:9,gamma:55,gan:16,gate:[9,10],gate_act:[9,10,32],gate_recurr:9,gather:[9,59],gauss:7,gcc:[20,27],gce:47,gcepersistentdisk:47,gdebi:29,gen:[9,62],gen_conf:62,gen_data:62,gen_result:62,gen_trans_fil:35,gender:[13,49,59],gener:[3,8,9,10,13,15,16,17,19,22,40,45,49,53,57,62],generated_word_embed:9,generatedinput:[9,34,35],genr:[49,59],gereat:8,get:[3,9,10,13,15,29,37,48,54,57,59,60,61],get_batch_s:60,get_best_pass:61,get_config_arg:[45,57,59,61],get_data:[48,57,60],get_dict:13,get_embed:13,get_imdb:61,get_input_lay:37,get_model:55,get_movie_title_dict:13,get_output_attr:10,get_sample_from_lin:24,get_shap:15,get_word_dict:13,getbatchs:37,getenv:[16,49],gethostbynam:49,gethostnam:49,getidmap:49,getinput:37,getinputgrad:37,getinputvalu:37,getoutputgrad:37,getoutputvalu:37,getparameterptr:37,getpodlist:49,getsiz:37,gettranspos:37,getw:37,getweight:37,getwgrad:37,gildea:60,gist:10,git:[23,27,36],github:[9,10,27,29,36,55],give:[3,17],given:[13,15,22,57],global:[7,16,17,40,59],global_learning_r:7,globalstat:40,globalstatinfo:40,globe:3,glusterf:47,godoc:20,goe:[9,10,17],good:[9,22],goodfellow13:9,googl:[16,24],googleapi:47,gpu:[7,9,12,23,24,27,29,40,45,55,60,61,62],gpu_id:[24,43,45],gpugpu_id:42,grab:17,grad:43,grad_share_block_num:[42,43],gradient:[7,8,9,11,15,17,43,49,57],gradient_clipping_threshold:[7,57,61],gradient_machin:[15,21],gradient_printer_evalu:8,gradient_serv:46,gradientmachin:[5,15,21,49,59,62],gradientserv:46,gram:53,graph:[9,15,17],grave:61,greater:9,grep:[27,61],groudtruth:35,ground:[8,9,62],group:10,group_id:59,group_input:[32,35],grouplen:[13,58],gru:[9,57],gru_attr:10,gru_bias_attr:10,gru_decod:35,gru_decoder_with_attent:35,gru_encoder_decod:[53,62],gru_memori:10,gru_siz:57,gru_step:[10,35],gru_step_lay:35,grumemori:[10,35],gserver:[9,37],gsizex:40,guid:48,gur_group:10,gzip:48,hadoop:16,handl:[16,22],handler:15,handwrit:61,harvest:57,has:[9,10,11,13,16,17,40,57,60],has_kei:15,hassubseq:32,have:[8,9,10,13,16,17,22],hdf:19,head:61,header:[21,25,55,59],height:[9,13,20,22,37],held:17,hello:16,help:[5,36],helper:[9,10],here:[7,9,10,13,16,17,22],heurist:[9,62],hidden1:46,hidden2:46,hidden:[9,10,24,59],hidden_a:24,hidden_b:24,hidden_dim:32,hidden_s:[10,59],hierach:34,hierarch:[9,32],high:7,highest:13,highli:13,him:16,hint:5,hl_dso_load:29,hl_get_sync_flag:37,hold:[16,17],home:[19,41,48,49],hook2:32,hook:[32,59,60],horizont:9,hors:54,horst:61,host:[41,48],hostnetwork:49,hostpath:[47,48,49],hot:57,hous:[3,13],how:[7,9,16,17],howardjohnson:32,howev:[10,22],howto:49,hpp:20,html:[13,54],htod:40,http:[9,10,13,27,29,36,47,48,54,55,58,62],huber:9,huge:9,huina:61,hyper:9,hyperplan:13,i0601:59,i0706:62,i0719:62,i1116:49,i1117:40,ib0:41,ics:13,icwsm:61,id_input:[8,35],idea:[9,22],identityoffsetproject:9,identityproject:9,idmap:49,ids:[8,9,24,57,59],idx:37,ieee:61,ignor:[3,8,9],ijcnlp:61,illustr:17,ilsvrc:55,imag:[12,13,16,22,36,48,49,50,51,54,55,62],image_a:22,image_b:22,image_classif:54,image_fil:22,image_lay:22,image_list_provid:55,image_nam:16,image_path:22,image_provid:54,image_reader_cr:22,image_s:55,imagenet:19,imagepullpolici:49,imageri:9,images_reader_cr:22,imdber:61,img:[3,9,46,54],img_conv:10,img_pool:10,img_siz:54,imgsiz:40,imgsizei:40,imgsizex:40,immutable_paramet:16,implement:[9,10,11,13,20,21],importerror:59,inarg:5,inc_path:24,includ:[9,10,13,16,20,21,36,40],incorrect:9,increas:[17,24],incupd:37,inde:[13,22],independ:9,index:[8,9,12,13,15,17,32,59],indexslot:9,indic:[8,9],individu:17,industri:17,infer:[16,17,20],infiniband:41,info:[8,9,13,32,37,41,49],inform:13,ininst:16,init:[7,37,49,59,60],init_hook:[32,57,59,60],init_model_path:[42,43,45,53,57,60],initi:[3,7,9,13,43,57],initial_max:[7,24],initial_mean:[7,9,24],initial_min:[7,24],initial_std:[7,9,24],initpaddl:5,inlcud:10,inner:[24,32],inner_:32,inner_mem:32,inner_param_attr:10,inner_rnn_output:32,inner_rnn_st:32,inner_rnn_state_:32,inner_step:32,inner_step_impl:32,input1:[9,10],input2:9,input:[3,6,8,9,10,12,13,15,22,24,25,31,32,34,35,37,45,46,49,53,54,57,59,60,62],input_data:37,input_data_target:37,input_featur:6,input_fil:[25,60],input_hassub_sequence_data:37,input_id:9,input_imag:[10,54],input_index:37,input_label:37,input_lay:37,input_nam:16,input_sequence_data:37,input_sequence_label:37,input_sparse_float_value_data:37,input_sparse_non_value_data:37,input_t:37,input_typ:[24,25,32,35,57,59],inputdef:37,inputlayers_:37,inputtyp:13,insert:36,insid:[8,9,17,22],instal:[24,29,36,41,48,54,59],instanc:[9,17,19],instanti:17,instead:[9,10,12,22],int32:[43,46],integ:[3,8,9,13,20,57],integer_sequ:24,integer_valu:[3,13,24,32,57],integer_value_sequ:[3,13,32,35,57,60],integer_value_sub_sequ:32,integr:60,inter:9,intercept:9,interfac:[7,9,10,41],intergr:9,intern:[9,10,13,15],internet:17,interpret:8,introduc:17,invalid:22,invok:[3,9,15,40,59],iob:8,ioe:8,ip_str:49,ipc:47,ips:49,ipt:[9,24,32,35],ipython:16,is_gener:[9,53,62],is_kei:59,is_loc:15,is_predict:[57,59,61],is_seq:[9,35,59],is_sequ:59,is_stat:7,is_test:[55,60,61],is_train:3,isinst:5,ispodallrun:49,isspars:37,item:[9,13,15,22,49],iter:[3,9,10,11,13,15,16,17,22],its:[3,8,9,10,16,17,29,40],itself:[10,17],java:20,jeremi:40,jie:[60,61],jmlr:9,job:[13,41,42,43,45,47,49,55,57,60,61,62],job_dispatch_packag:41,job_id:13,job_mod:53,job_nam:49,job_namespac:49,job_path:49,job_path_output:49,job_workspac:41,jobnam:49,jobpath:49,jobselector:49,johan:61,join:[17,32],joint:62,jointli:[10,62],journal:[60,61],jpg:55,json:[41,48,59],jth:10,jupyt:27,just:[6,8,9,10,13],jypyt:16,k8s:49,k8s_data:49,k8s_job:16,k8s_token:16,k8s_train:49,k8s_user:16,kafka:19,kaim:9,kaimingh:55,kebilinearinterpbw:40,kebilinearinterpfw:40,keep:[9,17],kei:[3,13,15,17,36,40,47,49,59],kernel:9,key1:43,key2:43,keyword:49,kill:17,kind:[16,17,47,48,49],kingsburi:60,know:[10,16],kriz:[13,54],ksimonyan:10,kube:47,kube_cluster_tl:16,kube_ctrl_start_job:16,kube_list_containers_in_job_and_return_current_containers_rank:16,kubeadm:47,kubectl:[47,48,49],kuberent:17,kubernet:[16,17,39,41,49,50,51],kubernetes_service_host:16,kwarg:[3,8,9,10,11,13,32,57,59,60],l1_rate:7,l2_rate:7,l2regular:[46,54,57,61],label:[3,8,9,11,13,15,22,24,25,32,46,48,54,55,57,59,60,61],label_dict:60,label_dim:[9,32,57],label_fil:[22,60],label_lay:22,label_list:60,label_path:22,label_slot:60,labeledbow:61,labelselector:49,lag:43,lake:3,lambdacost:9,lambdarank:9,languag:[9,13,53],larg:[12,13,36,62],larger:[7,8,9],last:[8,9,10,31,32],last_seq:32,last_time_step_output:9,lastseen:48,later:57,latest:[9,17,24,48,49],launcher:16,layer1:[9,10,31],layer2:[9,31],layer3:9,layer:[5,7,8,10,12,13,14,15,22,31,34,35,37,46,55,57,59,60],layer_0:37,layer_attr:[9,35,45,46],layer_num:[45,55],layer_s:9,layer_typ:9,layerbas:37,layerconfig:37,layergradutil:37,layermap:37,layeroutout:9,layeroutput:[9,46,59],layers_test:24,lbl:[8,54],ld_library_path:[29,41],learn:[7,8,9,10,11,13,16,22,27,40,54,55,60,61,62],learnabl:9,learning_method:[25,46,53,54,57,59,61,62],learning_r:[7,24,25,46,53,54,57,59,61,62],leas:17,least:[8,9,17],lecun:13,left:9,leman:62,len:[3,9,32,35,37,49,57,59,60],length:[9,10,13,48],less:[9,16],less_than:16,let02:48,let:[9,16],level:[7,9,34],lib64:[24,27,29,41,43],lib:[21,26,29],lib_path:24,libari:21,libcuda:[24,27],libjpeg:54