-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- 
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- 
- paddle.v2.activation.AbsAbs Activation.
-Forward: \(f(z) = abs(z)\)
-Derivative:
-paddle.v2.activation.ExpExponential Activation.
-paddle.v2.activation.Identityalias of Linear
paddle.v2.activation.LinearIdentity Activation.
-Just do nothing for output both forward/backward.
-paddle.v2.activation.LogLogarithm Activation.
-paddle.v2.activation.SquareSquare Activation.
-paddle.v2.activation.SigmoidSigmoid activation.
-paddle.v2.activation.SoftmaxSoftmax activation for simple input
-paddle.v2.activation.SequenceSoftmaxSoftmax activation for one sequence. The dimension of input feature must be -1 and a sequence.
-result = softmax(for each_feature_vector[0] in input_feature)
-for i, each_time_step_output in enumerate(output):
- each_time_step_output = result[i]
-paddle.v2.activation.ReluRelu activation.
-forward. \(y = max(0, z)\)
-derivative:
-paddle.v2.activation.BReluBRelu Activation.
-forward. \(y = min(24, max(0, z))\)
-derivative:
-paddle.v2.activation.SoftReluSoftRelu Activation.
-paddle.v2.activation.TanhTanh activation.
-paddle.v2.activation.STanhScaled Tanh Activation.
-paddle.v2.activation.SoftSignSoftSign Activation.
-
- paddle.v2.attr.Paramalias of ParameterAttribute
paddle.v2.attr.Extraalias of ExtraLayerAttribute
paddle.v2.attr.Hookalias of HookAttribute
paddle.v2.attr.HookAttralias of HookAttribute
paddle.v2.attr.ParamAttralias of ParameterAttribute
paddle.v2.attr.ExtraAttralias of ExtraLayerAttribute
paddle.v2.attr.ParameterAttribute(name=None, is_static=False, initial_std=None, initial_mean=None, initial_max=None, initial_min=None, l1_rate=None, l2_rate=None, learning_rate=None, momentum=None, gradient_clipping_threshold=None, sparse_update=False, update_hooks=None, initializer=None)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.
-NOTE: IT IS A HIGH LEVEL USER INTERFACE.
-| Parameters: |
|
-
|---|
set_default_parameter_name(name)Set default parameter name. If parameter not set, then will use default -parameter name.
-| Parameters: | name (basestring) – default parameter name. | -
|---|
paddle.v2.attr.ExtraLayerAttribute(error_clipping_threshold=None, drop_rate=None, device=None)Some high level layer attributes config. You can set all attributes here, -but some layer doesn’t support all attributes. If you set an attribute to a -layer that not support this attribute, paddle will print an error and core.
-| Parameters: |
|
-
|---|
- paddle.v2.evaluator.classification_error(*args, **xargs)Classification Error Evaluator. It will print error rate for classification.
-The classification error is:
-The simple usage is:
-eval = classification_evaluator.error(input=prob,label=lbl)
-| Parameters: |
|
-
|---|---|
| Returns: | None. - |
-
paddle.v2.evaluator.auc(*args, **xargs)Auc Evaluator which adapts to binary classification.
-The simple usage:
-eval = evaluator.auc(input, label)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.ctc_error(*args, **xargs)This evaluator is to calculate sequence-to-sequence edit distance.
-The simple usage is :
-eval = ctc_evaluator.error(input=input, label=lbl)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.chunk(*args, **xargs)Chunk evaluator is used to evaluate segment labelling accuracy for a -sequence. It calculates precision, recall and F1 scores for the chunk detection.
-To use chunk evaluator, several concepts need to be clarified firstly.
-We can name a label by combining tag type and chunk type. (ie. B-ORG for begining of an organization name)
-The construction of label dictionary should obey the following rules:
-Scheme Description
-plain Use the same label for the whole chunk.
-IOB Two labels for chunk type X, B-X for chunk begining and I-X for chunk inside.
-IOE Two labels for chunk type X, E-X for chunk ending and I-X for chunk inside.
-IOBES Four labels for chunk type X, B-X for chunk begining, I-X for chunk inside, E-X for chunk end and S-X for single word chunk.
-To make it clear, let’s illustrate by an NER example. -Assuming that there are three named entity types including ORG, PER and LOC which are called ‘chunk type’ here, -if ‘IOB’ scheme were used, the label set will be extended to a set including B-ORG, I-ORG, B-PER, I-PER, B-LOC, I-LOC and O, -in which B-ORG for begining of ORG and I-ORG for inside of ORG. -Prefixes which are called ‘tag type’ here are added to chunk types and there are two tag types including B and I. -Of course, the training data should be labeled accordingly.
-The following table are equations to extract tag type and chunk type from a label.
-tagType = label % numTagType
-chunkType = label / numTagType
-otherChunkType = numChunkTypes
-The following table shows the mapping rule between tagType and tag type in each scheme.
-Scheme Begin Inside End Single
-plain 0 - - -
-IOB 0 1 - -
-IOE - 0 1 -
-IOBES 0 1 2 3
-Continue the NER example, and the label dict should look like this to satify above equations:
-B-ORG 0
-I-ORG 1
-B-PER 2
-I-PER 3
-B-LOC 4
-I-LOC 5
-O 6
-In this example, chunkType has three values: 0 for ORG, 1 for PER, 2 for LOC, because the scheme is -“IOB” so tagType has two values: 0 for B and 1 for I. -Here we will use I-LOC to explain the above mapping rules in detail. -For I-LOC, the label id is 5, so we can get tagType=1 and chunkType=2, which means I-LOC is a part of NER chunk LOC -and the tag is I.
-The simple usage is:
-eval = evaluator.chunk(input, label, chunk_scheme, num_chunk_types)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.precision_recall(*args, **xargs)An Evaluator to calculate precision and recall, F1-score. -It is adapt to the task with multiple labels.
-The simple usage:
-eval = precision_evaluator.recall(input, label)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.pnpair(*args, **xargs)Positive-negative pair rate Evaluator which adapts to rank task like -learning to rank. This evaluator must contain at least three layers.
-The simple usage:
-eval = evaluator.pnpair(input, label, query_id)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.sum(*args, **xargs)An Evaluator to sum the result of input.
-The simple usage:
-eval = evaluator.sum(input)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.column_sum(*args, **xargs)This Evaluator is used to sum the last column of input.
-The simple usage is:
-eval = column_evaluator.sum(input, label)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.classification_error_printer(*args, **xargs)This Evaluator is used to print the classification error of each sample.
-The simple usage is:
-eval = classification_error_evaluator.printer(input)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.gradient_printer(*args, **xargs)This Evaluator is used to print the gradient of input layers. It contains -one or more input layers.
-The simple usage is:
-eval = gradient_evaluator.printer(input)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.maxid_printer(*args, **xargs)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.
-The simple usage is:
-eval = maxid_evaluator.printer(input)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.maxframe_printer(*args, **xargs)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. -It contains one or more input layers.
-Note
-The width of each frame is 1.
-The simple usage is:
-eval = maxframe_evaluator.printer(input)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.seqtext_printer(*args, **xargs)Sequence text printer will print text according to index matrix and a -dictionary. There can be multiple input to this layer:
-1. If there is no id_input, the input must be a matrix containing -the sequence of indices;
-The output format will be:
-id prob space_seperated_tokens_from_dictionary_according_to_seq
-id space_seperated_tokens_from_dictionary_according_to_seq
-id space_seperated_tokens_from_dictionary_according_to_sub_seq
- space_seperated_tokens_from_dictionary_according_to_sub_seq
-...
-Typically SequenceTextPrinter layer takes output of maxid or RecurrentGroup -with maxid (when generating) as an input.
-The simple usage is:
-eval = seqtext_evaluator.printer(input=maxid,
- id_input=sample_id,
- dict_file=dict_file,
- result_file=result_file)
-| Parameters: |
|
-
|---|---|
| Returns: | The seq_text_printer that prints the generated sequence to a file. - |
-
| Return type: | evaluator - |
-
paddle.v2.evaluator.value_printer(*args, **xargs)This Evaluator is used to print the values of input layers. It contains -one or more input layers.
-The simple usage is:
-eval = value_evaluator.printer(input)
-| Parameters: |
|
-
|---|
paddle.v2.evaluator.detection_map(*args, **xargs)Detection mAP Evaluator. It will print mean Average Precision (mAP) for detection.
-The detection mAP Evaluator based on the output of detection_output layer counts -the true positive and the false positive bbox and integral them to get the -mAP.
-The simple usage is:
-eval = detection_evaluator.map(input=det_output,label=lbl)
-| Parameters: |
|
-
|---|
- paddle.v2.layer.fcThe fully connected layer.
-The example usage is:
-fc = fc(input=layer,
- size=1024,
- act=paddle.v2.activation.Linear(),
- bias_attr=False)
-which is equal to:
-with mixed(size=1024) as fc:
- fc += full_matrix_projection(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.selective_fcSelectived fully connected layer. Different from fc, the output -of this layer can be sparse. It requires an additional input to indicate -several selected columns for output. If the selected columns is not -specified, selective_fc acts exactly like fc.
-The simple usage is:
-sel_fc = selective_fc(input=input, size=128, act=paddle.v2.activation.Tanh())
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.conv_operatorDifferent from img_conv, conv_op is an Operator, which can be used -in mixed. And conv_op takes two inputs to perform convolution. -The first input is the image and the second is filter kernel. It only -supports GPU mode.
-The example usage is:
-op = conv_operator(img=input1,
- filter=input2,
- filter_size=3,
- num_filters=64,
- num_channels=64)
-| Parameters: |
|
-
|---|---|
| Returns: | A ConvOperator Object. - |
-
| Return type: | ConvOperator - |
-
paddle.v2.layer.conv_projectionDifferent from img_conv and conv_op, conv_projection is a Projection, -which can be used in mixed and concat. It uses cudnn to implement -convolution and only supports GPU mode.
-The example usage is:
-proj = conv_projection(input=input1,
- filter_size=3,
- num_filters=64,
- num_channels=64)
-| Parameters: |
|
-
|---|---|
| Returns: | A Projection Object. - |
-
| Return type: | ConvTransProjection | ConvProjection - |
-
paddle.v2.layer.conv_shiftThe example usage is:
-conv_shift = conv_shift(a=layer1, b=layer2)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.img_convConvolution layer for image. Paddle can support both square and non-square -input currently.
-The details of convolution layer, please refer UFLDL’s convolution .
-Convolution Transpose (deconv) layer for image. Paddle can support both square -and non-square input currently.
-The details of convolution transpose layer, -please refer to the following explanation and references therein -<http://datascience.stackexchange.com/questions/6107/ -what-are-deconvolutional-layers/>`_ . -The num_channel means input image’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’s -num_filters.
-There are several groups of filters in PaddlePaddle implementation. -If the groups attribute is greater than 1, for example groups=2, -the input will be splitted into 2 parts along the channel axis, and -the filters will also be splitted into 2 parts. The first half of the filters -is only connected to the first half of the input channels, while the second -half of the filters is only connected to the second half of the input. After -the computation of convolution for each part of input, -the output will be obtained by concatenating the two results.
-The details of grouped convolution, please refer to: -ImageNet Classification with Deep Convolutional Neural Networks
-The example usage is:
-conv = img_conv(input=data, filter_size=1, filter_size_y=1,
- num_channels=8,
- num_filters=16, stride=1,
- bias_attr=False,
- act=paddle.v2.activation.Relu())
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.context_projectionContext Projection.
-It just reorganizes input sequence, combines “context_len” elements of the -sequence to one context from context_start. “context_start” will be set to --(context_len - 1) / 2 by default. When context position is out of sequence -length, padding will be filled as zero if padding_attr = False, otherwise -it is trainable.
-For example, origin sequence is [A B C D E F G], context len is 3, padding_attr -is not set, then after context projection, sequence will -be [ 0AB ABC BCD CDE DEF EFG FG0 ].
-| Parameters: |
|
-
|---|---|
| Returns: | Projection object. - |
-
| Return type: | Projection - |
-
paddle.v2.layer.row_convThe row convolution is called lookahead convolution. It is firstly -introduced in paper of Deep Speech 2: End-to-End Speech Recognition -in English and Mandarin .
-The bidirectional RNN that learns representation for a sequence by -performing a forward and a backward pass through the entire sequence. -However, unlike unidirectional RNNs, bidirectional RNNs are challenging -to deploy in an online and low-latency setting. The lookahead convolution -incorporates information from future subsequences in a computationally -efficient manner to improve unidirectional RNNs.
-The connection of row convolution is different from the 1D sequence -convolution. Assumed that, the future context-length is k, that is to say, -it can get the output at timestep t by using the the input feature from t-th -timestep to (t+k+1)-th timestep. Assumed that the hidden dim of input -activations are d, the activations r_t for the new layer at time-step t are:
-Note
-The context_len is k + 1. That is to say, the lookahead step -number plus one equals context_len.
-row_conv = row_conv(input=input, context_len=3)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.img_poolImage pooling Layer.
-The details of pooling layer, please refer to ufldl’s pooling .
-The example usage is:
-maxpool = img_pool(input=conv,
- pool_size=3,
- pool_size_y=5,
- num_channels=8,
- stride=1,
- stride_y=2,
- padding=1,
- padding_y=2,
- pool_type=MaxPooling())
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sppA layer performs spatial pyramid pooling.
- -The example usage is:
-spp = spp(input=data,
- pyramid_height=2,
- num_channels=16,
- pool_type=MaxPooling())
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.maxoutSo groups should be larger than 1, and the num of channels should be able -to be devided by groups.
-The simple usage is:
-maxout = maxout(input,
- num_channels=128,
- groups=4)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.roi_poolA layer used by Fast R-CNN to extract feature maps of ROIs from the last -feature map.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.padThis operation pads zeros to the input data according to pad_c,pad_h -and pad_w. pad_c, pad_h, pad_w specify the size in the corresponding -dimension. And the input data shape is NCHW.
-For example, pad_c=[2,3] means padding 2 zeros before the input data -and 3 zeros after the input data in the channel dimension. pad_h means -padding zeros in the height dimension. pad_w means padding zeros in the -width dimension.
-For example,
-input(2,2,2,3) = [
- [ [[1,2,3], [3,4,5]],
- [[2,3,5], [1,6,7]] ],
- [ [[4,3,1], [1,8,7]],
- [[3,8,9], [2,3,5]] ]
- ]
-
-pad_c=[1,1], pad_h=[0,0], pad_w=[0,0]
-
-output(2,4,2,3) = [
- [ [[0,0,0], [0,0,0]],
- [[1,2,3], [3,4,5]],
- [[2,3,5], [1,6,7]],
- [[0,0,0], [0,0,0]] ],
- [ [[0,0,0], [0,0,0]],
- [[4,3,1], [1,8,7]],
- [[3,8,9], [2,3,5]],
- [[0,0,0], [0,0,0]] ]
- ]
-The simply usage is:
-pad = pad(input=ipt,
- pad_c=[4,4],
- pad_h=[0,0],
- pad_w=[2,2])
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.img_cmrnormResponse normalization across feature maps.
- -The example usage is:
-norm = img_cmrnorm(input=net, size=5)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.batch_normBatch Normalization Layer. The notation of this layer is as follows.
-\(x\) is the input features over a mini-batch.
-The example usage is:
-norm = batch_norm(input=net, act=paddle.v2.activation.Relu())
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sum_to_one_normA layer for sum-to-one normalization, -which is used in NEURAL TURING MACHINE.
-where \(in\) is a (batchSize x dataDim) input vector, -and \(out\) is a (batchSize x dataDim) output vector.
-The example usage is:
-sum_to_one_norm = sum_to_one_norm(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cross_channel_normNormalize a layer’s output. This layer is necessary for ssd. This -layer applys normalization across the channels of each sample to -a convolutional layer’s output and scales the output by a group of -trainable factors whose dimensions equal to the channel’s number.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.row_l2_normA layer for L2-normalization in each row.
-where the size of \(in\) is (batchSize x dataDim) , -and the size of \(out\) is a (batchSize x dataDim) .
-The example usage is:
-row_l2_norm = row_l2_norm(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.recurrentSimple recurrent unit layer. It is just a fully connect layer through both -time and neural network.
-For each sequence [start, end] it performs the following computation:
-If reversed is true, the order is reversed:
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.lstmemoryLong Short-term Memory Cell.
-The memory cell was implemented as follow equations.
-NOTE: In PaddlePaddle’s implementation, the multiplications -\(W_{xi}x_{t}\) , \(W_{xf}x_{t}\), -\(W_{xc}x_t\), \(W_{xo}x_{t}\) are not done in the lstmemory layer, -so an additional mixed with full_matrix_projection or a fc must -be included in the configuration file to complete the input-to-hidden -mappings before lstmemory is called.
-NOTE: This is a low level user interface. You can use network.simple_lstm -to config a simple plain lstm layer.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.grumemoryGate Recurrent Unit Layer.
-The memory cell was implemented as follow equations.
-1. update gate \(z\): defines how much of the previous memory to -keep around or the unit updates its activations. The update gate -is computed by:
-2. reset gate \(r\): determines how to combine the new input with the -previous memory. The reset gate is computed similarly to the update gate:
-3. The candidate activation \(\tilde{h_t}\) is computed similarly to -that of the traditional recurrent unit:
-4. The hidden activation \(h_t\) of the GRU at time t is a linear -interpolation between the previous activation \(h_{t-1}\) and the -candidate activation \(\tilde{h_t}\):
-NOTE: In PaddlePaddle’s implementation, the multiplication operations -\(W_{r}x_{t}\), \(W_{z}x_{t}\) and \(W x_t\) are not performed -in gate_recurrent layer. Consequently, an additional mixed with -full_matrix_projection or a fc must be included before grumemory -is called.
- -The simple usage is:
-gru = grumemory(input)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.gated_unitThe gated unit layer implements a simple gating mechanism over the input. -The input \(X\) is first projected into a new space \(X'\), and -it is also used to produce a gate weight \(\sigma\). Element-wise -product between \(X'\) and \(\sigma\) is finally returned.
-The example usage is:
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.memoryThe memory takes a layer’s output at previous time step as its own output.
-If boot_bias, the activation of the bias is the initial value of the memory.
-If boot_with_const_id is set, then the memory’s output at the first time step
-is a IndexSlot, the Arguments.ids()[0] is this cost_id.
If boot is specified, the memory’s output at the first time step will -be the boot’s output.
-In other case, the default memory’s output at the first time step is zero.
-mem = memory(size=256, name='state')
-state = fc(input=mem, size=256, name='state')
-If you do not want to specify the name, you can also use set_input() -to specify the layer to be remembered as the following:
-mem = memory(size=256)
-state = fc(input=mem, size=256)
-mem.set_input(mem)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.recurrent_groupRecurrent 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 useful for attention-based models, or Neural -Turning Machine like models.
-The basic usage (time steps) is:
-def step(input):
- output = fc(input=layer,
- size=1024,
- act=paddle.v2.activation.Linear(),
- bias_attr=False)
- return output
-
-group = recurrent_group(input=layer,
- step=step)
-You can see following configs for further usages:
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.lstm_stepLSTM Step Layer. This function is used only in recurrent_group. -The lstm equations are shown as follows.
-The input of lstm step is \(Wx_t + Wh_{t-1}\), and user should use
-mixed and full_matrix_projection to calculate these
-input vectors.
The state of lstm step is \(c_{t-1}\). And lstm step layer will do
-This layer has two outputs. The default output is \(h_t\). The other
-output is \(o_t\), whose name is ‘state’ and users can use
-get_output to extract this output.
| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.gru_step| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.beam_searchBeam 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.
-The example usage is:
-def rnn_step(input):
- last_time_step_output = memory(name='rnn', size=512)
- with mixed(size=512, name='rnn') as simple_rnn:
- simple_rnn += full_matrix_projection(input)
- simple_rnn += last_time_step_output
- return simple_rnn
-
-generated_word_embedding = GeneratedInput(
- size=target_dictionary_dim,
- embedding_name="target_language_embedding",
- embedding_size=word_vector_dim)
-
-beam_gen = beam_search(name="decoder",
- step=rnn_step,
- input=[StaticInput(encoder_last),
- generated_word_embedding],
- bos_id=0,
- eos_id=1,
- beam_size=5)
-Please see the following demo for more details:
-| Parameters: |
|
-
|---|---|
| Returns: | The generated word index. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.get_outputGet layer’s output by name. In PaddlePaddle, a layer might return multiple -values, but returns one layer’s output. If the user wants to use another -output besides the default one, please use get_output first to get -the output from input.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.mixedMixed Layer. A mixed layer will add all inputs together, then activate the sum. -Each input is a projection or operator.
-There are two styles of usages.
-with mixed(size=256) as m:
- m += full_matrix_projection(input=layer1)
- m += identity_projection(input=layer2)
-m = mixed(size=256,
- input=[full_matrix_projection(input=layer1),
- full_matrix_projection(input=layer2)])
-| Parameters: |
|
-
|---|---|
| Returns: | MixedLayerType object. - |
-
| Return type: | MixedLayerType - |
-
paddle.v2.layer.embeddingDefine a embedding Layer.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.scaling_projectionscaling_projection multiplies the input with a scalar parameter.
-The example usage is:
-proj = scaling_projection(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | ScalingProjection object. - |
-
| Return type: | ScalingProjection - |
-
paddle.v2.layer.dotmul_projectionDotMulProjection takes a layer as input and performs -element-wise multiplication with weight.
-where \(.*\) means element-wise multiplication.
-The example usage is:
-proj = dotmul_projection(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | DotMulProjection object. - |
-
| Return type: | DotMulProjection - |
-
paddle.v2.layer.dotmul_operatorDotMulOperator takes two inputs and performs element-wise multiplication:
-where \(.*\) means element-wise multiplication, and -scale is a config scalar, its default value is 1.
-The example usage is:
-op = dotmul_operator(a=layer1, b=layer2, scale=0.5)
-| Parameters: |
|
-
|---|---|
| Returns: | DotMulOperator object. - |
-
| Return type: | DotMulOperator - |
-
paddle.v2.layer.full_matrix_projectionFull Matrix Projection. It performs full matrix multiplication.
-There are two styles of usage.
-with mixed(size=100) as m:
- m += full_matrix_projection(input=layer)
-proj = full_matrix_projection(input=layer,
- size=100,
- param_attr=ParamAttr(name='_proj'))
-| Parameters: |
|
-
|---|---|
| Returns: | FullMatrixProjection Object. - |
-
| Return type: | FullMatrixProjection - |
-
paddle.v2.layer.identity_projectionThe example usage is:
-proj = identity_projection(input=layer)
-The example usage is:
-proj = identity_projection(input=layer,
- offset=10)
-Note that neither of the projections have trainable parameter.
-| Parameters: |
|
-
|---|---|
| Returns: | IdentityProjection or IdentityOffsetProjection object - |
-
| Return type: | IdentityProjection | IdentityOffsetProjection - |
-
paddle.v2.layer.slice_projectionslice_projection slices the input value into multiple parts, -then selects and merges some of them into a new output.
-The example usage is:
-proj = slice_projection(input=layer, slices=[(0, 10), (20, 30)])
-Note that slice_projection has no trainable parameter.
-| Parameters: |
|
-
|---|---|
| Returns: | SliceProjection object. - |
-
| Return type: | SliceProjection - |
-
paddle.v2.layer.table_projectionTable Projection. It selects rows from parameter where row_id -is in input_ids.
-where \(out\) is output, \(table\) is parameter, \(ids\) is input_ids, -and \(i\) is row_id.
-There are two styles of usage.
-with mixed(size=100) as m:
- m += table_projection(input=layer)
-proj = table_projection(input=layer,
- size=100,
- param_attr=ParamAttr(name='_proj'))
-| Parameters: |
|
-
|---|---|
| Returns: | TableProjection Object. - |
-
| Return type: | TableProjection - |
-
paddle.v2.layer.trans_full_matrix_projectionDifferent from full_matrix_projection, this projection performs matrix -multiplication, using the transpose of weight.
-\(w^\mathrm{T}\) means the transpose of weight. -The simply usage is:
-proj = trans_full_matrix_projection(input=layer,
- size=100,
- param_attr=ParamAttr(
- name='_proj',
- initial_mean=0.0,
- initial_std=0.01))
-| Parameters: |
|
-
|---|---|
| Returns: | TransposedFullMatrixProjection Object. - |
-
| Return type: | TransposedFullMatrixProjection - |
-
paddle.v2.layer.AggregateLevelPaddlePaddle supports three sequence types:
-SequenceType.NO_SEQUENCE means the sample is not a sequence.SequenceType.SEQUENCE means the sample is a sequence.SequenceType.SUB_SEQUENCE means the sample is a nested sequence,
-each timestep of which is also a sequence.Accordingly, AggregateLevel supports two modes:
-AggregateLevel.TO_NO_SEQUENCE means the aggregation acts on each
-timestep of a sequence, both SUB_SEQUENCE and SEQUENCE will
-be aggregated to NO_SEQUENCE.AggregateLevel.TO_SEQUENCE means the aggregation acts on each
-sequence of a nested sequence, SUB_SEQUENCE will be aggregated to
-SEQUENCE.paddle.v2.layer.poolingPooling layer for sequence inputs, not used for Image.
-If stride > 0, this layer slides a window whose size is determined by stride, -and returns the pooling value of the sequence in the window as the output. Thus, -a long sequence will be shortened. Note that for sequence with sub-sequence, the -default value of stride is -1.
-The example usage is:
-seq_pool = pooling(input=layer,
- pooling_type=AvgPooling(),
- agg_level=AggregateLevel.TO_NO_SEQUENCE)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.last_seqGet Last Timestamp Activation of a sequence.
-If stride > 0, this layer will slide a window whose size is determined by stride, -and return the last value of the sequence in the window as the output. Thus, a -long sequence will be shortened. Note that for sequence with sub-sequence, the -default value of stride is -1.
-The simple usage is:
-seq = last_seq(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.first_seqGet First Timestamp Activation of a sequence.
-If stride > 0, this layer will slide a window whose size is determined by stride, -and return the first value of the sequence in the window as the output. Thus, a -long sequence will be shortened. Note that for sequence with sub-sequence, the -default value of stride is -1.
-The simple usage is:
-seq = first_seq(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sub_seqsub_seq will return sub-sequences from the input sequences. For each -sequence in the input sequence layer, sub_seq will slice it by given -offset and size. Please notice that, number of offset value and size value -both are equal to the number of sequence in the input layer.
-sub_seq = sub_seq(input=input_seq, offsets=offsets, sizes=sizes)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.concatConcatenate all input vectors to one vector. -Inputs can be a list of paddle.v2.config_base.Layer or a list of projection.
-The example usage is:
-concat = concat(input=[layer1, layer2])
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.seq_concatConcatenate sequence a and sequence b.
-Output: [a1, ..., am, b1, ..., bn]
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The example usage is:
-concat = seq_concat(a=layer1, b=layer2)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.seq_sliceseq_slice will return one or several sub-sequences from the -input sequence layer given start and end indices.
----
-- If only start indices are given, and end indices are set to None, -this layer slices the input sequence from the given start indices -to its end.
-- If only end indices are given, and start indices are set to None, -this layer slices the input sequence from its beginning to the -given end indices.
-- If start and end indices are both given, they should have the same -number of elements.
-
If start or end indices contains more than one elements, the input sequence -will be sliced for multiple times.
-seq_silce = seq_slice(input=input_seq,
- starts=start_pos, ends=end_pos)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sub_nested_seqThe sub_nested_seq accepts two inputs: the first one is a nested -sequence; the second one is a set of selceted indices in the nested sequence.
-Then sub_nest_seq trims the first nested sequence input according -to the selected indices to form a new output. This layer is useful in -beam training.
-The example usage is:
-sub_nest_seq = sub_nested_seq(input=data, selected_indices=selected_ids)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.block_expandThe expanding method is the same with ExpandConvLayer, but saved the transposed -value. After expanding, output.sequenceStartPositions will store timeline. -The number of time steps is outputH * outputW and the dimension of each -time step is block_y * block_x * num_channels. This layer can be used after -convolutional neural network, and before recurrent neural network.
-The simple usage is:
-block_expand = block_expand(input=layer,
- num_channels=128,
- stride_x=1,
- stride_y=1,
- block_x=1,
- block_x=3)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.ExpandLevelPlease refer to AggregateLevel first.
-ExpandLevel supports two modes:
-ExpandLevel.FROM_NO_SEQUENCE means the expansion acts on
-NO_SEQUENCE, which will be expanded to
-SEQUENCE or SUB_SEQUENCE.ExpandLevel.FROM_SEQUENCE means the expansion acts on
-SEQUENCE, which will be expanded to
-SUB_SEQUENCE.paddle.v2.layer.expandA layer for expanding dense data or (sequence data where the length of each -sequence is one) to sequence data.
-The example usage is:
-expand = expand(input=layer1,
- expand_as=layer2,
- expand_level=ExpandLevel.FROM_NO_SEQUENCE)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.repeatA layer for repeating the input for num_repeats times.
-If as_row_vector:
-If not as_row_vector:
-The example usage is:
-expand = repeat(input=layer, num_repeats=4)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.rotateA layer for rotating 90 degrees (clock-wise) for each feature channel, -usually used when the input sample is some image or feature map.
-where \(x\) is (M x N x C) input, and \(y\) is (N x M x C) output.
-The example usage is:
-rot = rotate(input=layer,
- height=100,
- width=100)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.seq_reshapeA 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.
-Note that T*M/N must be an integer.
-The example usage is:
-reshape = seq_reshape(input=layer, reshape_size=4)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.addtoAddtoLayer.
-where \(y\) is output, \(x\) is input, \(b\) is bias, -and \(f\) is activation function.
-The example usage is:
-addto = addto(input=[layer1, layer2],
- act=paddle.v2.activation.Relu(),
- bias_attr=False)
-This layer just simply adds all input layers together, then activates the -sum. All inputs should share the same dimension, which is also the dimension -of this layer’s output.
-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.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.linear_combwhere \(0 \le i \le N-1\)
-Or in the matrix notation:
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The simple usage is:
-linear_comb = linear_comb(weights=weight, vectors=vectors,
- size=elem_dim)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.interpolationThis layer performs linear interpolation on two inputs, -which is used in NEURAL TURING MACHINE.
-where \(x_1\) and \(x_2\) are two (batchSize x dataDim) inputs, -\(w\) is (batchSize x 1) weight vector, and \(y\) is -(batchSize x dataDim) output.
-The example usage is:
-interpolation = interpolation(input=[layer1, layer2], weight=layer3)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.bilinear_interpThis layer implements bilinear interpolation on convolutional layer’s output.
-Please refer to Wikipedia: https://en.wikipedia.org/wiki/Bilinear_interpolation
-The simple usage is:
-bilinear = bilinear_interp(input=layer1, out_size_x=64, out_size_y=64)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.dropoutThe example usage is:
-dropout = dropout(input=input, dropout_rate=0.5)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.dot_prodA layer for computing the dot product of two vectors.
-The example usage is:
-dot_prod = dot_prod(input1=vec1, input2=vec2)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.out_prodA layer for computing the outer product of two vectors -The result is a matrix of size(input1) x size(input2)
-The example usage is:
-out_prod = out_prod(input1=vec1, input2=vec2)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.powerThis layer applies a power function to a vector element-wise, -which is used in NEURAL TURING MACHINE.
-where \(x\) is an input vector, \(w\) is a scalar exponent, -and \(y\) is an output vector.
-The example usage is:
-power = power(input=layer1, weight=layer2)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.scalingA layer for multiplying input vector by weight scalar.
-where \(x\) is size=dataDim input, \(w\) is size=1 weight, -and \(y\) is size=dataDim output.
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The example usage is:
-scale = scaling(input=layer1, weight=layer2)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.clipA layer for clipping the input value by the threshold.
-clip = clip(input=input, min=-10, max=10)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.resizeThe resize layer resizes the input matrix with a shape of [Height, Width] -into the output matrix with a shape of [Height x Width / size, size], -where size is the parameter of this layer indicating the output dimension.
-| Parameters: |
|
-
|---|---|
| Returns: | A paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.slope_interceptThis layer for applying a slope and an intercept to the input.
-The simple usage is:
-scale = slope_intercept(input=input, slope=-1.0, intercept=1.0)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.tensorThis layer performs tensor operation on two inputs. -For example:
-The simple usage is:
-tensor = tensor(a=layer1, b=layer2, size=1000)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cos_simCosine Similarity Layer. The cosine similarity equation is here.
-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.
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The example usage is:
-cos = cos_sim(a=layer1, b=layer2, size=3)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.l2_distanceThis layer calculates and returns the Euclidean distance between two input -vectors x and y. The equation is as follows:
-The output size of this layer is fixed to be 1. Note that the above -computation is for one sample. Multiple samples are processed in one batch.
-The example usage is:
-l2_sim = l2_distance(x=layer1, y=layer2)
-| Parameters: |
|
-
|---|---|
| Returns: | The returned paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.transA layer for transposing a minibatch matrix.
-where \(x\) is (M x N) input, and \(y\) is (N x M) output.
-The example usage is:
-trans = trans(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.scale_shiftA layer applies a linear transformation to each element in each row of -the input matrix. For each element, the layer first re-scales it and then -adds a bias to it.
-This layer is very like the SlopeInterceptLayer, except the scale and -bias are trainable.
-scale_shift = scale_shift(input=input, bias_attr=False)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.factorization_machineThe Factorization Machine models pairwise feature interactions as inner -product of the learned latent vectors corresponding to each input feature. -The Factorization Machine can effectively capture feature interactions -especially when the input is sparse.
-This implementation only consider the 2-order feature interactions using -Factorization Machine with the formula:
-Note
-X is the input vector with size n. V is the factor matrix. Each row of V -is the latent vector corresponding to each input dimesion. The size of -each latent vector is k.
-For details of Factorization Machine, please refer to the paper: -Factorization machines.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.max_idA layer for finding the id which has the maximal value for each sample. -The result is stored in output.ids.
-The example usage is:
-maxid = maxid(input=layer)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sampling_idA layer for sampling id from a multinomial distribution from the input layer. -Sampling one id for one sample.
-The simple usage is:
-samping_id = sampling_id(input=input)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.multiplexThis layer multiplex multiple layers according to the indexes, -which are provided by the first input layer. -inputs[0]: the indexes of the layers to form the output of size batchSize. -inputs[1:N]; the candidate output data. -For each index i from 0 to batchSize - 1, the i-th row of the output is the -the same to the i-th row of the (index[i] + 1)-th layer.
-For each i-th row of output: -.. math:
-y[i][j] = x_{x_{0}[i] + 1}[i][j], j = 0,1, ... , (x_{1}.width - 1)
-where, y is output. \(x_{k}\) is the k-th input layer and -\(k = x_{0}[i] + 1\).
-The example usage is:
-maxid = multiplex(input=layers)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cross_entropy_costA loss layer for multi class entropy.
-The example usage is:
-cost = cross_entropy(input=input,
- label=label)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cross_entropy_with_selfnorm_costA loss layer for multi class entropy with selfnorm. -Input should be a vector of positive numbers, without normalization.
-The example usage is:
-cost = cross_entropy_with_selfnorm(input=input,
- label=label)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.multi_binary_label_cross_entropy_costA loss layer for multi binary label cross entropy.
-The example usage is:
-cost = multi_binary_label_cross_entropy(input=input,
- label=label)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.huber_regression_costIn statistics, the Huber loss is a loss function used in robust regression, -that is less sensitive to outliers in data than the squared error loss. -Given a prediction f(x), a label y and \(\delta\), the loss function -is defined as:
-The example usage is:
-cost = huber_regression_cost(input=input, label=label)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer. - |
-
paddle.v2.layer.huber_classification_costFor classification purposes, a variant of the Huber loss called modified Huber -is sometimes used. Given a prediction f(x) (a real-valued classifier score) and -a true binary class label \(y\in \{-1, 1 \}\), the modified Huber -loss is defined as:
-The example usage is:
-cost = huber_classification_cost(input=input, label=label)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.lambda_costlambdaCost for lambdaRank LTR approach.
-The example usage is:
-cost = lambda_cost(input=input,
- score=score,
- NDCG_num=8,
- max_sort_size=-1)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.square_error_costsum of square error cost:
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.rank_costA cost Layer for learning to rank using gradient descent.
-The example usage is:
-cost = rank_cost(left=out_left,
- right=out_right,
- label=label)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sum_costA loss layer which calculates the sum of the input as loss.
-The example usage is:
-cost = sum_cost(input=input)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer. - |
-
paddle.v2.layer.crfA layer for calculating the cost of sequential conditional random -field model.
-The example usage is:
-crf = crf(input=input,
- label=label,
- size=label_dim)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.crf_decodingA layer for calculating the decoding sequence of sequential conditional -random field model. The decoding sequence is stored in output.ids. -If the input ‘label’ is provided, it is treated as the ground-truth label, and -this layer will also calculate error. output.value[i] is 1 for an incorrect -decoding and 0 for the correct.
-The example usage is:
-crf_decoding = crf_decoding(input=input,
- size=label_dim)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.ctcConnectionist Temporal Classification (CTC) is designed for temporal -classication task. e.g. sequence labeling problems where the -alignment between the inputs and the target labels is unknown.
-Note
-Considering the ‘blank’ label needed by CTC, you need to use (num_classes + 1) -as the size of the input, where num_classes is the category number. -And the ‘blank’ is the last category index. So the size of ‘input’ layer (e.g. -fc with softmax activation) should be (num_classes + 1). The size of -ctc should also be (num_classes + 1).
-The example usage is:
-ctc = ctc(input=input,
- label=label,
- size=9055,
- norm_by_times=True)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.warp_ctcA layer intergrating the open-source warp-ctc library, which is used in
-Deep Speech 2: End-toEnd Speech Recognition in English and Mandarin, to compute Connectionist Temporal
-Classification (CTC) loss. Besides, another warp-ctc repository, which is forked from
-the official one, is maintained to enable more compiling options. During the
-building process, PaddlePaddle will clone the source codes, build and
-install it to third_party/install/warpctc directory.
Note
-The example usage is:
-ctc = warp_ctc(input=input,
- label=label,
- size=1001,
- blank=1000,
- norm_by_times=False)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.nceNoise-contrastive estimation.
- -The example usage is:
-cost = nce(input=[layer1, layer2], label=layer2,
- param_attr=[attr1, attr2], weight=layer3,
- num_classes=3, neg_distribution=[0.1,0.3,0.6])
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.hsigmoidOrganize the classes into a binary tree. At each node, a sigmoid function -is used to calculate the probability of belonging to the right branch.
-The example usage is:
-cost = hsigmoid(input=[layer1, layer2],
- label=data)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.smooth_l1_costThis is a L1 loss but more smooth. It requires that the -sizes of input and label are equal. The formula is as follows,
-in which
-The example usage is:
-cost = smooth_l1_cost(input=input,
- label=label)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.multibox_lossCompute the location loss and the confidence loss for ssd.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.detection_outputApply the NMS to the output of network and compute the predict bounding -box location. The output’s shape of this layer could be zero if there is -no valid bounding box.
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.eosA layer for checking EOS for each sample: -- output_id = (input_id == conf.eos_id)
-The result is stored in output_.ids. -It is used by recurrent layer group.
-The example usage is:
-eos = eos(input=layer, eos_id=id)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.preluThe Parametric Relu activation that actives outputs with a learnable weight.
-The example usage is:
-prelu = prelu(input=layers, partial_sum=1)
-| Parameters: |
|
-
|---|---|
| Returns: | paddle.v2.config_base.Layer object. - |
-
| Return type: | paddle.v2.config_base.Layer - |
-
- The v2.networks module contains pieces of neural network that combine multiple layers.
-paddle.v2.networks.sequence_conv_pool(*args, **kwargs)Text convolution pooling group.
-Text input => Context Projection => FC Layer => Pooling => Output.
-| Parameters: |
|
-
|---|---|
| Returns: | layer’s output. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.text_conv_pool(*args, **kwargs)Text convolution pooling group.
-Text input => Context Projection => FC Layer => Pooling => Output.
-| Parameters: |
|
-
|---|---|
| Returns: | layer’s output. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.img_conv_bn_pool(*args, **kwargs)Convolution, batch normalization, pooling group.
-Img input => Conv => BN => Pooling => Output.
-| Parameters: |
|
-
|---|---|
| Returns: | layer’s output - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.img_conv_group(*args, **kwargs)Image Convolution Group, Used for vgg net.
-| Parameters: |
|
-
|---|---|
| Returns: | layer’s output - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.simple_img_conv_pool(*args, **kwargs)Simple image convolution and pooling group.
-Img input => Conv => Pooling => Output.