,libnvidia:[24,27],libpaddl:[20,21,36],libpaddle_capi:21,libpaddle_gserv:21,libpaddle_math:21,libprotobuf:24,librari:[9,21,29,41,43],life:17,like:[8,9,13,17,22,55],limit:[9,13,24,40],line:[3,8,13,24,32,36,45,57,59,60],line_count:24,linear:9,linearactiv:[9,25],linguist:60,link:[9,10,34,61],linux:[27,47],lipeng:53,lipton:61,list:[2,3,8,9,13,15,16,25,41,45,46,54,55,57,59,60,61,62],lium:62,live:17,liwicki:61,load:[9,16,17,25,49,55,59,60,61,62],load_data_arg:5,load_featur:55,load_feature_c:55,load_feature_pi:55,load_missing_parameter_strategi:[42,43,45,53,60],loadparamet:5,loadsave_parameters_in_pserv:[42,43],local:[7,17,26,29,38,41,42,43,49],localhost:[27,47],localip:49,lock:17,log:[24,37,41,43,48,49,54,59,60,61,62],log_barrier_abstract:[42,43],log_barrier_lowest_nod:[42,43],log_barrier_show_log:[42,43],log_clip:[42,43],log_error_clip:[42,43],log_period:[43,45,48,49,54,57,59,60,61,62],log_period_serv:[42,43],logarithm:6,logger:[3,32],logist:57,look:[3,57],loop:22,loss:[9,57],low:9,lpaddle_capi_shar:21,lpaddle_capi_whol:21,lst:59,lstm:[9,32,35,48,57],lstm_attr:10,lstm_bias_attr:10,lstm_cell_attr:10,lstm_group:[10,32],lstm_group_input:32,lstm_input:32,lstm_last:32,lstm_nest_group:32,lstm_output:32,lstm_size:57,lstm_step:10,lstmemori:[10,32,35],lstmemory_group:[9,32],ltr:9,mac:[21,27],machan:10,machin:[9,10,13,15,34,61,62],made:17,mai:[9,22],main:5,maintain:9,major:62,make:[3,16,17,22,29,36,37,40,61],manag:17,mandarin:9,mani:[9,10],manufactur:62,mao:61,map:[9,13,15,16,59],map_read:13,mapreduc:16,marcu:61,mark:60,mark_slot:60,market:61,martha:60,mask:[7,9],master:[16,23,47,61],mat:[20,21],mat_param_attr:10,math:[10,20,37,40],matirx:9,matplotlib:54,matrix:[8,9,10,13,15,20,21,37],matrixptr:37,matrixtyp:21,max:[7,9,13,31,40,45,59],max_id:[9,15],max_job_id:13,max_length:[9,35],max_movie_id:13,max_sort_s:9,max_user_id:13,maxframe_printer_evalu:8,maxid:8,maxid_printer_evalu:8,maxim:9,maximum:[8,13],maxinum:12,maxpool:9,mayb:[9,10],md5:[13,18],mean:[7,8,9,10,11,12,13,15,22,24,43,55,57,59],mean_img_s:54,mean_meta:55,mean_meta_224:55,mean_valu:55,mechan:[9,10],meet:60,mem:[9,32],member:16,memcpi:40,memori:[10,35,40,48,57],memory_nam:9,memory_threshold_on_load_data:[42,43],mere:10,merg:36,mergedict:[53,62],messag:[36,48],meta:[41,54,55,59],meta_config:[41,59],meta_fil:59,meta_gener:[41,59],meta_path:54,meta_to_head:59,metadata:[48,49],metal:47,metaplotlib:16,method:[3,9,11,15,59,62],mfs:49,might:9,million:13,min:[7,40,45,59],min_pool_s:[3,24,46],mini:[9,13,15,17],mini_batch:22,minibatch:9,minibatch_data:13,minikub:47,minim:11,minimum:9,minut:17,miss:60,mix:[10,46],mixed_attr:10,mixed_bias_attr:10,mixed_lay:[9,32,35,46,60],mixedlayertyp:9,mkl:26,mkl_root:26,ml_data:[41,59],mnist:[3,5,19,22],mnist_model:5,mnist_provid:3,mnist_random_image_batch_read:22,mnist_train:[3,22],mnist_train_batch_read:22,mnt:49,mod:60,mode:[9,36,49,61],model:[9,10,13,17,18,45,46,53,54,57,59,60,61,62],model_config:5,model_list:[43,45,60,61],model_output:61,model_path:45,model_zoo:[53,55],modifi:36,modul:[3,10,13,15,24,25,46,54,55,57,59],modulo:9,momentum:[7,24,57],momentumoptim:[25,54],mon:48,mono:9,month:62,mood:61,moosef:47,more:[8,9,10,13,16,17,22,24,40],morin:9,mose:[61,62],moses_bleu:62,mosesdecod:61,most:[9,13,16,22],mountpath:[48,49],move:[9,17],movi:[3,13,59],movie_categori:13,movie_featur:59,movie_head:59,movie_id:[49,59],movie_info:13,movie_meta:59,movie_nam:59,movie_review:13,movieinfo:13,movielen:58,moving_average_fract:9,mpi:41,mse_cost:[25,59],much:[9,17,22],mul:37,multi:[9,55,62],multi_binary_label_cross_entropi:9,multi_crop:55,multinomi:9,multipl:[8,9,10,13,16],multipli:[8,9],must:[6,8,9,10,22,29,37],mypaddl:[48,49],name:[3,7,8,9,10,12,13,15,16,17,19,21,24,25,27,32,35,37,40,45,46,47,48,49,50,51,54,57,59,62],name_prefix:19,namespac:[20,37,47,48,49],nativ:9,nchw:9,ndarrai:[15,19],ndarri:15,ndcg:9,ndcg_num:9,necessari:[9,57],need:[9,10,13,16,24,40,49,57],neg:[3,8,9,57,60,61],neg_distribut:9,nest:13,net:[9,10,61],net_conf:61,net_diagram:55,network:[5,7,8,9,11,13,14,15,16,17,22,32,41,45,49,53,54,55,59,60,61,62],network_config:45,neural:[9,10,11,13,15,16,17,32,34,53,59,60,61,62],neuralnetwork:9,never:[13,22,48,49],next:[3,9,13,17],nfs:49,nfsdir:49,nginx:27,nic:[41,42,43,46,49],nine:13,nlp:9,nltk:13,nmt:62,nnz:37,no_cach:3,no_sequ:[3,59],noah:61,noavx:27,node0:49,node:[9,47,48,49],node_0:49,node_1:49,node_2:49,nodebook:27,nodefil:41,nois:[9,17],non:[9,17],none:[3,5,7,8,9,10,11,12,13,15,16,25,35,55,57],norm_by_tim:9,normal:[9,10,13,48,49,55],notat:9,note:[7,9,10,12,15,16,22,29,61],noth:[6,15,36],novel:61,now:[9,17,34],np_arrai:13,ntst1213:62,ntst14:62,nullptr:[29,37],num:[9,41,43,49,60,61,62],num_channel:[9,10,46,54],num_chunk_typ:8,num_class:[9,10,54],num_filt:[9,10,46],num_gradient_serv:[42,43,46],num_group:9,num_neg_sampl:9,num_parameter_serv:16,num_pass:[15,25,42,43,45,48,49,57,59,60,61,62],num_repeat:9,num_result:8,num_results_per_sampl:9,num_shard:19,number:[8,9,13,17,19,22,62],numchunktyp:8,numdevices_:45,numlogicaldevices_:45,numofallsampl:8,numofwrongpredict:8,numpi:[13,15,19,22,25,55],numsampl:40,numtagtyp:8,nvcc:27,nvidia:[24,27],obj:[3,24,25,46,54,55,57,59],object:[3,7,9,10,11,13,15,16,20,40,57,59],observ:11,occup:[49,59],occur:[13,15],oct:48,odd:9,off:[21,26,27,29,36],offlin:[17,19],offset:[9,59],ograd:37,omit:[24,57],on_init:3,onc:[9,17],one:[3,6,8,9,10,11,12,13,16,17,22,57,60,61],one_host_dens:59,one_hot_dens:59,onli:[8,9,10,12,13,15,16,32,34],onlin:[11,17,19,22],open:[3,9,16,22,24,25,32,55,57,59,60],openbla:26,openblas_root:26,oper:[9,10,11,46],opinion:61,opt:[16,26,49],optim:[7,14,15,24],option:[8,9,16],order:[9,10,13,22,49],org:[9,10,13,27,58],organ:9,origin:[9,13,36],other:[8,9,10,13,17,57,59],otherchunktyp:8,otherwis:[9,13,16,17,22],our:16,out:[9,15,16,32,34,35,46,54],out_dir:49,out_left:9,out_mem:35,out_right:9,out_size_i:9,out_size_x:9,outer:32,outer_mem:32,outer_rnn_st:32,outer_rnn_state_:32,outer_step:32,outptut_lay:15,output:[6,7,8,9,10,12,13,15,16,19,22,24,25,32,35,41,45,46,48,49,53,54,55,57,59,60,61],output_:9,output_dir:55,output_fil:60,output_id:9,output_lay:[15,55],output_max_index:12,output_mem:[9,35],output_path:19,outputh:9,outputw:9,outsid:[3,9,10],outv:37,over:[9,10,16],overrid:17,packag:[13,24],pad_c:9,pad_h:9,pad_w:9,padding_attr:9,padding_i:9,padding_x:9,paddl:[3,5,6,7,8,9,10,11,12,13,15,16,17,19,20,21,24,25,27,29,36,37,38,40,41,45,46,48,49,53,54,57,59,60,61,62],paddle_error:[20,21],paddle_matrix:[20,21],paddle_matrix_cr:21,paddle_matrix_get_shap:20,paddle_matrix_shap:20,paddle_n:[41,49],paddle_output:48,paddle_port:[41,49],paddle_ports_num:[41,49],paddle_ports_num_for_spars:41,paddle_ports_num_spars:49,paddle_process_by_paddl:49,paddle_pserver2:41,paddle_root:53,paddle_server_num:49,paddle_source_root:53,paddle_train:[21,23,41,49],paddledev:[24,27,48,49],paddlepaddl:[9,10,13,17,22,24,27,29,35,36,40,41,46,50,51,53,60],pair:8,palceholder_just_ignore_the_embed:53,palmer:60,paper:[9,62],para:53,paraconvert:53,parallel:[40,45,48,49,62],parallel_nn:[7,42,43],param:[7,9,59],param_attr:[9,10,24,25,35],paramattr:[7,9,24,25,35],paramet:[8,9,10,11,12,13,14,18,22,43,46,49,59,60,61,62],parameter_attribut:9,parameter_block_s:[42,43],parameter_block