-| Parameters: |
|
-
|---|---|
| Returns: | layer’s output - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.vgg_16_network(input_image, num_channels, num_classes=1000)Same model from https://gist.github.com/ksimonyan/211839e770f7b538e2d8
-| Parameters: |
|
-
|---|---|
| Returns: | layer’s output - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.lstmemory_unit(*args, **kwargs)lstmemory_unit defines the caculation process of a LSTM unit during a -single time step. This function is not a recurrent layer, so it can not be -directly used to process sequence input. This function is always used in -recurrent_group (see layers.py for more details) to implement attention -mechanism.
-Please refer to Generating Sequences With Recurrent Neural Networks -for more details about LSTM. The link goes as follows: -.. _Link: https://arxiv.org/abs/1308.0850
-The example usage is:
-lstm_step = lstmemory_unit(input=[layer1],
- size=256,
- act=TanhActivation(),
- gate_act=SigmoidActivation(),
- state_act=TanhActivation())
-| Parameters: |
|
-
|---|---|
| Returns: | The lstmemory unit name. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.lstmemory_group(*args, **kwargs)lstm_group is a recurrent_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 and merely use its outputs, -it is recommended to use the lstmemory, which is relatively faster than -lstmemory_group.
-NOTE: In PaddlePaddle’s implementation, the following input-to-hidden -multiplications: -\(W_{x_i}x_{t}\) , \(W_{x_f}x_{t}\), -\(W_{x_c}x_t\), \(W_{x_o}x_{t}\) 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.
-The example usage is:
-lstm_step = lstmemory_group(input=[layer1],
- size=256,
- act=TanhActivation(),
- gate_act=SigmoidActivation(),
- state_act=TanhActivation())
-| Parameters: |
|
-
|---|---|
| Returns: | the lstmemory group. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.simple_lstm(*args, **kwargs)Simple LSTM Cell.
-It just combines a mixed layer with fully_matrix_projection and a lstmemory -layer. The simple lstm cell was implemented with follow equations.
-Please refer to Generating Sequences With Recurrent Neural Networks for more -details about lstm. Link is here.
-| Parameters: |
|
-
|---|---|
| Returns: | layer’s output. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.bidirectional_lstm(*args, **kwargs)A bidirectional_lstm is a recurrent unit that iterates over the input -sequence both in forward and backward 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.
-Please refer to Neural Machine Translation by Jointly Learning to Align -and Translate for more details about the bidirectional lstm. -The link goes as follows: -.. _Link: https://arxiv.org/pdf/1409.0473v3.pdf
-The example usage is:
-bi_lstm = bidirectional_lstm(input=[input1], size=512)
-| Parameters: |
|
-
|---|---|
| Returns: | LayerOutput object. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.gru_unit(*args, **kwargs)gru_unit defines the calculation process of a gated recurrent unit during a single -time step. This function is not a recurrent layer, so it can not be -directly used to process sequence input. This function is always used in -the recurrent_group (see layers.py for more details) to implement attention -mechanism.
-Please see grumemory in layers.py for the details about the maths.
-| Parameters: |
|
-
|---|---|
| Returns: | the gru output layer. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.gru_group(*args, **kwargs)gru_group is a recurrent_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 and merely use the outputs of a GRU, it is recommended -to use the grumemory, which is relatively faster.
-Please see grumemory in layers.py for more detail about the maths.
-The example usage is:
-gru = gru_group(input=[layer1],
- size=256,
- act=TanhActivation(),
- gate_act=SigmoidActivation())
-| Parameters: |
|
-
|---|---|
| Returns: | the gru group. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.simple_gru(*args, **kwargs)You may 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 \(W x_t\) 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.
-The computational speed is that, grumemory is relatively better than -gru_group, and gru_group is relatively better than simple_gru.
-The example usage is:
-gru = simple_gru(input=[layer1], size=256)
-| Parameters: |
|
-
|---|---|
| Returns: | the gru group. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.simple_gru2(*args, **kwargs)simple_gru2 is the same with simple_gru, but using grumemory instead. -Please refer to grumemory in layers.py for more detail about the math. -simple_gru2 is faster than simple_gru.
-The example usage is:
-gru = simple_gru2(input=[layer1], size=256)
-| Parameters: |
|
-
|---|---|
| Returns: | the gru group. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.bidirectional_gru(*args, **kwargs)A bidirectional_gru is a recurrent unit that iterates over the input -sequence both in forward and backward 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.
-The example usage is:
-bi_gru = bidirectional_gru(input=[input1], size=512)
-| Parameters: |
|
-
|---|---|
| Returns: | LayerOutput object. - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.simple_attention(*args, **kwargs)Calculate and return a context vector with attention mechanism. -Size of the context vector equals to size of the encoded_sequence.
-where \(h_{j}\) is the jth element of encoded_sequence, -\(U_{a}h_{j}\) is the jth element of encoded_proj -\(s_{i-1}\) is decoder_state -\(f\) is weight_act, and is set to tanh by default.
-Please refer to Neural Machine Translation by Jointly Learning to -Align and Translate for more details. The link is as follows: -https://arxiv.org/abs/1409.0473.
-The example usage is:
-context = simple_attention(encoded_sequence=enc_seq,
- encoded_proj=enc_proj,
- decoder_state=decoder_prev,)
-| Parameters: |
|
-
|---|---|
| Returns: | a context vector - |
-
| Return type: | LayerOutput - |
-
paddle.v2.networks.dot_product_attention(*args, **kwargs)Calculate and return a context vector with dot-product attention mechanism. -The dimension of the context vector equals to that of the attended_sequence.
-where \(h_{j}\) is the jth element of encoded_sequence, -\(z_{j}\) is the jth element of attended_sequence, -\(s_{i-1}\) is transformed_state.
-The example usage is:
-context = dot_product_attention(encoded_sequence=enc_seq,
- attended_sequence=att_seq,
- transformed_state=state,)
-| Parameters: |
|
-
|---|---|
| Returns: | The context vector. - |
-
| Return type: | LayerOutput - |
-
- paddle.v2.optimizer.Momentum(momentum=None, sparse=False, **kwargs)Momentum Optimizer.
-When sparse=False, the momentum update formula is as follows:
-where, \(k\) is momentum, \(\lambda\) is decay rate, -\(\gamma_t\) is learning rate at the t’th iteration. -\(w_{t}\) is the weight as the t’th iteration. -And the \(v_{t}\) is the history momentum variable.
-When sparse=True, the update scheme:
-where \(k\) is momentum, \(\lambda\) is decay rate, -\(\gamma_t\) is learning rate at the t’th iteration.
-| Parameters: |
|
-
|---|
paddle.v2.optimizer.Adam(beta1=0.9, beta2=0.999, epsilon=1e-08, **kwargs)Adam optimizer. -The details of please refer Adam: A Method for Stochastic Optimization
-| Parameters: |
|
-
|---|
paddle.v2.optimizer.Adamax(beta1=0.9, beta2=0.999, **kwargs)Adamax optimizer.
-The details of please refer this Adam: A Method for Stochastic Optimization
-| Parameters: |
|
-
|---|
paddle.v2.optimizer.AdaGrad(**kwargs)Adagrad(for ADAptive GRAdient algorithm) optimizer.
-For details please refer this Adaptive Subgradient Methods for -Online Learning and Stochastic Optimization.
-paddle.v2.optimizer.DecayedAdaGrad(rho=0.95, epsilon=1e-06, **kwargs)AdaGrad method with decayed sum gradients. The equations of this method -show as follow.
-| Parameters: |
|
-
|---|
paddle.v2.optimizer.AdaDelta(rho=0.95, epsilon=1e-06, **kwargs)AdaDelta method. The details of adadelta please refer to this -ADADELTA: AN ADAPTIVE LEARNING RATE METHOD.
-| Parameters: |
|
-
|---|
paddle.v2.optimizer.RMSProp(rho=0.95, epsilon=1e-06, **kwargs)RMSProp(for Root Mean Square Propagation) optimizer. For details please -refer this slide.
-The equations of this method as follows:
-| Parameters: |
|
-
|---|
- paddle.v2.pooling.BasePool(name)Base Pooling Type. -Note these pooling types are used for sequence input, not for images. -Each PoolingType contains one parameter:
-| Parameters: | name (basestring) – pooling layer type name used by paddle. | -
|---|
paddle.v2.pooling.Avg(strategy='average')Average pooling.
-Return the average values for each dimension in sequence or time steps.
-paddle.v2.pooling.Max(output_max_index=None)Max pooling.
-Return the very large values for each dimension in sequence or time steps.
-| Parameters: | output_max_index (bool|None) – True if output sequence max index instead of max -value. None means use default value in proto. | -
|---|
paddle.v2.pooling.SumSum pooling.
-Return the sum values of each dimension in sequence or time steps.
-paddle.v2.pooling.SquareRootNSquare Root Pooling.
-Return the square root values of each dimension in sequence or time steps.
-paddle.v2.pooling.CudnnMaxCudnn max pooling only support GPU. Return the maxinum value in the -pooling window.
-
-
- paddle.v2.data_type.dense_array(dim, seq_type=0)Dense Array. It means the input feature is dense array with float type. -For example, if the input is an image with 28*28 pixels, the input of -Paddle neural network could be a dense vector with dimension 784 or a -numpy array with shape (28, 28).
-For the 2-D convolution operation, each sample in one mini-batch must have -the similarly size in PaddlePaddle now. But, it supports variable-dimension -feature across mini-batch. For the variable-dimension, the param dim is not -used. While the data reader must yield numpy array and the data feeder will -set the data shape correctly.
-| Parameters: |
|
-
|---|---|
| Returns: | An input type object. - |
-
| Return type: | InputType - |
-
paddle.v2.data_type.dense_vector(dim, seq_type=0)Dense Array. It means the input feature is dense array with float type. -For example, if the input is an image with 28*28 pixels, the input of -Paddle neural network could be a dense vector with dimension 784 or a -numpy array with shape (28, 28).
-For the 2-D convolution operation, each sample in one mini-batch must have -the similarly size in PaddlePaddle now. But, it supports variable-dimension -feature across mini-batch. For the variable-dimension, the param dim is not -used. While the data reader must yield numpy array and the data feeder will -set the data shape correctly.
-| Parameters: |
|
-
|---|---|
| Returns: | An input type object. - |
-
| Return type: | InputType - |
-
paddle.v2.data_type.dense_vector_sequence(dim)Data type of a sequence of dense vector.
-| Parameters: | dim (int) – dimension of dense vector. | -
|---|---|
| Returns: | An input type object | -
| Return type: | InputType | -
paddle.v2.data_type.integer_value(value_range, seq_type=0)Data type of integer.
-| Parameters: |
|
-
|---|---|
| Returns: | An input type object - |
-
| Return type: | InputType - |
-
paddle.v2.data_type.integer_value_sequence(value_range)Data type of a sequence of integer.
-| Parameters: | value_range (int) – range of each element. | -
|---|
paddle.v2.data_type.sparse_binary_vector(dim, seq_type=0)Sparse binary vector. It means the input feature is a sparse vector and the -every element in this vector is either zero or one.
-| Parameters: |
|
-
|---|---|
| Returns: | An input type object. - |
-
| Return type: | InputType - |
-
paddle.v2.data_type.sparse_binary_vector_sequence(dim)| Parameters: | dim (int) – dimension of sparse vector. | -
|---|---|
| Returns: | An input type object | -
| Return type: | InputType | -
paddle.v2.data_type.sparse_float_vector(dim, seq_type=0)Sparse vector. It means the input feature is a sparse vector. Most of the -elements in this vector are zero, others could be any float value.
-| Parameters: |
|
-
|---|---|
| Returns: | An input type object. - |
-
| Return type: | InputType - |
-
paddle.v2.data_type.sparse_float_vector_sequence(dim)Data type of a sequence of sparse vector, which most elements are zero, -others could be any float value.
-| Parameters: | dim (int) – dimension of sparse vector. | -
|---|---|
| Returns: | An input type object | -
| Return type: | InputType | -
paddle.v2.data_type.sparse_non_value_slot(dim, seq_type=0)Sparse binary vector. It means the input feature is a sparse vector and the -every element in this vector is either zero or one.
-| Parameters: |
|
-
|---|---|
| Returns: | An input type object. - |
-
| Return type: | InputType - |
-
paddle.v2.data_type.sparse_value_slot(dim, seq_type=0)Sparse vector. It means the input feature is a sparse vector. Most of the -elements in this vector are zero, others could be any float value.
-| Parameters: |
|
-
|---|---|
| Returns: | An input type object. - |
-
| Return type: | InputType - |
-
paddle.v2.data_type.InputType(dim, seq_type, tp)InputType is the base class for paddle input types.
-Note
-this is a base class, and should never be used by user.
-| Parameters: |
|
-
|---|
At training and testing time, PaddlePaddle programs need to read data. To ease -the users’ work to write data reading code, we define that
-Indeed, data reader doesn’t have to be a function that reads and yields data
-items. It can be any function with no parameter that creates a iterable
-(anything can be used in for x in iterable):
iterable = data_reader()
-Element produced from the iterable should be a single entry of data, -not a mini batch. That entry of data could be a single item, or a tuple of -items. -Item should be of supported type (e.g., numpy 1d -array of float32, int, list of int)
-An example implementation for single item data reader creator:
-def reader_creator_random_image(width, height):
- def reader():
- while True:
- yield numpy.random.uniform(-1, 1, size=width*height)
-return reader
-An example implementation for multiple item data reader creator:
-def reader_creator_random_image_and_label(width, height, label):
- def reader():
- while True:
- yield numpy.random.uniform(-1, 1, size=width*height), label
-return reader
-TODO(yuyang18): Should we add whole design doc here?
-paddle.v2.reader.map_readers(func, *readers)Creates a data reader that outputs return value of function using -output of each data readers as arguments.
-| Parameters: |
|
-
|---|---|
| Type: | callable - |
-
| Returns: | the created data reader. - |
-
| Return type: | callable - |
-
paddle.v2.reader.buffered(reader, size)Creates a buffered data reader.
-The buffered data reader will read and save data entries into a -buffer. Reading from the buffered data reader will proceed as long -as the buffer is not empty.
-| Parameters: |
|
-
|---|---|
| Returns: | the buffered data reader. - |
-
paddle.v2.reader.compose(*readers, **kwargs)Creates a data reader whose output is the combination of input readers.
-If input readers output following data entries: -(1, 2) 3 (4, 5) -The composed reader will output: -(1, 2, 3, 4, 5)
-| Parameters: |
|
-
|---|---|
| Returns: | the new data reader. - |
-
| Raises: | ComposeNotAligned – outputs of readers are not aligned. -Will not raise when check_alignment is set to False. - |
-
paddle.v2.reader.chain(*readers)Creates a data reader whose output is the outputs of input data -readers chained together.
-If input readers output following data entries: -[0, 0, 0] -[1, 1, 1] -[2, 2, 2] -The chained reader will output: -[0, 0, 0, 1, 1, 1, 2, 2, 2]
-| Parameters: | readers – input readers. | -
|---|---|
| Returns: | the new data reader. | -
| Return type: | callable | -
paddle.v2.reader.shuffle(reader, buf_size)Creates a data reader whose data output is shuffled.
-Output from the iterator that created by original reader will be -buffered into shuffle buffer, and then shuffled. The size of shuffle buffer -is determined by argument buf_size.
-| Parameters: |
|
-
|---|---|
| Returns: | the new reader whose output is shuffled. - |
-
| Return type: | callable - |
-
paddle.v2.reader.firstn(reader, n)Limit the max number of samples that reader could return.
-| Parameters: |
|
-
|---|---|
| Returns: | the decorated reader. - |
-
| Return type: | callable - |
-
paddle.v2.reader.xmap_readers(mapper, reader, process_num, buffer_size, order=False)Use multiprocess to map samples from reader by a mapper defined by user. -And this function contains a buffered decorator. -:param mapper: a function to map sample. -:type mapper: callable -:param reader: the data reader to read from -:type reader: callable -:param process_num: process number to handle original sample -:type process_num: int -:param buffer_size: max buffer size -:type buffer_size: int -:param order: keep the order of reader -:type order: bool -:return: the decarated reader -:rtype: callable
-paddle.v2.reader.PipeReader(command, bufsize=8192, file_type='plain')PipeReader read data by stream from a command, take it’s -stdout into a pipe buffer and redirect it to the parser to -parse, then yield data as your desired format.
-You can using standard linux command or call another program -to read data, from HDFS, Ceph, URL, AWS S3 etc:
-An example:
-def example_reader():
- for f in myfiles:
- pr = PipeReader("cat %s"%f)
- for l in pr.get_line():
- sample = l.split(" ")
- yield sample
-get_line(cut_lines=True, line_break='\n')---
-- - - - param cut_lines: - cut buffer to lines -- type cut_lines: bool -- param line_break: - - line break of the file, like -
| type line_break: | |
|---|---|
| string | -|
| return: | one line or a buffer of bytes | -
| rtype: | string | -
Creator package contains some simple reader creator, which could -be used in user program.
-paddle.v2.reader.creator.np_array(x)Creates a reader that yields elements of x, if it is a -numpy vector. Or rows of x, if it is a numpy matrix. -Or any sub-hyperplane indexed by the highest dimension.
-| Parameters: | x – the numpy array to create reader from. | -
|---|---|
| Returns: | data reader created from x. | -
paddle.v2.reader.creator.text_file(path)Creates a data reader that outputs text line by line from given text file. -Trailing new line (‘\n’) of each line will be removed.
-| Path: | path of the text file. | -
|---|---|
| Returns: | data reader of text file | -
paddle.v2.reader.creator.cloud_reader(paths, etcd_endpoints, timeout_sec=5, buf_size=64)| Paths: | path of recordio files, can be a string or a string list. | -
|---|---|
| Etcd_endpoints: | the endpoints for etcd cluster | -
| Returns: | data reader of recordio files. | -
paddle.v2.minibatch.batch(reader, batch_size)Create a batched reader.
-| Parameters: |
|
-
|---|---|
| Returns: | the batched reader. - |
-
| Return type: | callable - |
-
- Dataset package.
-MNIST dataset.
-This module will download dataset from http://yann.lecun.com/exdb/mnist/ and -parse training set and test set into paddle reader creators.
-paddle.v2.dataset.mnist.train()MNIST training set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| Returns: | Training reader creator | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.mnist.test()MNIST test set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| Returns: | Test reader creator. | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.mnist.convert(path)Converts dataset to recordio format
-CIFAR dataset.
-This module will download dataset from -https://www.cs.toronto.edu/~kriz/cifar.html and parse train/test set into -paddle reader creators.
-The CIFAR-10 dataset consists of 60000 32x32 colour images in 10 classes, -with 6000 images per class. There are 50000 training images and 10000 test -images.
-The CIFAR-100 dataset is just like the CIFAR-10, except it has 100 classes -containing 600 images each. There are 500 training images and 100 testing -images per class.
-paddle.v2.dataset.cifar.train100()CIFAR-100 training set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 99].
-| Returns: | Training reader creator | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.cifar.test100()CIFAR-100 test set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| Returns: | Test reader creator. | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.cifar.train10()CIFAR-10 training set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| Returns: | Training reader creator | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.cifar.test10()CIFAR-10 test set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| Returns: | Test reader creator. | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.cifar.convert(path)Converts dataset to recordio format
-Conll05 dataset. -Paddle semantic role labeling Book and demo use this dataset as an example. -Because Conll05 is not free in public, the default downloaded URL is test set -of Conll05 (which is public). Users can change URL and MD5 to their Conll -dataset. And a pre-trained word vector model based on Wikipedia corpus is used -to initialize SRL model.
-paddle.v2.dataset.conll05.get_dict()Get the word, verb and label dictionary of Wikipedia corpus.
-paddle.v2.dataset.conll05.get_embedding()Get the trained word vector based on Wikipedia corpus.
-paddle.v2.dataset.conll05.test()Conll05 test set creator.
-Because the training dataset is not free, the test dataset is used for -training. It returns a reader creator, each sample in the reader is nine -features, including sentence sequence, predicate, predicate context, -predicate context flag and tagged sequence.
-| Returns: | Training reader creator | -
|---|---|
| Return type: | callable | -
IMDB dataset.
-This module downloads IMDB dataset from -http://ai.stanford.edu/%7Eamaas/data/sentiment/. This dataset contains a set -of 25,000 highly polar movie reviews for training, and 25,000 for testing. -Besides, this module also provides API for building dictionary.
-paddle.v2.dataset.imdb.build_dict(pattern, cutoff)Build a word dictionary from the corpus. Keys of the dictionary are words, -and values are zero-based IDs of these words.
-paddle.v2.dataset.imdb.train(word_idx)IMDB training set creator.
-It returns a reader creator, each sample in the reader is an zero-based ID -sequence and label in [0, 1].
-| Parameters: | word_idx (dict) – word dictionary | -
|---|---|
| Returns: | Training reader creator | -
| Return type: | callable | -
paddle.v2.dataset.imdb.test(word_idx)IMDB test set creator.
-It returns a reader creator, each sample in the reader is an zero-based ID -sequence and label in [0, 1].
-| Parameters: | word_idx (dict) – word dictionary | -
|---|---|
| Returns: | Test reader creator | -
| Return type: | callable | -
paddle.v2.dataset.imdb.convert(path)Converts dataset to recordio format
-imikolov’s simple dataset.
-This module will download dataset from -http://www.fit.vutbr.cz/~imikolov/rnnlm/ and parse training set and test set -into paddle reader creators.
-paddle.v2.dataset.imikolov.build_dict(min_word_freq=50)Build a word dictionary from the corpus, Keys of the dictionary are words, -and values are zero-based IDs of these words.
-paddle.v2.dataset.imikolov.train(word_idx, n, data_type=1)imikolov training set creator.
-It returns a reader creator, each sample in the reader is a word ID -tuple.
-| Parameters: |
|
-
|---|---|
| Returns: | Training reader creator - |
-
| Return type: | callable - |
-
paddle.v2.dataset.imikolov.test(word_idx, n, data_type=1)imikolov test set creator.
-It returns a reader creator, each sample in the reader is a word ID -tuple.
-| Parameters: |
|
-
|---|---|
| Returns: | Test reader creator - |
-
| Return type: | callable - |
-
paddle.v2.dataset.imikolov.convert(path)Converts dataset to recordio format
-Movielens 1-M dataset.
-Movielens 1-M dataset contains 1 million ratings from 6000 users on 4000 -movies, which was collected by GroupLens Research. This module will download -Movielens 1-M dataset from -http://files.grouplens.org/datasets/movielens/ml-1m.zip and parse training -set and test set into paddle reader creators.
-paddle.v2.dataset.movielens.get_movie_title_dict()Get movie title dictionary.
-paddle.v2.dataset.movielens.max_movie_id()Get the maximum value of movie id.
-paddle.v2.dataset.movielens.max_user_id()Get the maximum value of user id.
-paddle.v2.dataset.movielens.max_job_id()Get the maximum value of job id.
-paddle.v2.dataset.movielens.movie_categories()Get movie categoriges dictionary.
-paddle.v2.dataset.movielens.user_info()Get user info dictionary.
-paddle.v2.dataset.movielens.movie_info()Get movie info dictionary.
-paddle.v2.dataset.movielens.convert(path)Converts dataset to recordio format
-paddle.v2.dataset.movielens.MovieInfo(index, categories, title)Movie id, title and categories information are stored in MovieInfo.
-paddle.v2.dataset.movielens.UserInfo(index, gender, age, job_id)User id, gender, age, and job information are stored in UserInfo.
-The script fetch and preprocess movie_reviews data set that provided by NLTK
-TODO(yuyang18): Complete dataset.
-paddle.v2.dataset.sentiment.get_word_dict()Sorted the words by the frequency of words which occur in sample -:return:
---words_freq_sorted
paddle.v2.dataset.sentiment.train()Default training set reader creator
-paddle.v2.dataset.sentiment.test()Default test set reader creator
-paddle.v2.dataset.sentiment.convert(path)Converts dataset to recordio format
-UCI Housing dataset.
-This module will download dataset from -https://archive.ics.uci.edu/ml/machine-learning-databases/housing/ and -parse training set and test set into paddle reader creators.
-paddle.v2.dataset.uci_housing.train()UCI_HOUSING training set creator.
-It returns a reader creator, each sample in the reader is features after -normalization and price number.
-| Returns: | Training reader creator | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.uci_housing.test()UCI_HOUSING test set creator.
-It returns a reader creator, each sample in the reader is features after -normalization and price number.
-| Returns: | Test reader creator | -
|---|---|
| Return type: | callable | -
WMT14 dataset. -The original WMT14 dataset is too large and a small set of data for set is -provided. This module will download dataset from -http://paddlepaddle.cdn.bcebos.com/demo/wmt_shrinked_data/wmt14.tgz and -parse training set and test set into paddle reader creators.
-paddle.v2.dataset.wmt14.train(dict_size)WMT14 training set creator.
-It returns a reader creator, each sample in the reader is source language -word ID sequence, target language word ID sequence and next word ID -sequence.
-| Returns: | Training reader creator | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.wmt14.test(dict_size)WMT14 test set creator.
-It returns a reader creator, each sample in the reader is source language -word ID sequence, target language word ID sequence and next word ID -sequence.
-| Returns: | Test reader creator | -
|---|---|
| Return type: | callable | -
paddle.v2.dataset.wmt14.convert(path)Converts dataset to recordio format
-ACL2016 Multimodal Machine Translation. Please see this website for more -details: http://www.statmt.org/wmt16/multimodal-task.html#task1
-If you use the dataset created for your task, please cite the following paper: -Multi30K: Multilingual English-German Image Descriptions.
-}
-paddle.v2.dataset.wmt16.train(src_dict_size, trg_dict_size, src_lang='en')WMT16 train set reader.
-This function returns the reader for train data. Each sample the reader -returns is made up of three fields: the source language word index sequence, -target language word index sequence and next word index sequence.
-NOTE: -The original like for training data is: -http://www.quest.dcs.shef.ac.uk/wmt16_files_mmt/training.tar.gz
-paddle.dataset.wmt16 provides a tokenized version of the original dataset by -using moses’s tokenization script: -https://github.com/moses-smt/mosesdecoder/blob/master/scripts/tokenizer/tokenizer.perl
-| Parameters: |
|
-
|---|---|
| Returns: | The train reader. - |
-
| Return type: | callable - |
-
paddle.v2.dataset.wmt16.test(src_dict_size, trg_dict_size, src_lang='en')WMT16 test set reader.
-This function returns the reader for test data. Each sample the reader -returns is made up of three fields: the source language word index sequence, -target language word index sequence and next word index sequence.
-NOTE: -The original like for test data is: -http://www.quest.dcs.shef.ac.uk/wmt16_files_mmt/mmt16_task1_test.tar.gz
-paddle.dataset.wmt16 provides a tokenized version of the original dataset by -using moses’s tokenization script: -https://github.com/moses-smt/mosesdecoder/blob/master/scripts/tokenizer/tokenizer.perl
-| Parameters: |
|
-
|---|---|
| Returns: | The test reader. - |
-
| Return type: | callable - |
-
paddle.v2.dataset.wmt16.validation(src_dict_size, trg_dict_size, src_lang='en')WMT16 validation set reader.
-This function returns the reader for validation data. Each sample the reader -returns is made up of three fields: the source language word index sequence, -target language word index sequence and next word index sequence.
-NOTE: -The original like for validation data is: -http://www.quest.dcs.shef.ac.uk/wmt16_files_mmt/validation.tar.gz
-paddle.dataset.wmt16 provides a tokenized version of the original dataset by -using moses’s tokenization script: -https://github.com/moses-smt/mosesdecoder/blob/master/scripts/tokenizer/tokenizer.perl
-| Parameters: |
|
-
|---|---|
| Returns: | The validation reader. - |
-
| Return type: | callable - |
-
paddle.v2.dataset.wmt16.get_dict(lang, dict_size, reverse=False)return the word dictionary for the specified language.
-| Parameters: |
|
-
|---|---|
| Returns: | The word dictionary for the specific language. - |
-
| Return type: | dict - |
-
paddle.v2.dataset.wmt16.fetch()download the entire dataset.
-paddle.v2.dataset.wmt16.convert(path, src_dict_size, trg_dict_size, src_lang)Converts dataset to recordio format.
-
- This file contains some common interfaces for image preprocess. -Many users are confused about the image layout. We introduce -the image layout as follows.
-CHW Layout
-Color format: RGB or BGR
-OpenCV use BGR color format. PIL use RGB color format. Both -formats can be used for training. Noted that, the format should -be keep consistent between the training and inference peroid.
-paddle.v2.image.batch_images_from_tar(data_file, dataset_name, img2label, num_per_batch=1024)¶Read images from tar file and batch them into batch file.
-| Parameters: |
|
-
|---|---|
| Returns: | path of list file containing paths of batch file - |
-
| Return type: | string - |
-
paddle.v2.image.load_image_bytes(bytes, is_color=True)¶Load an color or gray image from bytes array.
-Example usage:
-with open('cat.jpg') as f:
- im = load_image_bytes(f.read())
-| Parameters: |
|
-
|---|
paddle.v2.image.load_image(file, is_color=True)¶Load an color or gray image from the file path.
-Example usage:
-im = load_image('cat.jpg')
-| Parameters: |
|
-
|---|
paddle.v2.image.resize_short(im, size)¶Resize an image so that the length of shorter edge is size.
-Example usage:
-im = load_image('cat.jpg')
-im = resize_short(im, 256)
-| Parameters: |
|
-
|---|
paddle.v2.image.to_chw(im, order=(2, 0, 1))¶Transpose the input image order. The image layout is HWC format -opened by cv2 or PIL. Transpose the input image to CHW layout -according the order (2,0,1).
-Example usage:
-im = load_image('cat.jpg')
-im = resize_short(im, 256)
-im = to_chw(im)
-| Parameters: |
|
-
|---|
paddle.v2.image.center_crop(im, size, is_color=True)¶Crop the center of image with size.
-Example usage:
-im = center_crop(im, 224)
-| Parameters: |
|
-
|---|
paddle.v2.image.random_crop(im, size, is_color=True)¶Randomly crop input image with size.
-Example usage:
-im = random_crop(im, 224)
-| Parameters: |
|
-
|---|
paddle.v2.image.left_right_flip(im, is_color=True)¶Flip an image along the horizontal direction. -Return the flipped image.
-Example usage:
-im = left_right_flip(im)
-| Parameters: |
|
-
|---|
paddle.v2.image.simple_transform(im, resize_size, crop_size, is_train, is_color=True, mean=None)¶Simply data argumentation for training. These operations include -resizing, croping and flipping.
-Example usage:
-im = simple_transform(im, 256, 224, True)
-| Parameters: |
|
-
|---|
paddle.v2.image.load_and_transform(filename, resize_size, crop_size, is_train, is_color=True, mean=None)¶Load image from the input file filename and transform image for -data argumentation. Please refer to the simple_transform interface -for the transform operations.
-Example usage:
-im = load_and_transform('cat.jpg', 256, 224, True)
-| Parameters: |
|
-
|---|
-
-
- paddle.v2.fluid.evaluator.Accuracy(input, label, k=1, **kwargs)Average Accuracy for multiple mini-batches.
-paddle.v2.fluid.evaluator.ChunkEvaluator(input, label, chunk_scheme, num_chunk_types, excluded_chunk_types=None)Accumulate counter numbers output by chunk_eval from mini-batches and -compute the precision recall and F1-score using the accumulated counter -numbers.
-
- paddle.v2.fluid.executor.Executor(places)paddle.v2.fluid.executor.global_scope()paddle.v2.fluid.executor.scope_guard(*args, **kwds)paddle.v2.fluid.executor.switch_scope(scope)
- paddle.v2.fluid.initializer.Constantalias of ConstantInitializer
paddle.v2.fluid.initializer.Uniformalias of UniformInitializer
paddle.v2.fluid.initializer.Normalalias of NormalInitializer
paddle.v2.fluid.initializer.Xavieralias of XavierInitializer
- paddle.v2.fluid.io.save_vars(executor, dirname, main_program=None, vars=None, predicate=None, save_file_name=None)Save variables to directory by executor.
-| Parameters: |
|
-
|---|
program which fit predicate. Default default_main_program. -:param predicate: The Predicate describes a callable that returns a variable -as a bool. If it returns true, the corresponding input variable will be saved. -:param vars: variables need to be saved. If vars is specified, program & predicate -will be ignored -:param save_file_name: The name of a single file that all vars are saved to. -If it is None, save variables to separate files.
-| Returns: | None | -
|---|
paddle.v2.fluid.io.save_params(executor, dirname, main_program=None, save_file_name=None)Save all parameters to directory with executor.
-paddle.v2.fluid.io.save_persistables(executor, dirname, main_program=None, save_file_name=None)Save all persistables to directory with executor.
-paddle.v2.fluid.io.load_vars(executor, dirname, main_program=None, vars=None, predicate=None, load_file_name=None)Load variables from directory by executor.
-| Parameters: |
|
-
|---|
program which fit predicate. Default default_main_program(). -:param predicate: The Predicate describes a callable that returns a variable -as a bool. If it returns true, the corresponding input variable will be loaded. -:param vars: variables need to be loaded. If vars is specified, program & -predicate will be ignored -:param load_file_name: The name of the single file that all vars are loaded from. -If it is None, load variables from separate files.
-| Returns: | None | -
|---|
paddle.v2.fluid.io.load_params(executor, dirname, main_program=None, load_file_name=None)load all parameters from directory by executor.
-paddle.v2.fluid.io.load_persistables(executor, dirname, main_program=None, load_file_name=None)load all persistables from directory by executor.
-paddle.v2.fluid.io.save_inference_model(dirname, feeded_var_names, target_vars, executor, main_program=None, save_file_name=None)Build a model especially for inference, -and save it to directory by the executor.
-| Parameters: |
|
-
|---|
If it is None, save parameters to separate files.
-| Returns: | None | -
|---|
paddle.v2.fluid.io.load_inference_model(dirname, executor, load_file_name=None)Load inference model from a directory
-| Parameters: |
|
-
|---|
If it is None, load parameters from separate files.
-| Returns: | [program, feed_target_names, fetch_targets] -program: program especially for inference. -feed_target_names: Names of variables that need to feed data -fetch_targets: Variables from which we can get inference results. | -
|---|
paddle.v2.fluid.io.get_inference_program(target_vars, main_program=None)
- paddle.v2.fluid.layers.split_lod_tensor(input, mask, level=0)split_lod_tensor
-This function takes in an input that contains the complete lod information, -and takes in a mask which is used to mask certain parts of the input. -The output is the true branch and the false branch with the mask applied to -the input at a certain level in the tensor.
-| Parameters: |
|
-
|---|---|
| Returns: | The true branch of tensor as per the mask applied to input. -Variable: The false branch of tensor as per the mask applied to input. - |
-
| Return type: | Variable - |
-
Examples
-x = layers.data(name='x', shape=[1])
-x.persistable = True
-
-y = layers.data(name='y', shape=[1])
-y.persistable = True
-
-out_true, out_false = layers.split_lod_tensor(
- input=x, mask=y, level=level)
-paddle.v2.fluid.layers.merge_lod_tensor(in_true, in_false, x, mask, level=0)merge_lod_tensor
-This function takes in an input \(x\), the True branch, the False -branch and a binary \(mask\). Using this information, this function -merges the True and False branches of the tensor into a single Output -at a certain lod level indiacted by \(level\).
-| Parameters: |
|
-
|---|---|
| Returns: | The merged output tensor. - |
-
| Return type: | Variable - |
-
Examples
-x = layers.data(
- name='x', shape=[1], dtype='float32', stop_gradient=False)
-y = layers.data(
- name='y', shape=[1], dtype='bool', stop_gradient=False)
-
-level = 0
-
-out_true, out_false = layers.split_lod_tensor(
- input=x, mask=y, level=level)
-out = layers.merge_lod_tensor(
- in_true=out_true, in_false=out_false, mask=y, x=x, level=level)
-paddle.v2.fluid.layers.BlockGuard(main_program)BlockGuard class.
-BlockGuard class is used to create a sub-block in a program by -using the Python with keyword.
-paddle.v2.fluid.layers.BlockGuardWithCompletion(rnn)BlockGuardWithCompletion class.
-BlockGuardWithCompletion class is used to create an op with a block in a program.
-paddle.v2.fluid.layers.StaticRNNMemoryLink(init, pre_mem, mem=None)StaticRNNMemoryLink class.
-| Parameters: |
|
-
|---|
StaticRNNMemoryLink class is used to create a link between two -memory cells of a StaticRNN.
-paddle.v2.fluid.layers.WhileGuard(while_op)paddle.v2.fluid.layers.While(cond, name=None)paddle.v2.fluid.layers.lod_rank_table(x, level=0)LoD Rank Table Operator. Given an input variable x and a level number -of LoD, this layer creates a LodRankTable object. A LoDRankTable object -contains a list of bi-element tuples. Each tuple consists of an index and -a length, both of which are int type. Refering to specified level of LoD, -the index is the sequence index number and the length representes the -sequence length. Please note that the list is ranked in descending order by -the length. The following is an example:
-----x is a LoDTensor: - x.lod = [[0, 2, 3], - [0, 5, 6, 7]] - x.data = [a, b, c, d, e, f, g] - -1. set level to 0: - Create lod rank table: - lod_rank_table_obj = lod_rank_table(x, level=0) - - Get: - lod_rank_table_obj.items() = [(0, 2), (1, 1)] - -2. set level to 1: - Create lod rank table: - lod_rank_table_obj = lod_rank_table(x, level=1) - - Get: - lod_rank_table_obj.items() = [(0, 5), (1, 1), (2, 1)] -
| Parameters: |
|
-
|---|---|
| Returns: | The created LoDRankTable object. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10],
- dtype='float32', lod_level=1)
-out = layers.lod_rank_table(x=x, level=0)
-paddle.v2.fluid.layers.max_sequence_len(rank_table)Max Sequence Len Operator. Given a LoDRankTable object, this layer -returns the max length of a batch of sequences. In fact, a LoDRankTable -object contains a list of tuples(<sequence index, sequence length>) and -the list is already sorted by sequence length in descending order, so the -operator just returns the sequence length of the first tuple element.
-| Parameters: | rank_table (Variable) – Input variable which is a LoDRankTable object. | -
|---|---|
| Returns: | The max length of sequence. | -
| Return type: | Variable | -
Examples
-x = fluid.layers.data(name='x', shape=[10],
- dtype='float32', lod_level=1)
-rank_table = layers.lod_rank_table(x=x, level=0)
-max_seq_len = layers.max_sequence_len(rank_table)
-paddle.v2.fluid.layers.topk(input, k)topk
-This function performs the operation that selects the k entries in the input -vector and outputs their values and indices as vectors. Thus topk_out[j] is -the j-th largest entry in input, and its index is topk_indices[j]
-| Parameters: |
|
-
|---|---|
| Returns: |
|
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10])
-k = 5
-array = fluid.layers.topk(x, k)
-paddle.v2.fluid.layers.lod_tensor_to_array(x, table)Convert a LOD_TENSOR to an LOD_TENSOR_ARRAY.
-| Parameters: |
|
-
|---|---|
| Returns: |
|
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10])
-table = fluid.layers.lod_rank_table(x, level=0)
-array = fluid.layers.lod_tensor_to_array(x, table)
-paddle.v2.fluid.layers.array_to_lod_tensor(x, table)Convert a LoD_Tensor_Aarry to an LoDTensor.
-| Parameters: |
|
-
|---|---|
| Returns: |
|
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10])
-table = fluid.layers.lod_rank_table(x, level=0)
-array = fluid.layers.lod_tensor_to_array(x, table)
-lod_tensor = fluid.layers.array_to_lod_tensor(array, table)
-paddle.v2.fluid.layers.increment(x, value=1.0, in_place=True)This function performs an operation that increments each value in the -input \(x\) by an amount: \(value\) as mentioned in the input -parameter. This operation is performed in-place by default.
-| Parameters: |
|
-
|---|---|
| Returns: |
|
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.data(name='data', shape=[32, 32], dtype='float32')
-data = fluid.layers.increment(x=data, value=3.0, in_place=True)
-paddle.v2.fluid.layers.array_write(x, i, array=None)This function writes the given input variable to the specified position -indicating by the arrary index to an output LOD_TENSOR_ARRAY. If the -output LOD_TENSOR_ARRAY is not given(None), a new one will be created and -returned.
-| Parameters: |
|
-
|---|---|
| Returns: | The output LOD_TENSOR_ARRAY where the input tensor is written. - |
-
| Return type: | Variable - |
-
Examples
-paddle.v2.fluid.layers.create_array(dtype)This function creates an array of type \(LOD_TENSOR_ARRAY\) using the -LayerHelper.