_size_for_spars:[42,43],parameter_learning_r:7,parameter_nam:[15,16],parameter_serv:16,parameterattribut:[7,9,10],parameterclient2:49,parametermap:37,parameters_:37,parameterset:16,parametris:11,paramutil:59,paraphras:[53,62],paraphrase_data:53,paraphrase_model:53,paraphrase_modeldata:53,paraspars:37,pars:[13,59],parse_config:5,parse_known_arg:49,parsefromstr:24,parser:49,part:[59,61],partial:9,partit:[17,19],paserv:49,pass:[3,9,13,15,17,22,24,25,36,40,43,45,48,49,54,57,59,60,61,62],pass_id:15,pass_idx:22,passtyp:37,past:16,path:[8,13,15,17,22,29,41,43,47,48,49,60,61],pattern:[13,17,20,59,61],paul:60,paus:17,pave:62,pdf:[9,10],pem:16,pend:17,penn:60,per:[9,13,17,22],perform:[9,10,40,42],period:[17,43,60,61,62],perl:[61,62],peroid:9,persistentvolum:47,persistentvolumeclaim:[47,49],person:16,pfs:19,pickl:59,pid:47,piec:[9,10],pillow:54,pip:[24,36,41,54,59],pixel:[3,9,13,46],pixels_float:3,pixels_str:3,place:[3,17],plain:[8,9,15,21],plan:17,pleas:[7,9,10,11,16,17,22,24,29,49],plot:[16,54],plotcurv:54,png:54,pnpair_evalu:8,pnpairvalid:42,pod:[47,48,49],podip:49,podlist:49,point:[17,40],polar:13,poll:61,pool3:37,pool:[10,14,46,59],pool_attr:10,pool_bias_attr:10,pool_pad:10,pool_siz:[3,9,10,46],pool_size_i:9,pool_strid:10,pool_typ:[9,10],pooling_lay:[24,57],pooling_typ:[9,24,31,57],poolingtyp:12,port:[41,42,43,46,48,49],port_num:42,ports_num:[43,46,49],ports_num_for_spars:[42,43,45,46,49],pos:[59,61],pose:17,posit:[3,8,9,13,57],positive_label:8,posix:47,possibl:16,potenti:40,practic:9,pre:[9,10,13,16,53,61,62],pre_dictandmodel:53,precis:8,precision_recall_evalu:8,pred:[57,60],predetermin:9,predic:[13,60],predicate_dict:60,predicate_dict_fil:60,predicate_slot:60,predict:[3,5,8,9,11,15,24,41,46,53,54,55,57,59,60,61],predict_fil:[42,43],predict_output_dir:[42,43,57],predict_sampl:5,predin:54,prefer:47,prefetch:37,prefix:[17,19],pregrad:37,premodel:53,prepar:50,preprocess:[13,41,53,54,57,59,61,62],present:[16,36],prev_batch_st:[42,43],prevent:[11,16,17],previou:[9,10,17],previous:9,price:13,principl:16,print:[5,7,15,16,25],printallstatu:40,printer:8,printstatu:40,prite:8,privat:[21,36],prob:[8,15],probabilist:[9,53],probabl:[8,9,15],problem:[9,11,16],proc:27,proce:[13,17,22],proceed:[9,60],process2:32,process:[3,7,9,10,16,19,24,25,32,35,46,49,57,59,60],process_predict:57,processdata:[54,55],processor:40,prod:[27,36],produc:[10,13,17,22],productgraph:48,profil:40,proflier:40,prog:49,program:[13,16,19,22,40,49],progress:17,proj:9,project:[9,10,21,46],promis:[9,10],prone:16,prop:60,propag:11,properli:57,proposit:60,protect:37,proto:12,protobuf:24,provid:[9,13,16,24,25,32,42,46,57,59,60],prune:9,ps_desir:17,pserver:[41,42,43,46,49],pserver_id:18,pserver_num_thread:[42,43],pseudo:16,psize:37,ptr:21,pull:23,purpos:[17,40],push:49,push_back:37,put:[17,57],pwd:[27,36],py_paddl:[5,13,38],pydataprovid:[24,46],pydataprovider2:[3,5,25,46,49,59],pyramid:9,pyramid_height:9,python:[15,16,20,23,24,27,36,37,38,41,53,54,55,59,60,61,62],pythonpath:[24,54],pzo:61,queri:[9,62],question:[9,16],quick:48,quick_start:[48,49,50,57],quick_start_data:48,quickstart:48,rac:9,rais:13,ramnath:61,ran:40,rand:[40,43,45,60],random:[7,9,13,19,22,25],random_imag:19,rang:[9,13,19,22,49,57],rank:[9,16,55,57],rare:3,rate:[7,8,11,13,41,49,59],ratio:43,raw:9,raw_meta:59,rdma_tcp:[42,43],reach:17,read:[3,13,15,16,17,19,22,25,55,57,59],read_from_realistic_imag:16,read_from_rng:16,read_lock:18,read_mnist_imag:16,read_next_from_fil:24,read_ranking_model_data:16,reader:[15,19],reader_cr:19,reader_creator_bool:22,reader_creator_random_imag:[13,22],reader_creator_random_image_and_label:[13,22],readi:[17,48],readm:[21,36,59,61],real:22,real_process:3,realist:16,reason:[10,16,17,48],recal:8,receiv:17,recognit:[9,55,61],recommend:[10,16,41,49,59],record:[59,60],recordio:16,recov:17,rectangular:9,recurr:[32,33,60,61],recurrent_group:[10,32,34,35],recurrentgradientmachin:21,recurrentgroup:8,reduc:11,ref:59,refer:[7,9,10,11,17,26],referenc:9,reference_cblas_root:26,refine_unknown_arg:49,regex:59,register_gpu_profil:40,register_lay:37,register_timer_info:40,registri:48,regular:[7,46,54,57,61],rel:10,relat:[17,59],relationship:13,releas:[23,29,47,60],reliabl:17,relu:9,rememb:9,remot:[7,36,41,43,45],remov:[13,36],reorgan:9,repres:[9,13,57],represent:57,request:[17,23,48,62],requir:[8,9,16,17,41,59],res5_3_branch2c_bn:55,res5_3_branch2c_conv:55,res:60,research:[9,13,54],reserveoutput:37,reset:[9,17],reshap:22,reshape_s:9,residu:55,resnet_101:55,resnet_152:55,resnet_50:55,resolv:[36,48],respons:[9,48],rest:9,restart:[17,48],restartpolici:[48,49],resu:22,result:[3,6,8,9,15,40,57,62],result_fil:[8,35],ret_val:59,return_seq:10,reuqest:23,reus:22,reveal:16,revers:[9,10,34,35],review:[13,36,48,61],reviews_electronics_5:48,rewrit:62,rgb:9,rgen:61,rho:11,right:9,rmsprop:57,rmspropoptim:59,rnn:[9,10,34,35,42,61],rnn_bias_attr:35,rnn_layer_attr:35,rnn_out:35,rnn_state:32,rnn_state_:32,rnn_step:9,rnn_use_batch:[42,43],rnnlm:13,robot:54,role:[13,16,60,61],roman:61,root:[11,12,27,41,48,49,53],root_dir:41,rot:9,routin:59,routledg:61,row:[8,9,13],row_id:9,rstrip:49,rtype:[9,59],run:[16,17,24,27,36,40,41,48,49,50,51,59],runinitfunct:[40,49],runtim:[3,24],s_fusion:59,s_id:59,same:[3,8,9,10,16,32,57],samping_id:9,sampl:[3,8,13,57,59,61,62],sample_id:8,sample_num:8,santiago:61,save:[3,9,13,17,19,48,59,60,61,62],save_dir:[25,43,45,48,49,54,57,59,60,61,62],save_only_on:[42,43],saving_period:[42,43,49],saving_period_by_batch:[42,43,45],saw:3,scalar:9,scale:[6,59],scalingproject:9,scatter:9,scheduler_factor:7,scheme:8,schmidhub:61,schwenk:62,scienc:61,score:[8,9,59],script:[13,38],seaplane_s_000978:54,search:[9,17,35,62],seat:62,second:[9,13,16,22,57,59],sed:61,see:[9,10,16,17,24,57],seed:[40,43],segment:8,sel_fc:9,select:9,selectiv:9,selector:48,self:37,selfnorm:9,semant:[13,16,23,60,61],semantic_role_label:35,semat:16,sen_len:60,send:17,sent:[16,48],sent_id:35,sentanc:24,sentenc:[3,9,13,32,35,60],sentence_last_state1:32,sentence_last_state2:32,sentiment:[3,60,61],sentiment_data:61,sentiment_net:61,sentimental_provid:3,sentimental_train:3,separ:[8,57,59],seq:[9,32,59],seq_pool:[9,31],seq_text_print:8,seq_to_seq_data:[53,62],seq_typ:[13,59],seqlastin:32,seqtext_printer_evalu:[8,35],seqtoseq:[9,24,35,53,62],seqtoseq_net:[9,35,53,62],sequel:3,sequenc:[3,6,8,9,10,12,13,24,32,34,57,59,61,62],sequence_conv_pool:57,sequence_layer_group:[9,32],sequence_nest_layer_group:[9,32],sequencegen:32,sequencestartposit:9,sequencetextprint:8,sequencetyp:3,sequenti:[9,60],seri:[10,32,61],serial:15,server:[16,41,43,46,47,49],serverless:17,set:[3,7,8,9,10,13,15,16,17,24,25,32,35,40,41,46,48,53,54,55,57,59,60,61,62],set_active_typ:37,set_default_parameter_nam:7,set_drop_r:37,set_input:9,set_siz:37,set_typ:37,settotalbyteslimit:24,setup:[37,57],sever:9,sgd:[11,15,16,17,25,41,42,61],shape:[9,15],shard:[17,18,19],share:[9,21,60],shared_bia:10,shared_bias:9,shared_ptr:[20,21],ship:54,shold:61,shorten:9,should:[8,9,13,15,16,22,34],should_be_fals:16,should_be_tru:16,should_shuffl:[3,32,60],show:[11,13,17,60,61,62],show_check_sparse_distribution_log:[42,43],show_layer_stat:[42,43],show_parameter_stats_period:[42,43,45,48,60,61,62],shown:[8,9,16],shuf:[24,59],shuffl:[13,24],side:[9,15],sigint:41,sigmoid:[9,10],sigmoidactiv:32,similar:[9,22,59],similarli:9,simpl:[6,8,9,10,13,15,49],simple_attent:35,simple_gru:57,simple_img_conv_pool:46,simple_lstm:[9,57],simple_rnn:[9,35],simpli:[9,16],simplifi:16,sinc:[9,17,22],singl:[8,10,13,17],size:[3,8,9,10,11,13,17,19,22,24,25,32,35,37,46,54,55,57,59,61,62],size_a:9,size_b:9,size_t:37,skip:[22,25,36,55],sleep:49,slide:[9,11,13,17],slope:9,slot:[59,60],slot_dim:59,slot_nam:59,slottyp:59,small:13,small_messag:[42,43],small_vgg:54,smaller:[9,17],smith:61,smooth:9,snap:48,snapshot:18,sock_recv_buf_s:[42,43],sock_send_buf_s:[42,43],socket:49,softmax:[9,10,16,24,35,37,57],softmax_param:24,softmax_param_attr:10,softmax_selfnorm_alpha:9,softmaxactiv:[24,32,35,46,57],solv:16,some:[7,9,13,15,16],some_c_api_funct:21,some_inst:21,some_python_class:20,somecppclass:20,somedata:15,somegotyp:20,someth:9,sometim:[11,22],soon:17,sort:[9,13,49,59,61],sourc:[9,13,21,22,62],source_dict_dim:35,source_language_word:35,space:8,space_seperated_tokens_from_dictionary_according_to_seq:8,space_seperated_tokens_from_dictionary_according_to_sub_seq:8,spars:[7,9,11,13,24,37,41,43,45,49,57],sparse_binary_vector:[3,13,24,57],sparse_binary_vector_sequ:13,sparse_float_vector:3,sparse_non_value_slot:13,sparse_upd:[7,24],sparse_value_slot:13,sparse_vector:[13,24],sparse_vector_sequ:13,sparseparam:37,sparseprefetchrowcpumatrix:37,spatial:9,spec:[48,49],special:9,specifi:[8,9,13,16,29,57],speech:9,speed:10,sphinx:[20,27,38],split:[3,9,32,41,57,59,60],split_count:49,squar:[9,11,12],squarerootnpool:9,srand:43,src:[49,62],src_backward:35,src_dict:[24,35],src_dict_path:24,src_embed:35,src_forward:35,src_id:35,src_root:5,src_word_id:35,srl:[13,60],ssd:9,ssh:[27,41],sshd:27,sstabl:[16,19],stabl:[23,47],stacked_lstm_net:61,stacked_num:61,stackexchang:9,stage:36,stake:62,stale:17,standard:7,stanford:[13,48],start:[9,15,17,24,43,48,49],start_paddl:49,start_pass:[42,43],start_pserv:[42,43],startpaddl:49,startup:17,stat:[40,43,60,61,62],state:[9,10,17,34,48],state_act:[9,10,32],statfulset:49,staticinput:[9,34,35],statist:9,statset:40,statu:[36,40,48,49],status:48,std:[20,21,37,43],stderr:41,stdout:41,step:[9,10,12,17,32,34,35],stepout:32,stochast:[11,17],stock:61,stop:9,storag:47,store:[8,9,13,57,59],str:[15,45,49],strategi:[12,17,43,60],street:[9,60],strict:22,stride:9,stride_i:9,stride_x:9,string:[3,8,9,37,43,46],strip:[24,32,57,59,60],struct:21,structur:[13,57],stub:9,stuff:36,stun:3,style:9,sub:[8,9,13,16],sub_sequ:3,subgradi:11,subnet:16,subobjectpath:48,subseq:[31,34],subsequenceinput:[9,32],succeed:48,success:48,successfulcr:48,sudo:[29,54],suffic:22,suggest:9,sum:[9,11,18],sum_:9,sum_evalu:8,sumpool:[9,24],support:[7,8,9,12,13,17,22,32],sure:[29,61],swap_channel:55,swig:[20,21],swig_paddl:[5,13],symbol:[9,21],symlink:36,sync:17,syncflag:37,synchron:[11,17],syntax:22,sys:55,system:[17,24,61],t2b:53,t_i:9,tab:57,tabl:9,tableproject:9,tag:[8,13,27],tagtyp:8,tainer_id:49,take:[3,8,9,10,16],tanh:[9,10,37],tanhactiv:[32,35,46],target:[9,13,15,62],target_dict_dim:35,target_dictionary_dim:9,target_language_embed:9,target_language_word:35,targetinlink:[9,32],task:[8,9,60],tbd:32,tconf:61,tcp:[43,46],tcp_rdma:46,tear:40,technic:17,tee:[48,54,59,60,61,62],tell:17,tellig:61,templat:[48,49],tempor:9,term:[9,10,17],termin:48,tesh:60,test100:13,test10:13,test:[2,3,9,13,15,16,21,22,23,36,37,40,41,43,45,46,54,55,57,59,60,61,62],test_all_data_in_one_period:[48,54,59,60,61],test_compar:24,test_comparespars:24,test_comparetwonet:24,test_comparetwoopt:24,test_config_pars:24,test_data:[5,62],test_fcgrad:37,test_gpuprofil:40,test_layergrad:37,test_list:[3,24,25,46,54,57],test_networkcompar:24,test_part_000:61,test_pass:[42,43,45,62],test_period:[42,43,45],test_predict:24,test_pydataprovid:24,test_pydataprovider2:24,test_pydataproviderwrapp:24,test_ratio:59,test_recurrent_machine_gener:24,test_recurrentgradientmachin:[24,32],test_swig_api:24,test_train:24,test_traineronepass:24,test_wait:[42,43],testa:16,testb:16,testbilinearfwdbwd:40,testconfig:37,tester:[59,62],testfcgrad:37,testfclay:37,testlayergrad:37,testq:16,testresult:15,testutil:37,text:[3,8,10,13,16,53,57,61,62],text_conv:57,text_conv_pool:59,text_fil:[13,61],tflop:40,tgz:13,than:[7,8,9,10,17,24],thei:[16,17,40],them:[10,16,17,22,40,57,59],therein:9,thi:[3,7,8,9,10,11,13,15,16,17,22,40,57,59,61],thing:3,think:16,third:[9,17],those:[17,60],thread:40,thread_local_rand_use_global_se:[42,43],threadid:45,threadloc:40,three:[8,9,17,22,55],threshold:[7,8,17,43],through:[9,17],throughput:40,thu:9,tier:48,time:[9,10,12,13,16,17,22,32,40,43,48,49,61],timelin:9,timeout:17,timer:40,timestamp:[9,18],timestep:9,titl:[13,49,59],tmp:3,to_your_paddle_clone_path:38,todo:[8,10,13,17],toend:9,togeth:[9,10,13,15],token:[8,9,16,35,61],too:13,tool:[38,49],top:[8,55],top_k:8,topolog:[13,16,17],topolopi:15,toronto:[13,54],total:[8,15,17,22,40,48,62],total_pass:22,touch:61,tourist:62,track:[17,36],tractabl:9,tradit:9,trail:13,train100:13,train10:13,train:[2,3,7,9,11,13,19,24,25,29,41,43,45,46,48,49,50,51,53,54,55,57,59,60,61,62],train_arg:49,train_args_dict:49,train_args_list:49,train_conf:[53,62],train_config_dir:49,train_data:62,train_list:[3,24,25,46,54,55,57],train_part_000:61,trainabl:9,trainer:[3,5,16,18,19,25,37,43,45,46,49,59,60,61,62],trainer_config:[2,3,5,25,41,48,49,57,59,61],trainer_config_help:[3,24,25,37,46,54,59],trainer_count:[24,42,43,45,48,49,59,60,61,62],trainer_id:[43,49],trainerconfighelp:24,trainerid:49,trainerintern:[57,59,62],tran:37,transact:[17,61],transform:9,transform_param_attr:10,translat:[9,10,53,62],transpos:9,transposedfullmatrixproject:9,travel:3,travi:36,treat:9,tree:[9,49],trg:62,trg_dict:35,trg_dict_path:35,trg_embed:35,trg_id:35,trg_ids_next:35,tricki:20,trn:57,truck:54,true_imag:22,true_label:22,true_read:22,truth:[8,9,62],tst:57,tune:[7,42],tupl:[9,13,15,22],ture:9,turn:[9,22,34],tutori:[48,49,50,51,53,62],tweet:61,twitter:61,two:[9,10,16,22,40,57],txt:[3,37,41,47,57,59,61],type:[3,8,9,10,12,13,16,17,20,21,22,25,32,37,45,48,55,57,59,60],type_nam:59,typedef:[20,21],typic:8,typo_freq:13,ubuntu:23,ubyt:22,uci:13,ufldl:9,uid:48,uint64:20,uint64_t:20,unconstrain:61,undeterminist:40,uniform:[7,9,13,19,22],uninstal:24,uniqu:[16,17],unique_ptr:37,unit:[9,10],unittest:[21,24],univ:62,unix:41,unk:[53,62],unk_idx:[57,60],unknown:9,unseg:9,unsup:61,unsupbow:61,until:[17,49],untrack:36,unus:59,updat:[7,9,17,36,41,45],update_equ:15,updatecallback:37,upgrad:24,upload:17,upstream:36,url:[13,61],urls_neg:61,urls_po:61,urls_unsup:61,usag:[8,9,10,13,15,49,59],use:[7,8,9,10,12,13,15,16,17,36,40,49,55,57,59,60,61,62],use_global_stat:9,use_gpu:[5,24,42,43,45,48,49,54,55,57,59,60,61,62],use_jpeg:54,use_old_updat:[42,43],use_seq:[25,59],use_seq_or_not:59,used:[3,8,9,10,11,12,13,15,16,17,22,40,57,59,61],useful:[9,10],usegpu:37,user:[7,9,10,13,15,16,22,47,57,59],user_featur:59,user_head:59,user_id:[49,59],user_info:13,user_meta:59,user_nam:59,userinfo:13,usernam:36,uses:17,using:[7,9,10,13,16,17,22,29,60],usr:[24,26,27,29,41,43,49],usrdict:53,usrmodel:53,usual:[9,13,15,40],utc:58,util:[40,49,54,59],uuid:18,v28:9,valid:22,valu:[3,5,7,8,9,11,12,13,15,17,37,45,