-| Parameters: | dtype (int|float) – The data type of the elements in the array. | -
|---|---|
| Returns: | The tensor variable storing the elements of data type. | -
| Return type: | Variable | -
Examples
-data = fluid.layers.create_array(dtype='float32')
-paddle.v2.fluid.layers.less_than(x, y, cond=None, **ignored)Less than
-This layer returns the truth value of \(x < y\) elementwise.
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the output of less_than. - |
-
| Return type: | Variable - |
-
Examples
-less = fluid.layers.less_than(x=label, y=limit)
-paddle.v2.fluid.layers.array_read(array, i)This function performs the operation to read the data in as an -LOD_TENSOR_ARRAY. -:param array: The input tensor that will be written to an array. -:type array: Variable|list -:param i: The subscript index in tensor array, that points the
---place where data will be written to.
| Returns: | The tensor type variable that has the data written to it. | -
|---|---|
| Return type: | Variable | -
Examples
-paddle.v2.fluid.layers.shrink_memory(x, i, table)This function creates an operator to shrink_rnn_memory using the RankTable -as mentioned in the input parameter.
-paddle.v2.fluid.layers.array_length(array)This function performs the operation to find the length of the input -LOD_TENSOR_ARRAY.
-| Parameters: | array (LOD_TENSOR_ARRAY) – The input array that will be used -to compute the length. | -
|---|---|
| Returns: | The length of the input LoDTensorArray. | -
| Return type: | Variable | -
Examples
-paddle.v2.fluid.layers.IfElse(cond, name=None)paddle.v2.fluid.layers.DynamicRNN(name=None)paddle.v2.fluid.layers.ConditionalBlock(inputs, is_scalar_condition=False, name=None)paddle.v2.fluid.layers.StaticRNN(name=None)StaticRNN class.
-StaticRNN class is used to create a StaticRNN. The RNN will have its -own parameters like inputs, outputs, memories, status and length.
-memory(init=None, shape=None, batch_ref=None, init_value=0.0, init_batch_dim_idx=0, ref_batch_dim_idx=1)| Parameters: |
|
-
|---|
paddle.v2.fluid.layers.reorder_lod_tensor_by_rank(x, rank_table)ReorderLoDTensorByRankTable operator.
-Input(X) is a batch of sequences. Input(RankTable) stores new orders of the -input sequence batch. The reorder_lod_tensor_by_rank operator reorders the -Input(X) according to the information provided by Input(RankTable).
-For example:
-If the indices stored in the Input(RankTable) are [3, 0, 2, 1], the -Input(X) will be reordered that the fourth sequence in Input(X) will become the -first one, and then followed by the original first, third, and the second one.
-This is: -X = [Seq0, Seq1, Seq2, Seq3]. The indices in RankTable are [3, 0, 2, 1]. -Out = [Seq3, Seq0, Seq2, Seq1] with a new LoD information.
-If the LoD information of Input(X) is empty, this means Input(X) is not sequence -data. This is also identical to a batch of sequences where each sequence has a -fixed length 1. In this case, the reorder_lod_tensor_by_rank operator reorders -each slice of Input(X) along the first axis according to Input(RankTable).
-This is: -X = [Slice0, Slice1, Slice2, Slice3] and its LoD information is empty. The -indices in RankTable are [3, 0, 2, 1]. -Out = [Slice3, Slice0, Slice2, Slice1] with no LoD information is appended.
-NOTE: This operator sorts Input(X) according to a given LoDRankTable which does -not need to be calculated according to Input(X). It can be calculated according -to another different sequence, and then this operator sorts Input(X) according -to the given LoDRankTable.
-| Parameters: |
|
-
|---|---|
| Returns: | (LoDTensor), the reordered lod tensor. - |
-
paddle.v2.fluid.layers.ParallelDo(places, name=None)ParallelDo class.
-ParallelDo class is used to create a ParallelDo.
-paddle.v2.fluid.layers.Print(input, first_n=-1, message=None, summarize=-1, print_tensor_name=True, print_tensor_type=True, print_tensor_shape=True, print_tensor_lod=True, print_phase='both')Print operator
-This creates a print op that will print when a tensor is accessed.
-Wraps the tensor passed in so that whenever that a tensor is accessed, -the message message is printed, along with the current value of the -tensor t.
-| Parameters: |
|
-
|---|---|
| Returns: | Output tensor, same data with input tensor. - |
-
| Return type: | Variable - |
-
Examples
-
-value = some_layer(...) -Print(value, summarize=10,
---message=”The content of some_layer: ”)
paddle.v2.fluid.layers.get_places(device_count=None, device_type=None)Returns a list of places based on flags. The list will be used for parallel -execution.
-| Parameters: |
|
-
|---|---|
| Returns: | vector of Place - |
-
paddle.v2.fluid.layers.data(name, shape, append_batch_size=True, dtype='float32', lod_level=0, type=VarType.LOD_TENSOR, stop_gradient=True)Data Layer
-This function takes in the input and based on whether data has -to be returned back as a minibatch, it creates the global variable by using -the helper functions. The global variables can be accessed by all the -following operators in the graph.
-All the input variables of this function are passed in as local variables -to the LayerHelper constructor.
-| Parameters: |
|
-
|---|---|
| Returns: | The global variable that gives access to the data. - |
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.data(name='x', shape=[784], dtype='float32')
-paddle.v2.fluid.layers.BlockGuardServ(server)BlockGuardServ class.
-BlockGuardServ class is used to create an op with a block in a program.
-paddle.v2.fluid.layers.ListenAndServ(endpoint, fan_in=1, optimizer_mode=True)ListenAndServ class.
-ListenAndServ class is used to wrap listen_and_serv op to create a server -which can receive variables from clients and run a block.
-paddle.v2.fluid.layers.Send(endpoints, send_vars, get_vars)Send layer
-| Parameters: |
|
-
|---|
Send variables to the server side, and get vars from server -side when server have finished running server side program.
-paddle.v2.fluid.layers.fc(input, size, num_flatten_dims=1, param_attr=None, bias_attr=None, act=None, name=None)Fully Connected Layer
-The fully connected layer can take multiple tensors as its inputs. It -creates a variable (one for each input tensor) called weights for each -input tensor, which represents a fully connected weight matrix from -each input unit to each output unit. The fully connected layer -multiplies each input tensor with its coresponding weight to produce -an output Tensor. If multiple input tensors are given, the results of -multiple multiplications will be sumed up. If bias_attr is not None, -a biases variable will be created and added to the output. Finally, -if activation is not None, it will be applied to the output as well.
-This process can be formulated as follows:
-In the above equation:
-| Parameters: |
|
-
|---|---|
| Returns: | The output tensor variable. - |
-
| Return type: | Variable - |
-
| Raises: |
|
-
Examples
-data = fluid.layers.data(name="data", shape=[32, 32], dtype="float32")
-fc = fluid.layers.fc(input=data, size=1000, act="tanh")
-paddle.v2.fluid.layers.embedding(input, size, is_sparse=False, padding_idx=None, param_attr=None, dtype='float32')Embedding Layer
-This layer is used to lookup embeddings of IDs, provided by input, in
-a lookup table. The result of this lookup is the embedding of each ID in the
-input.
All the input variables are passed in as local variables to the LayerHelper -constructor.
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the embeddings of the supplied inputs. - |
-
| Return type: | Variable - |
-
Examples
-dict_size = len(dataset.ids)
-data = fluid.layers.data(name='ids', shape=[32, 32], dtype='float32')
-fc = fluid.layers.embedding(input=data, size=[dict_size, 16])
-paddle.v2.fluid.layers.dynamic_lstm(input, size, param_attr=None, bias_attr=None, use_peepholes=True, is_reverse=False, gate_activation='sigmoid', cell_activation='tanh', candidate_activation='tanh', dtype='float32', name=None)Dynamic LSTM Layer
-The defalut implementation is diagonal/peephole connection -(https://arxiv.org/pdf/1402.1128.pdf), the formula is as follows:
-where the \(W\) terms denote weight matrices (e.g. \(W_{xi}\) is -the matrix of weights from the input gate to the input), \(W_{ic}, W_{fc}, W_{oc}\) are diagonal weight matrices for peephole connections. In -our implementation, we use vectors to reprenset these diagonal weight -matrices. The \(b\) terms denote bias vectors (\(b_i\) is the input -gate bias vector), \(\sigma\) is the non-linear activations, such as -logistic sigmoid function, and \(i, f, o\) and \(c\) are the input -gate, forget gate, output gate, and cell activation vectors, respectively, -all of which have the same size as the cell output activation vector \(h\).
-The \(\odot\) is the element-wise product of the vectors. \(act_g\) -and \(act_h\) are the cell input and cell output activation functions -and tanh is usually used for them. \(\tilde{c_t}\) is also called -candidate hidden state, which is computed based on the current input and -the previous hidden state.
-Set use_peepholes to False to disable peephole connection. The formula -is omitted here, please refer to the paper -http://www.bioinf.jku.at/publications/older/2604.pdf for details.
-Note that these \(W_{xi}x_{t}, W_{xf}x_{t}, W_{xc}x_{t}, W_{xo}x_{t}\) -operations on the input \(x_{t}\) are NOT included in this operator. -Users can choose to use fully-connect layer before LSTM layer.
-| Parameters: |
|
-
|---|---|
| Returns: | The hidden state, and cell state of LSTM. The shape of both is (T x D), and lod is the same with the input. - |
-
| Return type: | tuple - |
-
Examples
-hidden_dim = 512
-forward_proj = fluid.layers.fc(input=input_seq, size=hidden_dim * 4,
- act=None, bias_attr=None)
-forward, _ = fluid.layers.dynamic_lstm(
- input=forward_proj, size=hidden_dim * 4, use_peepholes=False)
-paddle.v2.fluid.layers.dynamic_lstmp(input, size, proj_size, param_attr=None, bias_attr=None, use_peepholes=True, is_reverse=False, gate_activation='sigmoid', cell_activation='tanh', candidate_activation='tanh', proj_activation='tanh', dtype='float32', name=None)Dynamic LSTMP Layer
-LSTMP (LSTM with recurrent projection) layer has a separate projection -layer after the LSTM layer, projecting the original hidden state to a -lower-dimensional one, which is proposed to reduce the number of total -parameters and furthermore computational complexity for the LSTM, -espeacially for the case that the size of output units is relative -large (https://research.google.com/pubs/archive/43905.pdf).
-The formula is as follows:
-In the above formula:
-Set use_peepholes to False to disable peephole connection. The formula -is omitted here, please refer to the paper -http://www.bioinf.jku.at/publications/older/2604.pdf for details.
-Note that these \(W_{xi}x_{t}, W_{xf}x_{t}, W_{xc}x_{t}, W_{xo}x_{t}\) -operations on the input \(x_{t}\) are NOT included in this operator. -Users can choose to use fully-connected layer before LSTMP layer.
-| Parameters: |
|
-
|---|---|
| Returns: | The projection of hidden state, and cell state of LSTMP. The shape of projection is (T x P), for the cell state which is (T x D), and both LoD is the same with the input. - |
-
| Return type: | tuple - |
-
Examples
-hidden_dim, proj_dim = 512, 256
-fc_out = fluid.layers.fc(input=input_seq, size=hidden_dim * 4,
- act=None, bias_attr=None)
-proj_out, _ = fluid.layers.dynamic_lstmp(input=fc_out,
- size=hidden_dim * 4,
- proj_size=proj_dim,
- use_peepholes=False,
- is_reverse=True,
- cell_activation="tanh",
- proj_activation="tanh")
-paddle.v2.fluid.layers.dynamic_gru(input, size, param_attr=None, bias_attr=None, is_reverse=False, gate_activation='sigmoid', candidate_activation='tanh', h_0=None)Dynamic GRU Layer
-Refer to Empirical Evaluation of Gated Recurrent Neural Networks on -Sequence Modeling
-The formula is as follows:
-The \(\odot\) is the element-wise product of the vectors. \(act_g\) -is the update gate and reset gate activation function and \(sigmoid\) -is usually used for it. \(act_c\) is the activation function for -candidate hidden state and \(tanh\) is usually used for it.
-Note that these \(W_{ux}x_{t}, W_{rx}x_{t}, W_{cx}x_{t}\) operations on -the input \(x_{t}\) are NOT included in this operator. Users can choose -to use fully-connect layer before GRU layer.
-| Parameters: |
|
-
|---|---|
| Returns: | The hidden state of GRU. The shape is \((T \times D)\), and lod is the same with the input. - |
-
| Return type: | Variable - |
-
Examples
-hidden_dim = 512
-x = fluid.layers.fc(input=data, size=hidden_dim * 3)
-hidden = fluid.layers.dynamic_gru(input=x, dim=hidden_dim)
-paddle.v2.fluid.layers.gru_unit(input, hidden, size, weight=None, bias=None, activation='tanh', gate_activation='sigmoid')GRU unit layer. The equation of a gru step is:
----\[ \begin{align}\begin{aligned}u_t & = actGate(xu_{t} + W_u h_{t-1} + b_u)\\r_t & = actGate(xr_{t} + W_r h_{t-1} + b_r)\\m_t & = actNode(xm_t + W_c dot(r_t, h_{t-1}) + b_m)\\h_t & = dot((1-u_t), m_t) + dot(u_t, h_{t-1})\end{aligned}\end{align} \]-
The inputs of gru unit includes \(z_t\), \(h_{t-1}\). In terms -of the equation above, the \(z_t\) is split into 3 parts - -\(xu_t\), \(xr_t\) and \(xm_t\). This means that in order to -implement a full GRU unit operator for an input, a fully -connected layer has to be applied, such that \(z_t = W_{fc}x_t\).
-The terms \(u_t\) and \(r_t\) represent the update and reset gates -of the GRU cell. Unlike LSTM, GRU has one lesser gate. However, there is -an intermediate candidate hidden output, which is denoted by \(m_t\). -This layer has three outputs \(h_t\), \(dot(r_t, h_{t-1})\) -and concatenation of \(u_t\), \(r_t\) and \(m_t\).
-| Parameters: |
|
-
|---|---|
| Returns: | The hidden value, reset-hidden value and gate values. - |
-
| Return type: | tuple - |
-
Examples
-# assuming we have x_t_data and prev_hidden of size=10
-x_t = fluid.layers.fc(input=x_t_data, size=30)
-hidden_val, r_h_val, gate_val = fluid.layers.gru_unit(input=x_t,
- hidden = prev_hidden)
-paddle.v2.fluid.layers.linear_chain_crf(input, label, param_attr=None)paddle.v2.fluid.layers.crf_decoding(input, param_attr, label=None)paddle.v2.fluid.layers.cos_sim(X, Y, **kwargs)This function performs the cosine similarity between two tensors -X and Y and returns that as the output.
-paddle.v2.fluid.layers.cross_entropy(input, label, **kwargs)Cross Entropy Layer
-This layer computes the cross entropy between input and label. It -supports both standard cross-entropy and soft-label cross-entropy loss -computation.
-soft_label = False, Label[i, 0] indicates the class index for sample i:
-soft_label = True, Label[i, j] indicates the soft label of class j -for sample i:
-Please make sure that in this case the summation of each row of label -equals one.
-As a special case of 2), when each row of ‘label’ has only one -non-zero element which is equal to 1, soft-label cross-entropy degenerates -to a one-hot cross-entropy with one-hot label representation.
-| Parameters: |
|
-
|---|---|
| Returns: | A 2-D tensor with shape [N x 1], the cross entropy loss. - |
-
| Raises: | ValueError – 1) the 1st dimension of input and label are not equal. -2) when soft_label == True, and the 2nd dimension of ---
|
-
Examples
-predict = fluid.layers.fc(input=net, size=classdim, act='softmax')
-cost = fluid.layers.cross_entropy(input=predict, label=label)
-paddle.v2.fluid.layers.square_error_cost(input, label, **kwargs)Square error cost layer
-This layer accepts input predictions and target label and returns the -squared error cost.
-For predictions, \(X\), and target labels, \(Y\), the equation is:
-In the above equation:
----
-- \(X\): Input predictions, a tensor.
-- \(Y\): Input labels, a tensor.
-- \(Out\): Output value, same shape with \(X\).
-
| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the element-wise squared error difference of input and label. - |
-
| Return type: | Variable - |
-
Examples
-y = layers.data(name='y', shape=[1], dtype='float32')
-y_predict = layers.data(name='y_predict', shape=[1], dtype='float32')
-cost = layers.square_error_cost(input=y_predict, label=y)
-paddle.v2.fluid.layers.accuracy(input, label, k=1, correct=None, total=None, **kwargs)This function computes the accuracy using the input and label. -The output is the top_k inputs and their indices.
-paddle.v2.fluid.layers.chunk_eval(input, label, chunk_scheme, num_chunk_types, excluded_chunk_types=None, **kwargs)This function computes and outputs the precision, recall and -F1-score of chunk detection.
-paddle.v2.fluid.layers.sequence_conv(input, num_filters, filter_size=3, filter_stride=1, padding=None, bias_attr=None, param_attr=None, act=None)This function creates the op for sequence_conv, using the inputs and -other convolutional configurations for the filters and stride as given -in the input parameters to the function.
-paddle.v2.fluid.layers.conv2d(input, num_filters, filter_size, stride=None, padding=None, groups=None, param_attr=None, bias_attr=None, use_cudnn=True, act=None)Convlution2D Layer
-The convolution2D layer calculates the output based on the input, filter -and strides, paddings, dilations, groups parameters. Input(Input) and -Output(Output) are in NCHW format. Where N is batch size, C is the number of -channels, H is the height of the feature, and W is the width of the feature. -The details of convolution layer, please refer UFLDL’s convolution, . -If bias attribution and activation type are provided, bias is added to the -output of the convolution, and the corresponding activation function is -applied to the final result.
-For each input \(X\), the equation is:
-In the above equation:
-Example
-Input:
-Input shape: $(N, C_{in}, H_{in}, W_{in})$
-Filter shape: $(C_{out}, C_{in}, H_f, W_f)$
-Output: -Output shape: $(N, C_{out}, H_{out}, W_{out})$
-Where
-H_{out}&= frac{(H_{in} + 2 * paddings[0] - (dilations[0] * (H_f - 1) + 1))}{strides[0]} + 1 \ -W_{out}&= frac{(W_{in} + 2 * paddings[1] - (dilations[1] * (W_f - 1) + 1))}{strides[1]} + 1
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the convolution and non-linearity activation result. - |
-
| Return type: | Variable - |
-
| Raises: |
|
-
Examples
-data = fluid.layers.data(
- name='data', shape=[3, 32, 32], dtype='float32')
-conv2d = fluid.layers.conv2d(
- input=data, num_filters=2, filter_size=3, act="relu")
-paddle.v2.fluid.layers.sequence_pool(input, pool_type, **kwargs)This function add the operator for sequence pooling. -It pools features of all time-steps of each instance, and is applied -on top of the input using pool_type mentioned in the parameters.
-It supports four pool_type:
-x is a 1-level LoDTensor:
- x.lod = [[0, 2, 5, 7]]
- x.data = [1, 3, 2, 4, 6, 5, 1]
- x.dims = [7, 1]
-
-then output is a Tensor:
- out.dim = [3, 1]
- with condition len(x.lod[-1]) - 1 == out.dims[0]
-
-for different pool_type:
- average: out.data = [2, 4, 3], where 2=(1+3)/2, 4=(2+4+6)/3, 3=(5+1)/2
- sum : out.data = [4, 12, 6], where 4=1+3, 12=2+4+6, 6=5+1
- sqrt : out.data = [2.82, 6.93, 4.24], where 2.82=(1+3)/sqrt(2),
- 6.93=(2+4+6)/sqrt(3), 4.24=(5+1)/sqrt(2)
- max : out.data = [3, 6, 5], where 3=max(1,3), 6=max(2,4,6), 5=max(5,1)
-| Parameters: |
|
-
|---|---|
| Returns: | The sequence pooling variable which is a Tensor. - |
-
Examples
-x = fluid.layers.data(name='x', shape=[7, 1],
- dtype='float32', lod_level=1)
-avg_x = fluid.layers.sequence_pool(input=x, pool_type='average')
-sum_x = fluid.layers.sequence_pool(input=x, pool_type='sum')
-sqrt_x = fluid.layers.sequence_pool(input=x, pool_type='sqrt')
-max_x = fluid.layers.sequence_pool(input=x, pool_type='max')
-paddle.v2.fluid.layers.pool2d(input, pool_size, pool_type, pool_stride=None, pool_padding=None, global_pooling=False, use_cudnn=True, name=None)This function adds the operator for pooling in 2 dimensions, using the -pooling configurations mentioned in input parameters.
-paddle.v2.fluid.layers.batch_norm(input, act=None, is_test=False, momentum=0.9, epsilon=1e-05, param_attr=None, bias_attr=None, data_layout='NCHW', name=None, moving_mean_name=None, moving_variance_name=None)This function helps create an operator to implement -the BatchNorm layer using the configurations from the input parameters.
-paddle.v2.fluid.layers.layer_norm(input, scale=True, shift=True, begin_norm_axis=1, epsilon=1e-05, param_attr=None, bias_attr=None, act=None, name=None)Layer Normalization
-Assume feature vectors exist on dimensions
-begin_norm_axis ... rank(input) and calculate the moment statistics
-along these dimensions for each feature vector \(a\) with size
-\(H\), then normalize each feature vector using the corresponding
-statistics. After that, apply learnable gain and bias on the normalized
-tensor to scale and shift if scale and shift are set.
Refer to Layer Normalization
-The formula is as follows:
-| Parameters: |
|
-
|---|---|
| Returns: | A tensor variable with the same shape as the input. - |
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.data(
- name='data', shape=[3, 32, 32], dtype='float32')
-x = fluid.layers.layer_norm(input=data, begin_norm_axis=1)
-paddle.v2.fluid.layers.beam_search_decode(ids, scores, name=None)paddle.v2.fluid.layers.conv2d_transpose(input, num_filters, output_size=None, filter_size=None, padding=None, stride=None, dilation=None, param_attr=None, use_cudnn=True, name=None)Convlution2D transpose layer
-The convolution2D transpose layer calculates the output based on the input, -filter, and dilations, strides, paddings. Input(Input) and output(Output) -are in NCHW format. Where N is batch size, C is the number of channels, -H is the height of the feature, and W is the width of the feature. -Parameters(dilations, strides, paddings) are two elements. These two elements -represent height and width, respectively. The details of convolution transpose -layer, please refer to the following explanation and references -therein.
-For each input \(X\), the equation is:
-In the above equation:
-Example
-Input:
-Input shape: $(N, C_{in}, H_{in}, W_{in})$
-Filter shape: $(C_{in}, C_{out}, H_f, W_f)$
-Output:
-Output shape: $(N, C_{out}, H_{out}, W_{out})$
-Where
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the convolution transpose result. - |
-
| Return type: | Variable - |
-
| Raises: |
|
-
Examples
-data = fluid.layers.data(
- name='data', shape=[3, 32, 32], dtype='float32')
-conv2d_transpose = fluid.layers.conv2d_transpose(
- input=data, num_filters=2, filter_size=3)
-paddle.v2.fluid.layers.sequence_expand(x, y, name=None)Sequence Expand Layer. This layer will expand the input variable x -according to LoD information of y. And the following examples will -explain how sequence_expand works:
-* Case 1
- x is a LoDTensor:
- x.lod = [[0, 2, 3],
- [0, 1, 3, 4]]
- x.data = [a, b, c, d]
- x.dims = [4, 1]
-
- y is a LoDTensor:
- y.lod = [[0, 2, 4],
- [0, 3, 6, 7, 8]]
-
- with condition len(y.lod[-1]) - 1 == x.dims[0]
-
- then output is a 2-level LoDTensor:
- out.lod = [[0, 2, 4],
- [0, 3, 6, 7, 8]]
- out.data = [a, a, a, b, b, b, c, d]
- out.dims = [8, 1]
-
-* Case 2
- x is a Tensor:
- x.data = [a, b, c]
- x.dims = [3, 1]
-
- y is a LoDTensor:
- y.lod = [[0, 2, 3, 6]]
-
- with condition len(y.lod[-1]) - 1 == x.dims[0]
-
- then output is a 1-level LoDTensor:
- out.lod = [[0, 2, 3, 6]]
- out.data = [a, a, b, c, c, c]
- out.dims = [6, 1]
-| Parameters: |
|
-
|---|---|
| Returns: | The expanded variable which is a LoDTensor. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10], dtype='float32')
-y = fluid.layers.data(name='y', shape=[10, 20],
- dtype='float32', lod_level=1)
-out = layers.sequence_expand(x=x, y=y)
-paddle.v2.fluid.layers.lstm_unit(x_t, hidden_t_prev, cell_t_prev, forget_bias=0.0, param_attr=None, bias_attr=None, name=None)Lstm unit layer. The equation of a lstm step is:
----\[ \begin{align}\begin{aligned}i_t & = \sigma(W_{x_i}x_{t} + W_{h_i}h_{t-1} + b_i)\\f_t & = \sigma(W_{x_f}x_{t} + W_{h_f}h_{t-1} + b_f)\\c_t & = f_tc_{t-1} + i_t tanh (W_{x_c}x_t + W_{h_c}h_{t-1} + b_c)\\o_t & = \sigma(W_{x_o}x_{t} + W_{h_o}h_{t-1} + b_o)\\h_t & = o_t tanh(c_t)\end{aligned}\end{align} \]-
The inputs of lstm unit include \(x_t\), \(h_{t-1}\) and -\(c_{t-1}\). The 2nd dimensions of \(h_{t-1}\) and \(c_{t-1}\) -should be same. The implementation separates the linear transformation and -non-linear transformation apart. Here, we take \(i_t\) as an example. -The linear transformation is applied by calling a fc layer and the -equation is:
----\[L_{i_t} = W_{x_i}x_{t} + W_{h_i}h_{t-1} + b_i\]-
The non-linear transformation is applied by calling lstm_unit_op and the -equation is:
----\[i_t = \sigma(L_{i_t})\]-
This layer has two outputs including \(h_t\) and \(o_t\).
-| Parameters: |
|
-
|---|---|
| Returns: | The hidden value and cell value of lstm unit. - |
-
| Return type: | tuple - |
-
| Raises: |
|
-
Examples
-x_t = fluid.layers.fc(input=x_t_data, size=10)
-prev_hidden = fluid.layers.fc(input=prev_hidden_data, size=30)
-prev_cell = fluid.layers.fc(input=prev_cell_data, size=30)
-hidden_value, cell_value = fluid.layers.lstm_unit(x_t=x_t,
- hidden_t_prev=prev_hidden,
- cell_t_prev=prev_cell)
-paddle.v2.fluid.layers.reduce_sum(input, dim=None, keep_dim=False, name=None)Computes the sum of tensor elements over the given dimension.
-| Parameters: |
|
-
|---|---|
| Returns: | The reduced Tensor variable. - |
-
| Return type: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_sum(x) # [3.5]
-fluid.layers.reduce_sum(x, dim=0) # [0.3, 0.5, 1.1, 1.6]
-fluid.layers.reduce_sum(x, dim=-1) # [1.9, 1.6]
-fluid.layers.reduce_sum(x, dim=1, keep_dim=True) # [[1.9], [1.6]]
-paddle.v2.fluid.layers.reduce_mean(input, dim=None, keep_dim=False, name=None)Computes the mean of tensor elements over the given dimension.
-| Parameters: |
|
-
|---|---|
| Returns: | The reduced Tensor variable. - |
-
| Return type: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_mean(x) # [0.4375]
-fluid.layers.reduce_mean(x, dim=0) # [0.15, 0.25, 0.55, 0.8]
-fluid.layers.reduce_mean(x, dim=-1) # [0.475, 0.4]
-fluid.layers.reduce_mean(x, dim=1, keep_dim=True) # [[0.475], [0.4]]
-paddle.v2.fluid.layers.reduce_max(input, dim=None, keep_dim=False, name=None)Computes the maximum of tensor elements over the given dimension.
-| Parameters: |
|
-
|---|---|
| Returns: | The reduced Tensor variable. - |
-
| Return type: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_max(x) # [0.9]
-fluid.layers.reduce_max(x, dim=0) # [0.2, 0.3, 0.6, 0.9]
-fluid.layers.reduce_max(x, dim=-1) # [0.9, 0.7]
-fluid.layers.reduce_max(x, dim=1, keep_dim=True) # [[0.9], [0.7]]
-paddle.v2.fluid.layers.reduce_min(input, dim=None, keep_dim=False, name=None)Computes the minimum of tensor elements over the given dimension.
-| Parameters: |
|
-
|---|---|
| Returns: | The reduced Tensor variable. - |
-
| Return type: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_min(x) # [0.1]
-fluid.layers.reduce_min(x, dim=0) # [0.1, 0.2, 0.5, 0.7]
-fluid.layers.reduce_min(x, dim=-1) # [0.2, 0.1]
-fluid.layers.reduce_min(x, dim=1, keep_dim=True) # [[0.2], [0.1]]
-paddle.v2.fluid.layers.sequence_first_step(input, **kwargs)This funciton get the first step of sequence.
-x is a 1-level LoDTensor:
- x.lod = [[0, 2, 5, 7]]
- x.data = [1, 3, 2, 4, 6, 5, 1]
- x.dims = [7, 1]
-
-then output is a Tensor:
- out.dim = [3, 1]
- with condition len(x.lod[-1]) - 1 == out.dims[0]
- out.data = [1, 2, 5], where 1=first(1,3), 2=first(2,4,6), 5=first(5,1)
-| Parameters: | input (variable) – The input variable which is a LoDTensor. | -
|---|---|
| Returns: | The sequence’s first step variable which is a Tensor. | -
Examples
-x = fluid.layers.data(name='x', shape=[7, 1],
- dtype='float32', lod_level=1)
-x_first_step = fluid.layers.sequence_first_step(input=x)
-paddle.v2.fluid.layers.sequence_last_step(input, **kwargs)This funciton get the last step of sequence.
-x is a 1-level LoDTensor:
- x.lod = [[0, 2, 5, 7]]
- x.data = [1, 3, 2, 4, 6, 5, 1]
- x.dims = [7, 1]
-
-then output is a Tensor:
- out.dim = [3, 1]
- with condition len(x.lod[-1]) - 1 == out.dims[0]
- out.data = [3, 6, 1], where 3=last(1,3), 6=last(2,4,6), 1=last(5,1)
-| Parameters: | input (variable) – The input variable which is a LoDTensor. | -
|---|---|
| Returns: | The sequence’s last step variable which is a Tensor. | -
Examples
-x = fluid.layers.data(name='x', shape=[7, 1],
- dtype='float32', lod_level=1)
-x_last_step = fluid.layers.sequence_last_step(input=x)
-paddle.v2.fluid.layers.dropout(x, dropout_prob, is_test=False, seed=None, **kwargs)Computes dropout.
-Drop or keep each element of x independently. Dropout is a regularization -technique for reducing overfitting by preventing neuron co-adaption during -training. The dropout operator randomly set (according to the given dropout -probability) the outputs of some units to zero, while others are remain -unchanged.
-| Parameters: |
|
-
|---|---|
| Returns: | A tensor variable. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name="data", shape=[32, 32], dtype="float32")
-droped = fluid.layers.dropout(input=x, dropout_rate=0.5)
-paddle.v2.fluid.layers.split(input, num_or_sections, dim=-1, name=None)Split the input tensor into multiple sub-tensors.
-| Parameters: |
|
-
|---|---|
| Returns: | The list of segmented tensor variables. - |
-
| Return type: | List - |
-
Examples
-# x is a Tensor variable with shape [3, 9, 5]:
-x0, x1, x2 = fluid.layers.split(x, num_or_sections=3, dim=1)
-x0.shape # [3, 3, 5]
-x1.shape # [3, 3, 5]
-x2.shape # [3, 3, 5]
-x0, x1, x2 = fluid.layers.split(x, num_or_sections=[2, 3, 4], dim=1)
-x0.shape # [3, 2, 5]
-x1.shape # [3, 3, 5]
-x2.shape # [3, 4, 5]
-paddle.v2.fluid.layers.ctc_greedy_decoder(input, blank, name=None)This op is used to decode sequences by greedy policy by below steps: -1. Get the indexes of max value for each row in input. a.k.a.
---numpy.argmax(input, axis=0).
A simple example as below:
-Given:
-
-input.data = [[0.6, 0.1, 0.3, 0.1],
- [0.3, 0.2, 0.4, 0.1],
- [0.1, 0.5, 0.1, 0.3],
- [0.5, 0.1, 0.3, 0.1],
-
- [0.5, 0.1, 0.3, 0.1],
- [0.2, 0.2, 0.2, 0.4],
- [0.2, 0.2, 0.1, 0.5],
- [0.5, 0.1, 0.3, 0.1]]
-
-input.lod = [[0, 4, 8]]
-
-Then:
-
-output.data = [[2],
- [1],
- [3]]
-
-output.lod = [[0, 2, 3]]
-| Parameters: |
|
-
|---|---|
| Returns: | CTC greedy decode result. If all the sequences in result were -empty, the result LoDTensor will be [-1] with LoD [[0]] and dims [1, 1]. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[8], dtype='float32')
-
-cost = fluid.layers.ctc_greedy_decoder(input=x, blank=0)
-paddle.v2.fluid.layers.edit_distance(input, label, normalized=False, ignored_tokens=None, name=None)EditDistance operator computes the edit distances between a batch of -hypothesis strings and their references. Edit distance, also called -Levenshtein distance, measures how dissimilar two strings are by counting -the minimum number of operations to transform one string into anthor. -Here the operations include insertion, deletion, and substitution.
-For example, given hypothesis string A = “kitten” and reference -B = “sitting”, the edit distance is 3 for A will be transformed into B -at least after two substitutions and one insertion:
-“kitten” -> “sitten” -> “sittin” -> “sitting”
-Input(Hyps) is a LoDTensor consisting of all the hypothesis strings with -the total number denoted by batch_size, and the separation is specified -by the LoD information. And the batch_size reference strings are arranged -in order in the same way in the LoDTensor Input(Refs).
-Output(Out) contains the batch_size results and each stands for the edit -distance for a pair of strings respectively. If Attr(normalized) is true, -the edit distance will be divided by the length of reference string.
-| Parameters: |
|
-
|---|---|
| Returns: | sequence-to-sequence edit distance in shape [batch_size, 1]. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[8], dtype='float32')
-y = fluid.layers.data(name='y', shape=[7], dtype='float32')
-
-cost = fluid.layers.edit_distance(input=x,label=y)
-paddle.v2.fluid.layers.l2_normalize(x, axis, epsilon=1e-12, name=None)L2 normalize Layer
-The l2 normalize layer normalizes x along dimension axis using an L2 -norm. For a 1-D tensor (dim is fixed to 0), this layer computes
-output = x / sqrt(max(sum(x**2), epsilon))
-For x with more dimensions, this layer independently normalizes each 1-D -slice along dimension axis.
-| Parameters: |
|
-
|---|---|
| Returns: | The output tensor variable. - |
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.data(name="data",
- shape=(3, 17, 13),
- dtype="float32")
-normed = fluid.layers.l2_normalize(x=data, axis=1)
-paddle.v2.fluid.layers.matmul(x, y, transpose_x=False, transpose_y=False, name=None)Applies matrix multiplication to two tensors.
-Currently, the input tensors’ rank can be any, but when the rank of any -inputs is bigger than 3, this two inputs’ rank should be equal.
-The actual behavior depends on the shapes of \(x\), \(y\) and the
-flag values of transpose_x, transpose_y. Specifically:
Also note that if the raw tensor \(x\) or \(y\) is rank-1 and -nontransposed, the prepended or appended dimension \(1\) will be -removed after matrix multiplication.
-| Parameters: |
|
-
|---|---|
| Returns: | The product Tensor variable. - |
-
| Return type: | Variable - |
-
Examples
-# Examples to clarify shapes of the inputs and output
-# x: [B, ..., M, K], y: [B, ..., K, N]
-fluid.layers.matmul(x, y) # out: [B, ..., M, N]
-
-# x: [B, M, K], y: [B, K, N]
-fluid.layers.matmul(x, y) # out: [B, M, N]
-
-# x: [B, M, K], y: [K, N]
-fluid.layers.matmul(x, y) # out: [B, M, N]
-
-# x: [M, K], y: [K, N]
-fluid.layers.matmul(x, y) # out: [M, N]
-
-# x: [B, M, K], y: [K]
-fluid.layers.matmul(x, y) # out: [B, M]
-
-# x: [K], y: [K]
-fluid.layers.matmul(x, y) # out: [1]
-
-# x: [M], y: [N]
-fluid.layers.matmul(x, y, True, True) # out: [M, N]
-paddle.v2.fluid.layers.warpctc(input, label, blank=0, norm_by_times=False, **kwargs)An operator integrating the open source Warp-CTC library -(https://github.com/baidu-research/warp-ctc) -to compute Connectionist Temporal Classification (CTC) loss. -It can be aliased as softmax with CTC, since a native softmax activation is -interated to the Warp-CTC library, to to normlize values for each row of the -input tensor.
-| Parameters: |
|
-
|---|---|
| Returns: | The Connectionist Temporal Classification (CTC) loss, -which is a 2-D Tensor of the shape [batch_size, 1]. - |
-
| Return type: | Variable - |
-
Examples
-paddle.v2.fluid.layers.sequence_reshape(input, new_dim)Sequence Reshape Layer
-This layer will rearrange the input sequences. The new dimension is set by -user. Length of each sequence is computed according to original length, -original dimension and new dimension. The following example will help to -illustrate the function of this layer:
-x is a LoDTensor:
- x.lod = [[0, 2, 6]]
- x.data = [[1, 2], [3, 4],
- [5, 6], [7, 8], [9, 10], [11, 12]]
- x.dims = [6, 2]
-
-set new_dim = 4
-
-then out is a LoDTensor:
- out.lod = [[0, 1, 3]]
- out.data = [[1, 2, 3, 4],
- [5, 6, 7, 8], [9, 10, 11, 12]]
- out.dims = [3, 4]
-Currently, only 1-level LoDTensor is supported and please make sure -(original length * original dimension) can be divided by new dimension with -no remainder for each sequence.
-| Parameters: |
|
-
|---|---|
| Returns: | Reshaped LoDTensor according to new dimension. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[5, 20],
- dtype='float32', lod_level=1)
-x_reshaped = layers.sequence_reshape(input=x, new_dim=10)
-paddle.v2.fluid.layers.transpose(x, perm, name=None)transpose Layer
-Permute the dimensions of input according to perm.
-The i-th dimension of the returned tensor will correspond to the -perm[i]-th dimension of input.
-| Parameters: |
|
-
|---|---|
| Returns: | A transposed Tensor. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[5, 10, 15], dtype='float32')
-x_transposed = layers.transpose(x, perm=[1, 0, 2])
-paddle.v2.fluid.layers.im2sequence(input, filter_size=1, stride=1, padding=0, name=None)Extracts image patches from the input tensor to form a tensor of shape -{input.batch_size * output_height * output_width, filter_size_H * -filter_size_W * input.channels} which is similar with im2col. -This op use filter / kernel to scan images and convert these images to -sequences. After expanding, the number of time step are -output_height * output_width for an image, in which output_height and -output_width are calculated by below equation:
-And the dimension of each time step is block_y * block_x * input.channels.
-| Parameters: |
|
-
|---|---|
| Returns: | The output is a LoDTensor with shape -{input.batch_size * output_height * output_width, -filter_size_H * filter_size_W * input.channels}. -If we regard output as a matrix, each row of this matrix is -a step of a sequence. - |
-
| Return type: | output - |
-
Examples:
-As an example:
-----Given: - -x = [[[[ 6. 2. 1.] - [ 8. 3. 5.] - [ 0. 2. 6.]] - - [[ 2. 4. 4.] - [ 6. 3. 0.] - [ 6. 4. 7.]]] - - [[[ 6. 7. 1.] - [ 5. 7. 9.] - [ 2. 4. 8.]] - - [[ 1. 2. 1.] - [ 1. 3. 5.] - [ 9. 0. 8.]]]] - -x.dims = {2, 2, 3, 3} - -And: - -filter = [2, 2] -stride = [1, 1] -padding = [0, 0] - -Then: - -output.data = [[ 6. 2. 8. 3. 2. 4. 6. 3.] - [ 2. 1. 3. 5. 4. 4. 3. 0.] - [ 8. 3. 0. 2. 6. 3. 6. 4.] - [ 3. 5. 2. 6. 3. 0. 4. 7.] - [ 6. 7. 5. 7. 1. 2. 1. 3.] - [ 7. 1. 7. 9. 2. 1. 3. 5.] - [ 5. 7. 2. 4. 1. 3. 9. 0.] - [ 7. 9. 4. 8. 3. 5. 0. 8.]] - -output.dims = {8, 9} - -output.lod = [[0, 4, 8]] -The simple usage is:
---output = fluid.layers.im2sequence( - input=layer, stride=[1, 1], filter_size=[2, 2]) -
paddle.v2.fluid.layers.nce(input, label, num_total_classes, sample_weight=None, param_attr=None, bias_attr=None, num_neg_samples=None)Compute and return the noise-contrastive estimation training loss. -See [Noise-contrastive estimation: A new estimation principle for unnormalized statistical models](http://www.jmlr.org/proceedings/papers/v9/gutmann10a/gutmann10a.pdf). -By default this operator uses a uniform distribution for sampling.