49,55,57,60],value1:43,value2:43,value_printer_evalu:8,value_rang:13,vanilla:35,variabl:[9,13,16,48],varianc:9,vector:[9,10,13,16,57,59],verb:13,veri:[9,12,54],version:[9,10,23,27,29,40,42,43],versu:16,vertic:9,vgg:[10,54],vgg_16_cifar:54,via:[17,22,29,36],view:9,vision:54,visipedia:54,visual:9,vocab:61,volum:[27,47,48,49],volumemount:[48,49],vutbr:13,wai:[9,10,16,62],wait:[11,17,49],wall:60,want:[3,9,10,16,22],warn:[9,24,49],warp:9,watch:17,wbia:55,wei:[60,61],weight:[8,9,10,11,37,55],weight_act:10,weightlist:37,weights_:37,weights_t:37,wether:9,what:[7,9,10,11,36,57],when:[3,7,8,9,13,15,17,40],where:[9,10,11,16,17],whether:[8,9,10,22,61],which:[8,9,10,11,13,16,17,19,22,57,59],whole:[3,8,13,20,21],whole_cont:59,whose:[9,13,17],why:[10,21,40],width:[8,9,13,20,22,37,62],wiki:9,wikipedia:[9,13],wilder:3,window:[9,12,13,47],wise:9,with_avx:[26,27,29,36],with_doc:26,with_dock:38,with_doubl:[26,29,37],with_dso:26,with_gpu:[26,27,29,36],with_predict_sdk:29,with_profil:40,with_python:[26,29],with_rdma:[26,29],with_style_check:[26,36],with_swig_pi:26,with_test:[26,27,36],with_tim:[26,29,40],within:9,without:[8,9,17,22],wmt14:62,wmt14_data:62,wmt14_model:62,wmt_shrinked_data:13,woboq:27,won:32,wonder:3,word2vec:24,word:[3,8,9,13,24,32,34,57,60],word_dict:[32,57,60],word_dim:[32,57],word_id:[3,24],word_idx:13,word_slot:60,word_vector:57,word_vector_dim:[9,35,53],words_freq_sort:13,work:[13,16,17,22,27,32,36,48,49],workspac:41,would:[15,17,22,60],wrapper:[10,40],write:[13,16,17,22,60],write_lock:18,writelin:25,writer:16,wrong:22,wsj:60,wuyi:47,www:[9,13,54,62],xarg:[24,27,37],xgbe0:43,xgbe1:43,xiaojun:61,xrang:[22,25,37],xxbow:61,xxx:[16,55,62],xxxx:18,y_i:9,yaml:[48,49],yann:13,yapf:36,yeild:15,yield:[3,13,16,19,22,24,25,32,35,57,59,60],you:[3,7,9,10,29,55,61],your:[9,16,24],your_param_nam:24,your_repo:49,your_source_root:21,yuyang18:[10,13],zachari:61,zeng:61,zero:[7,9,11,13,17,43],zhou:[60,61],zip:[13,49,58],zoo:53},titles:["\u5173\u4e8ePaddlePaddle","API","DataProvider\u7684\u4ecb\u7ecd","PyDataProvider2\u7684\u4f7f\u7528","API\u4e2d\u6587\u624b\u518c","\u57fa\u4e8ePython\u7684\u9884\u6d4b","Activation","Parameter Attribute","Evaluators","Layers","Networks","Optimizer","Pooling","Data Reader Interface and DataSets","Model Configuration","Training and Inference","PaddlePaddle Design Doc","Design Doc: Distributed Training","\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\uff08Checkpointing\uff09","\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Python Data Reader Design Doc","Paddle\u53d1\u884c\u89c4\u8303","FAQ","\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u4efb\u52a1","PaddlePaddle\u7684\u7f16\u8bd1\u9009\u9879","PaddlePaddle\u7684Docker\u5bb9\u5668\u4f7f\u7528\u65b9\u5f0f","\u5b89\u88c5\u4e0e\u7f16\u8bd1","Ubuntu\u90e8\u7f72PaddlePaddle","\u65b0\u624b\u5165\u95e8","\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684Layer","\u5355\u53cc\u5c42RNN API\u5bf9\u6bd4\u4ecb\u7ecd","RNN\u76f8\u5173\u6a21\u578b","Recurrent Group\u6559\u7a0b","RNN\u914d\u7f6e","\u5982\u4f55\u8d21\u732e\u4ee3\u7801","\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42","\u5982\u4f55\u8d21\u732e/\u4fee\u6539\u6587\u6863","\u8fdb\u9636\u6307\u5357","GPU\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18","\u8fd0\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3","\u53c2\u6570\u6982\u8ff0","\u7ec6\u8282\u63cf\u8ff0","\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570","\u4f7f\u7528\u6848\u4f8b","\u57fa\u672c\u4f7f\u7528\u6982\u5ff5","Kubernetes \u7b80\u4ecb","Kubernetes\u5355\u673a\u8bad\u7ec3","Kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3","&lt;no title&gt;","&lt;no title&gt;","PaddlePaddle \u6587\u6863","\u4e2d\u6587\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u4f7f\u7528","\u56fe\u50cf\u5206\u7c7b\u6559\u7a0b","Model Zoo - ImageNet","\u5b8c\u6574\u6559\u7a0b","\u5feb\u901f\u5165\u95e8\u6559\u7a0b","MovieLens\u6570\u636e\u96c6","MovieLens\u6570\u636e\u96c6\u8bc4\u5206\u56de\u5f52\u6a21\u578b","\u8bed\u4e49\u89d2\u8272\u6807\u6ce8\u6559\u7a0b","\u60c5\u611f\u5206\u6790\u6559\u7a0b","\u6587\u672c\u751f\u6210\u6559\u7a0b"],titleterms:{"\u4e00\u4e9b\u7ec6\u8282\u7684\u8865\u5145":49,"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":19,"\u4e0b\u8f7d\u4e0e\u89e3\u538b\u7f29":62,"\u4e0b\u8f7d\u548c\u6570\u636e\u62bd\u53d6":53,"\u4e0b\u8f7d\u5e76\u89e3\u538b\u6570\u636e\u96c6":59,"\u4e0b\u8f7d\u6570\u636e":48,"\u4e0d\u4f7f\u7528":20,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":20,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":20,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":20,"\u4e2d\u6587\u5b57\u5178":53,"\u4e2d\u6587\u77ed\u8bed\u6539\u5199\u7684\u4f8b\u5b50":53,"\u4e2d\u6587\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u4f7f\u7528":53,"\u4e2d\u6587\u8bcd\u5411\u91cf\u7684\u9884\u8bad\u7ec3\u6a21\u578b":53,"\u4e3a\u4ec0\u4e48\u9700\u8981\u6027\u80fd\u5206\u6790":40,"\u4ec0\u4e48\u662f\u6027\u80fd\u5206\u6790":40,"\u4ec5\u4ec5\u4f7f\u7528void":20,"\u4ecb\u7ecd":[53,55],"\u4ece\u5feb\u7167\u6062\u590d":18,"\u4ee3\u7801\u8981\u6c42":36,"\u4efb\u52a1\u7b80\u4ecb":25,"\u4f18\u5316\u7b97\u6cd5":57,"\u4f18\u5316\u7b97\u6cd5\u914d\u7f6e":46,"\u4f7f\u7528":[36,48],"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":20,"\u4f7f\u7528\u6848\u4f8b":45,"\u4f7f\u7528\u6982\u8ff0":57,"\u4f7f\u7528\u6a21\u578b\u521d\u59cb\u5316\u7f51\u7edc":45,"\u4f7f\u7528\u73af\u5883\u53d8\u91cf":49,"\u4f7f\u7528\u7528\u6237\u6307\u5b9a\u7684\u8bcd\u5411\u91cf\u5b57\u5178":53,"\u4f7f\u7528\u8bf4\u660e":39,"\u4f7f\u7528\u8f6c\u6362\u5e93":19,"\u4f7f\u7528docker\u6784\u5efapaddlepaddle\u7684\u6587\u6863":38,"\u4f7f\u7528paddlepaddle\u751f\u6210\u6a21\u578b":62,"\u4f7f\u7528paddlepaddle\u8bad\u7ec3\u6a21\u578b":62,"\u4fdd\u6301\u672c\u5730\u4ed3\u5e93\u6700\u65b0":36,"\u4fee\u6539\u542f\u52a8\u811a\u672c":48,"\u4fee\u6539\u6587\u6863":38,"\u514b\u9686":36,"\u5173\u4e8epaddlepaddl":0,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":21,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":21,"\u5185\u5b58\u4e0d\u591f\u7528\u7684\u60c5\u51b5":3,"\u5185\u7f6e\u5b9a\u65f6\u5668":40,"\u5199\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":37,"\u51c6\u5907\u5de5\u4f5c\u7a7a\u95f4":41,"\u51c6\u5907\u5e8f\u5217\u6570\u636e":35,"\u51c6\u5907\u6570\u636e":[25,59],"\u51c6\u5907\u8bad\u7ec3\u6570\u636e":49,"\u51c6\u5907\u96c6\u7fa4\u4f5c\u4e1a\u914d\u7f6e":41,"\u51cf\u5c11\u6570\u636e\u8f7d\u5165\u7684\u8017\u65f6":24,"\u51cf\u5c11dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":24,"\u5206\u5272\u8bad\u7ec3":59,"\u5206\u5e03\u5f0f\u8bad\u7ec3":46,"\u5206\u652f\u89c4\u8303":23,"\u521b\u5efa\u672c\u5730\u5206\u652f":36,"\u521b\u5efajob":49,"\u521b\u5efapaddl":48,"\u5220\u9664\u672c\u5730\u5206\u652f":36,"\u5220\u9664\u8fdc\u7a0b\u5206\u652f":36,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":24,"\u5230\u8fdc\u7a0b\u4ed3\u5e93":36,"\u5236\u4f5c\u955c\u50cf":49,"\u5236\u4f5cdocker\u955c\u50cf":48,"\u524d\u63d0\u6761\u4ef6":41,"\u52a0\u901f\u6267\u884c":18,"\u52a0\u901f\u8bad\u7ec3\u901f\u5ea6":24,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":20,"\u52a8\u6001\u6269\u5bb9":18,"\u5355\u5143\u6d4b\u8bd5":43,"\u5355\u53cc\u5c42rnn":32,"\u5377\u79ef\u6a21\u578b":57,"\u5377\u79ef\u795e\u7ecf\u7f51\u7edc":54,"\u539f\u56e0":20,"\u539f\u56e0\u5217\u8868":20,"\u53c2\u6570\u4fe1\u606f":55,"\u53c2\u6570\u5185\u5b58":24,"\u53c2\u6570\u670d\u52a1\u5668\u548c\u5206\u5e03\u5f0f\u901a\u4fe1":43,"\u53c2\u6570\u6982\u8ff0":42,"\u53c2\u6570\u8bfb\u53d6":55,"\u53c2\u8003":3,"\u53c2\u8003\u6587\u6863":61,"\u53c2\u8003\u8d44\u6599":40,"\u53cc\u5411lstm":61,"\u53cc\u5c42rnn":32,"\u53cc\u5c42rnn\u4ecb\u7ecd":34,"\u53cc\u5c42rnn\u7684\u4f7f\u7528":34,"\u53ef\u80fd\u7684\u5185\u5b58\u6cc4\u9732\u95ee\u9898":3,"\u53ef\u80fd\u9047\u5230\u7684\u95ee\u9898":29,"\u53ef\u9009\u529f\u80fd":53,"\u5411\u7cfb\u7edf\u4f20\u9001\u6570\u636e":57,"\u5411\u91cf":43,"\u542f\u52a8\u4efb\u52a1":49,"\u542f\u52a8\u96c6\u7fa4\u4f5c\u4e1a":41,"\u547d\u4ee4\u884c\u53c2\u6570":57,"\u548c":31,"\u56fe\u50cf\u5206\u7c7b\u6559\u7a0b":54,"\u5728\u4e0d\u540c\u8bbe\u5907\u4e0a\u6307\u5b9a\u5c42":45,"\u5728paddlepaddle\u5e73\u53f0\u8bad\u7ec3\u6a21\u578b":53,"\u57fa\u4e8epython\u7684\u9884\u6d4b":5,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":46,"\u57fa\u672c\u539f\u7406":34,"\u57fa\u672c\u8981\u6c42":20,"\u5982\u4f55\u4e66\u5199paddlepaddle\u7684\u6587\u6863":38,"\u5982\u4f55\u5171\u4eab\u53c2\u6570":24,"\u5982\u4f55\u51cf\u5c11\u5185\u5b58\u5360\u7528":24,"\u5982\u4f55\u521d\u59cb\u5316\u53c2\u6570":24,"\u5982\u4f55\u52a0\u901fpaddlepaddle\u7684\u8bad\u7ec3\u901f\u5ea6":24,"\u5982\u4f55\u6307\u5b9agpu\u8bbe\u5907":24,"\u5982\u4f55\u66f4\u65b0www":38,"\u5982\u4f55\u6784\u5efapaddlepaddle\u7684\u6587\u6863":38,"\u5982\u4f55\u8d21\u732e":38,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":36,"\u5982\u4f55\u8fdb\u884c\u6027\u80fd\u5206\u6790":40,"\u5982\u4f55\u9009\u62e9sgd\u7b97\u6cd5\u7684\u5b66\u4e60\u7387":24,"\u5b50\u5e8f\u5217\u95f4\u65e0memori":32,"\u5b50\u5e8f\u5217\u95f4\u6709memori":32,"\u5b57\u6bb5\u914d\u7f6e\u6587\u4ef6":59,"\u5b89\u88c5":[29,57],"\u5b89\u88c5\u4e0e\u7f16\u8bd1":28,"\u5b89\u88c5\u6d41\u7a0b":28,"\u5b89\u88c5kubectl":47,"\u5b8c\u6574\u6559\u7a0b":56,"\u5b9e\u73b0":20,"\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42":37,"\u5b9e\u73b0\u65b9\u5f0f":21,"\u5b9e\u73b0c":37,"\u5b9e\u73b0python\u5c01\u88c5":37,"\u5bfc\u51fac":20,"\u5c06\u547d\u4ee4\u53c2\u6570\u4f20\u7ed9\u7f51\u7edc\u914d\u7f6e":45,"\u5c0f\u7ed3":3,"\u5de5\u5177":40,"\u5e38\u7528\u6a21\u578b":56,"\u5e76\u5b8c\u6210":36,"\u5efa\u7acb":36,"\u5f00\u53d1\u6807\u51c6":39,"\u5f00\u59cb\u5f00\u53d1":36,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":43,"\u5f15\u7528":60,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":18,"\u5feb\u901f\u5165\u95e8\u6559\u7a0b":57,"\u6027\u80fd\u4f18\u5316":39,"\u6027\u80fd\u5206\u6790\u5c0f\u6280\u5de7":40,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u4ecb\u7ecd":40,"\u6027\u80fd\u8c03\u4f18":43,"\u603b\u4f53\u6548\u679c\u603b\u7ed3":57,"\u60c5\u611f\u5206\u6790\u6559\u7a0b":61,"\u6216\u8005\u662f":24,"\u627e\u5230\u7684pythonlibs\u548cpythoninterp\u7248\u672c\u4e0d\u4e00\u81f4":24,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":20,"\u63a5\u53e3":55,"\u63a8\u5bfc\u65b9\u7a0b":37,"\u63a8\u6d4b\u6267\u884c":18,"\u63d0\u4ea4":36,"\u63d0\u4ea4\u955c\u50cf":48,"\u63d0\u53d6\u7535\u5f71\u6216\u7528\u6237\u7684\u7279\u5f81\u5e76\u751f\u6210python\u5bf9\u8c61":59,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":31,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":19,"\u6570\u636e\u51c6\u5907":[54,59,62],"\u6570\u636e\u63cf\u8ff0":60,"\u6570\u636e\u63d0\u4f9b":60,"\u6570\u636e\u63d0\u4f9b\u5668":46,"\u6570\u636e\u63d0\u4f9b\u811a\u672c":59,"\u6570\u636e\u652f\u6301":43,"\u6570\u636e\u683c\u5f0f\u51c6\u5907":57,"\u6570\u636e\u6e90\u914d\u7f6e":46,"\u6570\u636e\u7684\u51c6\u5907\u548c\u9884\u5904\u7406":53,"\u6570\u636e\u96c6\u7279\u5f81":58,"\u6570\u636e\u9884\u5904\u7406":62,"\u6570\u6910\u4ecb\u7ecd":61,"\u6570\u6910\u51c6\u5907":61,"\u6574\u4f53\u65b9\u6848":49,"\u6587\u4ef6\u9884\u5904\u7406":19,"\u6587\u672c\u751f\u6210":62,"\u6587\u672c\u751f\u6210\u6559\u7a0b":62,"\u6587\u6863":[27,52],"\u65b0\u624b\u5165\u95e8":30,"\u65f6\u5e8f\u6a21\u578b":57,"\u65f6\u5e8f\u6a21\u578b\u7684\u4f7f\u7528\u573a\u666f":3,"\u65f6\u95f4\u5e8f\u5217":32,"\u65f6\u95f4\u6b65":32,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":21,"\u672c\u5730\u6d4b\u8bd5":45,"\u672c\u5730\u8bad\u7ec3":45,"\u672f\u8bed":18,"\u6784\u5efa\u548c\u6d4b\u8bd5":36,"\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":48,"\u67e5\u770b\u8f93\u51fa":49,"\u6837\u4f8b\u6570\u636e":3,"\u6848\u4f8b\u4e00":45,"\u6848\u4f8b\u4e8c":45,"\u68c0\u67e5\u6a21\u578b\u8f93\u51fa":41,"\u68c0\u67e5\u96c6\u7fa4\u8bad\u7ec3\u7ed3\u679c":41,"\u6982\u8ff0":[31,34],"\u6a21\u578b":55,"\u6a21\u578b\u4e0b\u8f7d":55,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":18,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":21,"\u6a21\u578b\u68c0\u9a8c":25,"\u6a21\u578b\u7f51\u7edc\u7ed3\u6784":57,"\u6a21\u578b\u8bad\u7ec3":[54,62],"\u6a21\u578b\u8bc4\u4f30\u548c\u9884\u6d4b":59,"\u6a21\u578b\u914d\u7f6e":[32,39],"\u6a21\u578b\u914d\u7f6e\u7684\u6a21\u578b\u914d\u7f6e":32,"\u6ce8\u610f\u4e8b\u9879":3,"\u6d41\u7a0b\u4ecb\u7ecd":19,"\u6d4b\u8bd5":[43,60],"\u6d4b\u8bd5\u6587\u4ef6":59,"\u6d4b\u8bd5\u6a21\u578b":61,"\u7279\u5f81":60,"\u7279\u5f81\u63d0\u53d6":55,"\u751f\u6210\u5e8f\u5217":35,"\u751f\u6210\u6a21\u578b\u7684\u547d\u4ee4\u4e0e\u7ed3\u679c":62,"\u751f\u6210\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":34,"\u7528\u6237\u6587\u4ef6\u63cf\u8ff0":58,"\u7528\u6237\u81ea\u5b9a\u4e49\u6570\u636e\u96c6":62,"\u7528\u6237\u81ea\u5b9a\u4e49\u6570\u6910\u9884\u5904\u7406":61,"\u7535\u5f71\u6587\u4ef6\u63cf\u8ff0":58,"\u76ee\u5f55\u7ed3\u6784":21,"\u76f4\u63a5\u6784\u5efapaddlepaddle\u7684\u6587\u6863":38,"\u76f8\u5173\u6982\u5ff5":34,"\u77e9\u9635":43,"\u793a\u4f8b1":32,"\u793a\u4f8b2":32,"\u793a\u4f8b3":32,"\u793a\u4f8b4":32,"\u793a\u4f8b\u7a0b\u5e8f":19,"\u795e\u7ecf\u5143\u6fc0\u6d3b\u5185\u5b58":24,"\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e":59,"\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":60,"\u7a00\u758f\u8bad\u7ec3":45,"\u7b26\u53f7":20,"\u7b80\u4ecb":[47,62],"\u7b80\u5355\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":35,"\u7c7b":[20,37],"\u7cfb\u7edf\u6846\u56fe":46,"\u7ec3\u4e60":54,"\u7ec6\u8282\u63a2\u7a76":54,"\u7ec6\u8282\u63cf\u8ff0":43,"\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":41,"\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u4efb\u52a1":25,"\u7f16\u5199yaml\u6587\u4ef6":48,"\u7f16\u8bd1\u6d41\u7a0b":28,"\u7f16\u8bd1\u9009\u9879":21,"\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":26,"\u7f29\u5bb9":18,"\u7f51\u7edc\u53ef\u89c6\u5316":55,"\u7f51\u7edc\u7ed3\u6784\u914d\u7f6e":46,"\u7f51\u7edc\u914d\u7f6e\u4e2d\u7684\u8c03\u7528":3,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":20,"\u80cc\u666f":20,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":43,"\u81f4\u8c22":0,"\u89c2\u6d4b\u8bcd\u5411\u91cf":53,"\u8bad\u7ec3":[43,59,60],"\u8bad\u7ec3\u5668\u914d\u7f6e\u6587\u4ef6":59,"\u8bad\u7ec3\u56e0\u6b64\u9000\u51fa\u600e\u4e48\u529e":24,