-| Parameters: |
|
-
|---|---|
| Returns: | (Tensor) A tensor of shape [batch_size, 1]. Cost of samples. - |
-
paddle.v2.fluid.layers.beam_search(pre_ids, ids, scores, beam_size, end_id, level=0)This function implements the beam search algorithm.
-paddle.v2.fluid.layers.row_conv(input, future_context_size, param_attr=None, act=None)Row Conv Operator. This layer will apply lookahead convolution to -input. The input variable should be a 2D LoDTensor with shape [T, D]. -Parameters with shape [future_context_size + 1, D] will be created. The math -equation of row convolution is as follows:
-In the above equation:
-More details about row_conv please refer to the paper (http://www.cs.cmu.edu/~dyogatam/papers/wang+etal.iclrworkshop2016.pdf) and -the design document (https://github.com/PaddlePaddle/Paddle/issues/2228#issuecomment-303903645).
-| Parameters: |
|
-
|---|---|
| Returns: | The output tensor with same shape as input tensor. - |
-
| Return type: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[16],
- dtype='float32', lod_level=1)
-out = fluid.layers.row_conv(input=x, future_context_size=2)
-paddle.v2.fluid.layers.multiplex(inputs, index)Multiplex Layer
-Referring to the given index variable, this layer selects rows from the -input variables to construct a multiplex variable. Assuming that there are -\(m\) input variables and \(I_i\) represents the i-th input -variable and \(i\) is in [0, \(m\)). All input variables are -tensors with same shape [\(d_0\), \(d_1\), ..., \(d_R\)]. -Please note that rank of the input tensor should be at least 2. Each input -variable will be treated as a 2-D matrix with shape [\(M\), \(N\)] -where \(M\) for \(d_0\) and \(N\) for \(d_1\) * \(d_2\) -* ... * \(d_R\). Let \(I_i[j]\) be the j-th row of the i-th input -variable. The given index variable should be a 2-D tensor with shape -[\(M\), 1]. Let ID[i] be the i-th index value of the index variable. -Then the output variable will be a tensor with shape [\(d_0\), -\(d_1\), ..., \(d_R\)]. If we treat the output tensor as a 2-D -matrix with shape [\(M\), \(N\)] and let \(O[i]\) be the i-th -row of the matrix, then O[i] is equal to \(I_{ID[i]}[i]\).
-| Parameters: |
|
-
|---|---|
| Returns: | Multiplex variable gathered from input variables. - |
-
| Return type: | Variable - |
-
Examples
-x1 = fluid.layers.data(name='x1', shape=[4], dtype='float32')
-x2 = fluid.layers.data(name='x2', shape=[4], dtype='float32')
-index = fluid.layers.data(name='index', shape=[1], dtype='int32')
-out = fluid.layers.multiplex(inputs=[x1, x2], index=index)
-paddle.v2.fluid.layers.mean(**kwargs)Mean Operator.
-Out is a scalar which is the mean of all elements in X.
-| Parameters: | x – The input of mean op -Duplicable: False Optional: False | -
|---|---|
| Returns: | The output of mean op | -
paddle.v2.fluid.layers.mul(**kwargs)Mul Operator.
-This operator is used to perform matrix multiplication for input $X$ and $Y$.
-The equation is:
-$$Out = X * Y$$
-Both the input $X$ and $Y$ can carry the LoD (Level of Details) information, -or not. But the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | (Tensor), The output tensor of mul op. - |
-
paddle.v2.fluid.layers.reshape(**kwargs)Reshape Operator.
-Reshape Input(X) into the shape specified by Attr(shape).
-An example: -Given a 2-D tensor X with 2 rows and 2 columns : [[1, 2], [3, 4]]
-and target shape = [1, 4], the reshape operator will transform -the tensor X into a 2-D tensor: [[1, 2, 3, 4]]
-One dimension in the target shape can be set -1, representing that its -size is unknown. In this case, the real dimension will be infered from -the original shape of Input(X) and other dimensions in the target shape.
-| Parameters: |
|
-
|---|---|
| Returns: | The output tensor of reshape operator. - |
-
paddle.v2.fluid.layers.scale(**kwargs)Scale operator
-$$Out = scale*X$$
-| Parameters: |
|
-
|---|---|
| Returns: | (Tensor) Output tensor of scale operator. - |
-
paddle.v2.fluid.layers.sigmoid_cross_entropy_with_logits(**kwargs)SigmoidCrossEntropyWithLogits Operator.
-This measures the element-wise probability error in classification tasks -in which each class is independent. This can be thought of as predicting labels -for a data-point, where labels are not mutually exclusive. -For example, a news article can be about politics, technology or sports -at the same time or none of these.
-The logistic loss is given as follows:
---$$loss = -Labels * log(sigma(X)) - (1 - Labels) * log(1 - sigma(X))$$
We know that $$sigma(X) = (1 / (1 + exp(-X)))$$. By substituting this we get:
---$$loss = X - X * Labels + log(1 + exp(-X))$$
For stability and to prevent overflow of $$exp(-X)$$ when X < 0, -we reformulate the loss as follows:
---$$loss = max(X, 0) - X * Labels + log(1 + exp(-|X|))$$
Both the input X and Labels can carry the LoD (Level of Details) information. -However the output only shares the LoD with input X.
-| Parameters: |
|
-
|---|---|
| Returns: | (Tensor, default Tensor<float>), a 2-D tensor with shape N x D of elementwise logistic losses. - |
-
paddle.v2.fluid.layers.elementwise_add(**kwargs)Limited Elementwise Add Operator.
-The equation is:
-$$Out = X + Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_div(**kwargs)Limited Elementwise Div Operator.
-The equation is:
-$$Out = X / Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_sub(**kwargs)Limited Elementwise Sub Operator.
-The equation is:
-$$Out = X - Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_mul(**kwargs)Limited Elementwise Mul Operator.
-The equation is:
-$$Out = X odotY$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_max(**kwargs)Limited Elementwise Max Operator.
-The equation is:
-$$Out = max(X, Y)$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_min(**kwargs)Limited Elementwise Max Operator.
-The equation is:
-$$Out = min(X, Y)$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_pow(**kwargs)Limited Elementwise Pow Operator.
-The equation is:
-$$Out = X ^ Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.clip(**kwargs)Clip Operator.
-The clip operator limits the value of given input within an interval. The -interval is specified with arguments ‘min’ and ‘max’:
-$$ -Out = min(max(X, min), max) -$$
-| Parameters: |
|
-
|---|---|
| Returns: | (Tensor)The output of clip op with shape as input(X) - |
-
paddle.v2.fluid.layers.clip_by_norm(**kwargs)ClipByNorm Operator.
-This operator limits the L2 norm of the input $X$ within $max_norm$. -If the L2 norm of $X$ is less than or equal to $max_norm$, $Out$ will be -the same as $X$. If the L2 norm of $X$ is greater than $max_norm$, $X$ will -be linearly scaled to make the L2 norm of $Out$ equal to $max_norm$, as -shown in the following formula:
-$$ -Out = frac{max_norm * X}{norm(X)}, -$$
-where $norm(X)$ represents the L2 norm of $X$.
-| Parameters: |
|
-
|---|---|
| Returns: | (Tensor) The output of clip_by_norm op with shape as input(X) - |
-
paddle.v2.fluid.layers.sequence_softmax(**kwargs)Sequence Softmax Operator.
-SequenceSoftmaxOp computes the softmax activation among all time-steps for each -sequence. The dimension of each time-step should be 1. Thus, the shape of -input Tensor can be either [N, 1] or [N], where N is the sum of the length -of all sequences.
-The algorithm works as follows:
---for i-th sequence in a mini-batch:
$$ -Out(X[lod[i]:lod[i+1]], :) = frac{exp(X[lod[i]:lod[i+1], :])} {sum(exp(X[lod[i]:lod[i+1], :]))} -$$
-For example, for a mini-batch of 3 sequences with variable-length, -each containing 2, 3, 2 time-steps, the lod of which is [0, 2, 5, 7], -then softmax will be computed among X[0:2, :], X[2:5, :], X[5:7, :] -and N turns out to be 7.
-| Parameters: | x – (LoDTensor) 1-D or 2-D input LoDTensor with the 2-nd dimension of length 1. -Duplicable: False Optional: False | -
|---|---|
| Returns: | (LoDTensor) 1-D or 2-D output LoDTensor with the 2-nd dimension of length 1. | -
paddle.v2.fluid.layers.sigmoid(**kwargs)Sigmoid Activation Operator
-$$out = frac{1}{1 + e^{-x}}$$
-| Parameters: | x – Input of Sigmoid operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Sigmoid operator | -
paddle.v2.fluid.layers.logsigmoid(**kwargs)Logsigmoid Activation Operator
-$$out = log frac{1}{1 + e^{-x}}$$
-| Parameters: | x – Input of LogSigmoid operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of LogSigmoid operator | -
paddle.v2.fluid.layers.exp(**kwargs)Exp Activation Operator.
-$out = e^x$
-| Parameters: | x – Input of Exp operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Exp operator | -
paddle.v2.fluid.layers.relu(**kwargs)Relu Activation Operator.
-$out = max(x, 0)$
-| Parameters: | x – Input of Relu operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Relu operator | -
paddle.v2.fluid.layers.tanh(**kwargs)Tanh Activation Operator.
-$$out = frac{e^{x} - e^{-x}}{e^{x} + e^{-x}}$$
-| Parameters: | x – Input of Tanh operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Tanh operator | -
paddle.v2.fluid.layers.tanh_shrink(**kwargs)TanhShrink Activation Operator.
-$$out = x - frac{e^{x} - e^{-x}}{e^{x} + e^{-x}}$$
-| Parameters: | x – Input of TanhShrink operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of TanhShrink operator | -
paddle.v2.fluid.layers.softshrink(**kwargs)Softshrink Activation Operator.
-$$ -out = begin{cases}
---x - lambda, text{if } x > lambda \ -x + lambda, text{if } x < -lambda \ -0, text{otherwise} -end{cases}
$$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of Softshrink operator - |
-
paddle.v2.fluid.layers.sqrt(**kwargs)Sqrt Activation Operator.
-$out = sqrt{x}$
-| Parameters: | x – Input of Sqrt operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Sqrt operator | -
paddle.v2.fluid.layers.abs(**kwargs)Abs Activation Operator.
-$out = |x|$
-| Parameters: | x – Input of Abs operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Abs operator | -
paddle.v2.fluid.layers.ceil(**kwargs)Ceil Activation Operator.
-$out = ceil(x)$
-| Parameters: | x – Input of Ceil operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Ceil operator | -
paddle.v2.fluid.layers.floor(**kwargs)Floor Activation Operator.
-$out = floor(x)$
-| Parameters: | x – Input of Floor operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Floor operator | -
paddle.v2.fluid.layers.round(**kwargs)Round Activation Operator.
-$out = [x]$
-| Parameters: | x – Input of Round operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Round operator | -
paddle.v2.fluid.layers.reciprocal(**kwargs)Reciprocal Activation Operator.
-$$out = frac{1}{x}$$
-| Parameters: | x – Input of Reciprocal operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Reciprocal operator | -
paddle.v2.fluid.layers.log(**kwargs)Log Activation Operator.
-$out = ln(x)$
-Natural logarithm of x.
-| Parameters: | x – Input of Log operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Log operator | -
paddle.v2.fluid.layers.square(**kwargs)Square Activation Operator.
-$out = x^2$
-| Parameters: | x – Input of Square operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Square operator | -
paddle.v2.fluid.layers.softplus(**kwargs)Softplus Activation Operator.
-$out = ln(1 + e^{x})$
-| Parameters: | x – Input of Softplus operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Softplus operator | -
paddle.v2.fluid.layers.softsign(**kwargs)Softsign Activation Operator.
-$$out = frac{x}{1 + |x|}$$
-| Parameters: | x – Input of Softsign operator -Duplicable: False Optional: False | -
|---|---|
| Returns: | Output of Softsign operator | -
paddle.v2.fluid.layers.brelu(**kwargs)BRelu Activation Operator.
-$out = max(min(x, t_{min}), t_{max})$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of BRelu operator - |
-
paddle.v2.fluid.layers.leaky_relu(**kwargs)LeakyRelu Activation Operator.
-$out = max(x, alpha * x)$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of LeakyRelu operator - |
-
paddle.v2.fluid.layers.soft_relu(**kwargs)SoftRelu Activation Operator.
-$out = ln(1 + exp(max(min(x, threshold), threshold))$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of SoftRelu operator - |
-
paddle.v2.fluid.layers.elu(**kwargs)ELU Activation Operator.
-Applies the following element-wise computation on the input according to -https://arxiv.org/abs/1511.07289.
-$out = max(0, x) + min(0, alpha * (e^x - 1))$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of ELU operator - |
-
paddle.v2.fluid.layers.relu6(**kwargs)Relu6 Activation Operator.
-$out = min(max(0, x), 6)$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of Relu6 operator - |
-
paddle.v2.fluid.layers.pow(**kwargs)Pow Activation Operator.
-$out = x^{factor}$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of Pow operator - |
-
paddle.v2.fluid.layers.stanh(**kwargs)STanh Activation Operator.
-$$out = b * frac{e^{a * x} - e^{-a * x}}{e^{a * x} + e^{-a * x}}$$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of STanh operator - |
-
paddle.v2.fluid.layers.hard_shrink(**kwargs)HardShrink Activation Operator.
-$$ -out = begin{cases}
---x, text{if } x > lambda \ -x, text{if } x < -lambda \ -0, text{otherwise} -end{cases}
$$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of HardShrink operator - |
-
paddle.v2.fluid.layers.thresholded_relu(**kwargs)ThresholdedRelu Activation Operator.
-$$ -out = begin{cases}
---x, text{if } x > threshold \ -0, text{otherwise} -end{cases}
$$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of ThresholdedRelu operator - |
-
paddle.v2.fluid.layers.hard_sigmoid(**kwargs)HardSigmoid Activation Operator.
-Segment-wise linear approximation of sigmoid(https://arxiv.org/abs/1603.00391), -which is much faster than sigmoid.
-$out = max(0, min(1, slope * x + shift))$
-The slope should be positive. The offset can be either positive or negative. -The default slope and shift are set according to the above reference. -It is recommended to use the defaults for this activation.
-| Parameters: |
|
-
|---|---|
| Returns: | Output of HardSigmoid operator - |
-
paddle.v2.fluid.layers.swish(**kwargs)Swish Activation Operator.
-$$out = frac{x}{1 + e^{- beta x}}$$
-| Parameters: |
|
-
|---|---|
| Returns: | Output of Swish operator - |
-
paddle.v2.fluid.layers.create_tensor(dtype, name=None, persistable=False)paddle.v2.fluid.layers.create_parameter(shape, dtype, name=None, attr=None, is_bias=False, default_initializer=None)Create a parameter -:param shape: shape of the parameter -:type shape: list[int] -:param dtype: element type of the parameter -:type dtype: string -:param attr: attributes of the parameter -:type attr: ParamAttr -:param is_bias: This can affect which default initializer is chosen
---when default_initializer is None. If is_bias, -initializer.Constant(0.0) will be used. Otherwise, -Xavier() will be used.
| Parameters: | default_initializer (Initializer) – initializer for the parameter | -
|---|---|
| Returns: | the created parameter | -
| Return type: | Parameter | -
paddle.v2.fluid.layers.create_global_var(shape, value, dtype, persistable=False, force_cpu=False, name=None)Create a global variable. such as global_step -:param shape: shape of the variable -:type shape: list[int] -:param value: the value of the variable -:type value: float -:param dtype: element type of the parameter -:type dtype: string -:param persistable: if this variable is persistable -:type persistable: bool -:param force_cpu: force this variable to be on CPU -:type force_cpu: bool
-| Returns: | the created Variable | -
|---|---|
| Return type: | Variable | -
paddle.v2.fluid.layers.cast(x, dtype)This function takes in the input with input_dtype -and casts it to the output_dtype as the output.
-paddle.v2.fluid.layers.concat(input, axis=0)Concat
-This function concatenates the input along the axis mentioned -and returns that as the output.
-| Parameters: |
|
-
|---|---|
| Returns: | Output variable of the concatenation - |
-
| Return type: | Variable - |
-
Examples
-paddle.v2.fluid.layers.sums(input, out=None)This function performs the sum operation on the input and returns the -result as the output.
-| Parameters: | input (Variable|list) – The input tensor that has the elements -that need to be summed up. | -
|---|---|
| Returns: |
|
-
| Return type: | Variable | -
Examples
-paddle.v2.fluid.layers.assign(input, output)Assign
-This function copies the input Variable to the output Variable.
-| Parameters: |
|
-
|---|---|
| Returns: | The destination variable that was supplied as the output. - |
-
| Return type: | Variable - |
-
Examples
-paddle.v2.fluid.layers.fill_constant_batch_size_like(input, shape, dtype, value, input_dim_idx=0, output_dim_idx=0)fill_constant_batch_size_like
-This function creates a tensor of specified shape, dtype and batch size, -and initializes this with a constant supplied in value. The batch size is -obtained from the input tensor.
-It also sets stop_gradient to True.
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the output - |
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.fill_constant_batch_size_like(
- input=like, shape=[1], value=0, dtype='int64')
-paddle.v2.fluid.layers.fill_constant(shape, dtype, value, force_cpu=False, out=None)fill_constant
-This function creates a tensor with specified shape and dtype, and -initializes it with a constant specifed by value.
-The attribute stop_gradient of the created tensor is set to True.
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the output. - |
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.fill_constant(shape=[1], value=0, dtype='int64')
-paddle.v2.fluid.layers.ones(shape, dtype, force_cpu=False)ones
-This function creates a tensor of specified shape and -dtype, and initializes this with 1.
-It also sets stop_gradient to True.
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the output - |
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.ones(shape=[1], dtype='int64')
-paddle.v2.fluid.layers.zeros(shape, dtype, force_cpu=False)zeros
-This function creates a tensor of specified shape and -dtype, and initializes this with 0.
-It also sets stop_gradient to True.
-| Parameters: |
|
-
|---|---|
| Returns: | The tensor variable storing the output - |
-
| Return type: | Variable - |
-
Examples
-data = fluid.layers.zeros(shape=[1], dtype='int64')
-
- paddle.v2.fluid.nets.simple_img_conv_pool(input, num_filters, filter_size, pool_size, pool_stride, act, param_attr=None, pool_type='max', use_cudnn=True)paddle.v2.fluid.nets.sequence_conv_pool(input, num_filters, filter_size, param_attr=None, act='sigmoid', pool_type='max')paddle.v2.fluid.nets.glu(input, dim=-1)The gated linear unit composed by split, sigmoid activation and elementwise -multiplication. Specifically, Split the input into two equal sized parts -\(a\) and \(b\) along the given dimension and then compute as -following:
----\[{GLU}(a, b)= a \otimes \sigma(b)\]-
Refer to Language Modeling with Gated Convolutional Networks.
-| Parameters: |
|
-
|---|---|
| Returns: | The Tensor variable with half the size of input. - |
-
| Return type: | Variable - |
-
Examples
-# x is a Tensor variable with shape [3, 6, 9]
-fluid.nets.glu(input=x, dim=1) # shape of output: [3, 3, 9]
-paddle.v2.fluid.nets.scaled_dot_product_attention(queries, keys, values, num_heads=1, dropout_rate=0.0)The dot-product attention.
-Attention mechanism can be seen as mapping a query and a set of key-value -pairs to an output. The output is computed as a weighted sum of the values, -where the weight assigned to each value is computed by a compatibility -function (dot-product here) of the query with the corresponding key.
-The dot-product attention can be implemented through (batch) matrix -multipication as follows:
----\[Attention(Q, K, V)= softmax(QK^\mathrm{T})V\]-
Refer to Attention Is All You Need.
-| Parameters: |
|
-
|---|---|
| Returns: | A 3-D Tensor computed by multi-head scaled dot product attention. - |
-
| Return type: | Variable - |
-
| Raises: |
|
-
Note
-1. When num_heads > 1, three linear projections are learned respectively -to map input queries, keys and values into queries’, keys’ and values’. -queries’, keys’ and values’ have the same shapes with queries, keys -and values.
-1. When num_heads == 1, scaled_dot_product_attention has no learnable -parameters.
-Examples
-# Suppose q, k, v are Tensors with the following shape:
-# q: [3, 5, 9], k: [3, 6, 9], v: [3, 6, 10]
-
-contexts = fluid.nets.scaled_dot_product_attention(q, k, v)
-contexts.shape # [3, 5, 10]
-
- paddle.v2.fluid.optimizer.SGDalias of SGDOptimizer
paddle.v2.fluid.optimizer.Momentumalias of MomentumOptimizer
paddle.v2.fluid.optimizer.Adagradalias of AdagradOptimizer
paddle.v2.fluid.optimizer.Adamalias of AdamOptimizer
paddle.v2.fluid.optimizer.Adamaxalias of AdamaxOptimizer
paddle.v2.fluid.optimizer.DecayedAdagradalias of DecayedAdagradOptimizer
- paddle.v2.fluid.param_attr.ParamAttr(name=None, initializer=None, learning_rate=1.0, regularizer=None, trainable=True, gradient_clip=None)paddle.v2.fluid.param_attr.WeightNormParamAttr(dim=None, **kwargs)Used for weight normalization. Any field in ParamAttr can also be set here. -Besides, an extra field dim can be set to indicate the dimension except -which to normalize.
-
- paddle.v2.fluid.profiler.cuda_profiler(*args, **kwds)The CUDA profiler. -This fuctions is used to profile CUDA program by CUDA runtime application -programming interface. The profiling result will be written into -output_file with Key-Value pair format or Comma separated values format. -The user can set the output mode by output_mode argument and set the -counters/options for profiling by config argument. The default config -is [‘gpustarttimestamp’, ‘gpustarttimestamp’, ‘gridsize3d’, -‘threadblocksize’, ‘streamid’, ‘enableonstart 0’, ‘conckerneltrace’].
-| Parameters: |
|
-
|---|
paddle.v2.fluid.profiler.reset_profiler()The profiler clear interface. -reset_profiler will clear the previous time record.
-paddle.v2.fluid.profiler.profiler(*args, **kwds)The profiler interface. -Different from cuda_profiler, this profiler can be used to profile both CPU -and GPU program. By defalut, it records the CPU and GPU operator kernels, -if you want to profile other program, you can refer the profiling tutorial -to add more records.
-| Parameters: |
|
-
|---|
- paddle.v2.fluid.regularizer.append_regularization_ops(parameters_and_grads, regularization=None)Create and add backward regularization Operators
-Creates and adds backward regularization operators in the BlockDesc. -This will add gradients of the regularizer function to the gradients -of the parameters and return these modified gradients. This is the -same as implementing weight decay in optimizers for regularization.
-| Parameters: |
|
-
|---|---|
| Returns: | list of (parameters, gradients) pair with the regularized gradient - |
-
| Raises: |
|
-
paddle.v2.fluid.regularizer.L1Decayalias of L1DecayRegularizer
paddle.v2.fluid.regularizer.L2Decayalias of L2DecayRegularizer
-
- paddle.v2.parameters.ParametersParameters manages all the learnable parameters in a neural network. -It stores parameters’ information in an OrderedDict. The key is -the name of a parameter, and value is a parameter’s configuration(in -protobuf format), such as initialization mean and std, its size, whether it -is a static parameter, and so on.
-| Parameters: |
|
-
|---|
Basically usage is
-data = paddle.layers.data(...)
-...
-out = paddle.layers.fc(...)
-
-parameters = paddle.parameters.create(out)
-
-parameter_names = parameters.names()
-fc_mat = parameters.get('fc')
-print fc_mat
-keys()keys are the names of each parameter.
-| Returns: | list of parameter name | -
|---|---|
| Return type: | list | -
names()names of each parameter.
-| Returns: | list of parameter name | -
|---|---|
| Return type: | list | -
has_key(key)has_key return true if there are such parameter name == key
-| Parameters: | key (basestring) – Parameter name | -
|---|---|
| Returns: | True if contains such key | -
get_shape(key)get shape of the parameter.
-| Parameters: | key (basestring) – parameter name | -
|---|---|
| Returns: | parameter’s shape | -
| Return type: | tuple | -
get(parameter_name)Get parameter by parameter name.
-| Note: | It will always copy the parameter from C++ side. | -
|---|---|
| Parameters: | parameter_name (basestring) – parameter name | -
| Returns: | The parameter matrix. | -
| Return type: | np.ndarray | -
get_grad(key)Get grandient by parameter name.
-| Note: | It will always copy the parameter from C++ side. | -
|---|---|
| Parameters: | key (basestring) – parameter name | -
| Returns: | The grandient matrix. | -
| Return type: | np.ndarray | -
set(parameter_name, value)Set parameter by parameter name & matrix.
-| Parameters: |
|
-
|---|---|
| Returns: | Nothing. - |
-
append_gradient_machine(gradient_machine)append gradient machine to parameters. This method is used internally in -Trainer.train.
-| Parameters: | gradient_machine (api.GradientMachine) – PaddlePaddle C++ GradientMachine object. | -
|---|---|
| Returns: | - |
serialize(name, f)| Parameters: |
|
-
|---|---|
| Returns: | - | -
deserialize(name, f)| Parameters: |
|
-
|---|---|
| Returns: | - | -
to_tar(f)Save parameters to a tar file.
-| Parameters: | f (file) – | -
|---|---|
| Returns: | - |
from_tar(f)Create a Parameters object from the given file. And -the Parameters only contains the parameters in this -file. It is adapted the parameters are same in the -defined network and the given file. For example, it -can be used in the inference.
-| Parameters: | f (tar file) – the initialized model file. | -
|---|---|
| Returns: | A Parameters object. | -
| Return type: | Parameters. | -
init_from_tar(f, exclude_params=[])Different from from_tar, this interface can be used to -init partial network parameters from another saved model.
-| Parameters: |
|
-
|---|---|
| Returns: | Nothing. - |
-
Module Trainer
-paddle.v2.trainer.SGD(cost, parameters, update_equation, extra_layers=None, is_local=True, pserver_spec=None, use_etcd=True)Simple SGD Trainer. -SGD Trainer combines data reader, network topolopy and update_equation together -to train/test a neural network.
-| Parameters: |
|
-
|---|
train(reader, num_passes=1, event_handler=None, feeding=None)Training method. Will train num_passes of input data.
-| Parameters: |
|
-
|---|---|
| Returns: | - | -
test(reader, feeding=None)Testing method. Will test input data.
-| Parameters: |
|
-
|---|---|
| Returns: | - | -
Testing and training events.
-There are:
-paddle.v2.event.TestResult(evaluator, cost)Result that trainer.test return.
-paddle.v2.event.BeginPass(pass_id)Event On One Pass Training Start.
-paddle.v2.event.EndPass(pass_id, evaluator, gm)Event On One Pass Training Complete. -To get the output of a specific layer, add “event.gm.getLayerOutputs(‘predict_layer’)” -in your event_handler call back
-paddle.v2.event.BeginIteration(pass_id, batch_id)Event On One Batch Training Start.
-paddle.v2.event.EndForwardBackward(pass_id, batch_id, gm)Event On One Batch ForwardBackward Complete.
-paddle.v2.event.EndIteration(pass_id, batch_id, cost, evaluator, gm)Event On One Batch Training Complete. -To get the output of a specific layer, add “event.gm.getLayerOutputs(‘predict_layer’)” -in your event_handler call back
-paddle.v2.infer(output_layer, parameters, input, feeding=None, field='value')Infer a neural network by given neural network output and parameters. The -user should pass either a batch of input data or reader method.
-Example usage for sinlge output_layer:
-result = paddle.infer(output_layer=prediction,
- parameters=parameters,
- input=SomeData)
-print result
-Example usage for multiple outout_layers and fields:
-result = paddle.infer(output_layer=[prediction1, prediction2],
- parameters=parameters,
- input=SomeData,
- field=[id, value]])
-print result
-| Parameters: |
|
-
|---|---|
| Returns: | The prediction result. If there are multiple outout_layers and fields, -the return order is outout_layer1.field1, outout_layer2.field1, ..., -outout_layer1.field2, outout_layer2.field2 ... - |
-
| Return type: | numpy.ndarray - |
-
| - |
| - |
| - | - |
|
-
| - | - |
| - |
| - |
- 
IfElse Operator","Design Doc: InferVarType","Problem","Memory Optimization","Intel\u00ae MKL Packed on PaddlePaddle: Design Doc","Intel\u00ae MKL-DNN on PaddlePaddle: Design Doc","Design Doc: Add MKLDNN Kernel in Fluid Operator","Design Doc: Model Format","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Design Doc: The Keys of Operator Kernel Type","RNNOp design","Design: Sequence Decoder Generating LoDTensors","Optimizer Design","Design Doc: NCCL support in Paddle Fluid","Averaging Parameter in PaddlePaddle","Design Doc: The C++ Class Parameters","Introduction","Design Doc: PaddlePaddle Programs","Prune","Design Doc: Python API","Python Data Reader Design Doc","Design Doc: Refactorization Overview","Design Doc: Gradient Operators Registration","Regularization in PaddlePaddle","PaddlePaddle\u53d1\u884c\u89c4\u8303","Design of Scope in Paddle","Design Doc: Selected Rows","Interaction between C++ and Python","DeepSpeech2 on PaddlePaddle: Design Doc","Design Doc: Supporting new Device/Library","Design Doc: Switch","Background","Design for TensorArray","Background","Contribute Code","Development","Write New Layers","How to write a new operator","Add Kernels for a New Device","How to use Eigen in Paddle","Contribute Documentation","GET STARTED","Quick Start","Command-line arguments","Fluid Distributed Training","Distributed Training","Cluster Training Using Fabric","Use different clusters","Distributed PaddlePaddle Training on AWS with Kubernetes","PaddlePaddle On Kubernetes","Cluster Training Using OpenMPI","<no title>","<no title>","Preparations","Argument Outline","Detail Description","Set Command-line Parameters","Use Case","HOW TO","Profiling the Python Code","Tune GPU Performance","PaddlePaddle Fluid Source Code Overview","RNN Models","RNN Configuration","PaddlePaddle Documentation","Build PaddlePaddle for Android","Build PaddlePaddle for iOS","Build PaddlePaddle for Raspberry Pi","Cluster bootstrapping tool survey"],titleterms:{"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":38,"\u4e0d\u4f7f\u7528":71,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":71,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":71,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":71,"\u4ec5\u4ec5\u4f7f\u7528void":71,"\u4ece\u5feb\u7167\u6062\u590d":37,"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":71,"\u4f7f\u7528\u8f6c\u6362\u5e93":38,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":72,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":72,"\u5206\u5757\u6587\u4ef6\u4f20\u8f93":54,"\u5206\u652f\u89c4\u8303":88,"\u52a0\u901f\u6267\u884c":37,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":71,"\u52a8\u6001\u6269\u5bb9":37,"\u539f\u56e0":71,"\u539f\u56e0\u5217\u8868":71,"\u53c2\u8003\u6587\u6863":54,"\u53d1\u5e03docker\u955c\u50cf":88,"\u53d1\u5e03wheel\u5305\u5230pypi":88,"\u540d\u8bcd\u89e3\u91ca":54,"\u57fa\u672c\u8981\u6c42":71,"\u5b9e\u73b0":71,"\u5b9e\u73b0\u65b9\u5f0f":72,"\u5bfc\u51fac":71,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":37,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":71,"\u63a8\u6d4b\u6267\u884c":37,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":38,"\u6587\u4ef6\u4f20\u8f93\u4f18\u5316":54,"\u6587\u4ef6\u8bbf\u95ee\u65b9\u5f0f":38,"\u6587\u4ef6\u8bbf\u95ee\u7684\u6743\u9650":38,"\u6587\u4ef6\u9884\u5904\u7406":38,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":72,"\u672f\u8bed":37,"\u67b6\u6784\u56fe":54,"\u6846\u67b6\u751f\u6210":54,"\u6982\u5ff5\u89e3\u91ca":38,"\u6a21\u5757":54,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":37,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":72,"\u6d41\u7a0b\u4ecb\u7ecd":38,"\u751f\u6210sparse\u6587\u4ef6":54,"\u7528\u6237\u4f7f\u7528\u6d41\u7a0b":54,"\u76ee\u5f55\u7ed3\u6784":72,"\u76ee\u6807":54,"\u793a\u4f8b\u7a0b\u5e8f":38,"\u7b26\u53f7":71,"\u7c7b":71,"\u7f16\u8bd1\u9009\u9879":72,"\u7f29\u5bb9":37,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":71,"\u80cc\u666f":71,"\u8986\u76d6\u4e0d\u4e00\u81f4\u7684\u90e8\u5206":54,"\u8bad\u7ec3\u6570\u636e\u5b58\u50a8":38,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":38,"\u8f6c\u6362\u5e93":38,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":71,"\u8fdb\u884c\u8bad\u7ec3":38,"abstract":[48,49,50,77,132],"book\u4e2d\u6240\u6709\u7ae0\u8282":88,"case":[33,121],"class":[60,79,83,100],"filemanager\u8bbe\u8ba1\u6587\u6863":54,"final":65,"function":[35,59,60,83],"new":[93,100,101,102],"paddle\u52a8\u6001\u5e93\u4e2d":71,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":71,"paddlepaddle\u53d1\u884c\u89c4\u8303":88,"paddlepaddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":88,"return":[83,84],"switch":[69,93,94],"tensor\u5230eigentensor\u7684\u8f6c\u6362":103,"while":18,AWS:112,Abs:1,DNS:112,E2E:108,EFS:112,For:[35,113],KMS:112,Not:26,The:[34,41,45,53,57,60,61,63,64,73,76,79,85,86,94,131],Use:[34,81,104,106,111,113,121],Using:[30,35,41,110,114,129],With:44,about:60,abs:18,absolut:75,access:112,account:112,accuraci:[14,18],action:[67,68],activ:[1,4,68],actor:47,adadelta:6,adagrad:[6,20],adam:[6,20],adamax:[6,20],add:[66,69,102,112],address:112,addto:4,advanc:93,aggreg:4,aggregatelevel:4,alalysi:39,algorithm:[32,36,48,74,82],all:[89,96],analog:45,analysi:[48,66],android:129,api:[0,48,67,68,72,76,78,83,87,91],append_regularization_op:23,appendix:132,approach:124,arbitrari:57,architectur:[48,80,127],argument:[55,84,107,118,121,129,131],arrai:32,array_length:18,array_read:18,array_to_lod_tensor:18,array_writ:18,asset:112,assign:18,associ:[89,112],assumpt:132,async:119,attent:127,attribut:[2,66,87],auc:3,auto:32,averag:78,avg:7,aws:112,background:[32,50,67,93,94,95,96,97,101,102],backward:[33,57,61,85,101],base:[44,75],basepool:7,basic:[66,93,132],batch:84,batch_norm:[4,18],batch_siz:84,beam:[75,92],beam_search:[4,18],beam_search_decod:18,benchmark:[67,68],benefit:[50,85],between:[31,47,83,85,91,93],bidirectional_gru:5,bidirectional_lstm:5,bilinear_interp:4,binari:34,bind:101,bla:27,block:[34,58,60,61,81,83,85],block_expand:4,blockdesc:81,blockguard:18,blockguardserv:18,blockguardwithcomplet:18,book:28,bool:27,bootstrap:132,bottleneck:123,brelu:[1,18],bring:132,bucket:112,build:[26,27,29,33,60,85,104,113,129,130,131],built:124,can:89,capi:72,capi_priv:72,cast:18,ceil:18,challeng:[33,50,82],chang:75,channel:47,check:[4,32,100,110],checkpoint:[36,37,43],choic:65,choos:[35,112],chunk:3,chunk_ev:18,chunkevalu:14,cifar:10,classif:3,classification_error:3,classification_error_print:3,client:41,clip:[4,18,51],clip_by_norm:18,close:32,cloudform:112,cluster:[42,108,110,111,112,114,121,132],cmake:[35,67,68,131],code:[44,58,83,98,123,125],column_sum:3,command:[107,120,121],commit:113,common:119,commun:119,compar:132,comparis:83,compat:57,compil:[27,34,56,58,81,85,101,125,129,130,131],complet:57,compos:84,comput:[34,61,69,85,87,103],con:132,concat:[4,18],concept:[83,85,112],concern:68,conclus:[43,62,132],concurr:[45,47],condit:60,conditionalblock:18,config:121,configur:[24,112,127,130],conll05:10,connect:4,constant:16,construct:61,contain:[28,113],content:[67,68,72,92,112,124],context:102,context_project:4,contribut:[98,104],control:[66,85],control_flow:18,contruct:66,conv2d:18,conv2d_transpos:18,conv:4,conv_oper:4,conv_project:4,conv_shift:4,convert:43,convolut:92,core:[32,83,112],corner:33,cos_sim:[4,18],cost:4,cpu:[49,121],creat:[33,47,84,85,89,112,113],create_arrai:18,create_global_var:18,create_paramet:18,create_tensor:18,createreaderop:46,creation:[40,78,87],creator:84,credenti:112,crf:4,crf_decod:[4,18],cross:[129,130,131],cross_channel_norm:4,cross_entropi:18,cross_entropy_cost:4,cross_entropy_with_selfnorm_cost:4,csp:47,ctc:[4,92],ctc_error:3,ctc_greedy_decod:18,cuda:[27,56],cuda_profil:22,cudnn:27,cudnnavg:7,cudnnmax:7,current:[56,86],custom:84,data:[4,8,9,18,36,46,48,84,95,112,113],data_feed:13,datafeed:[9,13],dataflow:66,dataprovid:119,dataset:[8,10,36,40,107],datatyp:[9,73],decayedadagrad:[6,20],decod:75,decor:84,decoratedread:46,deep:[34,57],deepspeech2:92,defin:[101,112],definit:97,delet:112,demo:[60,108,112],dens:43,dep:30,depend:[27,30,60,92],deploi:44,deriv:100,describ:[57,76],descript:[55,85,119],design:[31,32,34,36,40,41,42,43,45,47,48,49,50,52,53,56,57,59,60,61,64,67,68,69,70,73,74,75,76,77,79,81,83,84,85,86,89,90,92,93,94,96],destroi:[89,112],detail:[39,92,119],detect:3,detection_map:3,detection_output:4,develop:[26,85,99],devic:[18,93,102,121],devicecontext:93,dictionari:84,differ:[85,93,111,121],directori:112,discrimin:60,discuss:[50,60],dispatch:[36,40],distribut:[31,36,39,44,48,50,108,109,112,119],dnn:68,doc:[31,34,36,40,41,42,43,45,47,48,49,50,53,56,57,59,61,64,67,68,69,70,73,77,79,81,83,84,85,86,90,92,93,94],docker:[26,28,44,113,129],document:[104,128],doe:84,dot_prod:4,dot_product_attent:5,dotmul_oper:4,dotmul_project:4,down:112,download:[112,113],dropout:[4,18],dure:[75,84],dylib:72,dynam:[36,96],dynamic_gru:18,dynamic_lstm:18,dynamic_lstmp:18,dynamicrnn:18,dynet:62,each:30,ec2:112,edit_dist:18,eigen:103,elast:112,elect:43,elementwise_add:18,elementwise_div:18,elementwise_max:18,elementwise_min:18,elementwise_mul:18,elementwise_pow:18,elementwise_sub:18,els:34,elu:18,embed:[4,18],engin:60,enough:32,entri:84,environ:[44,129],eos:4,equat:100,error:51,evalu:[3,14,52],event:[25,31,80],evolut:57,examin:123,exampl:[31,35,45,47,63,72],execut:[34,49,57,81,85],executor:[15,53],exp:[1,18],expand:4,expandlevel:4,explan:32,extern:112,fabric:110,factorization_machin:4,faq:[29,30],fault:36,feed:46,file:[34,112,113,123],fileread:46,fill_const:18,fill_constant_batch_size_lik:18,find:112,first_seq:4,float16:56,floor:18,flow:66,fluid:[12,45,47,57,58,69,77,108,125],format:[34,36,70],forward:[61,101],frame:34,framework:[32,102,103],from:[27,29,31,43,91],full_matrix_project:4,fulli:4,functor:93,futur:[57,92],gan:60,gate:127,gated_unit:4,gener:[26,58,60,75,123,127,132],get:[105,108,113],get_inference_program:17,get_output:4,get_plac:18,give:84,global:[81,83],global_scop:15,glu:19,gotcha:26,gpu:[28,119,121,124],grad_op:33,gradient:[32,33,41,68,86,100],gradient_print:3,graph:[61,62,66,85,87],group:[4,112],gru:[5,119],gru_group:5,gru_step:4,gru_unit:[5,18],grumemori:4,hand:124,handler:[31,71],happen:43,hard_shrink:18,hard_sigmoid:18,hardwar:56,have:108,helper:83,hierarchi:34,high:[76,78,87,91],how:[27,32,39,78,84,85,93,101,103,104,122,124],hsigmoid:4,huber_classification_cost:4,huber_regression_cost:4,iOS:130,iam:112,ident:1,identifi:123,identity_project:4,ifels:[18,63],ifelseop:34,im2sequ:18,imag:[4,5,11,28,44,113,129],imdb:10,img_cmrnorm:4,img_conv:4,img_conv_bn_pool:5,img_conv_group:5,img_pool:4,imikolov:10,implement:[32,33,35,39,49,51,52,56,70,74,77,78,83,84,85,86,87,100,101,102,103],increment:18,infer:[25,129],infershap:[81,90],infervartyp:64,ingredi:31,ingress:54,initi:[16,41,60,112,121],insid:89,inspect:112,instal:[29,30,106,108,112,129,130,131,132],instanc:112,instead:84,integr:[93,112],intel:[67,68],interact:91,interfac:[8,9,11,32,36,41,42,53,76,84,89],intermedi:85,interpol:4,introduc:[75,96,108],introduct:[80,87,108],isn:84,issu:56,job:[36,44,110,112,113,114],kei:[67,73,112],kernel:[69,73,85,102],kill:110,kmax_sequence_scor:4,kube:112,kubectl:112,kubernet:[44,112,113],l1decai:23,l2_distanc:4,l2_normal:18,l2decai:23,lambda_cost:4,languag:[34,58],larg:39,last_seq:4,launch:[28,110,114],layer:[4,18,31,59,67,68,83,100,121],layer_norm:18,layout:73,leaky_relu:18,learn:[34,57],less_than:18,leval:91,level:[76,78,87,91],libpaddle_capi_shar:72,libpaddle_capi_whol:72,librari:[41,56,73,85,93,129],limit:48,line:[107,120],linear:1,linear_chain_crf:18,linear_comb:4,linux:[110,129],list:[37,84],listenandserv:18,live:66,load:47,load_inference_model:17,load_param:17,load_persist:17,load_var:17,local:[48,89,112,121],lod:75,lod_rank_t:18,lod_tensor_to_arrai:18,lodtensor:[74,75,96],lodtensordesc:97,log:[1,18,98],logic:40,logsigmoid:18,look:123,low:[78,87,91],lstm:[5,119],lstm_step:4,lstm_unit:18,lstmemori:4,lstmemory_group:5,lstmemory_unit:5,machin:75,macro:85,main:60,make:66,manag:35,map:[84,85],master:[36,40,44,45],math:[4,93],mathemat:32,matmul:18,matrix:[68,119],max:7,max_sequence_len:18,maxframe_print:3,maxid:4,maxid_print:3,maxout:4,mean:18,member:60,memori:[4,66,74,93],merge_lod_tensor:18,messag:91,method:75,might:60,migrat:85,mileston:85,mini:84,minibatch:[9,47],mix:[4,121],mkl:[67,68],mkldnn:69,mkldnn_helper:69,mkldnndevicecontext:69,mnist:10,model:[24,31,39,41,43,47,57,60,70,75,110,121,126,127],modifi:113,modul:[85,93,103],momentum:[6,20],more:60,motiv:[33,47,53,70,77,82],movielen:10,mul:18,multi:[49,58],multi_binary_label_cross_entropy_cost:4,multibox_loss:4,multipl:84,multiplex:[4,18],mxnet:62,name:[89,112],nativ:58,nccl:77,nce:[4,18],necess:83,necessari:85,need:[26,84,124],nest:74,net:19,network:[5,85,121,127],neural:127,nlp:[5,119],non:108,norm:[4,87],normal:16,note:32,numer:32,numpi:32,nvprof:124,nvvp:124,object:36,offset:75,ones:18,onli:[84,89],op_mak:85,openmpi:114,oper:[46,59,63,66,69,73,78,81,83,85,86,90,96,101],opinfomap:85,opkernel:[85,93,101,102],opproto:91,ops:[18,87],optim:[6,20,36,41,61,66,76,83],option:[27,55],opwithkernel:85,order:55,org:104,origin:85,orthogon:89,other:68,out_prod:4,outlin:118,output:[110,112],overview:[43,51,53,67,68,85,89,92,125],pack:[67,75],packag:[30,35],pad:4,paddl:[39,77,84,89,103],paddlejob:44,paddlepaddl:[27,28,30,31,34,47,57,58,67,68,78,81,87,88,92,104,108,112,113,125,128,129,130,131],pair:112,paradigm:57,parallel_nn:121,paralleldo:18,param_attr:21,paramattr:21,paramet:[2,25,31,36,41,42,44,47,50,68,78,79,83,87,107,108,112,119,120],parameteraverageoptim:78,parent:89,part:61,partit:41,pass:[27,121],path:[43,55],penalti:87,perform:[78,119,123,124],persist:40,pfsclient:[54,55],pfsserver:54,pip:30,place:[66,73,93,102],placement:48,pnpair:3,point:[67,112],polici:66,pool2d:18,pool:[4,7],pose:[64,86],potenti:65,pow:18,power:4,precision_recal:3,prefetch:84,prelu:4,prepar:[107,108,110,112,114,117,130],principl:69,print:[3,18],privat:112,pro:132,problem:[52,64,65,66,73,76,86,95],procedur:132,process:[26,36,41,44,76,85],profil:[22,123,124],program:[28,34,45,47,49,57,58,81,83,107,108],programdesc:[58,81],project:35,propos:[64,86,87],protobuf:90,protomak:101,provid:84,prune:82,pserver:43,pull:28,python:[32,44,48,67,68,74,76,78,83,84,87,91,97,100,101,123],qualiti:85,queue:[36,40],quick:106,randomnumb:119,rank:3,rank_cost:4,raspberri:131,reader:[8,9,31,46,84],readerbas:46,readerhold:46,readi:108,readop:46,realiz:85,reciproc:18,recoveri:36,recurr:[4,5,127],recurrent_group:4,recv:47,reduce_max:18,reduce_mean:18,reduce_min:18,reduce_sum:18,refactor:85,refer:[32,48,50,66,67,68,92,124],region:112,regist:[64,85,91,101,102],registr:[85,86],registri:85,regular:[23,41,87],rel:75,relat:[46,85,96],relu6:18,relu:[1,18],remark:101,remot:42,remoteexecutor:48,render:112,reorder_lod_tensor_by_rank:18,repeat:4,represent:[34,85],requir:[35,60],reset_profil:22,reshap:[4,18],resiz:4,result:[110,113],retri:40,reus:83,review:98,rmsprop:6,rnn:[74,96,119,126,127],rnnop:[34,74,85],roi_pool:4,rotat:4,round:18,route53:112,row:[90,92],row_conv:[4,18],row_l2_norm:4,rpc:47,run:[27,28,53,101,113,125],runtim:[30,44],sampl:4,sampling_id:4,save:43,save_inference_model:17,save_param:17,save_persist:17,save_var:17,scale:[4,18,36],scale_shift:4,scaled_dot_product_attent:19,scaling_project:4,scope:[34,74,85,89],scope_guard:15,script:[108,113],search:[75,92],secur:112,select:[41,47,90],selectedrow:90,selective_fc:4,semant:94,send:[18,47],sentiment:10,separ:85,seq_concat:4,seq_reshap:4,seq_slic:4,seqtext_print:3,sequenc:[75,127],sequence_conv:18,sequence_conv_pool:[5,19],sequence_expand:18,sequence_first_step:18,sequence_last_step:18,sequence_pool:18,sequence_reshap:18,sequence_softmax:18,sequencesoftmax:1,server:[36,40,41,44,47,50,107,108,112,119],servic:112,set:120,setup:[112,129],sextant:132,sgd:[20,119],shape:75,share:[31,33,66,89],should:89,shrink_memori:18,shuffl:84,sigmoid:[1,18],sigmoid_cross_entropy_with_logit:18,simpl:[75,127],simple_attent:5,simple_gru2:5,simple_gru:5,simple_img_conv_pool:[5,19],simple_lstm:5,singl:84,slice:108,slice_project:4,slope_intercept:4,small_vgg:5,smooth_l1_cost:4,soft_relu:18,softmax:1,softplu:18,softrelu:1,softshrink:18,softsign:[1,18],solut:[64,65,66,67,73,82,86,95],some:26,sourc:[27,29,125],spars:[41,42,43,90,121],specifi:121,split:18,split_lod_tensor:18,spp:4,sqrt:18,squar:[1,18],square_error_cost:[4,18],squarerootn:7,stack:34,standard:98,stanh:[1,18],start:[31,105,106,107,112,113],startup:113,statement:52,staticrnn:18,staticrnnmemorylink:18,step:[29,74],storag:87,store:36,strategi:66,style:98,sub_nested_seq:4,sub_seq:4,subcommond:55,submit:44,suffici:84,suitabl:35,sulut:69,sum:[3,7,18],sum_cost:4,sum_to_one_norm:4,summar:[31,45],summari:70,support:[56,77,93,96],survei:[56,62,87,132],swish:18,switch_scop:15,synopsi:55,syntax:47,system:[57,112],tabl:[72,92],table_project:4,tanh:[1,18],tanh_shrink:18,task:[36,40,92],tear:112,tecton:132,templat:112,tensor:[4,18,85,93,103],tensorarrai:[75,96],tensordesc:97,tensorflow:62,test:[27,67,68,69,98,100,101,119,121],text_conv_pool:5,theori:32,thi:89,think:60,three:96,thresholded_relu:18,time:125,timelin:43,timer:124,tip:124,todo:[37,38,49],togeth:89,toler:36,tool:[35,104,124,132],toolchain:131,topic:93,topk:18,toward:58,train:[25,28,31,36,39,42,44,48,76,84,107,108,109,110,112,113,114,119,121],trainer:[25,36,41,43,44,47,107,112],tran:4,trans_full_matrix_project:4,transform:95,translat:75,transpil:[48,49,50,58,66,77],transpos:18,tune:[119,124],ture:57,two:32,type:[27,47,73,101],uci_h:10,uniform:[16,96],unit:[67,68,69,98,100,101,119],unpack:75,updat:[31,42,43,104,108,112],usag:[33,51,74,75,84,103],use:[39,84,103],user:36,using:26,util:3,valu:83,value_print:3,vardesc:97,variabl:[33,66,83,85,89,97],vector:119,verifi:112,version:[30,45,56],vgg_16_network:5,volum:112,vpc:112,warp_ctc:4,warpctc:18,weightnormparamattr:21,what:[26,39,43,124],when:[43,89],whileguard:18,whl:30,why:[56,57,78,84,85,96,124],wmt14:10,wmt16:10,work:92,worker:45,workflow:98,wrapper:100,write:[98,100,101,102,104],www:104,xavier:16,yaml:113,your:[28,102],zero:18}})
\ No newline at end of file
+Search.setIndex({docnames:["build_and_install/build_en","build_and_install/build_from_source_en","build_and_install/docker_install_en","build_and_install/index_en","build_and_install/pip_install_en","design/api","design/auto_gradient_check","design/backward","design/block","design/build_system/README","design/cluster_train/README","design/cluster_train/checkpointing","design/cluster_train/data_dispatch","design/cluster_train/large_model_dist_train","design/cluster_train/master_server","design/cluster_train/pserver_client","design/cluster_train/remote_parameter_updater","design/cluster_train/save_model","design/cluster_train/submit-job","design/concurrent_programming","design/cpp_data_feeding","design/csp","design/dist_refactor/distributed_architecture","design/dist_refactor/multi_cpu","design/dist_refactor/parameter_server","design/error_clip","design/evaluator","design/executor","design/file_manager/README","design/file_manager/pfs/pfsclient","design/float16","design/fluid","design/fluid_compiler","design/functions_operators_layers","design/gan_api","design/graph","design/graph_survey","design/if_else_op","design/infer_var_type","design/kernel_hint_design","design/memory_optimization","design/mkl/mkl_packed","design/mkl/mkldnn","design/mkl/mkldnn_fluid","design/model_format","design/multi_language_interface/00.why_plain_c","design/multi_language_interface/01.inference_implementation","design/operator_kernel_type","design/ops/rnn","design/ops/sequence_decoder","design/optimizer","design/paddle_nccl","design/parameter_average","design/parameters_in_cpp","design/profiler","design/program","design/prune","design/python_api","design/reader/README","design/refactorization","design/register_grad_op","design/regularization","design/releasing_process","design/scope","design/selected_rows","design/simple_op_design","design/speech/deep_speech_2","design/support_new_device","design/switch","design/switch_kernel","design/tensor_array","design/var_desc","dev/contribute_to_paddle_en","dev/index_en","dev/new_layer_en","dev/new_op_en","dev/new_op_kernel_en","dev/use_eigen_en","dev/write_docs_en","getstarted/index_en","getstarted/quickstart_en","howto/cluster/cmd_argument_en","howto/cluster/fluid_cluster_train_en","howto/cluster/index_en","howto/cluster/multi_cluster/fabric_en","howto/cluster/multi_cluster/index_en","howto/cluster/multi_cluster/k8s_aws_en","howto/cluster/multi_cluster/k8s_en","howto/cluster/multi_cluster/openmpi_en","howto/cluster/multi_cluster/src/k8s_data/README","howto/cluster/multi_cluster/src/k8s_train/README","howto/cluster/preparations_en","howto/cmd_parameter/arguments_en","howto/cmd_parameter/detail_introduction_en","howto/cmd_parameter/index_en","howto/cmd_parameter/use_case_en","howto/index_en","howto/optimization/cpu_profiling_en","howto/optimization/gpu_profiling_en","howto/read_source","howto/rnn/index_en","howto/rnn/rnn_config_en","index_en","mobile/cross_compiling_for_android_en","mobile/cross_compiling_for_ios_en","mobile/cross_compiling_for_raspberry_en","survey/cluster_bootstrapping_tools"],envversion:50,filenames:["build_and_install/build_en.md","build_and_install/build_from_source_en.rst","build_and_install/docker_install_en.rst","build_and_install/index_en.rst","build_and_install/pip_install_en.rst","design/api.md","design/auto_gradient_check.md","design/backward.md","design/block.md","design/build_system/README.md","design/cluster_train/README.md","design/cluster_train/checkpointing.md","design/cluster_train/data_dispatch.md","design/cluster_train/large_model_dist_train.md","design/cluster_train/master_server.md","design/cluster_train/pserver_client.md","design/cluster_train/remote_parameter_updater.md","design/cluster_train/save_model.md","design/cluster_train/submit-job.md","design/concurrent_programming.md","design/cpp_data_feeding.md","design/csp.md","design/dist_refactor/distributed_architecture.md","design/dist_refactor/multi_cpu.md","design/dist_refactor/parameter_server.md","design/error_clip.md","design/evaluator.md","design/executor.md","design/file_manager/README.md","design/file_manager/pfs/pfsclient.md","design/float16.md","design/fluid.md","design/fluid_compiler.md","design/functions_operators_layers.md","design/gan_api.md","design/graph.md","design/graph_survey.md","design/if_else_op.md","design/infer_var_type.md","design/kernel_hint_design.md","design/memory_optimization.md","design/mkl/mkl_packed.md","design/mkl/mkldnn.md","design/mkl/mkldnn_fluid.md","design/model_format.md","design/multi_language_interface/00.why_plain_c.md","design/multi_language_interface/01.inference_implementation.md","design/operator_kernel_type.md","design/ops/rnn.md","design/ops/sequence_decoder.md","design/optimizer.md","design/paddle_nccl.md","design/parameter_average.md","design/parameters_in_cpp.md","design/profiler.md","design/program.md","design/prune.md","design/python_api.md","design/reader/README.md","design/refactorization.md","design/register_grad_op.md","design/regularization.md","design/releasing_process.md","design/scope.md","design/selected_rows.md","design/simple_op_design.md","design/speech/deep_speech_2.md","design/support_new_device.md","design/switch.md","design/switch_kernel.md","design/tensor_array.md","design/var_desc.md","dev/contribute_to_paddle_en.md","dev/index_en.rst","dev/new_layer_en.rst","dev/new_op_en.md","dev/new_op_kernel_en.md","dev/use_eigen_en.md","dev/write_docs_en.rst","getstarted/index_en.rst","getstarted/quickstart_en.rst","howto/cluster/cmd_argument_en.md","howto/cluster/fluid_cluster_train_en.md","howto/cluster/index_en.rst","howto/cluster/multi_cluster/fabric_en.md","howto/cluster/multi_cluster/index_en.rst","howto/cluster/multi_cluster/k8s_aws_en.md","howto/cluster/multi_cluster/k8s_en.md","howto/cluster/multi_cluster/openmpi_en.md","howto/cluster/multi_cluster/src/k8s_data/README.md","howto/cluster/multi_cluster/src/k8s_train/README.md","howto/cluster/preparations_en.md","howto/cmd_parameter/arguments_en.md","howto/cmd_parameter/detail_introduction_en.md","howto/cmd_parameter/index_en.rst","howto/cmd_parameter/use_case_en.md","howto/index_en.rst","howto/optimization/cpu_profiling_en.md","howto/optimization/gpu_profiling_en.rst","howto/read_source.md","howto/rnn/index_en.rst","howto/rnn/rnn_config_en.rst","index_en.rst","mobile/cross_compiling_for_android_en.md","mobile/cross_compiling_for_ios_en.md","mobile/cross_compiling_for_raspberry_en.md","survey/cluster_bootstrapping_tools.md"],objects:{},objnames:{},objtypes:{},terms:{"00m":98,"03m":98,"0424m":98,"055ee37d":86,"0630u":98,"06u":98,"0810u":98,"0957m":98,"0_cudnn5":1,"0_cudnn5_avx_mkl":[2,4],"0_cudnn7_avx_mkl":4,"0rc":91,"0rc1":62,"0rc2":62,"0x10f256d50":36,"0x7ffe4de00110":36,"100gb":98,"100gi":86,"10g":18,"10m":98,"1150u":98,"11\u5b9e\u73b0\u4e86c":46,"11e6":87,"124n":98,"12gb":40,"13m":87,"1490u":98,"1550u":98,"16u":98,"173n":98,"1770u":98,"18ad":86,"18e457ce3d362ff5f3febf8e7f85ffec852f70f3b629add10aed84f930a68750":87,"197u":98,"1gb":98,"210u":98,"215n":98,"228u":98,"2520u":98,"2680u":98,"279n":98,"27m":98,"285m":98,"2863m":98,"28m":98,"2977m":98,"2cbf7385":86,"302n":98,"30u":98,"328n":98,"32u":98,"331n":98,"3320u":98,"365e":86,"36u":98,"3710m":98,"3768m":98,"387u":98,"38u":98,"3920u":98,"39u":98,"3rd":104,"4035m":98,"4090u":98,"4096mb":93,"4279m":98,"43u":98,"448a5b355b84":87,"4560u":98,"4563m":98,"45u":98,"4650u":98,"4726m":98,"473m":87,"4gb":93,"50bd":86,"50gi":86,"514u":98,"525n":98,"526u":98,"536u":98,"5460u":98,"5470u":98,"54u":98,"5690m":98,"573u":98,"578n":98,"5798m":98,"586u":98,"58s":87,"5969m":98,"5_cudnn5_avx_mkl":4,"5_cudnn5_avx_openbla":[4,80],"6080u":98,"6140u":98,"6305m":98,"639u":98,"64m":44,"655u":98,"6780u":98,"6810u":98,"682u":98,"6970u":98,"6ce9":86,"704u":98,"7090u":98,"72u":98,"73u":98,"75u":98,"760u":98,"767u":98,"783n":98,"784u":98,"78m":98,"7kb":87,"8250u":98,"8300u":98,"830n":98,"849m":98,"861u":98,"8661m":98,"892m":98,"901n":98,"90u":98,"918u":98,"9247m":98,"924n":98,"9261m":98,"9330m":98,"94u":98,"9530m":98,"983m":98,"988u":98,"997u":98,"99u":98,"9a235":104,"9f18":87,"\u4e00\u4e2a\u5178\u578b\u7684chunk\u5982\u4e0b\u6240\u793a":28,"\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u6a21\u578b\u7531\u5927\u91cf\u7684\u53c2\u6570\u7ec4\u6210":11,"\u4e00\u4e2achunk\u7531\u6240\u5728\u7684\u6587\u4ef6\u504f\u79fb":28,"\u4e00\u4e2aposix\u517c\u5bb9\u7684\u6587\u4ef6\u7cfb\u7edf":28,"\u4e00\u822c\u4e0d\u5141\u8bb8\u518d\u4ece":62,"\u4e00\u822c\u7531mkl":42,"\u4e0a\u4f20\u5230cloud\u6216\u8005\u4e0b\u8f7d\u5230\u672c\u5730\u7684\u65f6\u95f4\u53ef\u80fd\u6bd4\u8f83\u957f":28,"\u4e0a\u4f20\u65b9\u6cd5":62,"\u4e0a\u6ce8\u518c\u4e00\u4e0b":28,"\u4e0a\u8ff0paddlepaddl":62,"\u4e0b\u4e00\u4e2awheel\u5305\u9700\u8981\u66f4\u65b0\u7248\u672c\u53f7\u624d\u53ef\u4ee5\u4e0a\u4f20":62,"\u4e0b\u5b58\u653e\u516c\u5171\u6570\u636e\u96c6\u5408":12,"\u4e0b\u62c9\u6846\u4e2d\u627e\u5230\u751f\u6210\u76843\u4e2a\u4e8c\u8fdb\u5236\u6587\u4ef6":62,"\u4e0b\u8f7d":28,"\u4e0b\u8f7d\u5230\u672c\u5730":28,"\u4e0b\u8f7d\u5f97\u5230":62,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u67d0\u4e00\u7c7b\u6587\u4ef6\u7684\u5b9e\u73b0\u65b9\u5f0f":46,"\u4e0d\u4e00\u81f4\u7684\u7531pfsclient\u4e0b\u8f7d\u6216\u8005\u4f20\u8f93chunk\u5b8c\u6210":28,"\u4e0d\u4f7f\u7528\u9759\u6001\u5e93":45,"\u4e0d\u4f7f\u7528c":45,"\u4e0d\u4f7f\u7528swig":45,"\u4e0d\u53ef\u4ee5\u66f4\u6539":62,"\u4e0d\u540c":42,"\u4e0d\u540c\u7248\u672c\u7684\u7f16\u8bd1\u5668\u4e4b\u95f4":45,"\u4e0d\u540c\u8bed\u8a00\u7684\u63a5\u53e3\u9002\u5e94\u4e0d\u540c\u8bed\u8a00\u7684\u7279\u6027":45,"\u4e0d\u5728":46,"\u4e0d\u5bb9\u6613\u51fa\u9519":28,"\u4e0d\u5d4c\u5165\u5176\u4ed6\u8bed\u8a00\u89e3\u91ca\u5668":45,"\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":45,"\u4e0d\u663e\u793a\u7684\u5199\u6bcf\u4e2a\u7c7b\u5177\u4f53\u5305\u542b\u4ec0\u4e48":45,"\u4e0d\u7528mount\u7684\u65b9\u5f0f\u6765\u8bbf\u95ee\u6570\u636e":12,"\u4e0e":42,"\u4e0e\u4e4b\u76f8\u5bf9\u7684\u662flocal":28,"\u4e0e\u5176\u4ed6\u7b2c\u4e09\u