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8":19,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":19,"\u8bad\u7ec3\u6a21\u578b":[25,57,61],"\u8bad\u7ec3\u6a21\u578b\u7684\u547d\u4ee4\u4e0e\u7ed3\u679c":62,"\u8bad\u7ec3\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":34,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0":24,"\u8bad\u7ec3\u914d\u7f6e\u6587\u4ef6":46,"\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570":44,"\u8bc4\u5206\u6587\u4ef6\u63cf\u8ff0":58,"\u8bcd\u5411\u91cf\u6a21\u578b":57,"\u8bcd\u5411\u91cf\u6a21\u578b\u7684\u4fee\u6b63":53,"\u8bcd\u6c47\u8868":32,"\u8be6\u7ec6\u6559\u7a0b":40,"\u8bed\u4e49\u89d2\u8272\u6807\u6ce8\u6559\u7a0b":60,"\u8bfb\u53d6\u53cc\u5c42\u5e8f\u5217\u6570\u636e":32,"\u8f6c\u6362\u5e93":19,"\u8f93\u5165":34,"\u8f93\u5165\u4e0d\u7b49\u957f":32,"\u8f93\u5165\u793a\u4f8b":34,"\u8f93\u51fa":34,"\u8f93\u51fa\u65e5\u5fd7":57,"\u8fd0\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":41,"\u8fd0\u884c\u5bb9\u5668":48,"\u8fd0\u884cdocker":24,"\u8fd0\u884cpaddlepaddl":27,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":20,"\u8fdb\u884c\u8bad\u7ec3":[19,48],"\u8fdb\u9636\u6307\u5357":39,"\u9009\u62e9\u5b58\u50a8\u65b9\u6848":47,"\u901a\u7528":43,"\u901a\u8fc7docker\u5bb9\u5668\u5f00\u53d1paddlepaddl":27,"\u903b\u8f91\u56de\u5f52\u6a21\u578b":57,"\u9047\u5230":24,"\u90e8\u7f72kubernetes\u96c6\u7fa4":47,"\u914d\u7f6e\u4e2d\u7684\u6570\u636e\u52a0\u8f7d\u5b9a\u4e49":57,"\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":35,"\u914d\u7f6ekubectl":47,"\u914d\u7f6ekubectl\u8bbf\u95ee\u4f60\u7684kubernetes\u96c6\u7fa4":47,"\u94a9\u5b50":36,"\u9644\u5f55":57,"\u968f\u673a\u6570":43,"\u96c6\u7fa4\u8bad\u7ec3":45,"\u975e\u6cd5\u6307\u4ee4":24,"\u9884\u5904\u7406":54,"\u9884\u5904\u7406\u547d\u4ee4\u548c\u7ed3\u679c":62,"\u9884\u5904\u7406\u5de5\u4f5c\u6d41\u7a0b":62,"\u9884\u6d4b":[54,55,57,60,61],"\u9884\u6d4b\u6d41\u7a0b":5,"\u9884\u6d4bdemo":5,"\u9884\u8bad\u7ec3\u7684\u6a21\u578b":62,"api\u4e2d\u6587\u624b\u518c":4,"api\u5bf9\u6bd4\u4ecb\u7ecd":32,"beam_search\u7684\u751f\u6210":32,"blas\u8def\u5f84\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":26,"bleu\u8bc4\u4f30":62,"book\u4e2d\u6240\u6709\u7ae0\u8282":23,"bool\u578b\u7684\u7f16\u8bd1\u9009\u9879":26,"cmake\u6e90\u7801\u7f16\u8bd1":24,"cudnn\u7684\u7f16\u8bd1\u9009\u9879":26,"dataprovider\u7684\u4ecb\u7ecd":2,"dataprovider\u7684\u4f7f\u7528":3,"float":24,"gpu\u548ccpu\u6df7\u5408\u4f7f\u7528":45,"gpu\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18":40,"gpu\u955c\u50cf\u51fa\u73b0":24,"group\u6559\u7a0b":34,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":49,"kubernetes\u5355\u673a\u8bad\u7ec3":48,"meta\u6587\u4ef6":59,"meta\u914d\u7f6e\u6587\u4ef6":59,"mnist\u7684\u4f7f\u7528\u573a\u666f":3,"movielens\u6570\u636e\u96c6":58,"movielens\u6570\u636e\u96c6\u8bc4\u5206\u56de\u5f52\u6a21\u578b":59,"org\u6587\u6863":38,"paddle\u52a8\u6001\u5e93\u4e2d":20,"paddle\u53d1\u884c\u89c4\u8303":23,"paddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":23,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":20,"paddlepaddle\u53d1\u5e03\u7684docker\u955c\u50cf\u4f7f\u7528\u8bf4\u660e":27,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":26,"paddlepaddle\u7684docker\u5bb9\u5668\u4f7f\u7528\u65b9\u5f0f":27,"pod\u95f4\u901a\u4fe1":49,"pydataprovider2\u7684\u4f7f\u7528":3,"python\u63a5\u53e3":55,"python\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86":24,"python\u811a\u672c\u8bfb\u53d6\u6570\u636e":57,"return":22,"rnn\u76f8\u5173\u6a21\u578b":33,"rnn\u914d\u7f6e":35,"so\u627e\u4e0d\u5230":29,"ubuntu\u90e8\u7f72paddlepaddl":29,Abs:6,activ:6,adadelta:11,adagrad:11,adam:11,adamax:11,addto:9,aggreg:9,algorithm:17,api:[1,4,21],applic:4,argument:22,async:43,attent:35,attribut:7,auc:8,avg:12,basepool:12,batch:22,batch_norm:9,batch_siz:22,beam_search:9,becaus:24,between:16,bidirectional_gru:10,bidirectional_lstm:10,big:24,bilinear_interp:9,bla:26,block_expand:9,book:27,brelu:6,cach:3,capi:21,capi_priv:21,check:9,checkpoint:[17,18],chunk:8,cifar:13,classif:8,classification_error:8,classification_error_print:8,clone:36,column_sum:8,commit:36,compos:22,concat:9,config:4,configur:14,conll05:13,connect:9,content:[3,21,24,31,40,46],context_project:9,conv:9,conv_oper:9,conv_project:9,conv_shift:9,cos_sim:9,cost:9,cp27mu:24,creat:22,creator:22,crf:9,crf_decod:9,cross_channel_norm:9,cross_entropy_cost:9,cross_entropy_with_selfnorm_cost:9,ctc:9,ctc_error:8,cuda:[24,26],cudnn:26,cudnnavg:12,cudnnmax:12,custom:22,dat:58,data:[9,13,17,22],datafeed:13,dataprovid:[4,43],dataset:[13,17],datatyp:13,decayedadagrad:11,decor:22,design:[16,17,22],dictionari:22,dispatch:17,distribut:[16,17],doc:[16,17,22],dotmul_oper:9,dotmul_project:9,driver:24,dropout_lay:10,dylib:21,dynam:17,embed:9,entri:22,eos:9,evalu:8,event:[15,16],exampl:[16,21],except:24,exp:6,expand:[9,31],faq:24,fault:17,first_seq:[9,31],fork:36,format:17,from:16,full_matrix_project:9,fulli:9,gate:35,get_output:9,gpu:43,gradient_print:8,group:9,gru:[10,43],gru_group:10,gru_step:9,gru_unit:10,grumemori:9,handler:[16,20],how:22,hsigmoid:9,huber_cost:9,ident:6,identity_project:9,illeg:24,imag:[9,10],imagenet:55,imdb:[13,61],img_cmrnorm:9,img_conv:9,img_conv_bn_pool:10,img_conv_group:10,img_pool:9,imikolov:13,implement:22,infer:15,ingredi:16,init_hook:3,input_typ:3,instruct:24,insuffici:24,interfac:[13,17,22],interpol:9,isn:22,issu:36,job:[17,48],join:9,kubernet:[47,48],lambda_cost:9,last_seq:[9,31],layer:[9,16],libcudart:29,libcudnn:29,libpaddle_capi_shar:21,libpaddle_capi_whol:21,linear:6,linear_comb:9,linux_x86_64:24,list:[18,22],log:6,lstm:[10,43,60,61],lstm_step:9,lstmemori:9,lstmemory_group:10,lstmemory_unit:10,map:22,master:17,math:9,max:12,maxframe_print:8,maxid:9,maxid_print:8,maxout:9,memori:[9,32,34],messag:24,mini:22,minibatch:13,misc:10,mix:9,mnist:13,model:[4,14,16,35,55],momentum:11,movi:58,movielen:13,mse_cost:9,multi_binary_label_cross_entropy_cost:9,multipl:22,nce:9,need:22,network:[10,35],neural:35,nlp:[10,43],norm:9,nvprof:40,nvvp:40,object:17,onli:22,optim:[11,17],pad:9,paddl:[22,23],paddlepaddl:[16,38,52],parallel_nn:45,paramet:[7,15,16,17],perform:43,platform:24,pnpair:8,point:24,pool:[9,12,31],power:9,pre:36,precision_recal:8,prefetch:22,print:8,process:17,protocol:24,provid:[3,22],pull:36,push:36,python:22,queue:17,rank:8,rank_cost:9,rate:58,reader:[13,16,22],recoveri:17,recurr:[9,10,34,35],recurrent_group:9,refer:3,reject:24,relu:6,repeat:9,request:36,reshap:9,resnet:55,rmsprop:11,rnn:[32,43],rotat:9,sampl:9,sampling_id:9,scale:[9,17],scaling_project:9,selective_fc:9,sentiment:13,seq_concat:9,seq_reshap:9,seqtext_print:8,sequenc:35,sequence_conv_pool:10,sequencesoftmax:6,server:17,sgd:43,share:16,shuffl:22,sigmoid:6,simple_attent:10,simple_gru2:10,simple_gru:10,simple_img_conv_pool:10,simple_lstm:10,singl:22,slice:9,slope_intercept:9,small_vgg:10,smooth_l1_cost:9,softmax:6,softrelu:6,spp:9,squar:6,squarerootn:12,stack:61,stanh:6,start:16,store:17,suffici:22,sum:[8,12],sum_cost:9,sum_to_one_norm:9,summar:16,support:24,tabl:21,table_project:9,take:22,tanh:6,task:17,tensor:9,text_conv_pool:10,thi:24,todo:[18,19],toler:17,too:24,train:[15,16,17,22],trainer:[15,17],tran:9,trans_full_matrix_project:9,tune:43,uci_h:13,updat:16,usag:22,use:22,user:[17,58],util:8,value_print:8,version:24,vgg_16_network:10,warp_ctc:9,wheel:24,whl:24,why:22,wmt14:13,zoo:55}})