65b9\u5e93\u4e00\u6837":42,"\u4e0e\u529f\u80fd\u5206\u652f\u4e0d\u540c\u7684\u662f":62,"\u4e0e\u53ef\u80fd\u6709\u7684":62,"\u4e0ebatch":41,"\u4e14\u589e\u52a0\u4e00\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00":45,"\u4e14\u8c03\u7528\u65f6\u4e0d\u80fd\u629b\u51fa\u5f02\u5e38\u6216\u51fa\u73b0\u8fd0\u884c\u65f6\u9519\u8bef":46,"\u4e14c99\u652f\u6301bool\u7c7b\u578b\u548c\u5b9a\u957f\u6574\u6570":45,"\u4e14c99\u76f8\u5bf9\u4e8ec11\u4f7f\u7528\u66f4\u52a0\u5e7f\u6cdb":45,"\u4e25\u683c\u7684\u547d\u540d\u89c4\u8303pep":62,"\u4e2a\u6027\u5316\u63a8\u8350":62,"\u4e2d":[41,42,45,46],"\u4e2d\u4f1a\u63d0\u4f9b\u4e00\u4e9b\u5fc5\u8981\u7684\u63a5\u53e3\u548c\u51fd\u6570":42,"\u4e2d\u5199\u5165json\u5185\u5bb9":11,"\u4e2d\u5b8c\u5168\u4e00\u81f4":45,"\u4e2d\u5b9e\u73b0\u4e86\u4e00\u4e2amerge\u7684\u65b9\u6cd5":42,"\u4e2d\u5b9e\u73b0\u7684\u7ed3\u6784\u4f53":46,"\u4e2d\u5bf9\u5e94\u7684layer\u5904":41,"\u4e2d\u5f15\u5165\u7684":41,"\u4e2d\u63d0\u4f9b\u4e00\u4e2a\u4e0emkl\u6709\u5173\u7684\u603b\u5f00\u5173":42,"\u4e2d\u6839\u636e":41,"\u4e2d\u6dfb\u52a0":41,"\u4e2d\u6dfb\u52a0\u4e00\u4e2a":42,"\u4e2d\u7684\u7248\u672c\u4fe1\u606f":62,"\u4e2d\u8fd0\u884c\u4efb\u52a1\u7684\u89d2\u5ea6":12,"\u4e3a":[41,42],"\u4e3a\u4e86\u5c3d\u53ef\u80fd\u5c11\u7684\u5728\u7236\u7c7blayer\u4e2d\u6dfb\u52a0\u53d8\u91cf\u6216\u8005\u51fd\u6570":42,"\u4e3a\u4e86\u5e94\u5bf9\u4ee5\u4e0a\u7684\u95ee\u9898":28,"\u4e3a\u4e86\u66b4\u9732\u7684\u63a5\u53e3\u5c3d\u91cf\u7b80\u5355":46,"\u4e3a\u4e86\u66f4\u597d\u7684\u7b26\u5408paddlepaddle\u7684\u4ee3\u7801\u98ce\u683c":42,"\u4e3a\u4e86\u6700\u5927\u7a0b\u5ea6\u51cf\u5c11\u591a\u6b21\u8c03\u7528":41,"\u4e3a\u4e86\u8fdb\u4e00\u6b65\u63d0\u5347paddlepaddle\u5728\u57fa\u672c\u6570\u5b66\u8fd0\u7b97\u7684\u8ba1\u7b97\u901f\u5ea6":42,"\u4e3b\u8981\u529f\u80fd\u5305\u62ec":28,"\u4e3b\u8981\u5305\u62ec":42,"\u4e3b\u8981\u5305\u62ec\u4e86\u6df1\u5ea6\u5b66\u4e60\u76f8\u5173\u7684\u6570\u5b66\u539f\u8bed\u4e0e\u64cd\u4f5c":42,"\u4e3b\u8981\u9488\u5bf9paddlepaddle\u5728\u91cd\u6784\u4e4b\u524d\u7684\u4ee3\u7801\u6846\u67b6\u4ee5\u53cav1\u7684api":42,"\u4e4b\u5916\u7684\u6240\u6709\u5934\u6587\u4ef6":46,"\u4e5f\u4e0d\u4f7f\u7528\u5176\u4ed6\u52a8\u6001\u5e93":45,"\u4e5f\u4e0d\u5e94\u8be5\u62a5\u9519":46,"\u4e5f\u4e0d\u751f\u6210":46,"\u4e5f\u53ef\u4ee5\u4f7f\u7528\u8fd9\u4e9b\u955c\u50cf":62,"\u4e5f\u5c31\u662f\u8bf4\u8f93\u51fa\u7684\u7ed3\u679c\u4e0d\u4f1a\u5728\u539f\u6765\u7684\u6570\u636e\u4e0a\u7d2f\u52a0":42,"\u4e66\u5199":45,"\u4eba\u8138\u8bc6\u522b":12,"\u4ec5\u4ec5\u4f7f\u7528":45,"\u4ec5\u4f1a\u5728\u652f\u6301avx2\u6307\u4ee4\u96c6\u53ca\u4ee5\u4e0a\u7684\u673a\u5668\u624d\u4f7f\u7528mkl":42,"\u4ece":62,"\u4ece\u78c1\u76d8\u6587\u4ef6\u4e2d\u52a0\u8f7duuid\u6587\u4ef6\u540d\u7684\u68c0\u67e5\u70b9\u5feb\u7167\u6587\u4ef6":11,"\u4ece\u800c\u907f\u514d\u4e86packing\u5197\u4f59":41,"\u4eceetcd\u4e2d\u8bfb\u53d6\u8282\u70b9":11,"\u4ed6\u4e3b\u8981\u5305\u542b\u4e86\u5b9e\u9645\u66b4\u9732\u7684\u7c7b\u578b\u7ed3\u6784":46,"\u4ed6\u662f\u5c06":46,"\u4ed6\u7684\u76ee\u6807\u662f\u4f7f\u7528c":45,"\u4ee3\u7801\u751f\u6210\u7684\u7b26\u53f7\u53ef\u80fd\u4e0d\u4e00\u81f4":45,"\u4ee3\u8868\u8fd9\u4e2alayer\u662f\u7528\u4e8e\u8dd1\u5728mkl":42,"\u4ee3\u8868\u8fd9\u4e2ashard\u7684\u6700\u5927index":12,"\u4ee3\u8868shard\u7684index":12,"\u4ee5\u4e0a\u4ee3\u7801\u7684reader\u8f93\u51fa\u7684data":12,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u4e0b\u751f\u6210100\u4e2a\u6587\u4ef6":12,"\u4ee5\u4e0b":12,"\u4ee5\u4e0b\u7b80\u79f0rnn":41,"\u4ee5\u4fbf\u6211\u4eec\u53ef\u4ee5\u628a\u66f4\u591a\u7684\u7cbe\u529b\u653e\u5230\u903b\u8f91\u672c\u8eab\u4e0a":28,"\u4ee5\u53ca":41,"\u4ee5\u53canumpi":12,"\u4ee5\u6b64\u8fbe\u5230\u6700\u597d\u7684\u6027\u80fd":42,"\u4ee5\u793a\u533a\u5206":[41,42],"\u4efb\u610f\u65f6\u523b\u53ea\u53ef\u80fd\u540c\u65f6\u6709\u4e00\u53f0\u670d\u52a1\u5668\u6545\u969c":11,"\u4f18\u5316\u524d":41,"\u4f18\u5316\u540e":41,"\u4f1a\u4ee5":[41,42],"\u4f1a\u4f7f\u7528\u76f8\u540c\u7684\u539f\u6570\u636e":41,"\u4f1a\u5148\u4e34\u65f6\u4fdd\u5b58\u5728":42,"\u4f1a\u5728":42,"\u4f1a\u5728\u7f16\u8bd1paddlepaddle\u7684\u65f6\u5019\u4e0b\u8f7d\u5e76\u7f16\u8bd1mkl":42,"\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7248\u672cpython\u5728\u4e00\u4e2a\u8fdb\u7a0b\u91cc\u7684bug":45,"\u4f1a\u5f15\u5165":42,"\u4f1a\u628acpu\u7684buffer\u5bf9\u9f50\u4e3a4096":42,"\u4f1a\u6dfb\u52a0\u76f8\u5e94\u7684\u811a\u672c\u5728":42,"\u4f1a\u6dfb\u52a0\u76f8\u5e94\u7684\u811a\u672c\u7528\u4e8e\u6d4b\u8bd5\u548c\u5bf9\u6bd4\u5728\u4f7f\u7528mkl":41,"\u4f1a\u76f4\u63a5\u62a5\u9519\u9000\u51fa":45,"\u4f1a\u81ea\u52a8\u4f7f\u7528mklml\u5e93\u4f5c\u4e3apaddlepaddle\u7684cblas\u548clapack\u5e93":42,"\u4f1a\u81ea\u52a8\u6839\u636e\u786c\u4ef6\u914d\u7f6e":42,"\u4f1a\u88abpickle\u5e8f\u5217\u5316\u6210\u5b57\u7b26\u4e32":12,"\u4f20\u5165":12,"\u4f46":46,"\u4f46\u4e0d\u66b4\u9732":46,"\u4f46\u5e76\u6ca1\u6709\u7ecf\u8fc7\u56de\u5f52\u6d4b\u8bd5":62,"\u4f46\u6240\u6709fork\u7684\u7248\u672c\u5e93\u7684\u6240\u6709\u5206\u652f\u90fd\u76f8\u5f53\u4e8e\u7279\u6027\u5206\u652f":62,"\u4f46\u662f\u53c8\u8fc7\u4e8e\u7410\u788e":46,"\u4f46\u662f\u5728mkl":42,"\u4f46\u662f\u5728paddlepaddle\u4e2d":42,"\u4f46\u662f\u6574\u4e2a\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4e0d\u9700\u8981\u4efb\u4f55\u8f6c\u6362":42,"\u4f46\u662f\u6ce8\u610f\u7684\u662f":42,"\u4f46\u662f\u89e3\u91ca\u6027\u8bed\u8a00":45,"\u4f5c\u4e3a\u53e6\u4e00\u4e2a\u7b2c\u4e09\u65b9\u5e93\u96c6\u6210\u8fdbpaddlepaddl":42,"\u4f5c\u4e3a\u5b58\u50a8\u7cfb\u7edf":12,"\u4f5c\u4e3a\u7c7b\u53e5\u67c4":45,"\u4f7f\u7528":[42,46,62],"\u4f7f\u7528\u4e0b\u9762\u547d\u4ee4":12,"\u4f7f\u7528\u52a8\u6001\u5e93":45,"\u4f7f\u7528\u540c\u6837\u7684\u8bad\u7ec3\u6570\u636eblock":11,"\u4f7f\u7528\u667a\u80fd\u6307\u9488\u7684\u539f\u56e0\u662f":46,"\u4f7f\u7528\u7684\u53c2\u6570\u4e0epaddlepaddle\u7533\u8bf7\u7684buffer\u5171\u7528\u4e00\u5757\u5185\u5b58":42,"\u4f7f\u7528\u76f8\u5bf9\u8def\u5f84\u7684\u5f15\u7528\u65b9\u5f0f":46,"\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u5b8c\u6210\u5bf9\u65b0\u6570\u636e\u7684\u9884\u6d4b":11,"\u4f7f\u7528\u9759\u6001\u5e93\u548c\u52a8\u6001\u5e93\u96be\u5ea6\u5dee\u4e0d\u591a":45,"\u4f7f\u7528c":46,"\u4f7f\u7528c99\u505a\u63a5\u53e3":45,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c11\u7684\u539f\u56e0\u662f":45,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c89":45,"\u4f7f\u7528regress":62,"\u4f7f\u7528swig\u53ea\u652f\u6301cpython\u89e3\u91ca\u5668":45,"\u4f7f\u7528swig\u9700\u8981\u591a\u8bed\u8a00\u7ed1\u5b9a\u7684\u5f00\u53d1\u4eba\u5458\u719f\u7ec3\u638c\u63e1swig\u914d\u7f6e":45,"\u4f7f\u7528void":45,"\u4f8b\u5982":[12,45,46,62],"\u4f8b\u5982\u5728deepspeech2":41,"\u4f8b\u5982\u5bf9\u4e8ejava\u6216\u8005python":45,"\u4f8b\u5982\u5bf9\u4e8ejava\u6765\u8bf4":45,"\u4f8b\u5982\u5bf9\u4e8epython":45,"\u4f8b\u5982c":45,"\u4f8b\u5982java\u4e0epython\u7684\u9519\u8bef\u5904\u7406\u662f\u76f4\u63a5\u6254\u51fa\u6765except":45,"\u4f8b\u5982python\u53ef\u4ee5\u4f7f\u7528":45,"\u4f8b\u5982python\u7684":45,"\u4f8b\u5982rnn":41,"\u4f9d\u6b21\u7c7b\u63a8":62,"\u4fbf\u662f\u5c06\u9759\u6001\u5e93\u52a0\u5165jvm\u4e2d":45,"\u4fee\u590d\u6240\u6709bug\u540e":62,"\u4fee\u590ddocker\u7f16\u8bd1\u955c\u50cf\u95ee\u9898":62,"\u4fee\u6539":[42,62],"\u4fee\u6539\u6210":62,"\u505a\u53ea\u8bfb\u6302\u8f7d":12,"\u505a\u5982\u4e0b\u51e0\u4e2a\u64cd\u4f5c":62,"\u505a\u63a5\u53e3":45,"\u505c\u6b62\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":11,"\u5145\u5206\u53d1\u6325\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u4f18\u52bf":41,"\u5145\u5206\u5c55\u73b0\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u4f18\u52bf":42,"\u5148\u5b8c\u6210\u5bf9\u6743\u91cd\u7684packing\u64cd\u4f5c":41,"\u5148\u5b9e\u73b0\u6a21\u578b\u63a8\u65ad\u7684api":46,"\u5171\u4eab\u5185\u5b58":42,"\u5171\u4eab\u540c\u4e00\u4e2a\u6743\u91cd":41,"\u5176\u4e2d":[45,62],"\u5176\u4ed6\u51fd\u6570\u5747\u8fd4\u56de":46,"\u5176\u4ed6\u7528\u6237\u7684fork\u7248\u672c\u5e93\u5e76\u4e0d\u9700\u8981\u4e25\u683c\u9075\u5b88":62,"\u5176\u8f6c\u6362\u6b21\u6570\u51cf\u5c11\u81f3":41,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":46,"\u5177\u4f53\u539f\u56e0\u53c2\u8003":46,"\u5177\u4f53\u53ef\u4ee5\u53c2\u8003mkl":42,"\u5177\u4f53\u5b9e\u73b0\u65b9\u5f0f\u6bd4\u5982":[41,42],"\u5177\u4f53\u7684\u5b8c\u6210\u72b6\u6001\u53ef\u4ee5\u53c2\u89c1":42,"\u5177\u4f53\u8bf7\u53c2\u8003":46,"\u5185\u90e8\u5b58\u50a8":42,"\u5185\u90e8\u9a71\u52a8python\u89e3\u91ca\u5668\u8fdb\u884c\u6a21\u578b\u914d\u7f6e\u89e3\u6790\u548c\u6570\u636e\u8bfb\u53d6":45,"\u518d\u5728\u6bcf\u4e00\u4e2aapi\u4e2d\u81ea\u5df1\u68c0\u67e5\u7c7b\u578b":45,"\u518d\u57fa\u4e8e":62,"\u518d\u628a\u5df2\u8f6c\u6362\u4e3apacked\u683c\u5f0f\u7684\u6570\u636e\u4f20\u9012\u7ed9\u90a3\u4e9b\u590d\u7528\u540c\u4e00\u6570\u636e\u7684gemm":41,"\u5199\u4ee3\u7801":45,"\u5199\u5165\u5feb\u7167\u6570\u636e":11,"\u51fd\u6570":[41,42],"\u51fd\u6570\u5373\u53ef\u5b8c\u6210\u8f6c\u6362":12,"\u51fd\u6570\u540d\u4e3a":46,"\u51fd\u6570\u547d\u540d":45,"\u5206\u522b\u4ee3\u8868\u8f93\u5165\u6570\u636e":42,"\u5206\u522b\u5bf9\u5e94capi":62,"\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":11,"\u5206\u652f":62,"\u5206\u652f\u4e00\u65e6\u5efa\u7acb":62,"\u5206\u652f\u4e2d":62,"\u5206\u652f\u4e3a\u5f00\u53d1":62,"\u5206\u652f\u4e3a\u6bcf\u4e00\u6b21release\u65f6\u5efa\u7acb\u7684\u4e34\u65f6\u5206\u652f":62,"\u5206\u652f\u4e3a\u7a33\u5b9a":62,"\u5206\u652f\u529f\u80fd\u7684\u5c01\u95ed":62,"\u5206\u652f\u5408\u5165":62,"\u5206\u652f\u5408\u5165master\u5206\u652f":62,"\u5206\u652f\u540c\u6b65\u4e3b\u7248\u672c\u5e93\u7684":62,"\u5206\u652f\u540d\u4e3a":62,"\u5206\u652f\u5b58\u5728\u7684\u65f6\u5019":62,"\u5206\u652f\u6d3e\u751f\u51fa\u65b0\u7684\u5206\u652f":62,"\u5206\u652f\u7684\u7248\u672c\u90fd\u662f\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5\u548c\u56de\u5f52\u6d4b\u8bd5\u7684\u7248\u672c":62,"\u5206\u652f\u7684\u7248\u672c\u90fd\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5":62,"\u5206\u7247":11,"\u5219\u4f7f\u7528\u542f\u52a8\u53c2\u6570\u5b9a\u4e49\u7684\u521d\u59cb\u5316\u65b9\u6cd5\u521d\u59cb\u5316\u53c2\u6570":11,"\u5219\u5ffd\u7565":11,"\u5219\u628a\u53e6\u4e00\u4e2a\u6162\u901f\u7684kill\u6389":11,"\u5219\u76f4\u63a5\u5f15\u5165\u53e6\u4e00\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5219\u9700\u8981\u56de\u6eda\u5230\u4e0a\u4e00\u4e2a\u68c0\u67e5\u70b9":11,"\u521b\u5efa":42,"\u5220\u9664\u78c1\u76d8\u76ee\u5f55\u4e2d\u4e0d\u662f\u5f53\u524duuid\u7684\u5feb\u7167\u6587\u4ef6":11,"\u5230":11,"\u5230\u7b2c\u4e8c\u6b65":62,"\u524d\u540e\u7684\u7f51\u7edc\u6027\u80fd":41,"\u529f\u80fd":28,"\u529f\u80fd\u7684\u6b63\u786e\u6027\u5305\u62ec\u9a8c\u8bc1paddlepaddle\u76ee\u524d\u7684":62,"\u52a8\u6001\u5e93":45,"\u5305\u542b\u4e86\u67d0\u79cd\u7c7b\u578b\u7684\u7c7b\u578b\u5b9a\u4e49\u548c\u66b4\u9732\u7684\u5168\u90e8\u51fd\u6570":46,"\u5305\u62ec":[12,41,42],"\u5305\u62ec\u6743\u91cdw\u548c\u504f\u7f6eb":11,"\u5305\u62ecmkl":42,"\u534f\u540c\u5b8c\u6210releas":62,"\u5355\u4e2a\u503c":12,"\u5355\u70b9\u6545\u969c":11,"\u5373":46,"\u5373\u4f7f\u7528":46,"\u5373\u4f7f\u7528\u6237\u76f4\u63a5\u5f15\u7528\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5373\u4f7fc":46,"\u5373\u4f8b\u5982":46,"\u5373\u4fbfpaddl":46,"\u5373\u5b8c\u6210\u67d0\u4e00\u4e2a\u4efb\u52a1\u7684\u6700\u5c11\u51fd\u6570":46,"\u5373\u66b4\u9732":46,"\u5373\u8868\u793a\u4e0d\u9700\u8981\u8f6c\u6362":42,"\u5373\u8fd9\u4e2a\u52a8\u6001\u5e93\u662f\u4e0d\u4f9d\u8d56\u4e8e\u5176\u4ed6\u4efb\u4f55\u6587\u4ef6\u7684":45,"\u539f\u6765\u7684\u65b9\u6848":42,"\u53c2\u6570":45,"\u53c2\u8003":[28,45],"\u53c2\u8003\u4e0b\u56fe":62,"\u53c8\u53ef\u4ee5\u907f\u514d\u4e0d\u5fc5\u8981\u7684\u8f6c\u6362":42,"\u53cc\u5411\u9a8c\u8bc1":28,"\u53d1\u578b\u7248":62,"\u53d1\u5e03\u5230dockerhub":62,"\u53d1\u5e03docker\u955c\u50cf\u53ea\u9700\u8981\u5bf9\u81ea\u52a8push\u7684\u955c\u50cf\u6253\u4e0a":62,"\u53d8\u91cf\u6765\u533a\u5206layer\u7684\u5c5e\u6027":42,"\u53ea\u5bf9\u7279\u6b8a\u5728\u7ebf\u7cfb\u7edf\u8003\u8651\u4e24\u53f0\u4ee5\u4e0a\u540c\u65f6\u6545\u969c\u7684\u5bb9\u707e":11,"\u53ea\u66b4\u9732\u6982\u5ff5\u7684\u63a5\u53e3":46,"\u53ea\u80fd\u8c03\u7528paddle\u7684\u52a8\u6001\u5e93":45,"\u53ea\u9700\u8981\u6062\u590d\u8fd9\u53f0\u8282\u70b9":11,"\u53ef\u4ee5":62,"\u53ef\u4ee5\u51cf\u5c0f\u7cfb\u7edf\u590d\u6742\u6027":11,"\u53ef\u4ee5\u5728\u4efb\u4f55\u673a\u5668\u4e0a\u6267\u884c\u7684":45,"\u53ef\u4ee5\u5728\u6b64\u9875\u9762\u7684":62,"\u53ef\u4ee5\u628a\u672c\u5730\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5b58\u50a8\u96c6\u7fa4\u4e2d":12,"\u53ef\u4ee5\u6709\u6548\u7684\u907f\u514dparamet":11,"\u53ef\u4ee5\u7528":28,"\u53ef\u4ee5\u7528\u4ee5\u4e0b\u6307\u4ee4":12,"\u53ef\u4ee5\u7ee7\u7eed\u5728\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f\u63d0\u4ea4\u4ee3\u7801":62,"\u53ef\u4ee5\u901a\u8fc7\u9636\u6bb5\u6027\u7684\u4fdd\u5b58\u6bcf\u4e2aparamet":11,"\u53ef\u80fd\u4f1a\u9020\u6210\u7f51\u7edc\u62e5\u585e":11,"\u53f3\u4fa7\u7684":62,"\u5404\u6b21\u524d\u5411\u4e4b\u95f4\u4e5f\u90fd\u4f7f\u7528\u4e86\u76f8\u540c\u7684\u6743\u91cd":41,"\u540c\u4e00\u6b21\u524d\u5411":41,"\u540c\u65f6":[41,42],"\u540c\u65f6\u4f1a\u5f00\u542fintel":42,"\u540c\u65f6\u518d\u5c06":62,"\u540c\u65f6\u53c8\u5c3d\u53ef\u80fd\u5c11\u7684\u727a\u7272mkl":42,"\u540c\u65f6\u63d0\u8d77":62,"\u540c\u65f6\u6570\u636e\u683c\u5f0f\u5c31\u662f":42,"\u540d\u5b57\u4fee\u9970":45,"\u540e\u5411":41,"\u540e\u5411\u65f6\u590d\u7528\u5df2\u7ecf\u8f6c\u6362\u8fc7\u7684\u6743\u91cd":41,"\u5411\u6307\u5b9a\u7684\u76ee\u5f55\u4e2d\u4e00\u4e2a\u65b0\u7684\u6587\u4ef6":11,"\u5411paddlepaddle\u7684\u4e3b\u7248\u672c\u5e93\u63d0\u4ea4":62,"\u5426\u5219\u5f97\u628apaddle\u9759\u6001\u5e93\u94fe\u63a5\u5230\u89e3\u91ca\u5668\u91cc":45,"\u542f\u52a8\u4e00\u4e2a\u65b0\u7684\u7ebf\u7a0b\u5f00\u59cb\u4fdd\u5b58\u68c0\u67e5\u70b9":11,"\u548c":[12,41,42,45,46,62],"\u548c\u672a\u6765\u53ef\u80fd\u8fd8\u4f1a\u7528\u5230":42,"\u548c\u79bb\u7ebf\u6570\u636e\u7684\u65b9\u5f0f":12,"\u54ea\u4e2atrainer\u5148\u5b8c\u6210block\u7684\u8bad\u7ec3":11,"\u56e0\u4e3a\u8fd9\u6837\u505a\u4e5f\u6ca1\u6cd5\u4fdd\u8bc1\u6d88\u9664\u968f\u673a\u6027":11,"\u56e0\u4e3aswig\u5728\u7b2c\u4e09\u65b9\u8bed\u8a00\u4e2d\u66b4\u9732\u7684\u51fd\u6570\u540d":45,"\u56e0\u6b64":41,"\u56fe\u50cf\u5206\u7c7b":62,"\u5728":[41,42,46,62],"\u5728\u4e00\u4e2a\u4e0d\u53ef\u4e2d\u65ad\u5e76\u7f3a\u5c11\u5907\u4efd\u7684\u8bad\u7ec3\u4efb\u52a1\u4e2d":11,"\u5728\u4e0a\u56fe\u4e2d\u663e\u793a\u4e86\u5728\u4e00\u4e2a\u5b9e\u9645\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u5e94\u7528":12,"\u5728\u4f7f\u7528twine\u4e0a\u4f20\u4e4b\u524d":62,"\u5728\u51fa\u73b0\u5355\u70b9\u6545\u969c\u65f6":11,"\u5728\u5b9e\u73b0\u6bcf\u4e2a\u5b50\u7c7b\u7684\u65f6\u5019\u5c31\u4e0d\u9700\u8981\u5173\u5fc3\u5206\u652f\u7684\u4e8b\u60c5\u4e86":42,"\u5728\u5b9e\u73b0\u8fc7\u7a0b\u4e2d":46,"\u5728\u5bf9\u5e94\u7684":41,"\u5728\u5c42\u521d\u59cb\u5316\u7684\u65f6\u5019":41,"\u5728\u5f00\u59cb\u8bad\u7ec3\u4e4b\u524d":12,"\u5728\u5f02\u6784\u96c6\u7fa4\u4e2d":11,"\u5728\u5f15\u5165\u5176\u4ed6\u7c7b\u578b\u7684\u5934\u6587\u4ef6\u65f6":46,"\u5728\u5feb\u7167\u5199\u5165\u5b8c\u6210\u540e":11,"\u5728\u60a8\u7684\u5b9e\u9645\u73af\u5883\u4e2d":11,"\u5728\u6709\u666e\u901a\u7684cpu":42,"\u5728\u672c\u6587\u6863\u4e2d":28,"\u5728\u673a\u7fa4\u4e0a\u8fd0\u884c\u8f6c\u6362\u7a0b\u5e8f":12,"\u5728\u6837\u4f8b\u4e2d":46,"\u5728\u7528\u6237\u4f7f\u7528c":46,"\u5728\u7b2c\u4e8c\u4e2atab":62,"\u5728\u7ebf\u6a21\u578b\u9884\u6d4b\u670d\u52a1":12,"\u5728\u8bad\u7ec3\u7ed3\u675f\u7684\u65f6\u5019\u518d\u4fdd\u5b58\u4e3apaddlepaddle\u7684\u683c\u5f0f":42,"\u5728\u8bc4\u5ba1\u8fc7\u7a0b\u4e2d":62,"\u5728\u8fd9\u4e2a":62,"\u5728\u8fd9\u4e2a\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":45,"\u5728\u8fd9\u4e2a\u9636\u6bb5\u7684\u4ee3\u7801\u6b63\u5728\u7ecf\u5386\u56de\u5f52\u6d4b\u8bd5":62,"\u5728\u8fd9\u4e9b\u5934\u6587\u4ef6\u4e2d":46,"\u5728\u8fd9\u4e9b\u6587\u4ef6\u4e2d":46,"\u5728\u91cd\u6784\u524d\u7684paddlepaddle\u4e2d":42,"\u5728\u95ee\u9898\u672c\u8eab\u7684\u8ba1\u7b97\u91cf\u6bd4\u8f83\u5c0f\u7684\u65f6\u5019":41,"\u5728batch":41,"\u5728c":45,"\u5728c\u7684\u5934\u6587\u4ef6":45,"\u5728packing\u4e0a\u7684\u8017\u65f6":41,"\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":12,"\u5728paramet":11,"\u5728rnn\u7684\u60c5\u51b5\u4e0b":41,"\u5747\u4f1a\u88ab\u5b89\u88c5\u5230includ":46,"\u5747\u662f\u5728":46,"\u57fa\u4e8e\u7c98\u6027\u4f1a\u8bdd\u7684\u8d1f\u8f7d\u5747\u8861\u529f\u80fd":28,"\u5916\u90e8\u5b58\u50a8":42,"\u591a\u4e2a\u503c":12,"\u591a\u4e2aparamet":11,"\u591a\u6b21\u8c03\u7528":41,"\u5927\u591a\u6570\u8bed\u8a00\u90fd\u652f\u6301\u4f7f\u7528c\u8bed\u8a00api":45,"\u5982\u56fe\u4e2dtrainer":11,"\u5982\u679c\u4e0a\u9762\u4e24\u6b65\u51fa\u73b0\u9519\u8bef":11,"\u5982\u679c\u4e0d\u9700\u8981\u5916\u90e8\u5b58\u50a8\u7528\u4e8e\u8f6c\u6362":42,"\u5982\u679c\u4f7f\u7528swig\u6211\u4eec\u9700\u8981\u5c06\u5728interface\u6587\u4ef6\u91cc":45,"\u5982\u679c\u5728\u4f7f\u7528mkl":42,"\u5982\u679c\u5931\u8d25":62,"\u5982\u679c\u5b58\u5728\u6570\u636e\u6392\u5217\u683c\u5f0f\u4e0d\u4e00\u6837\u7684\u60c5\u51b5\u65f6":42,"\u5982\u679c\u5b58\u5728\u67d0\u4e9btrainer\u6267\u884c\u901f\u5ea6\u8fc7\u6162\u4f1a\u5f71\u54cd\u6574\u4f53\u96c6\u7fa4\u7684\u901f\u5ea6":11,"\u5982\u679c\u5df2\u7ecf\u6b63\u5728\u6267\u884c\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":11,"\u5982\u679c\u662f\u5176\u5b83\u7c7b\u578b":12,"\u5982\u679c\u6709bugfix\u7684\u884c\u4e3a":62,"\u5982\u679c\u67d0\u4e00\u4e2a\u7c7b\u578b\u9700\u8981\u5f15\u7528\u53e6\u4e00\u4e2a\u7c7b\u578b":46,"\u5982\u679c\u67d0\u4e00\u4e2apaddl":46,"\u5982\u679c\u67d0\u4e00\u4e2apaddle\u6982\u5ff5\u5fc5\u987b\u8981\u66b4\u9732":46,"\u5982\u679c\u6ee1\u8db3\u6761\u4ef6":11,"\u5982\u679c\u7528\u6237\u8981\u628apaddle\u7684\u9759\u6001\u5e93":45,"\u5982\u679c\u8981\u4e0a\u4f20gpu\u7248\u672c\u7684\u5305":62,"\u5982\u679c\u8c03\u7528\u9759\u6001\u5e93\u53ea\u80fd\u5c06\u9759\u6001\u5e93\u4e0e\u89e3\u91ca\u5668\u94fe\u63a5":45,"\u5982\u679c\u9700\u8981\u624b\u52a8\u7f16\u8bd1":62,"\u5982\u679cmkl":42,"\u5982\u679cparamet":11,"\u5b50\u7c7b\u53ea\u9700\u8981\u4f7f\u7528\u5b9a\u4e49\u597d\u7684\u63a5\u53e3":42,"\u5b57\u6bb5\u8bbe\u4e3a":62,"\u5b57\u7b26\u4e32":12,"\u5b58\u50a8":12,"\u5b66\u4e60\u6210\u672c\u9ad8":45,"\u5b83\u4eec\u4e3b\u8981\u662f\u7528\u4e8e":42,"\u5b83\u4eec\u7684\u6587\u4ef6\u540d\u662f":12,"\u5b83\u53ea\u4f1a\u5305\u62ec\u751f\u6210\u597d\u7684\u52a8\u6001\u5e93\u548c\u5934\u6587\u4ef6":42,"\u5b83\u8d1f\u8d23\u51b3\u5b9a\u7f16\u8bd1\u65f6\u662f\u5426\u4f7f\u7528mklml\u548cmkl":42,"\u5b89\u88c5\u540e\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":46,"\u5b8c\u6210\u4e00\u4e2a\u4f20\u8f93\u52a8\u4f5c\u5b8c\u6210\u7684\u65f6\u95f4\u4e5f\u6bd4\u8f83\u77ed":28,"\u5b8c\u6210\u5e38\u7528layer\u7684mkl":42,"\u5b8c\u6210\u5e38\u89c1\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edcvgg":42,"\u5b8c\u6210\u6570\u636e\u7684\u9884\u5904\u7406":12,"\u5b8c\u6210\u81ea\u52a8\u5316\u4e8c\u8fdb\u5236\u7f16\u8bd1":62,"\u5b9a\u4e49":42,"\u5b9a\u4e49\u4e00\u4e9b\u9664\u4e86layer\u548cmemory\u76f8\u5173\u7684\u7c7b\u548c\u51fd\u6570":42,"\u5b9e\u73b0\u5177\u4f53\u7684\u51fd\u6570\u529f\u80fd\u5373\u53ef":42,"\u5b9e\u73b0\u7b80\u5355":45,"\u5bf9\u4e8e\u4e0d\u540c\u8bed\u8a00":45,"\u5bf9\u4e8e\u540c\u4e00\u6bb5c":45,"\u5bf9\u4e8e\u540c\u6837\u8bbe\u7f6e\u7684\u7f51\u7edc\u6a21\u578b":41,"\u5bf9\u4e8e\u591a\u8bed\u8a00\u63a5\u53e3":45,"\u5bf9\u4e8e\u5927\u591a\u6570\u8bed\u8a00":45,"\u5bf9\u4e8e\u5e8f\u5217\u957f\u5ea6":41,"\u5bf9\u4e8e\u6709\u53c2\u6570\u7684\u5c42":42,"\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u65b0\u52a0\u7684rnn":41,"\u5bf9\u4e8e\u6bcf\u79cd\u7c7b\u578b":46,"\u5bf9\u4e8e\u6bcf\u79cdc":46,"\u5bf9\u65b0\u7684\u6743\u91cd\u8fdb\u884c\u8f6c\u6362\u7528\u4e8e\u4e0b\u6b21\u8fed\u4ee3":41,"\u5bf9\u6bd4":45,"\u5bf9\u6bd4\u4f18\u5316\u540elayer\u4e0e\u76f8\u5bf9\u5e94\u7684paddlepaddle\u539f\u6709lay":41,"\u5bf9\u6bd4\u4f18\u5316\u540elayer\u81ea\u8eab":41,"\u5bf9\u8f93\u5165\u53c2\u6570\u7684\u5b89\u5168\u6027\u8fdb\u884c\u4e86\u5fc5\u8981\u7684\u5224\u65ad":46,"\u5bf9\u8fd9\u4e2a\u7248\u672c\u7684\u63d0\u4ea4":62,"\u5bfb\u627e\u6709\u6ca1\u6709\u5176\u4ed6\u53ef\u4ee5\u4f18\u5316\u7684\u53ef\u80fd":42,"\u5bfc\u51fa\u8fd9\u4e9b\u63a5\u53e3":46,"\u5c06":62,"\u5c06\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u62c6\u5206\u6210\u591a\u4efd":11,"\u5c06\u5927\u91cf\u7684":45,"\u5c06\u65b0\u5206\u652f\u7684\u7248\u672c\u6253\u4e0atag":62,"\u5c06master\u5206\u652f\u7684\u5408\u5165commit\u6253\u4e0atag":62,"\u5c0f\u4e8e\u67d0\u4e2a\u6bd4\u8f83\u5c0f\u7684\u9608\u503c\u8ba4\u4e3a\u901a\u8fc7":42,"\u5c31\u9700\u8981\u5bf9\u8fd9\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00\u589e\u52a0\u4e00\u4e9b\u5b9a\u4e49":45,"\u5de5\u5177\u4e0a\u4f20\u5373\u53ef":62,"\u5e73\u5747\u6545\u969c\u4fee\u590d\u65f6\u95f4":11,"\u5e73\u5747\u6545\u969c\u7387":11,"\u5e76\u4e14\u4f1a\u5199\u597d":42,"\u5e76\u4e14\u4f7f\u7528":46,"\u5e76\u4e14\u53ea\u9700\u8981\u5728\u5fc5\u8981\u7684\u65f6\u5019\u8f6c\u6362\u8fd9\u79cd\u683c\u5f0f":42,"\u5e76\u4e14\u5728\u5e38\u89c1\u7684\u5e73\u53f0\u4e0a":45,"\u5e76\u4e14\u5f53\u7f16\u8bd1\u65f6":41,"\u5e76\u4e14\u628a\u7cfb\u7edf\u751f\u6210\u7684ca":28,"\u5e76\u4e14\u628a\u7ed3\u679c\u8fd4\u56depfsclient\u7aef":28,"\u5e76\u4e14\u8ba9\u63a5\u53e3\u8131\u79bb\u5b9e\u73b0\u7ec6\u8282":45,"\u5e76\u4e14\u8f93\u5165\u8f93\u51fa\u90fd\u662f\u5171\u7528\u4e00\u5757\u5185\u5b58":42,"\u5e76\u5220\u9664":62,"\u5e76\u5220\u9664\u66f4\u65e9\u7684\u5feb\u7167":11,"\u5e76\u52a0\u8f7d\u5176\u4e2d\u7684\u53c2\u6570":11,"\u5e76\u53d1\u5e03\u5230pypi":62,"\u5e76\u5728\u6bcf\u6b21\u6743\u91cd\u66f4\u65b0\u540e":41,"\u5e76\u5728\u96c6\u7fa4\u4e2d\u8fd0\u884c\u591a\u4e2a\u5206\u5e03\u5f0f\u6570\u636e\u5904\u7406\u4efb\u52a1":12,"\u5e76\u5c06":62,"\u5e76\u5c06c":46,"\u5e76\u628a\u5feb\u7167\u4fdd\u5b58\u5230\u8fd9\u4e2a\u76ee\u5f55\u4e0b":11,"\u5e76\u628a\u7ed3\u679c\u653e\u5230\u5f53\u524d\u5c42\u7684":42,"\u5e76\u6ca1\u6709paddle\u7279\u522b\u9700\u8981\u7684\u7279\u6027":45,"\u5e76\u6dfb\u52a0\u5934\u6587\u4ef6":41,"\u5e76\u88ab\u5b58\u50a8\u5728\u8bf8\u5982hadoop":12,"\u5e76\u9002\u5e94github\u7684\u7279\u6027\u505a\u4e86\u4e00\u4e9b\u533a\u522b":62,"\u5e76\u91cd\u65b0\u6253\u5305wheel\u5305":62,"\u5efa\u8bae":62,"\u5f00\u53d1\u4e86\u6a21\u578b\u9884\u6d4b\u7684\u6837\u4f8b\u4ee3\u7801":46,"\u5f00\u53d1\u8005\u4fee\u6539\u81ea\u5df1\u7684\u4ee3\u7801":62,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4e2d":62,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4f7f\u7528":62,"\u5f00\u5934":[41,42],"\u5f00\u59cb\u63d0\u4f9b\u670d\u52a1":11,"\u5f15\u5165\u4e86\u4ee5\u4e0b\u56db\u4e2aapi":41,"\u5f15\u5165\u4e86\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5f39\u51fa\u4e0b\u9762\u7684\u9009\u62e9\u6846":62,"\u5f53\u529f\u80fd\u5206\u652f\u5f00\u53d1\u5b8c\u6bd5\u540e":62,"\u5f53\u53ea\u505a\u63a8\u65ad":41,"\u5f53\u5f00\u542f":42,"\u5f53\u6253\u5f00":42,"\u5f53\u6570\u636e\u683c\u5f0f\u4e0epaddlepaddle\u9ed8\u8ba4\u7684":42,"\u5f53\u7136\u8fd9\u4e24\u8005\u4e5f\u53ef\u4ee5\u76f8\u7b49":42,"\u5f53\u7528\u6237\u4f7f\u7528\u5b8c\u8fd9\u4e2a\u53c2\u6570\u540e":46,"\u5f53\u7f51\u7edc\u51fa\u73b0\u5206\u652f\u4e14\u5728":42,"\u5f53destination\u6587\u4ef6\u4e0d\u5b58\u5728\u6216\u8005\u5927\u5c0f\u548csource\u6587\u4ef6\u4e0d\u4e00\u81f4\u65f6":28,"\u5f88\u96be\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":45,"\u5f97\u4f7f\u7528":45,"\u5fc5\u8981":46,"\u5fc5\u987b\u5206\u522b\u4e0e":42,"\u60c5\u611f\u5206\u6790":62,"\u6211\u4eec\u4e5f\u53ef\u4ee5\u786e\u5b9a\u6bcf\u4e00\u4e2a\u53c2\u6570\u7684\u7c7b\u578b":46,"\u6211\u4eec\u4e5f\u5c06mklml\u5373":42,"\u6211\u4eec\u4f1a\u4fdd\u8bc1":42,"\u6211\u4eec\u4f1a\u5728\u7f51\u7edc\u8bad\u7ec3\u4e4b\u524d\u628a\u683c\u5f0f\u8f6c\u6362\u4e3amkl":42,"\u6211\u4eec\u4f1a\u5bf9\u6bd4\u5982\u4e0b2\u4e2a\u65b9\u9762":41,"\u6211\u4eec\u4f1a\u628amkl":42,"\u6211\u4eec\u4f1a\u6dfb\u52a0":[41,42],"\u6211\u4eec\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":45,"\u6211\u4eec\u51b3\u5b9a\u4f7f\u7528\u5df2\u6709\u7684":42,"\u6211\u4eec\u53ef\u4ee5\u5148\u5b8c\u6210\u5bf9\u539f\u6570\u636e\u7684packing\u64cd\u4f5c":41,"\u6211\u4eec\u603b\u7ed3\u51fa\u4e00\u4e9b\u7279\u522b\u9700\u8981\u6ce8\u610f\u7684\u70b9":42,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u8f6c\u6362\u65b9\u5f0f":12,"\u6211\u4eec\u63d0\u51fa\u4e86chunk\u7684\u6982\u5ff5":28,"\u6211\u4eec\u6700\u7ec8\u7684\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165python\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u8bed\u8a00\u7684\u89e3\u91ca\u5668":45,"\u6211\u4eec\u8ba1\u5212\u5c06":41,"\u6211\u4eec\u8ba1\u5212\u5c06\u82f1\u7279\u5c14\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u6570\u5b66\u5e93":42,"\u6211\u4eec\u8bbe\u8ba1\u8bf4\u660e\u4e86\u540d\u4e3afilemanager\u7cfb\u7edf":28,"\u6211\u4eec\u9009\u62e9":12,"\u6211\u4eec\u901a\u8fc7\u4f7f\u7528\u65b0\u5f15\u5165\u7684gemm":41,"\u6211\u4eec\u90fd\u63d0\u4f9bpython\u7684\u8f6c\u6362\u5e93":12,"\u6216\u8005":[42,45,46],"\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":11,"\u6216\u8005\u7528tuple\u8868\u793a\u7684\u591a\u4e2a\u503c":12,"\u6216\u8005\u7531\u5b83\u4eec\u7ec4\u6210\u7684list":12,"\u6216activ":42,"\u6240\u4ee5":[42,62],"\u6240\u4ee5\u4e00\u4e2a\u7248\u672c\u53f7\u7684wheel\u5305\u53d1\u5e03\u4e4b\u540e":62,"\u6240\u4ee5\u4e0d\u5b58\u5728\u8fd9\u4e2a\u95ee\u9898":42,"\u6240\u4ee5\u5728":42,"\u6240\u4ee5\u5728\u5199\u5165\u5feb\u7167\u7684\u8fc7\u7a0b\u4e2d":11,"\u6240\u4ee5\u6211\u4eec\u5b9a\u4e49\u4e86\u4e00\u4e2a":42,"\u6240\u4ee5\u6574\u4f53\u4e0a":42,"\u6240\u4ee5\u6dfb\u52a0\u4e86\u5bf9\u5e94\u7684":42,"\u6240\u4ee5\u7528\u6237\u9700\u8981\u9996\u5148\u5728":28,"\u6240\u4ee5\u9700\u8981\u5f15\u5165\u4e00\u4e2a\u8f6c\u6362\u65b9\u6cd5":42,"\u6240\u6709\u4e0e\u7c7b\u578b\u76f8\u5173\u7684\u51fd\u6570":46,"\u6240\u6709\u5916\u90e8\u7684\u8f6c\u6362\u5de5\u4f5c\u90fd\u4f1a\u5728reset\u7cfb\u5217\u51fd\u6570\u4e2d\u90fd\u51c6\u5907\u597d":42,"\u6240\u6709\u7684":41,"\u6240\u6709\u7684\u63a5\u53e3\u5747\u4e3ac\u63a5\u53e3":46,"\u6240\u6709\u76f8\u5173\u7684":41,"\u6240\u6709\u7c7b\u578b\u540d\u4e3a":46,"\u6240\u6709mkl":42,"\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":45,"\u624d\u80fd\u66f4\u597d\u7684\u53d1\u6325mkl":42,"\u6253\u5f00\u8fd9\u4e2a\u7f16\u8bd1\u9009\u9879":46,"\u6267\u884c":62,"\u628a":12,"\u628a\u4e4b\u524d\u793a\u4f8b\u4e2d\u8f6c\u6362\u5b8c\u6bd5\u7684random":12,"\u6307\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u4e4b\u540e\u5f97\u5230\u7684\u6240\u6709\u53c2\u6570":11,"\u6309\u94ae":62,"\u63a5\u53e3":[45,46],"\u63a5\u53e3\u5c42\u505a\u8fc7\u591a\u5c01\u88c5":46,"\u63a5\u53e3\u662f":12,"\u63a5\u6536\u5904\u7406pfsclient\u7aef\u7684\u6587\u4ef6\u7ba1\u7406\u8bf7\u6c42":28,"\u63a7\u5236\u662f\u5426\u4f7f\u7528mkl":42,"\u63a7\u5236\u662f\u5426\u4f7f\u7528mklml\u5e93":42,"\u63a7\u5236\u7528\u6237\u6743\u9650":12,"\u63d0\u4f9b\u4e03\u5c42\u534f\u8bae\u7684\u53cd\u5411\u4ee3\u7406":28,"\u63d0\u4f9b\u5e38\u7528\u7684\u547d\u4ee4\u884c\u7ba1\u7406\u547d\u4ee4\u7ba1\u7406\u6587\u4ef6\u548c\u76ee\u5f55":28,"\u63d0\u4f9b\u7528\u6237\u7ba1\u7406\u6587\u4ef6\u7684\u547d\u4ee4":28,"\u63d0\u4f9b\u7ed9paddle\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e":12,"\u652f\u6301\u5927\u6587\u4ef6\u7684\u65ad\u70b9\u4e0a\u4f20":28,"\u6570\u636e":28,"\u6570\u636e\u8bfb\u53d6\u5747\u4ea4\u7531\u5176\u4ed6\u8bed\u8a00\u5b8c\u6210":45,"\u6570\u636e\u957f\u5ea6\u53ca\u6821\u9a8c\u503c\u7ec4\u6210":28,"\u6570\u636e\u96c6\u9700\u8981\u9884\u5148\u88ab\u8f6c\u6362\u6210paddlepaddle\u5206\u5e03\u5f0f\u8bad\u7ec3\u4f7f\u7528\u7684\u5b58\u50a8\u683c":12,"\u6570\u636e\u9884\u5904\u7406\u4efb\u52a1":12,"\u6587\u4ef6":45,"\u6587\u4ef6\u4f20\u8f93\u7684\u7684\u5173\u952e\u5728\u4e8e\u9700\u8981pfsclient\u7aef\u5bf9\u6bd4source\u548cdestination\u7684\u6587\u4ef6chunks\u7684checksum\u662f\u5426\u4fdd\u6301\u4e00\u81f4":28,"\u6587\u4ef6\u5185\u5bb9\u4e3a":45,"\u6587\u4ef6\u540d\u4e3a\u6b64uuid":11,"\u6587\u4ef6\u5bf9\u5e94\u7684data":12,"\u6587\u4ef6\u7684\u4e0a\u4f20\u548c\u4e0b\u8f7d\u90fd\u662f\u901a\u8fc7\u5bf9chunk\u7684\u64cd\u4f5c\u6765\u5b9e\u73b0\u7684":28,"\u65b0\u624b\u5165\u95e8\u7ae0\u8282":62,"\u65b0\u7248\u672c":42,"\u65b9\u4fbf\u6d4b\u8bd5\u4eba\u5458\u6d4b\u8bd5paddlepaddle\u7684\u884c\u4e3a":62,"\u65b9\u4fbf\u7528\u6237\u4e0a\u4f20\u81ea\u5df1\u7684\u8bad\u7ec3\u6570\u636e\u4ee5\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":28,"\u65b9\u4fbf\u7528\u6237\u5728python\u7aef\u9009\u62e9\u662f\u5426\u542f\u7528\u8fd9\u4e2a\u529f\u80fd":41,"\u65b9\u4fbf\u7528\u6237\u9009\u62e9\u4f7f\u7528mkl":42,"\u65b9\u5f0f\u7c7b\u4f3c\u4e8e":42,"\u65e0\u6cd5\u505a\u5230\u5bf9\u4e8e\u5404\u79cd\u8bed\u8a00\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u7684\u9002\u914d":45,"\u65e0\u8bba\u5728\u672c\u5730\u8fd8\u662f\u5728\u4e91\u7aef":12,"\u65e0\u8bba\u662f\u4ece":12,"\u65e0\u8bba\u662f\u5728\u672c\u5730\u6216\u662f\u4e91\u7aef\u8f6c\u6362":12,"\u65e0\u8bba\u662f\u91cd\u6784\u524d\u7684layer\u8fd8\u662f\u91cd\u6784\u540e\u7684op":42,"\u65f6":[11,41,42],"\u65f6\u4e00\u8d77\u66f4\u65b0":42,"\u662f":[28,42],"\u662f\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3\u7684\u4ee3\u7801\u751f\u6210\u5668":45,"\u662f\u4e00\u4e2a\u7c7b\u578b\u7684\u6807\u5fd7":46,"\u662f\u4e0d\u5e38\u89c1\u7684\u505a\u6cd5":45,"\u662f\u5404\u4e2a\u5b9e\u73b0\u4e2d\u5171\u4eab\u7684\u5934\u6587\u4ef6":46,"\u662f\u5426\u6253\u5f00":41,"\u662f\u56e0\u4e3ac99\u652f\u6301":45,"\u662f\u5bf9\u7528\u6237\u6587\u4ef6\u5b58\u50a8\u7a7a\u95f4\u7684\u62bd\u8c61":28,"\u662f\u6307":46,"\u662f\u7528\u6237\u4f7f\u7528c":46,"\u662fc":46,"\u663e\u5f97\u76f8\u5bf9\u6765\u8bf4\u8f83\u4e3a\u8017\u65f6":41,"\u6682\u65f6\u4e0d\u8003\u8651\u591a\u4e2aparamet":11,"\u66b4\u9732\u8fd9\u4e2a\u6982\u5ff5\u5fc5\u8981\u51fd\u6570":46,"\u6700\u540e\u5220\u9664":62,"\u6700\u5e38\u89c1\u7684\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662fexcept":45,"\u6709\u6548\u63d0\u5347paddlepaddle\u5728\u82f1\u7279\u5c14\u67b6\u6784\u4e0a\u7684\u6027\u80fd":[41,42],"\u6709\u6807\u51c6\u7684":45,"\u6709\u7684\u65f6\u5019":45,"\u672c\u5217\u8868\u8bf4\u660epaddlepaddle\u53d1\u7248\u4e4b\u524d\u9700\u8981\u6d4b\u8bd5\u7684\u529f\u80fd\u70b9":62,"\u672c\u6587\u6863\u63cf\u8ff0paddl":46,"\u673a\u5668\u7ffb\u8bd1":62,"\u6765\u4fdd\u8bc1\u8bad\u7ec3\u8fc7\u7a0b\u53ef\u4ee5\u4ece\u4e2d\u95f4\u72b6\u6001\u91cd\u65b0\u542f\u52a8":11,"\u6765\u51b3\u5b9a\u662f\u5426\u5f00\u542fmkl":41,"\u6765\u5b9e\u73b0":42,"\u6765\u786e\u4fdd\u628a":45,"\u6765\u8868\u793apaddle\u5185\u90e8\u7c7b":45,"\u6765\u8bbf\u95ee\u7528\u6237\u81ea\u5df1\u7684\u6570\u636e":12,"\u6765\u8fdb\u884c\u8ba8\u8bba":46,"\u67e5\u770blatest":62,"\u6807\u51c6\u8868\u793apaddlepaddle\u7248\u672c\u53f7":62,"\u683c\u5f0f\u4e0d\u5339\u914d\u65f6":42,"\u68c0\u67e5\u70b9\u4fdd\u5b58\u7a0b\u5e8f\u6d41\u7a0b":11,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\u901a\u8fc7\u5b9a\u671f\u5411\u78c1\u76d8\u4e0a\u4fdd\u5b58\u4e00\u4efd\u5b58\u50a8\u5728paramet":11,"\u6a21\u578b\u6570\u636e\u68c0\u67e5\u70b9\u7684\u5b9e\u73b0":11,"\u6a21\u578b\u914d\u7f6e\u89e3\u6790":45,"\u6b21\u8fed\u4ee3\u6267\u884c\u7684\u8f6c\u6362\u6b21\u6570\u4e3a":41,"\u6b64\u65f6\u6bcf\u4e2a\u5c0f\u5206\u652f\u7684":42,"\u6b64\u65f6master\u5c06\u8d1f\u8d23\u542f\u52a8\u4e00\u4e2a\u65b0\u7684train":11,"\u6bcf\u4e00\u4e2a":62,"\u6bcf\u4e00\u4e2a\u6587\u4ef6\u662f\u6570\u636e\u96c6\u7684\u4e00\u4e2ashard":12,"\u6bcf\u4e2a":42,"\u6bcf\u4e2a\u503c\u7684\u7c7b\u578b\u53ef\u4ee5\u662f\u6574\u5f62":12,"\u6bcf\u4e2a\u6d4b\u8bd5\u4f1a\u5bf9\u6bd4paddlepaddle\u4e2dcpu\u7b97\u51fa\u7684\u7ed3\u679c\u4e0emkl":42,"\u6bcf\u4e2adata":12,"\u6bcf\u4e2amkldnnlayer\u90fd\u5305\u542b\u7528\u4e8e\u5185\u90e8\u5b58\u50a8\u548c\u5916\u90e8\u5b58\u50a8\u7684\u4e00\u7cfb\u5217mkldnnmatrix":42,"\u6bcf\u4e2aparamet":11,"\u6bcf\u4e2ashard\u5206\u522b\u5b58\u50a8\u5728\u5176\u4e2d\u4e00\u53f0paramet":11,"\u6bcf\u6b21\u8c03\u7528\u65f6\u5bf9\u539f\u6570\u636e\u7684\u91cd\u590dpacking\u4fbf\u6210\u4e3a\u4e86\u5197\u4f59":41,"\u6bcf\u6b21\u8f93\u51fa\u4e00\u4e2adata":12,"\u6bcf\u969410\u5206\u949f":11,"\u6bd4\u5982":[12,42],"\u6bd4\u5982\u53ef\u80fd\u4f1a\u7528openmp\u6539\u8fdbsgd\u7684\u66f4\u65b0\u6027\u80fd":42,"\u6bd4\u5982\u5c06":62,"\u6bd4\u5982\u6bcf\u969410\u5206\u949f\u6700\u65b0\u7684\u5feb\u7167":11,"\u6bd4\u5982\u6d41\u5f0f\u6570\u636e\u5904\u7406":12,"\u6bd4\u5982imagenet\u8fd9\u4e2a\u6570\u636e\u96c6\u53ef\u80fd\u88ab\u5206\u62101000\u4e2ashard":12,"\u6ca1\u6709\u5fc5\u8981\u5728\u6bcf\u6b21\u524d\u5411\u4e2d\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8ba1\u7b97\u65f6\u5bf9\u6743\u91cd\u8fdb\u884c\u91cd\u590d\u7684packing\u64cd\u4f5c":41,"\u6ce8":[11,62],"\u6ce8\u518clayer\u7684\u65f6\u5019\u4fdd\u8bc1":[41,42],"\u6ce8\u610f":42,"\u6d4b\u8bd5\u5206\u4e3a\u6bcf\u4e2alayer":42,"\u6d4b\u8bd5\u672c\u6b21release\u7684\u6b63\u786e\u6027":62,"\u6d4b\u8bd5\u7684\u6027\u80fd\u5bf9\u6bd4\u7ed3\u679c\u4f1a\u5728":42,"\u6d6e\u70b9\u578b\u6570\u636e":12,"\u6df1\u5165paddlepaddl":42,"\u6dfb\u52a0":41,"\u6dfb\u52a0\u7684\u76f8\u5173\u6587\u4ef6\u548c\u76ee\u5f55\u7ed3\u6784\u5982\u4e0b":[41,42],"\u6fc0\u6d3b\u51fd\u6570\u662f\u72ec\u7acb\u4e8e":42,"\u70b9\u51fb":62,"\u7136\u540e\u5728\u524d\u5411":41,"\u7136\u540e\u5728etcd\u7684":11,"\u7136\u540e\u5c31\u53ef\u4ee5\u5e76\u53d1\u5199\u5165\u591a\u4e2achunk":28,"\u7136\u540e\u624d\u80fd\u4f7f\u7528pfsclient":28,"\u7136\u540e\u6309\u7167\u4e0a\u8ff0\u7684\u65b9\u6cd5":62,"\u7136\u540e\u70b9\u51fb":62,"\u7248\u672c\u5206\u652f":62,"\u7248\u672c\u53f7":62,"\u7248\u672c\u53f7\u5bf9\u5e94\u7684tag\u5373\u53ef":62,"\u7248\u672c\u53f7rc":62,"\u7248\u672cfork\u51fa\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f":62,"\u7279\u6709\u7684\u8bbe\u5907id":42,"\u73b0\u9636\u6bb5\u7684\u4f18\u5316\u4e3b\u8981\u9488\u5bf9":41,"\u73b0\u9636\u6bb5paddle\u6709\u4e00\u4e2a\u95ee\u9898\u662f":45,"\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u8bad\u7ec3\u6570\u636e\u96c6\u901a\u5e38\u4f53\u79ef\u5f88\u5927":12,"\u751f\u4ea7\u73af\u5883\u7684\u65e5\u5fd7\u6570\u636e\u4f1a\u901a\u8fc7\u5b9e\u65f6\u6d41\u7684\u65b9\u5f0f":12,"\u751f\u6210\u5404\u79cd\u8bed\u8a00\u7684\u7ed1\u5b9a\u4ee3\u7801":45,"\u751f\u6210\u6587\u6863":45,"\u751f\u6210\u7684":12,"\u751f\u6210\u7ed9\u5b9a":12,"\u751f\u6210api\u6587\u6863":45,"\u751f\u6210pfsclient\u548cpfsserver\u7684\u6846\u67b6\u90e8\u5206":28,"\u7528":28,"\u7528\u4e8e\u6d4b\u8bd5\u548c\u5bf9\u6bd4\u5728\u4f7f\u7528mkl":42,"\u7528\u4e8e\u7ba1\u7406mkl":42,"\u7528\u4e8e\u9009\u62e9\u662f\u5426\u4f7f\u7528\u76f8\u5173\u529f\u80fd":41,"\u7528\u4e8e\u9009\u62e9\u662f\u5426\u4f7f\u7528mkl":42,"\u7528\u4e8emkl":[41,42],"\u7528\u6237\u4e0a\u4f20\u6570\u636e\u540e":12,"\u7528\u6237\u4e5f\u53ef\u4ee5\u4e0a\u4f20label":12,"\u7528\u6237\u53ef\u4ee5\u5b89\u5168\u7684\u91ca\u653e\u67d0\u4e2ac":46,"\u7528\u6237\u53ef\u4ee5\u628a\u81ea\u5df1\u7684\u6570\u636e\u5206\u4eab\u7ed9\u522b\u4eba":12,"\u7528\u6237\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u8fd9\u4e2a\u52a8\u6001\u5e93\u6765\u5f15\u5165paddl":46,"\u7528\u6237\u5728\u672c\u5730\u8f6c\u6362\u597d\u518d\u4e0a\u4f20":12,"\u7528\u6237\u6587\u4ef6\u53ef\u80fd\u662f\u6bd4\u8f83\u5927\u7684":28,"\u7528\u6237\u901a\u8fc7c":46,"\u7531\u4e8e\u5728\u73b0\u6709\u7684\u67d0\u4e9b\u60c5\u51b5\u4e0b":41,"\u7531\u4e8e\u5bf9parameters\u7684\u66f4\u65b0\u9700\u8981\u83b7\u53d6parameters\u5185\u5b58\u7684":11,"\u7531\u4e8e\u96c6\u7fa4\u4e2d\u540c\u65f6\u5b58\u5728\u4e24\u53f0\u673a\u5668\u6545\u969c\u7684\u6982\u7387\u6781\u4f4e":11,"\u7531\u4e8ec":45,"\u7531\u4e8echunk\u6bd4\u8f83\u5c0f":28,"\u7531\u4e8emkl":42,"\u7531\u4e8epypi":62,"\u7531\u5206\u652f\u5904\u7684layer\u8d1f\u8d23\u6c42\u548c":42,"\u7533\u8bf7\u7528\u6237\u7a7a\u95f4":28,"\u7684\u4e00\u4e2a\u5b50\u96c6":42,"\u7684\u4fe1\u606f":42,"\u7684\u5355\u5143\u6d4b\u8bd5\u548c\u7b80\u5355\u7f51\u7edc\u7684\u6574\u4f53\u6d4b\u8bd5":42,"\u7684\u547d\u540d\u98ce\u683c\u5e76\u4e0d\u80fd\u9002\u5e94\u5176\u4ed6\u7b2c\u4e09\u65b9\u8bed\u8a00":45,"\u7684\u57fa\u672c\u903b\u8f91":42,"\u7684\u5934\u6587\u4ef6":45,"\u7684\u5b50\u7c7b\u53ea\u9700\u8981\u4f7f\u7528\u5185\u90e8\u5b58\u50a8\u5c31\u53ef\u4ee5\u4e86":42,"\u7684\u60c5\u51b5\u4e0b":41,"\u7684\u63a5\u53e3\u6837\u5f0f":45,"\u7684\u6570\u636e\u6d41\u56fe":12,"\u7684\u65f6\u5019":42,"\u7684\u683c\u5f0f\u59cb\u7ec8\u662f":42,"\u7684\u683c\u5f0f\u5b58\u50a8":42,"\u7684\u6982\u5ff5":42,"\u7684\u6e90\u7801\u91cc\u4f7f\u7528\u4e86":45,"\u7684\u7248\u672c":62,"\u7684\u7ed3\u679c":41,"\u7684\u7f29\u5199":28,"\u7684\u7f51\u7edc\u6a21\u578b":41,"\u7684\u89c4\u8303":45,"\u7684\u89d2\u5ea6":12,"\u7684\u914d\u7f6e\u5199\u5230\u914d\u7f6e\u6587\u4ef6\u4e2d":12,"\u7684flag":[41,42],"\u7684vanilla":41,"\u76ee\u524d\u53ea\u8003\u8651":42,"\u76ee\u524d\u53ea\u8003\u8651\u52a8\u6001\u6269\u5bb9trainer\u6570\u91cf":11,"\u76ee\u524d\u5728paddlepaddle\u4e2d":42,"\u76ee\u524d\u5728paddlepaddle\u4e2d\u6570\u636e\u90fd\u662f\u4ee5":42,"\u76ee\u524d\u5d4c\u5165python\u89e3\u91ca\u5668":45,"\u76ee\u524d\u6211\u4eec\u7528cephfs\u6765\u642d\u5efa":28,"\u76ee\u524d\u7684\u4f18\u5316":42,"\u76ee\u524dpaddle\u7684\u8fdb\u7a0b\u6a21\u578b\u662fc":45,"\u76ee\u524dpaddlepaddle\u91c7\u7528\u4e86":41,"\u76ee\u5f55\u4e0b":46,"\u76ee\u5f55\u4e0b\u5bf9\u5e94\u7684\u5730\u65b9":42,"\u76f4\u63a5\u4f7f\u7528c\u8bed\u8a00\u7684":45,"\u76f4\u63a5\u5220\u9664\u8fd9\u4e2a\u53c2\u6570\u5373\u53ef":46,"\u76f4\u63a5\u5bfc\u51fa\u5230c\u7684\u63a5\u53e3\u6bd4\u8f83\u56f0\u96be":45,"\u76f8\u5173\u5c42":41,"\u77e9\u9635\u5927\u5c0f\u662f":41,"\u793e\u533a\u53c2\u4e0e\u56f0\u96be":45,"\u793e\u533a\u8d21\u732e\u4ee3\u7801\u5b66\u4e60\u6210\u672c\u9ad8":45,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u7684\u53c2\u6570":11,"\u79bb\u7ebf\u6279\u5904\u7406":12,"\u7b2c\u4e00\u4e2atag\u4e3a":62,"\u7b2c\u4e09\u6b65\u5b8c\u6210\u540e":62,"\u7b2c\u4e8c\u4e2a\u4e3a":62,"\u7b49":[42,46],"\u7b49\u5168\u90e8\u9759\u6001\u5e93\u4e2d\u7684\u76ee\u6807\u6587\u4ef6\u5168\u90e8\u6253\u5305\u540e\u4ea7\u751f\u7684\u6587\u4ef6":46,"\u7b49\u5f85\u7f16\u8bd1\u5b8c\u6210\u540e":62,"\u7b49\u6587\u4ef6":46,"\u7c7b\u4f3c":46,"\u7c7b\u540d\u548cc":45,"\u7c7b\u578b":45,"\u7ed3\u8bba":45,"\u7edf\u4e00\u7528":12,"\u7f16\u8bd1\u5668\u6ca1\u6709":45,"\u7f16\u8bd1\u578b\u8bed\u8a00":45,"\u7f16\u8bd1\u65f6\u4f1a\u628a\u5bf9\u5e94\u7684\u5934\u6587\u4ef6\u548c\u5e93\u653e\u5728":42,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684docker\u53d1\u884c\u955c\u50cf":62,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684python":62,"\u7f16\u8bd1c":46,"\u800c\u4e0d\u5fc5\u5728\u610fpaddl":46,"\u800c\u4e0d\u652f\u6301pypy\u89e3\u91ca\u5668":45,"\u800c\u4e0d\u66b4\u9732\u6982\u5ff5\u7684\u5b9e\u73b0":46,"\u800c\u4e14\u5728\u4f20\u8f93\u7684\u8fc7\u7a0b\u4e2d\u4e5f\u53ef\u80fd\u51fa\u73b0\u7f51\u7edc\u4e0d\u7a33\u5b9a\u7684\u60c5\u51b5":28,"\u800c\u51fa\u73b0\u9636\u6bb5\u6027\u7684\u8fd0\u884c\u505c\u6ede":11,"\u800c\u5728cpp\u91cc\u9762\u5b9e\u73b0\u8fd9\u4e2ac\u7684\u63a5\u53e3":45,"\u800c\u591a\u8bed\u8a00\u63a5\u53e3\u9700\u8981\u76f4\u63a5\u8bfb\u53d6\u751f\u6210\u7684\u4e8c\u8fdb\u5236":45,"\u800c\u5bf9\u4e8egolang":45,"\u800c\u5bf9\u4e8egolang\u9519\u8bef\u5904\u7406\u5e94\u8be5\u4f7f\u7528\u8fd4\u56de\u503c":45,"\u800c\u662f\u76f4\u63a5\u4fee\u6539paddl":46,"\u800c\u662f\u76f4\u63a5\u7528api\u7684\u63a5\u53e3\u8fdc\u7a0b\u8bbf\u95ee":12,"\u800cswig\u53ea\u80fd\u7b80\u5355\u7684\u66b4\u9732c":45,"\u81ea\u52a8\u6302\u8f7d\u5206\u5e03\u5f0f\u5b58\u50a8\u76ee\u5f55":11,"\u81f3\u4e8e\u4e3a\u4ec0\u4e48\u9700\u8981c":46,"\u826f\u597d\u7684\u6587\u6863":45,"\u8282\u7701\u4e86\u4e0d\u5fc5\u8981\u7684\u64cd\u4f5c":42,"\u83b7\u53d6\u6700\u65b0\u7684\u68c0\u67e5\u70b9\u7684\u6587\u4ef6uuid":11,"\u867d\u7136\u4e0d\u9f13\u52b1\u8fd9\u6837":46,"\u8868\u793a\u5bf9\u8f93\u5165\u6570\u636e":42,"\u89e3\u91ca\u578b\u8bed\u8a00\u53ea\u80fd\u8c03\u7528\u52a8\u6001\u5e93":45,"\u89e3\u91ca\u6027\u8bed\u8a00\u5b9e\u9645\u8fd0\u884c\u7684\u4e8c\u8fdb\u5236\u662f\u89e3\u91ca\u5668\u672c\u8eab":45,"\u8ba1\u5212\u5728":[41,42],"\u8ba1\u7b97\u8fd9\u4e2a\u6587\u4ef6\u7684md5":11,"\u8ba9paddle\u6838\u5fc3\u4e2d":46,"\u8bad\u7ec3\u4efb\u52a1\u7684\u8fd0\u884c\u53ef\u80fd\u4f1a\u5360\u6ee1trainer\u548cparamet":11,"\u8bad\u7ec3\u548c\u7eaf\u4f7f\u7528":62,"\u8bad\u7ec3\u6a21\u578b\u6b63\u786e\u6027":62,"\u8bb0\u5f55\u4e0b\u6240\u6709\u5931\u8d25\u7684\u4f8b\u5b50":62,"\u8bbe\u7f6e":46,"\u8bc6\u522b\u6570\u5b57":62,"\u8bcd\u5411\u91cf":62,"\u8be5\u6587\u4ef6\u5bf9\u76f8\u5173gemm":41,"\u8be5\u7c7b\u7ee7\u627f\u4e8epaddlepaddle\u7684\u57fa\u7c7b":42,"\u8be6\u7ec6\u8bbe\u8ba1":28,"\u8bed\u610f\u89d2\u8272\u6807\u6ce8":62,"\u8bf4\u660e":11,"\u8bf7\u53c2\u8003":46,"\u8f6c\u6362\u5185\u5b58\u7684\u5de5\u4f5c":42,"\u8f6c\u6362\u5197\u4f59":41,"\u8f6c\u6362\u51fd\u6570":42,"\u8f6c\u6362\u751f\u6210\u7684\u6587\u4ef6\u540d\u4f1a\u662f\u4ee5\u4e0b\u683c\u5f0f":12,"\u8f6c\u6362\u8017\u65f6":41,"\u8f93\u5165\u68af\u5ea6":42,"\u8f93\u51fa\u6570\u636e\u548c\u8f93\u51fa\u68af\u5ea6":42,"\u8f93\u51fa\u6570\u636e\u548c\u8f93\u51fa\u68af\u5ea6\u7684\u8f6c\u6362":42,"\u8fbe\u5230\u5bb9\u707e\u7684\u76ee\u7684":11,"\u8fc7\u7a0b\u4e2d\u6240\u6709\u65f6\u95f4\u6b65":41,"\u8fd1\u671f\u76ee\u6807":42,"\u8fd4\u56de\u7b2c\u4e8c\u6b65":62,"\u8fd8\u662f\u4ece":12,"\u8fd9\u4e00\u5c42\u8fdb\u884c\u5c01\u88c5":46,"\u8fd9\u4e00\u6570\u636e\u683c\u5f0f\u7684\u8f6c\u6362\u64cd\u4f5c":41,"\u8fd9\u4e00\u6982\u5ff5\u4e0d\u518d\u7410\u788e":46,"\u8fd9\u4e09\u4e2a\u5206\u652f":62,"\u8fd9\u4e2a\u51fd\u6570\u672c\u8eab\u4f1a\u5728\u8ba1\u7b97\u524d\u5c06\u539f\u6570\u636e\u8f6c\u6362\u4e3a\u66f4\u9002\u5408\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u5185\u90e8\u683c\u5f0f":41,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u7684\u8fde\u63a5\u53c2\u6570\u4e0epaddle\u7684\u5176\u4ed6\u4e8c\u8fdb\u5236":46,"\u8fd9\u4e2a\u53c2\u6570\u4e5f\u4e0d\u4f1a\u4e00\u5e76\u5220\u9664":46,"\u8fd9\u4e2a\u5934\u6587\u4ef6\u4e0d\u5047\u8bbe\u5176\u4ed6\u6587\u4ef6\u7684\u5f15\u7528\u987a\u5e8f":46,"\u8fd9\u4e2a\u63a5\u53e3\u9700\u8981\u505a\u5230":45,"\u8fd9\u4e2a\u6587\u4ef6\u5177\u6709\u72ec\u7279\u7684\u8bed\u6cd5":45,"\u8fd9\u4e2a\u76ee\u5f55\u4e2d\u9664\u4e86":46,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u4e2d\u7684\u53e6\u4e00\u4e2a\u9879\u76ee\u662f":46,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u5305\u542b\u4e24\u4e2a\u9879\u76ee":46,"\u8fd9\u4e2a\u9009\u62e9":[41,42],"\u8fd9\u4e2a\u9759\u6001\u5e93\u5305\u542b\u4e86paddle\u7684\u5168\u90e8\u7b26\u53f7":46,"\u8fd9\u4e2ainstance\u53ef\u4ee5\u662f\u5355\u4e2a\u503c":12,"\u8fd9\u4e9b\u4f1a\u5728":[41,42],"\u8fd9\u4e9b\u51fd\u6570\u4f1a\u6839\u636e\u8f93\u5165\u53c2\u6570\u91cd\u65b0\u8bbe\u7f6e\u5185\u90e8\u548c\u5916\u90e8\u5b58\u50a8":42,"\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":12,"\u8fd9\u4e9b\u955c\u50cf\u4e5f\u53ef\u4ee5\u4ece":62,"\u8fd9\u5bf9\u4e8e\u901a\u5e38\u7684java\u7684\u5f00\u53d1\u8005\u6765\u8bf4":45,"\u8fd9\u662f\u56e0\u4e3a":45,"\u8fd9\u6837":46,"\u8fd9\u6837\u4e0b\u4e00\u4e2acpu":42,"\u8fd9\u6837\u4fdd\u8bc1":62,"\u8fd9\u6837\u5c31\u53ef\u4ee5\u5728\u4e91\u7aef\u6267\u884c\u591a\u79cd\u6570\u636e\u7c7b\u8ba1\u7b97\u4efb\u52a1":12,"\u8fd9\u6837\u5df2\u7ecf\u4f20\u8f93\u6210\u529f\u7684\u90e8\u5206\u5c31\u4e0d\u7528\u91cd\u65b0\u4f20\u8f93\u4e86":28,"\u8fd9\u6837\u5e26\u6765\u7684\u597d\u5904\u5c31\u662f\u4e0d\u9700\u8981\u4e00\u76f4\u6e05\u7a7amemori":42,"\u8fd9\u6837\u65e2\u4f7f\u5f97\u6700\u7ec8\u4fdd\u5b58\u7684\u53c2\u6570\u683c\u5f0f\u4e0epaddlepaddle\u4e00\u81f4":42,"\u8fd9\u90fd\u9700\u8981\u8fd9\u4e2a\u63a5\u53e3\u6309\u7167\u7ea6\u5b9a\u4fd7\u6210\u7684\u89c4\u5219\u6765\u6ce8\u91ca\u5b8c\u5907":45,"\u8fd9\u91cc":42,"\u8fd9\u91cc\u7684dockerimage\u4f5c\u4e3a\u7f16\u8bd1\u73af\u5883\u4ee5\u652f\u6301\u66f4\u591a\u7684linux":62,"\u8fd9\u91cc\u9009\u62e90":62,"\u8fd9\u91cc\u9700\u8981\u7528\u6237\u989d\u5916\u6ce8\u610f":11,"\u8fdb\u4e00\u6b65\u4f18\u5316":42,"\u8fdb\u5165":62,"\u8fdb\u800c\u8fdb\u884c\u4ee3\u7801\u8bc4\u5ba1":62,"\u9009\u62e9\u662f\u5426\u7f16\u8bd1mkl":42,"\u9009\u62e9\u9700\u8981\u53d1\u5e03\u7684\u7248\u672c":62,"\u900f\u4f20\u7528\u6237\u8eab\u4efd\u7684\u529e\u6cd5":28,"\u901a\u5e38":46,"\u901a\u5e38\u5305\u542b\u4e00\u4e2acpu\u7248\u672c\u548c\u4e00\u4e2agpu\u7248\u672c":62,"\u901a\u5e38\u6307\u5c06\u4e00\u4e2a\u6574\u4f53\u62c6\u5206\u6210\u591a\u4efd\u7684\u5176\u4e2d\u7684\u4e00\u4efd":11,"\u901a\u8fc7\u4f7f\u7528\u8fd9\u4e9bapi":41,"\u901a\u8fc7\u6a21\u578b\u63a8\u65adapi\u7684\u5b9e\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u6837\u4f8b":46,"\u903b\u8f91\u5212\u4e0a\u6587\u4ef6\u5206\u5757\u7684\u5355\u4f4d":28,"\u9075\u5faa\u4ee5\u4e0b\u6d41\u7a0b":62,"\u90a3\u4e48":46,"\u90a3\u4e48\u5bf9\u5e94\u7684\u5185\u90e8\u5b58\u50a8\u4e5f\u4f1a\u4e0e\u5b83\u4eec\u5171\u4eab\u5185\u5b58":42,"\u90a3\u4e48\u5c31\u4f1a\u4f7f":42,"\u90fd\u4e0d\u4f1a\u60f3\u8981\u77e5\u9053next":42,"\u90fd\u662f\u4e94\u4f4d\u7684\u6570\u5b57":12,"\u90fd\u662f\u4ee5ext\u5f00\u5934":42,"\u90fd\u662fabi\u8c03\u7528\u6807\u51c6\u7684":45,"\u90fd\u7ee7\u627f\u4e8epaddlepaddle\u7684\u57fa\u7c7b":41,"\u914d\u7f6e\u7684\u65b9\u6cd5\u53c2\u8003":28,"\u91ca\u653e\u5bf9paramters\u5185\u5b58\u7684\u9501\u5b9a":11,"\u91cc\u6240\u6709\u7684\u7b26\u53f7\u90fd\u5199\u5165\u81ea\u5df1\u7684\u7a0b\u5e8f\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cc":45,"\u91cc\u9009\u62e9\u9700\u8981\u53d1\u5e03\u7684\u5206\u652f":62,"\u91cc\u9762\u6dfb\u52a0":42,"\u91cd\u5199\u7236\u7c7blayer\u7684":42,"\u91cd\u547d\u540d\u6210":45,"\u94fe\u63a5\u5230\u81ea\u5df1\u7684\u7a0b\u5e8f\u91cc":45,"\u9519\u8bef\u5904\u7406":45,"\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662f\u8fd4\u56de\u503c":45,"\u9519\u8bef\u5904\u7406\u7684\u65b9\u5f0f\u4e5f\u4e0d\u5c3d\u76f8\u540c":45,"\u9664\u6784\u9020\u67d0\u79cd\u7c7b\u578b\u7684\u51fd\u6570":46,"\u96c6\u6210\u5230":41,"\u96c6\u6210\u5230paddlepaddl":42,"\u9700\u8981":12,"\u9700\u8981\u4fee\u6539build":62,"\u9700\u8981\u53ef\u4ee5\u8de8\u5e73\u53f0\u6267\u884c":28,"\u9700\u8981\u5728cmake\u7684\u65f6\u5019":46,"\u9700\u8981\u5c06bugfix\u7684\u5206\u652f\u540c\u65f6merge\u5230":62,"\u9700\u8981\u5f15\u7528":46,"\u9700\u8981\u6709\u7a33\u5b9a\u7684\u5bfc\u51fa\u7b26\u53f7":45,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[42,62],"\u9700\u8981\u7d2f\u52a0\u4e0d\u540clayer\u4f20\u8fc7\u6765\u7684\u68af\u5ea6":42,"\u9700\u8981\u88ab\u66b4\u9732\u5230\u5176\u4ed6\u8bed\u8a00":46,"\u9700\u8981\u91cd\u547d\u540dwheel\u5305\u4e2dplatform\u76f8\u5173\u7684\u540e\u7f00":62,"\u9ed8\u8ba4256k":28,"\u9ed8\u8ba4\u8bbe\u7f6e\u4e3a":41,"abstract":[19,20,27,31,52,61,63,74,93],"api\u4e2d\u4f7f\u7528":45,"api\u5bfc\u51fa\u7684\u52a8\u6001\u5e93":46,"api\u5bfc\u51fa\u7684\u9759\u6001\u5e93":46,"api\u63a5\u53d7\u7684\u7c7b\u578b\u5168\u662f":46,"api\u63a5\u53e3":28,"api\u63a5\u53e3\u7684\u53c2\u6570\u8f6c\u53d1\u7ed9":46,"api\u65f6":46,"api\u65f6\u6240\u552f\u4e00\u9700\u8981\u5f15\u5165\u7684\u5934\u6587\u4ef6":46,"api\u662f\u591a\u8bed\u8a00api\u7684\u57fa\u7840\u90e8\u5206":46,"api\u66b4\u9732\u7684\u7c7b\u578b":46,"api\u751f\u6210\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u4f1a\u88ab\u5b89\u88c5\u5230":46,"api\u7684\u5b9e\u4f8b":46,"api\u7684\u5b9e\u73b0\u7ec6\u8282":46,"api\u7684\u63a5\u53e3":46,"api\u7684\u65f6\u5019\u63a8\u8350paddle\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":46,"api\u7684\u7f16\u8bd1\u9009\u9879\u9ed8\u8ba4\u5173\u95ed":46,"api\u76ee\u5f55\u7ed3\u6784\u5982\u4e0a\u56fe\u8868\u6240\u793a":46,"api\u83b7\u5f97\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u5b9e\u4f8b":46,"apis\u505a\u4e86\u5c01\u88c5":41,"block\u6784\u6210\u4e00\u4e2amodel":11,"book\u4e2d\u6240\u6709\u7ae0\u8282\u529f\u80fd\u7684\u6b63\u786e\u6027":62,"boolean":[27,29,37,45,68],"break":[9,66,68,70,71],"bugfix\u5206\u652f\u4e5f\u662f\u5728\u5f00\u53d1\u8005\u81ea\u5df1\u7684fork\u7248\u672c\u5e93\u7ef4\u62a4":62,"bugfix\u5206\u652f\u9700\u8981\u5206\u522b\u7ed9\u4e3b\u7248\u672c\u5e93\u7684":62,"byte":[28,44],"c99\u662f\u76ee\u524dc\u6700\u5e7f\u6cdb\u7684\u4f7f\u7528\u6807\u51c6":45,"c\u6709\u6807\u51c6\u7684abi":45,"c\u8bed\u8a00\u662f\u6709\u5bfc\u51fa\u7b26\u53f7\u7684\u6807\u51c6\u7684":45,"case":[13,19,21,22,27,31,40,46,52,56,58,59,68,74,75,82,86,94,98,101,106],"char":15,"ci\u73af\u5883\u4f7f\u7528":62,"ci\u7f16\u8bd1wheel\u5b8c\u6210\u540e\u4f1a\u81ea\u52a8\u5c06docker\u955c\u50cfpush\u5230dockerhub":62,"class":[5,8,19,20,21,22,25,26,30,31,32,33,35,36,38,40,45,49,50,54,55,59,60,61,63,64,65,67,70,75,76,77,92,99],"compute\u51fd\u6570":41,"const":[8,13,15,20,30,32,38,39,53,54,56,60,63,65,67,69,70,74,75,76,77],"core\u4e2d\u7684\u6a21\u578b\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u53c2\u6570":46,"core\u4e2d\u8fd9\u4e00\u7c7b\u578b\u63a5\u53e3\u7684\u667a\u80fd\u6307\u9488":46,"core\u662f\u5426\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u5b9e\u4f8b":46,"core\u6982\u5ff5":46,"data\u5230\u5206\u5e03\u5f0f\u5b58\u50a8\u8865\u5145\u8bad\u7ec3\u6570\u636e":12,"default":[0,1,2,4,5,8,9,19,21,25,34,38,44,47,48,56,57,63,64,65,68,71,72,75,76,81,84,86,87,93,95,97,103,104,105,106],"device\u5c31\u80fd\u62ff\u5230\u6b63\u786e\u7684\u6570\u636e":42,"dnn\u4e09\u8005\u5173\u7cfb\u5982\u4e0b\u8868":42,"dnn\u4e2d\u7684":42,"dnn\u4e2d\u7684\u6392\u5217\u65b9\u5f0f\u4e0d\u6b62\u8fd9\u4e00\u79cd":42,"dnn\u4f1a\u4f5c\u4e3a\u7b2c\u4e09\u65b9\u5e93\u96c6\u6210\u8fdbpaddlepaddl":42,"dnn\u4f1a\u7528\u5230":42,"dnn\u5171\u540c\u4f7f\u7528":42,"dnn\u524d\u540e\u7684cnn\u7f51\u7edc\u6027\u80fd":42,"dnn\u5728\u53d1\u5e03":42,"dnn\u5b9e\u73b0":42,"dnn\u5e0c\u671b\u7684\u683c\u5f0f":42,"dnn\u6570\u636e\u7684\u4e0d\u540c\u683c\u5f0f\u4ee5\u53ca\u76f8\u4e92\u4e4b\u95f4\u7684\u8f6c\u6362":42,"dnn\u7684":42,"dnn\u7684\u5e93\u76ee\u524d\u53ea\u6709\u52a8\u6001\u5e93":42,"dnn\u7684\u6027\u80fd":42,"dnn\u7684\u60c5\u51b5\u4e0b":42,"dnn\u7684\u64cd\u4f5c\u90fd\u662f\u76f4\u63a5\u8986\u76d6\u7684\u5f62\u5f0f":42,"dnn\u7684\u6d4b\u8bd5":42,"dnn\u7684\u73af\u5883\u4e0b":42,"dnn\u7684\u76f8\u5173\u529f\u80fd":42,"dnn\u7684\u7ed3\u679c":42,"dnn\u7684\u9ad8\u6027\u80fd\u683c\u5f0f\u4e0epaddlepaddle\u539f\u6709\u7684":42,"dnn\u7684layer":42,"dnn\u7684layers\u90fd\u4f1a\u7ee7\u627f\u4e8e":42,"enum":[13,15,21,47,54,55,64,65,71,76],"export":[2,31,36,78,81],"final":[6,7,22,36,48,49,66,70,74,75],"float":[25,30,38,65,67,69,74,75,76,77,95,98],"function":[5,7,8,10,14,15,16,18,19,20,21,22,25,26,30,32,35,38,43,48,49,52,53,54,55,56,58,59,60,61,63,65,70,74,75,76,77,81,82,93,97,98,101,106],"golang\u53ef\u4ee5\u4f7f\u7528":45,"golang\u7684":45,"gpu\u7b49":62,"h\u5e76\u4e0d\u56f0\u96be":45,"images\u6570\u636e\u96c6\u4e0a\u4f20\u5230\u4e91\u7aef\u7684":12,"import":[4,5,8,9,19,21,24,32,34,36,37,43,48,49,55,63,66,75,80,81,82,86,98,101,105],"ingress\u9700\u8981\u628apfsclient\u7684\u8eab\u4efd\u4fe1\u606f\u4f20\u7ed9pfsserv":28,"instance\u4e0e\u751f\u6210\u6570\u636e\u96c6\u65f6":12,"instance\u5305\u6db5\u4e24\u4e2a\u503c":12,"instance\u662f\u4e00\u6a21\u4e00\u6837\u7684":12,"int":[7,8,13,14,15,18,19,21,23,37,38,41,42,43,45,46,54,55,57,58,64,65,67,70,71,74,76,77,81,95],"interface\u6587\u4ef6\u7684\u5199\u6cd5\u975e\u5e38":45,"layer\u65f6":42,"layer\u7684\u540e\u9762\u63a5\u6709cpu":42,"list\u4f5c\u4e3a\u68c0\u67e5\u5217\u8868":62,"long":[21,98],"mkl\u5e93\u7684":41,"mklml\u4ee5\u53camkl":42,"mklml\u53ef\u4ee5\u4e0emkl":42,"mklml\u7684\u5e93\u76ee\u524d\u90fd\u662f\u52a8\u6001\u5e93":42,"mode\u4e0b\u7684\u7ed3\u679c":41,"model\u505a\u5206\u652f\u7ba1\u7406":62,"ndarray\u7c7b\u578b\u7684\u503c\u548c\u6574\u578b\u7684\u503c":12,"new":[0,6,7,8,9,10,13,14,15,16,17,20,21,22,25,30,31,40,41,43,47,49,52,57,58,59,61,65,66,70,72,73,80,86,87,106],"note\u7684\u4e66\u5199":62,"null":[36,74,93],"op\u7684\u4fe1\u606f":42,"openmp\u7528\u4e8e\u63d0\u9ad8mklml\u7684\u6027\u80fd":42,"org\u76ee\u524d\u9075\u5faa":62,"packed\u4f18\u5316\u540elayer\u7684\u6d4b\u8bd5":41,"packed\u76f8\u5173\u529f\u80fd":41,"paddle\u4e00\u4e2a\u52a8\u6001\u5e93\u53ef\u4ee5\u5728\u4efb\u4f55linux\u7cfb\u7edf\u4e0a\u8fd0\u884c":45,"paddle\u5185\u5d4c\u7684python\u89e3\u91ca\u5668\u548c\u5916\u90e8\u4f7f\u7528\u7684python\u5982\u679c\u7248\u672c\u4e0d\u540c":45,"paddle\u5185\u90e8\u7684\u7c7b\u4e3ac":45,"paddle\u7684\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0\u5305\u62ec\u4e00\u4e0b\u51e0\u4e2a\u65b9\u9762":45,"paddle\u7684\u7c7b\u578b\u5168\u90e8\u9000\u5316\u6210":46,"paddle\u7684\u94fe\u63a5\u65b9\u5f0f\u6bd4\u8f83\u590d\u6742":45,"paddle\u7684c":46,"paddle\u8bad\u7ec3\u4efb\u52a1":12,"paddle\u8def\u5f84\u4e0b":46,"paddle\u9700\u8981\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3":45,"paddle\u9700\u8981\u66b4\u9732\u7684api\u5f88\u591a":46,"paddle\u9759\u6001\u5e93\u94fe\u63a5\u590d\u6742":45,"paddle_\u7c7b\u578b\u540d":46,"paddle_\u7c7b\u578b\u540d_\u51fd\u6570\u540d":46,"paddlepaddle\u4e2d\u7684cudnn\u90e8\u5206\u4f7f\u7528\u7684\u4e5f\u662f":42,"paddlepaddle\u4f7f\u7528git":62,"paddlepaddle\u5f00\u53d1\u8fc7\u7a0b\u4f7f\u7528":62,"paddlepaddle\u63d0\u4f9b\u4e13\u7528\u7684":12,"paddlepaddle\u6bcf\u6b21\u53d1\u65b0\u7684\u7248\u672c":62,"paddlepaddle\u6bcf\u6b21\u53d1\u7248\u672c\u9996\u5148\u8981\u4fdd\u8bc1paddlepaddl":62,"paddlepaddle\u7684\u4e3b\u7248\u672c\u5e93\u9075\u5faa":62,"paddlepaddle\u7684activation\u4f1a\u76f4\u63a5\u4f7f\u7528":42,"patch\u53f7":62,"patch\u53f7\u52a0\u4e00":62,"pfsclient\u9700\u8981\u548cingress\u4e4b\u95f4\u505a\u53cc\u5411\u9a8c\u8bc1":28,"pfsclient\u9700\u8981\u5728\u4f20\u8f93\u5b8c\u6bd5\u6700\u540e\u4e00\u4e2achunk\u7684\u65f6\u5019\u68c0\u67e5destination\u6587\u4ef6\u7684md5\u503c\u662f\u5426\u548csource\u6587\u4ef6\u4e00\u81f4":28,"pfsserver\u63d0\u4f9brest":28,"public":[8,20,30,33,38,54,60,63,65,66,67,70,74,75,76,77,81,86,87],"py\u4e2d":62,"pypi\u4e0a\u7684package\u540d\u79f0\u4e3apaddlepaddle\u548cpaddlepaddl":62,"pypi\u4e0d\u652f\u6301\u8986\u76d6\u4e0a\u4f20":62,"reader\u7684\u4f7f\u7528\u65b9\u5f0f\u90fd\u662f\u4e00\u81f4\u7684":12,"reader\u8f93\u51fa\u7684data":12,"resnet\u7684mkl":42,"return":[5,6,7,8,12,13,15,18,19,20,21,26,30,32,33,34,36,38,39,40,43,48,49,50,54,55,56,60,63,65,67,70,74,75,76,77,82,86,101],"rnn\u90e8\u5206\u4e2d":41,"s3\u4e4b\u7c7b\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u4e4b\u4e0a":12,"server\u4e4b\u4e0a":11,"server\u4e4b\u95f4\u7684\u7f51\u7edc\u5e26\u5bbd":11,"server\u4f1a\u6682\u505c\u53c2\u6570\u66f4\u65b0\u5e76\u7b49\u5f85":11,"server\u4f1a\u83b7\u53d6parameters\u5185\u5b58\u7684":11,"server\u5185\u5b58\u4e2d\u7684\u6a21\u578b\u6570\u636e\u7684\u5b8c\u6574\u955c\u50cf":11,"server\u540c\u6b65\u7684\u4fdd\u5b58\u4e00\u4e2a\u7279\u5b9a\u65f6\u95f4\u70b9\u7684\u5168\u5c40\u68c0\u67e5\u70b9":11,"server\u5728\u96c6\u7fa4\u4e2d\u542f\u52a8\u540e":11,"server\u6545\u969c\u540e\u88abkubernetes\u91cd\u65b0\u542f\u52a8":11,"server\u6b64\u65f6\u8fd8\u9700\u8981\u901a\u8fc7\u7f51\u7edc\u8bbf\u95ee\u5206\u5e03\u5f0f\u5b58\u50a8\u4ee5\u4fdd\u5b58\u5feb\u7167":11,"server\u751f\u6210\u4e00\u4e2auuid":11,"server\u7684\u5355\u70b9\u6216\u591a\u70b9\u540c\u65f6\u6545\u969c":11,"server\u7684\u6570\u636e\u5feb\u7167":11,"server\u7684\u68c0\u67e5\u70b9\u5404\u81ea\u72ec\u7acb\u4fdd\u5b58":11,"server\u7b2c\u4e00\u6b21\u542f\u52a8\u6216\u4efb\u610f\u65f6\u95f4paramet":11,"short":[30,34,57,63,66,70,75],"static":[15,46,63,65,86,104,106],"super":[57,74],"swig\u652f\u6301\u7684\u8bed\u8a00\u6216\u8005\u89e3\u91ca\u5668\u6709\u5c40\u9650":45,"swig\u66b4\u9732\u7684\u63a5\u53e3\u4fdd\u7559\u4e86c":45,"swig\u751f\u6210\u7684\u4ee3\u7801\u4e0d\u80fd\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":45,"swig\u76f4\u63a5\u8bfb\u53d6c":45,"swig\u9700\u8981\u5199\u4e00\u4e2ainterface\u6587\u4ef6":45,"switch":[8,46,86],"tag\u4e3a":62,"tag\u53ef\u4ee5\u662flatest\u6216latest":62,"tag\u7684\u66f4\u65b0\u65f6\u95f4\u662f\u5426\u5728\u4e0a\u8ff0\u7f16\u8bd1wheel\u5305\u5b8c\u6210\u540e\u662f\u5426\u6700\u65b0":62,"throw":86,"true":[5,7,8,13,31,37,41,50,55,56,57,58,62,65,70,74,81,86,93,95,101],"try":[0,2,4,9,10,13,14,15,31,36,40,43,58,63,66,82,98,104],"type\u5b57\u6bb5\u5747\u4e0d\u5c3d\u76f8\u540c":46,"var":[7,8,19,20,21,25,32,33,35