\ No newline at end of file
diff --git a/develop/doc_cn/tutorials/embedding_model/index_cn.html b/develop/doc_cn/tutorials/embedding_model/index_cn.html
index 8e196a1e5a76aeb053f9c05a59250acf2e0b4b2e..c2c91bd95854049ed7ab400969cc82293ea4dca6 100644
--- a/develop/doc_cn/tutorials/embedding_model/index_cn.html
+++ b/develop/doc_cn/tutorials/embedding_model/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/tutorials/image_classification/index_cn.html b/develop/doc_cn/tutorials/image_classification/index_cn.html
index 456b3f581d63d922c5b7d3dbb319414859e7f3bd..a0cce10cc81baebaee1601d488c7932d124ce43b 100644
--- a/develop/doc_cn/tutorials/image_classification/index_cn.html
+++ b/develop/doc_cn/tutorials/image_classification/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/tutorials/imagenet_model/resnet_model_cn.html b/develop/doc_cn/tutorials/imagenet_model/resnet_model_cn.html
index 080003bec81f4d9ce4b7cf191def232b4fcafd80..476abba41e51c2ba57a1096bc6d198764a89bfe8 100644
--- a/develop/doc_cn/tutorials/imagenet_model/resnet_model_cn.html
+++ b/develop/doc_cn/tutorials/imagenet_model/resnet_model_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/tutorials/index_cn.html b/develop/doc_cn/tutorials/index_cn.html
index d09170fb458c66a540b262ac19406da2adb54d05..73167d0e71f0e0170f6f3fe2a19f41c6e393902f 100644
--- a/develop/doc_cn/tutorials/index_cn.html
+++ b/develop/doc_cn/tutorials/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/tutorials/quick_start/index_cn.html b/develop/doc_cn/tutorials/quick_start/index_cn.html
index 32bccfc187e4c5f387e4fa203604d9cee438e979..e0a892702062e2944d745e069b7f37d9869f501d 100644
--- a/develop/doc_cn/tutorials/quick_start/index_cn.html
+++ b/develop/doc_cn/tutorials/quick_start/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
@@ -413,7 +414,7 @@
 <blockquote>
 <div><a class="reference internal image-reference" href="../../_images/PipelineNetwork_cn.jpg"><img alt="../../_images/PipelineNetwork_cn.jpg" class="align-center" src="../../_images/PipelineNetwork_cn.jpg" style="width: 544.8px; height: 44.8px;" /></a>
 </div></blockquote>
-<p>我们将以最基本的逻辑回归网络作为起点，并逐渐展示更加深入的功能。更详细的网络配置连接请参考 <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers"><span class="std std-ref">Layers</span></a> 。
+<p>我们将以最基本的逻辑回归网络作为起点，并逐渐展示更加深入的功能。更详细的网络配置连接请参考 <span class="xref std std-ref">api_trainer_config_helpers_layers</span> 。
 所有配置都能在 <a class="reference external" href="https://github.com/PaddlePaddle/Paddle">源代码</a> 的 <code class="docutils literal"><span class="pre">demo/quick_start</span></code> 目录下找到。</p>
 <div class="section" id="id11">
 <h3>逻辑回归模型<a class="headerlink" href="#id11" title="永久链接至标题">¶</a></h3>
diff --git a/develop/doc_cn/tutorials/rec/ml_dataset_cn.html b/develop/doc_cn/tutorials/rec/ml_dataset_cn.html
index 165dba8ed90e443745003ad7cfd6d5323c3ae62e..bc5fec96fe288ac1e5f75cc13fedd1701fd5a989 100644
--- a/develop/doc_cn/tutorials/rec/ml_dataset_cn.html
+++ b/develop/doc_cn/tutorials/rec/ml_dataset_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/tutorials/rec/ml_regression_cn.html b/develop/doc_cn/tutorials/rec/ml_regression_cn.html
index 224f393116572676cf19eef95b2e35e53cca7701..11538f3329c368448da6e7f9d7ce1f5669356fe1 100644
--- a/develop/doc_cn/tutorials/rec/ml_regression_cn.html
+++ b/develop/doc_cn/tutorials/rec/ml_regression_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
@@ -618,14 +619,14 @@ cp ml-1m/ratings.dat.test .
 并且对用户的特征做同样的操作，也得到一个用户特征。然后我们求这两个特征的余弦相似度。</p>
 <p>在这些网络中，我们用以下的一些:ref:<cite>api_trainer_config</cite> 中的接口。</p>
 <ul class="simple">
-<li>数据层， <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-data-layer"><span class="std std-ref">data_layer</span></a></li>
-<li>全连接层， <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-fc-layer"><span class="std std-ref">fc_layer</span></a></li>
-<li>嵌入层， <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-embedding-layer"><span class="std std-ref">embedding_layer</span></a></li>
-<li>文本投影层， <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-context-projection"><span class="std std-ref">context_projection</span></a></li>
-<li>采样层， <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-pooling-layer"><span class="std std-ref">pooling_layer</span></a></li>
-<li>余弦相似度层， <a class="reference internal" href="../../api/v1/trainer_config_helpers/layers.html#api-trainer-config-helpers-layers-cos-sim"><span class="std std-ref">cos_sim</span></a></li>
+<li>数据层， <span class="xref std std-ref">api_trainer_config_helpers_layers_data_layer</span></li>
+<li>全连接层， <span class="xref std std-ref">api_trainer_config_helpers_layers_fc_layer</span></li>
+<li>嵌入层， <span class="xref std std-ref">api_trainer_config_helpers_layers_embedding_layer</span></li>
+<li>文本投影层， <span class="xref std std-ref">api_trainer_config_helpers_layers_context_projection</span></li>
+<li>采样层， <span class="xref std std-ref">api_trainer_config_helpers_layers_pooling_layer</span></li>
+<li>余弦相似度层， <span class="xref std std-ref">api_trainer_config_helpers_layers_cos_sim</span></li>
 <li>文本卷积采样层， <a class="reference internal" href="../../api/v2/config/networks.html#api-trainer-config-helpers-network-text-conv-pool"><span class="std std-ref">text_conv_pool</span></a></li>
-<li>声明Python数据源， <a class="reference internal" href="../../api/v1/trainer_config_helpers/data_sources.html#api-trainer-config-helpers-data-sources"><span class="std std-ref">DataSources</span></a></li>
+<li>声明Python数据源， <span class="xref std std-ref">api_trainer_config_helpers_data_sources</span></li>
 </ul>
 </div>
 <div class="section" id="id9">
diff --git a/develop/doc_cn/tutorials/semantic_role_labeling/index_cn.html b/develop/doc_cn/tutorials/semantic_role_labeling/index_cn.html
index 270d55341e8892f41ae3d07d30d80c1738ef999d..9f79366d778ab6111729405579fc20ee671e8a43 100644
--- a/develop/doc_cn/tutorials/semantic_role_labeling/index_cn.html
+++ b/develop/doc_cn/tutorials/semantic_role_labeling/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/tutorials/sentiment_analysis/index_cn.html b/develop/doc_cn/tutorials/sentiment_analysis/index_cn.html
index e5b6922308fcb2013194610fa00df68e04c53173..f9a157241bfb961637b67f477d4cf058857e8b5a 100644
--- a/develop/doc_cn/tutorials/sentiment_analysis/index_cn.html
+++ b/develop/doc_cn/tutorials/sentiment_analysis/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
diff --git a/develop/doc_cn/tutorials/text_generation/index_cn.html b/develop/doc_cn/tutorials/text_generation/index_cn.html
index 9b72175886448cf81c35e19fde171c9238463da0..d9196876060581d18166662846529b6a9dfa6f72 100644
--- a/develop/doc_cn/tutorials/text_generation/index_cn.html
+++ b/develop/doc_cn/tutorials/text_generation/index_cn.html
@@ -142,6 +142,7 @@
 <li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">模型配置</a><ul>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
+<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
 <li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>