,37,50,55,56,57,59,63,68,69,70,78],"void":[8,13,15,20,27,30,32,33,38,43,44,45,46,55,56,64,65,67,69,74,75,76,77],"wheel\u5305":62,"while":[8,17,20,21,31,36,39,40,49,52,53,58,61,63,67,75,77,93,101],AGE:[86,87],AWS:[12,85,89,90],Added:66,And:[13,17,18,27,34,36,47,51,54,58,63,67,86,95,101],But:[4,6,33,39,47,54,63,72,83,106],For:[1,2,5,7,8,14,15,16,18,19,20,22,25,26,31,32,33,35,38,40,44,47,48,49,52,53,54,55,56,57,58,59,60,61,64,65,66,67,69,71,72,74,75,76,77,80,82,92,93,95,97,98,101,103,105,106],IDE:[0,104],IDs:[17,21,49],IPs:81,IRs:22,Into:86,Its:[32,64,75,86,101],K8s:106,NOT:[57,75],Not:[4,5,10,40,66,106],OPs:[22,24,82],One:[6,17,44,47,63,66,69,74,93,101],Ops:[59,61,65,75],PFS:28,PRs:78,QoS:87,Such:[38,57,66,70],TLS:[5,28,86],That:[2,20,51,93,95],The:[0,2,4,5,6,7,9,10,14,16,17,18,20,21,22,24,25,26,29,30,32,36,39,40,43,44,46,48,49,51,52,54,55,56,57,58,61,63,64,65,66,67,69,70,71,72,74,75,76,77,78,82,83,84,85,86,87,91,93,95,97,98,99,101,103],Their:10,Then:[1,2,4,19,22,33,38,40,51,54,56,74,82,83,86,87,88,97,98,101],There:[5,8,9,10,15,17,18,21,22,29,30,31,36,40,47,48,49,52,53,54,57,61,63,64,67,75,86,98,103,104],These:[1,7,8,20,25,30,35,50,61,64,65,66,85,95],Use:[1,5,23,29,58,59,66,74,81,83,86,93,94,97,98,105],Used:[59,67],Uses:[40,104],Using:[0,3,10,31,52,58,59,61,63,77,85,87],VMs:0,VPS:86,WITH:105,With:[19,20,25,31,51,55,66,70],YES:18,Yes:[0,2,42],___fc_layer_0__:86,__align__:30,__cuda_align__:30,__device__:30,__doc__:65,__file__:18,__forceinline__:30,__fp16:30,__global__:30,__hadd:30,__half:30,__half_raw:30,__impl__:65,__init__:[25,26,34,40,50,57,70,74,97],__main__:34,__metaclass__:75,__name__:34,__rnn_step__:101,__va_args__:60,__x:30,_addup_repetitive_outputs_:7,_append_backward_ops_:[7,25],_append_backward_vars_:7,_binari:9,_create_global_var:57,_def:40,_dtype:36,_filer:43,_filter:43,_fwd:43,_fwd_pd:43,_input:43,_librari:9,_live_in:40,_live_out:40,_loss:34,_op:[36,75],_output:43,_presucessor:40,_program:40,_remove_no_grad_branch_:7,_reorder_input:43,_source_language_embed:101,_src_input:43,_sucessor:40,_target_language_embed:101,_test:9,_update_op:26,_use:40,_value_index:36,a75:30,a_op:75,a_prev:66,aaaaa:12,aaaaaaaaaaaaa:86,abi:103,abil:34,abl:[5,7,20,21,22,38,50,54,57,106],about:[2,4,6,8,9,18,24,29,32,40,48,58,63,65,66,67,75,80,86,92,93,97,98,103,105],abov:[0,2,3,5,6,7,8,9,10,14,19,21,22,30,31,32,33,35,43,48,49,50,51,52,54,55,57,65,66,68,69,70,72,75,76,77,80,82,86,87,97,98,103,106],abs:[6,34],abs_numerical_grad:6,absolut:[2,6,103,105],acc:22,acceler:[2,11,42,51,52,95],accept:[5,59,103],access:[4,5,9,14,17,18,19,21,22,57,81,101],accessmod:86,accessor:57,accord:[6,7,15,22,24,35,49,59,70,75,81,92,93,95],accordingli:74,account:[59,72,106],accoust:66,accrodingli:13,accumul:[10,15,26,51,52,66],accur:[6,17,54],accuraci:[26,66,74],achiev:[20,24,51,52,66,67,98],ack:93,acquir:31,across:[7,22,48,66],act1:36,act2:36,act:[8,20,22,36,49,57,70,80,82,99,101],act_output:65,act_typ:36,action:[21,86],activ:[9,36,40,49,54,57,61,65,74,77,78,80,93,101],actual:[13,25,31,34,36,43,47,52,65,67,69,77],actual_layout:43,adagrad:[52,64],adagradoptim:50,adam:[5,15,22,34],adapt:[54,69],add:[0,1,6,7,8,9,13,17,20,21,22,24,26,30,33,37,39,50,52,56,57,59,61,63,67,68,72,74,75,77,95,98,99,105],add_activ:57,add_bia:57,add_depend:9,add_execut:9,add_input:[48,74],add_memori:48,add_output:48,add_scalar:[8,49,55],add_sum:57,add_test:[9,74],add_two:[8,48],add_unittest_without_exec:74,addattr:[65,75],addbia:74,addcom:[65,75],added:[8,24,25,30,47,51,52,61,72,74,75,76],adding:[61,72],addinput:[65,75],addit:[7,21,51,54,59,61,70,75,77,82],addition:48,addmemori:43,addop:[33,77],addoutput:75,addprimit:43,addprimitivedesc:43,addr:10,address:[2,10,15,19,22,82,84,88,91,93,98,106],addrow:74,addtyp:65,adjust:[7,25],admin:106,administr:[0,17,106],adopt:[30,34],advanc:[6,93,98,101],advantag:[6,30,31,52,58,91],adversari:[34,58],advic:98,affect:8,affili:49,afford:14,aforement:9,after:[1,2,7,8,9,14,15,17,22,23,24,25,27,29,30,40,43,51,53,54,57,66,72,74,75,76,77,84,86,87,91,93,95,97,101,103,104],aftern:54,again:[5,6,10,52,98],against:86,aggreg:[26,51,86],ago:9,agre:72,agreement:72,ahead:66,aid:98,alexnet_pass1:95,alexnet_pass2:95,algo:43,algorithm:[7,14,25,40,43,49,52,61,66,69,101],all:[0,2,5,7,8,9,10,13,15,17,18,19,20,21,22,23,25,27,29,31,34,35,36,39,40,43,44,46,47,48,49,50,51,52,54,55,57,59,65,66,67,74,77,81,82,83,84,86,87,92,93,95,98,99,101,103,104,105,106],all_output_nam:7,alloc:[15,18,40,43,67,74,77,95,99],allow:[5,15,19,20,22,25,31,52,61,72,74,86,93,98],allow_only_one_model_on_one_gpu:[92,93,95],allreduc:51,almost:[0,84],alpha:[9,61],alreadi:[2,9,10,31,43,57,63,84,86,93,98,105],alreali:92,also:[0,1,2,4,5,7,8,9,13,16,20,21,22,30,31,33,34,35,36,39,40,47,48,49,52,53,54,55,56,57,58,61,63,65,66,67,70,71,74,75,78,82,83,87,91,98,101,103,106],altern:[75,97],although:[7,51],altogeth:106,alwai:[9,21,44,64,86,93],amazon:[86,87],amazonaw:86,amazonec2fullaccess:86,amazonelasticfilesystemfullaccess:86,amazonroute53domainsfullaccess:86,amazonroute53fullaccess:86,amazons3fullaccess:86,amazonvpcfullaccess:86,ambigu:[58,66],amd64:86,amd:47,amend:72,amodei:66,among:86,amort:51,amount:98,analys:54,analysi:[54,97,98],analyz:40,ancestor:[55,57],andd:86,andrew:40,android:105,android_abi:103,android_api:103,android_arm_neon:103,android_native_api_level:103,android_standalone_toolchain:103,android_toolchain:103,ani:[0,1,5,9,10,15,17,18,19,21,22,27,30,31,38,40,44,49,51,52,57,58,60,61,66,69,75,77,82,86,88,98,105],annoi:84,announc:30,anoth:[0,5,7,8,18,20,21,31,32,43,49,57,63,65,67,69,86,93],anroid_arm_mod:103,ans:86,answer:[19,31,72,86],anymor:51,anyth:[49,58,86],anytim:34,anywai:[97,103],apach:42,api:[1,4,5,7,9,15,16,18,19,20,26,28,33,34,36,48,53,54,59,62,70,71,74,75,81,82,85,86,97,98,103,104,105,106],api_shar:9,api_test:9,api_trainer_config_helpers_lay:101,apiserv:86,apivers:[86,87],app:104,appar:7,appear:[19,31,35,67],appel:40,append:[7,25,26,49,57,58,66,72,74,81,101],append_backward:[7,50,97],append_clip_op:25,append_op:[25,39,57],append_oper:57,appl:104,appleyard:98,appli:[34,35,51,54,63,74,101],applic:[19,21,30,31,32,35,57,59,72,75,85,86,87,97,98,106],applyl1:13,appoint:75,appreci:[66,72],approach:[22,23,24,51,52,53,61,66,103,105,106],approxim:52,apt:[2,97],arbitrari:[22,44,77],arch:103,archetectur:66,architectur:[1,30,66,83,103,104],archiv:[45,46],area:34,arg:[7,36,50,65,75,92],argu:56,argument:[1,7,8,13,14,22,50,53,56,57,72,74,83,93,94,101],arithmet:30,arm64:[103,104],arm64_standalone_toolchain:103,arm:[30,103,104,105],arm_standalone_toolchain:103,armeabi:103,armv7:[30,104],armv8:30,arn:86,around:[17,40,57,86,106],arrai:[15,19,21,35,49,55,57,58,59,70,75],arrang:70,array_to_lod_tensor:40,arrow:34,articl:[0,32,35,82,87,105],artifact:[4,62,86],artifici:40,arxiv:[34,66],as_step_input:8,asap:76,asgd:52,ask:[0,7,10,17],asr:66,assgin:40,assign:[7,14,19,23,25,30,32,51,66,77,81,82,86,91,93,106],assigne:66,assignmemt:40,associ:[53,60,77],assum:[2,8,22,43,82,95,101],assumpt:22,ast:19,astyp:[58,75],asyc:10,async:[10,24,92],async_count:93,async_lagged_grad_discard_ratio:93,async_lagged_ratio_default:[92,93],async_lagged_ratio_min:[92,93],asynchron:[10,21,51,54,83,93],atom:23,attr:[8,19,36,39,43,55,56,57,65,75,101],attr_map:65,attrdesc:55,attribu:43,attribut:[7,8,24,25,39,55,57,59,63,65,70,74,75],attributemap:75,attrproto:65,attrtyp:[55,65,75],attrvalu:65,auc:[26,92],aucvalidationlay:93,audio:66,augment:66,authent:86,author:[28,66,86],auto:[1,8,13,23,32,43,45,56,59,63,69,70,72,74,75,77,98],autom:[85,86],automat:[1,5,7,15,22,24,25,33,50,59,65,66,72,74,75,82,84,86,92,93,97,101,104],avail:[4,10,15,24,30,31,40,86,106],averag:[14,93],average_test_period:[92,93],avg:[82,98],avg_cost:[22,82,99],avg_loss_valu:82,avoid:[1,6,8,10,21,22,39,43,51,52,53,98],avx2:1,avx:[1,2],awai:31,await:87,awar:[0,5,19,21,26,32,48,57,86,97],awk:88,awni:66,aws:28,aws_account_id:86,awsaccountid:86,awskeymanagementservicepowerus:86,axiom:21,b363:87,b8561f5c79193550d64fa47418a9e67ebdd71546186e840f88de5026b8097465:87,ba5f:86,back:[2,7,10,22,30,34,52,75],background:[61,66,81],backpropag:[6,7],backward:[6,8,13,15,25,34,41,42,50,52,53,56,60,61,68,74,93,95,99,101],backward_first:101,backward_op:6,backwardactiv:74,bacth:20,baidu:[31,66,87],bake:22,balanc:[24,51,86,93],bandwidth:[30,51],bare:[82,87,106],barrier:[83,93],barrierstatset:98,basci:36,base:[0,5,14,20,25,26,30,31,38,43,47,50,51,52,54,59,60,61,67,70,74,75,86,93,97,98,99,101,103,104],baseerrorclipattr:25,baselin:66,basematrix:74,bash:[0,1,2,86,87,91],basic:[22,36,43,54,55,59,60,66,70,74],batch:[5,8,10,12,13,20,21,22,26,27,31,34,37,47,48,49,51,52,66,70,72,74,82,86,87,93],batch_id:34,batch_im:34,batch_label:34,batch_norm:[34,66],batch_read:[12,20,58],batch_siz:[22,34,41,49,82],batch_szi:34,batch_z:34,batchnorm:[34,66],batchread:20,batchsiz:74,bazel:9,bbbbb:12,bcm2708:105,bdist_wheel:62,beacus:36,beam:[93,101],beam_gen:101,beam_search:[49,101],beam_siz:[49,92,93,95,101],becaus:[0,5,6,8,9,10,15,30,49,53,57,58,61,63,64,68,69,70,71,74,77,82,86,95,97,101,103],becom:[23,24,63,67,98],been:[0,7,9,14,21,31,72],befor:[1,2,7,10,17,21,25,29,32,35,47,52,53,54,58,61,72,75,76,86,97,103,104,106],begin:[7,13,15,20,26,29,35,49,51,74,81],beginiter:5,beginn:101,beginpass:5,begintrain:5,behavior:[21,98],behind:[31,70,82],being:[7,17,25,31,56,58,77,97],belong:[20,22,63],below:[1,4,8,10,15,22,24,30,31,44,53,58,61,70,71,74,81,83,86,91,98,101,103,104],benchmark:[44,66],benefit:[17,18,49],besid:[4,22,40,47,51],best:[0,2,9,43,93],besteffort:87,beta:34,better:[9,31,40,43,49,69,86,104,106],between:[2,7,9,10,15,22,24,30,31,43,46,51,53,60,63,75,77,86],bia:[49,57,74,101],bias:74,bias_attr:[57,101],biases_:74,biasparameter_:74,biassiz:74,bidi:87,bidirect:[66,101],big:[19,24,40,98,106],bigger:10,bilinearfwdbwd:98,bin:[2,81,86,87,91],binari:[0,1,3,4,9,18,22,30,32,34,44,86,97,98],bind:[19,21,30,33,63,67],bit:30,bitcod:104,bla:[2,104],black:[34,104],blank:[86,104],block0:40,block1:40,block2:40,block:[7,11,13,15,19,21,22,23,24,25,26,27,31,38,40,47,48,50,67,71,74,75,77,93,98],block_expand:66,block_id:[19,27],blockdesc:[8,35,57,59],blockdescbind:38,blockingcount:23,blueprint:49,book:[59,66,78,99,101],book_distribut:82,bool:[8,20,30,37,39,41,42,43,56,57,64,65,70,71,74,76,93,95],boost:[47,66,67,76],boot:[101,106],boot_lay:101,boot_stat:70,bootstrapp:106,borrow:[34,70],bos_id:101,both:[5,8,9,10,17,20,21,22,24,30,31,34,38,40,47,49,51,54,56,64,66,67,74,75,77,78,82,83,86,98,101,104],bottl:51,bottleneck:[54,98],bottom:66,bound:40,boundari:22,box:[34,98],brace:[8,35],brain:17,branch:[4,5,8,9,22,31,37,55,62,68,72,75,78],breadth:93,break_if:70,brief:[9,15,30,67,77],briefli:98,bring:[31,40,77],broadcast:[10,51,59,106],broken:72,browser:[2,78,86,97],bsd:[0,21,51],bsp:21,bucket_nam:86,buddy_alloc:72,buf:13,buf_siz:[22,82],buffer:[13,21,43,44,52,58,63,93,99],buffer_s:21,buffered_read:58,bug:[72,86],build:[2,4,9,18,22,35,36,40,42,52,61,62,65,66,72,75,80,82,86,89,90,91,93,97,102],build_android:103,build_model:34,builder:72,buildtool:62,built:[0,1,4,9,19,22,30,32,40,47,51,65,66,70,97,103,105,106],bulk:21,bunch:[44,91,98],button:[72,78,86],c11:45,c703c041:72,c99:46,c99e:86,cach:[1,30],cacul:[26,81],caff:[8,31],caffe2:[8,19,21,31],calcul:[2,6,7,10,15,23,26,30,40,74,83,93,95,98,101],calcut:40,calendar:54,call:[0,5,6,7,8,13,14,15,16,18,19,21,22,25,32,34,35,40,48,49,50,54,57,59,60,63,65,67,70,74,75,76,77,80,82,86,91,93,97,98,101],callback:[25,74],caller:[6,86,97],can:[0,1,2,3,4,5,6,7,8,9,10,13,14,17,18,19,20,21,22,24,25,27,30,31,32,33,34,35,36,38,39,40,43,47,48,49,50,51,52,54,55,56,57,58,59,60,61,65,67,69,70,71,72,74,75,76,77,78,80,81,82,83,84,85,86,87,88,91,92,93,95,97,98,101,103,104,105,106],cancel:17,candid:[49,66],cannot:[20,59,63,69,70,74,75],cantain:36,capabl:[30,53,59],capac:[61,86],capi:[1,45],capi_prvi:46,caption:49,captur:84,card:[51,82,91],care:[2,18,40,58,66,67,92,93,106],carefulli:[66,93],carpedm20:34,cast:[30,69],cast_to_op_attr:65,cat:[2,88],categor:75,categori:10,categoryfil:87,caus:[4,10,29,75],caution:[86,87],cbla:41,cc_:9,cc_binari:9,cc_test:9,cclient:16,cduadevicecontext:[47,67],cento:[1,4,80,106],central:[61,82],ceph:[12,87],cephf:[12,18,28],cer:66,certain:[20,39,47,50,54,63,67,76,92],certif:[5,28,86],cffi:45,cfg:[40,87],cgo:45,ch1:21,ch2:21,chain:[7,35,74],challeng:[6,10,31,37,67],chan:21,chanc:[5,30,74],chang:[1,9,14,18,22,31,43,53,55,58,60,62,63,66,72,74,75,77,81,82,86,93,98,101,103],changes:43,channel:[19,98],chapter:[48,49,66,82],chapter_data:48,chapter_out:48,charact:[21,66],characterist:95,check:[1,4,7,8,9,25,43,56,59,68,72,75,78,86,91,93,95],check_attr:65,check_eq:74,check_grad:[6,75],check_l:74,check_output:75,check_sparse_distribution_batch:[92,93],check_sparse_distribution_in_pserv:[92,93],check_sparse_distribution_ratio:[92,93],check_sparse_distribution_unbalance_degre:[92,93],check_styl:72,checker:[6,59],checkgrad:93,checkgrad_ep:93,checkmark:106,checkout:72,checkpoint:[24,56],checksum:28,child:8,china:2,chines:78,chip:31,chmod:86,choic:[2,9,31],choos:[0,1,2,3,21,39,76,93,103],chosen:[34,47],chunk:[14,28],circl:35,circular:21,circumst:67,claim:86,claimnam:86,clang:[30,45,72,103],clariti:49,classic:[40,66],classif:[35,95],classifi:34,claster:86,clean:[0,1,8,9,27,53,59,72],clear:[9,49,53,63,69,105],clearer:[53,57],clearli:63,cli:86,click:[4,72,78,86,97,98],client:[13,16,59],clip:93,clip_op:25,clip_op_desc:25,clone:[0,1,72,78,97,103,105],close:[21,58,72],close_channel:21,cloud:[9,10,18,28,29,59,106],cludform:86,cluster:[5,8,10,15,22,66,81,83,87,91,92,93],cluster_test_fil:81,cluster_train:84,cluster_train_fil:81,cluster_train_v2:[84,85,88],cm469:86,cmake:[0,1,46,72,74,75,78,82,97,98,103,104],cmake_build_typ:[97,103,104,105],cmake_c:104,cmake_install_prefix:103,cmake_system_nam:[103,104,105],cmakelist:[9,41,42,74],cmatrix:[45,46],cmd:87,cname:86,cnn:87,coars:33,code:[0,1,2,4,5,7,9,17,20,21,22,24,27,30,33,34,35,39,44,47,50,52,53,54,56,58,59,60,61,65,70,73,74,75,76,77,78,80,81,82,86,87,98,101],codebas:[59,72],collabor:10,collect:54,collectbia:74,column:[35,58,74,97],com:[0,1,2,9,34,62,72,78,86,87,97,99,103,105,106],combin:[40,50,59,63],come:[22,26,40,55,66,70],comma:[15,82,93],command:[0,1,2,4,9,13,18,29,72,74,75,78,80,82,83,84,86,87,88,89,90,91,96,97,98,103,104,105],commandlin:98,comment:[9,36,65,66,72,75],commit:[9,72],common:[12,61,67,74,92,101],commonli:[29,61,95,97,98,101,105],commun:[10,15,16,21,22,24,51,72,74,81,82,86],compani:31,compar:[0,6,9,19,59,74,75],comparison:[9,31,76],compat:[30,33,51],compil:[0,9,22,31,36,38,40,47,51,60,64,65,71,74,78,91],complaint:9,complet:[3,7,8,10,14,15,25,28,35,44,47,59,74,75,77,82,86,87,97,106],complex:[17,21,40,49,59,98,101],complic:[22,33,58,69,70],compon:[21,22,36,66,70,71,74,76],compos:[5,21,33,36,48,57,59],composit:33,compress:14,compris:7,compromis:0,comput:[0,2,5,6,10,21,22,24,27,30,31,32,36,40,44,47,50,51,52,54,60,63,66,67,69,71,72,74,75,76,82,85,86,91,95,97,98,99,101,103,104,105],computation:101,computationgraph:36,con:51,concat:[34,101],concaten:[34,48,70],concentr:59,concept:[5,19,20,21,31,33,34,36,43,48,49,52,53,55,63,70,71,101],conceptu:[21,27,31,34,36],concern:[5,21,26,104],concis:[34,70],conclud:75,concret:[59,67,75],concurr:[10,17,24,54,83],concurrentremoteparameterupdat:93,cond:[8,31,37,55],condit:[14,22,31,37,43,66,68,87,101],condtion:34,conduct:98,conf:84,conf_paddle_gradient_num:86,conf_paddle_n:86,conf_paddle_port:86,conf_paddle_ports_num:86,conf_paddle_ports_num_spars:86,config:[12,29,49,74,86,87,92,93,106],config_:[13,93],config_arg:[92,93,95],config_lay:74,config_len:15,config_pars:[41,42,74],config_proto:15,configmap:22,configur:[0,7,13,15,17,18,22,24,31,36,39,57,66,67,69,72,74,75,77,80,81,82,93,98,100,103,105,106],confirm:29,conflict:[63,72],confus:[34,39],congest:93,connect:[18,19,22,24,66,74,81,82,86,87,91,106],consid:[0,7,56,67,95,98,106],consider:[47,66],consist:[14,20,21,32,44,55,58,59,60,65,66,71,75],consol:[86,98],consolid:[8,78],constant:[36,38,39,47,74,75],constraint:63,construct:[5,27,36,40,48,57,59,63,65,76,101],constructbackwardgraph:35,constructoptimizationgraph:35,constructor:[25,30,54,57,59,63,65,74,75],consum:[10,97],consumpt:40,contact:17,contain:[0,3,4,5,7,8,14,27,34,36,43,44,47,53,54,57,59,60,63,64,65,66,70,71,75,81,82,86,88,91,101,104],container:85,containerport:86,content:[15,29,44,49,78,87],content_dir:78,content_len:15,context:[25,43,63,64,67,75,77,99,101],contin:86,continu:[7,10,44,66,83,93,103],contrib:61,contribut:[0,61,66,73],contributor:[59,72],control:[8,19,21,68,86,87,93,106],controlflowgraph:40,conv2d:[34,76],conv:[34,43,69],conv_fwd:43,conv_pool_2:22,conveni:[5,7,36,50,65,66],convent:[7,15,72,75],converg:84,convers:[30,31],convert:[12,22,23,24,30,31,32,43,58,60,66,82],convlut:66,convolut:[34,47,57,67],convolution_algorithm_opt:43,cool:72,cooper:66,coordin:[10,15],copi:[1,5,14,17,29,35,48,49,51,52,70,72,81,86,88],copy_from:25,copyvariablewithtensor:69,core:[0,7,36,39,46,52,53,70,93,99],coreo:[86,106],corner:59,corpu:66,correct:[6,7,30,51,74,75,76,86],correctli:[7,30,34,74],corresond:30,correspoind:5,correspond:[5,7,8,9,25,30,36,37,43,47,48,49,57,59,60,61,65,67,68,74,75,76,77,97,104],correspondingli:104,corss_entropi:5,cortex:30,cos:65,cosin:65,cosineop:65,cosineopproto:65,cosineopprotomak:65,cost:[5,7,22,35,50,51,55,56,69,82,93,99],cost_np:56,could:[0,5,6,14,19,21,22,23,24,30,31,32,48,50,52,53,55,57,58,60,68,76,84,86,97,98,103],count:[10,18,26,56,58,66,81,82,87,93,95,98],counter:[10,14,23,35],cours:[0,18,47],cover:[31,66,77],cp27:4,cp27m:[4,62],cp27mu:[4,62],cpp:[6,13,33,41,42,45,46,53,59,71,74,76,98],cprofil:97,cprofilev:97,cpu:[0,1,2,6,18,30,39,47,52,53,54,59,61,62,67,69,75,76,77,87,93,97,98,99],cpu_avx_mkl:[2,4],cpu_avx_openbla:[4,80],cpu_kernel:39,cpu_noavx_openbla:4,cpu_ns_:54,cpu_per_pserv:22,cpu_per_train:22,cpudevicecontext:[47,67,75,76],cpuelapsedu:54,cpuengin:42,cpuinfo:2,cpuplac:[22,39,43,47,67,69,75,76,77,82,99],cpusparsematrix:46,crash:[10,84,93,98],creat:[0,2,5,6,8,10,15,19,20,23,25,26,27,28,29,30,31,33,34,35,43,47,48,50,51,52,53,57,60,61,66,72,74,75,78,80,81,82,88,93,103,106],create_backward_pass:50,create_bias_paramet:74,create_block:57,create_doc_str:65,create_input_paramet:74,create_local_scop:27,create_oper:33,create_optimization_pass:50,create_paramet:57,create_python_ops_creatation_funct:65,create_rnn:8,create_rnn_op:48,create_tmp_var:57,create_tmp_vari:57,create_var:57,create_whileloop:70,creategradientoper:60,creatememori:43,createop:65,createoper:8,createprimitivedesc:43,createstack:86,createvari:8,creation:[33,86],creationd:86,creator:[12,59,60],creator_:60,credenti:29,crf:[67,69],critic:[34,97],crlf:72,crop:67,crop_grad:67,cropgradkernel:67,cropkernel:67,cross:[57,75],cross_compil:105,cross_entropi:[5,22,34,40,69],crt:28,csc:74,csr:74,ctc_error_evalu:66,ctest:[0,1,75],ctor:57,ctrl:[0,84],ctx:[43,69,75,77],cubla:[47,76],cublas_handle_:67,cublashandle_t:67,cuda7:[4,80],cuda8:[1,2,4],cuda:[0,2,4,9,32,47,54,59,67,75,76,82,91,93,98],cuda_context:32,cuda_dir:[92,93],cuda_fp16:30,cuda_so:2,cudaconfigurecal:98,cudadevicecontext:[32,47,67,75],cudadevicegetattribut:98,cudaelapsedu:54,cudaevent_t:54,cudaeventcr:98,cudaeventcreatewithflag:98,cudafre:98,cudagetdevic:98,cudagetdevicecount:98,cudagetdeviceproperti:98,cudagetlasterror:98,cudahostalloc:98,cudalaunch:98,cudamalloc:98,cudamemcpi:98,cudaplac:[47,67,69,76],cudaprofilerstart:98,cudaprofilerstop:98,cudaruntimegetvers:98,cudasetdevic:98,cudasetupargu:98,cudastream_t:67,cudastreamcr:98,cudastreamcreatewithflag:98,cudastreamsynchron:98,cudeviceget:98,cudevicegetattribut:98,cudevicegetcount:98,cudevicegetnam:98,cudevicetotalmem:98,cudnn:[9,39,43,47,67,69,76,93],cudnn_conv_workspace_limit_in_mb:[92,93],cudnn_dir:[92,93],cudnn_kernel:39,cudnnconvopkernel:76,cudnnv5:1,cudrivergetvers:98,cuinit:98,cumtim:97,cur_mem:49,curl:86,curli:[8,35],current:[0,2,4,7,8,9,10,13,15,19,24,26,31,39,47,48,49,52,53,54,57,63,69,70,74,78,81,82,84,86,93,101,104],current_block:[55,57],current_endpoint:82,current_oper:55,current_word:101,curv:5,custom:[5,18,30,34,49,52,59,66,74,86],custom_batch_read:58,cut:70,cxx:104,cxx_compil:[103,104,105],cxx_flag:104,cxxabi_1:4,cycl:10,cython:45,d3e0:86,d_b0:34,d_b1:34,d_b2:34,d_block:34,d_f:34,d_g:34,d_h0:34,d_h0_bn:34,d_h0_relu:34,d_h1:34,d_h1_bn:34,d_h1_relu:34,d_h2:34,d_loss:34,d_loss_fak:34,d_loss_real:34,d_optim:34,d_step:34,d_t:34,d_w0:34,d_w1:34,d_w2:34,daili:72,dandroid_abi:103,dandroid_arm_mod:103,dandroid_arm_neon:103,dandroid_standalone_toolchain:103,dangl:0,dario:66,darwin:86,dash:34,dat:12,data:[5,6,8,12,13,14,21,24,26,28,30,31,34,35,36,38,39,40,43,44,47,48,49,50,51,52,53,55,57,59,61,63,64,65,66,67,70,71,74,75,77,80,81,82,83,88,89,92,93,95,98,99,101],data_i:34,data_lay:13,data_layout_:69,data_read:58,data_reader_creator_random_imag:58,data_shar:70,data_typ:[44,64,66,69,71,76,80,101],data_type_:[39,47,69],data_x:34,datacent:[12,29],datacenter1:12,datacenter2:12,datacenter_1:12,datacenter_2:12,datacenter_nam:12,datafeed:[82,99],dataflow:36,dataflow_analysi:40,datalayout:69,dataparallel:22,dataprovider_convert:66,dataset:[12,18,22,52,58,66,80,82,83,93,97,101],datatransform:69,datatyp:[39,43,64,66,69,71],date:81,dcgan:34,dcmake_install_prefix:[103,104,105],dcmake_system_nam:[103,104,105],dcuda_arch_nam:1,dcudnn_root:1,ddim:[20,47,67,77],dead:10,deal:[7,106],debug:[2,6,7,22,29,31,57,72,97],debug_str:36,decayr:13,decent:14,decid:[5,17,34,44,52,60,61,64],declar:[8,34,48],decod:[66,101],decoder_boot:101,decoder_dim:49,decoder_group_nam:101,decoder_input:[49,101],decoder_mem:[49,101],decoder_s:101,decoder_st:101,deconv:34,decor:[20,74],decrement:23,decrementcount:23,decrypt:86,deduc:59,deep:[2,7,17,21,27,34,35,40,42,54,59,61,66,67,82,98,104],deeper:[2,32],deepspeech2:41,def:[5,6,7,12,18,25,26,33,34,36,39,40,48,49,50,57,58,65,70,74,75,101],def_block:34,defalut:[93,95],default_block:34,default_devic:95,default_main_program:[82,99],default_param_attr:57,default_st:70,default_startup_program:[82,99],default_valu:95,defaultdict:40,defaultinfervartyp:38,defect:53,defer:17,defin:[5,7,8,9,10,17,19,20,23,24,25,30,31,32,33,34,36,39,40,47,48,51,55,57,58,59,63,65,67,70,74,77,81,82,84,93,97,99,101],definit:[2,7,8,10,14,22,27,32,39,55,60,65,70,75,97,99],definiton:33,deivc:76,delai:[52,67,77,93],delet:[18,28,72],deletestack:86,deliv:106,delta:6,demand:[10,67],demo:[59,84,87,89],demolish:87,demonstr:[77,101],denot:[75,95],dens:[15,16,64,66,74,86],dense_vector:80,densescann:66,dep:9,depart:66,depend:[0,2,8,9,10,18,20,21,22,24,36,51,56,64,75,81,83,95,103,104,105,106],dependent_var:56,deploi:[84,95,106],deploy:[36,44,59,84,85,86,104,106],deprec:66,depth:[8,31,66],dequeu:24,deriv:[5,20,22,25,37,50,103],desc:[8,25,43,44,57,65,70],desc_:8,descend:70,descent:[10,52,83],descproto:44,describ:[5,7,8,9,14,19,22,27,32,39,43,44,48,49,53,55,57,59,64,65,69,71,74,75,76,77,86,87],describestack:86,describestackev:86,describestackresourc:86,descripotor:43,descript:[1,4,8,9,38,42,44,47,60,64,66,69,71,72,75,81,86,94],descriptor:[21,43,69],deseri:[44,53],deserializ:59,desgin:35,design:[7,13,20,39,40,45,52,54,61,75,106],desir:[10,22,52,86,87],destin:[15,29],destroi:[8,27],destruct:63,destructor:[54,74],detail:[0,1,6,7,14,18,22,24,29,31,34,36,40,43,44,47,48,54,57,61,63,67,69,70,71,74,75,76,77,78,80,82,84,86,87,94,95,97,98,101,105,106],detect:[1,38,72,103],determin:[8,22,40,47,59,74],dev:[0,2,59,97,103,106],dev_ctx:[8,43,54],devel:62,develop:[2,4,7,9,31,38,53,54,57,60,62,66,72,77,78,81,92,93,97,99,102,104,105],deverlop:93,devic:[2,6,19,22,26,30,36,42,43,47,51,53,54,59,69,75,77,93,99,104],device_:54,device_context:[43,75],devicecontext:[8,47,54,75,76],deviceid:[42,95],deviceid_:42,deviceplac:67,devid:93,devot:66,devtools2:1,dhcp:106,diagnos:84,diagram:[48,83],diamond:34,dict:[7,57,81,88],dict_siz:[13,49],dictionari:[5,6,57,95],did:[2,53],diff_mat:6,differ:[2,6,7,8,9,10,15,17,20,22,23,24,25,26,27,30,31,34,36,37,40,43,47,49,51,52,54,56,60,63,66,68,69,70,71,74,75,76,77,81,83,84,86,87,93,97,101,104],differenti:33,difficult:[0,6,31],dig:[2,86,98],digit:81,digraph:36,dilat:43,dim0:75,dim1:75,dim:[13,43,44,48,59,64,67,71,74,75,77],dim_:[67,77],dimens:[34,59,64,66,67,70,74,75,77,95],dimension:[74,77,101],dios_arch:104,dios_enable_bitcod:104,dios_platform:104,dios_use_veclib_for_bla:104,dir:103,direcit:66,direct:[31,40,52,66,97],directli:[1,3,9,16,18,20,22,30,39,53,65,69,70,84,87,103],director:75,directori:[0,1,2,9,12,17,28,29,67,77,78,81,84,87,88,93,98,103,104,105],disabl:54,disadvantag:[52,57],discard:[10,14,49,93],discov:10,discoveri:86,discrep:98,discrim:34,discuss:[5,8,14,15,16,22,43,66],disk:[0,44,87],dispatch:[22,53,84,85,93],displai:[18,29,72],dist:[1,62],dist_train:[5,18],distinguish:[9,84],distribut:[4,8,14,15,16,17,19,21,26,32,51,59,66,71,81,84,85,87,89,90,96,106],distribute_test:[92,93],distributedli:[22,74],distributetranspil:82,disucss:5,divid:[7,26,65,71,92,97],diy_beam_search_prob_so:[92,93],django:78,dnn:[1,43,66],dns:86,do_forward_backward:58,doc:[36,48,70,75,77,78,81],doc_cn:78,dockefil:0,docker:[1,3,62,72,78,86,89,90,91,106],docker_build:5,docker_clust:[84,88],docker_push:5,dockerfil:[0,103,105],dockerhub:2,document:[0,1,6,20,22,28,35,48,49,54,59,66,72,73,75,76,77,82,91,95,104],doe:[0,4,10,14,15,17,18,19,20,22,24,27,30,36,40,48,53,57,59,60,61,74,75,77,98,99],doesn:[0,2,5,8,19,21,58,72,87,97,98,103,104],doing:[13,17,22,35,98],domain:86,don:[0,1,2,5,9,33,35,40,58,66,72,75,78,86],done:[7,9,10,14,15,22,23,38,40,44,52,60,61,66,72,86,97,98],dot:[75,93],dot_period:[93,95],doubl:[1,22,30,35,54,69,75,76,93],down:[66,98],download:[1,2,4,9,10,13,17,28,81,83,104,106],dozen:9,draw:49,drive:63,driver:[2,82,91],drop:[20,49],dropout:74,drpi_arm_neon:105,drpi_toolchain:105,drwxr:87,ds2:66,dst:[15,43],dst_primitive_desc:43,dtoh:98,dtype:[21,22,36,57,82,99],due:[14,17,34,40,49,57,97],dummi:14,dump:44,duplic:24,durat:[14,98],dure:[7,8,10,14,17,18,26,31,40,51,52,54,57,59,66,71,74,75,86,92,93,106],duse_eigen_for_bla:103,dwith_c_api:[46,103,104,105],dwith_distribut:82,dwith_doc:82,dwith_gpu:[1,82,105],dwith_profil:98,dwith_python:[46,82,105],dwith_swig_pi:[46,82,103,104,105],dwith_test:[1,75,104],dwith_tim:98,dynam:[1,15,46,48,57,58,93,98],dynamic_cast:74,dynamic_recurrent_op:70,e2e:106,each:[6,7,9,10,13,14,15,17,18,19,20,21,22,25,26,27,32,35,38,40,43,47,48,49,51,53,54,56,57,58,59,60,63,64,65,66,67,68,69,70,71,74,76,81,82,83,84,86,91,93,95,97,101,106],eager:31,earli:[30,32,72,75],eas:[38,75],easi:[6,7,49,52,58,59,61,72,74,84],easier:[0,5,24,30,31,58,68,70,72,74],easili:[5,34,51,54,58,60,63,67],echo:2,edg:40,edit:[0,2,21,86],editor:[0,57],edu:[86,87],eeoi3ezpr86c:86,effect:[1,86,93,104],effici:[0,22,44,58,66,67,74,101],effort:[22,66],efs:86,efs_dns_nam:86,efsvol:86,egd:40,eigen:[30,47,52,59,61,67,75,103,104],eigen_device_:67,eigen_test:77,eigen_use_gpu:75,eigenmatrix:77,eigenscalar:77,eigentensor:77,eigenvector:77,either:[3,5,22,34,37,38,48,52,61,98,104],elabor:66,elb:86,elbapis:86,electr:40,electron:87,element:[6,14,21,24,36,49,59,75,77],element_typ:[15,76],elementari:59,elif:[5,65,68],els:[0,1,2,5,13,18,21,22,24,25,31,34,37,38,39,40,63,65,68,74,75],elsewher:54,emac:0,email:72,emailweixu:9,emb1:13,emb2:13,emb:87,embed:[5,8,13,24,38,49,64,70,81,101],embedding_lay:13,embedding_nam:101,embedding_s:101,emphas:98,emplace_back:74,emploi:[7,25,65,101],empti:[7,10,20,21,49,75],emul:30,enabl:[0,8,9,14,19,24,25,36,54,72,81,86,93,98,104],enable_grad_shar:[92,93],enable_parallel_vector:93,enc_proj:101,enc_vec:101,encapsul:15,encod:[14,49,101],encoded_proj:101,encoded_sequ:101,encoded_vector:101,encoder_ctx:49,encoder_ctx_expand:49,encoder_ctx_proj:49,encoder_dim:49,encoder_out_seq:49,encoder_s:101,encount:13,encourag:[22,27],encrypt:86,encrypt_decrypt:86,end2end:106,end:[0,4,7,8,22,25,32,36,40,49,53,54,58,63,66,68,72,93,101],end_pass:5,endian:44,endif:[47,54],enditer:5,endpass:5,endpoint:[12,86],endtrain:5,enforc:104,engin:[18,42,43,66,98],english:[66,78],enjoi:2,enough:[0,7,8,39,40,47],enqueu:24,ensur:[0,2,4,10,43,51,63,74,104],enter:[8,27],enterpris:59,entir:[15,17,75],entiti:[8,63],entranc:27,entri:[0,14,18,38,72,74,86,103],entropi:57,entry_point:18,enumer:47,env:[78,86,97],environ:[0,1,4,5,20,22,72,81,85,86,87,92,93,97,98,104],environmenterror:81,eos_id:101,epoch:34,epol:21,equal:[10,70,75,76,93],equat:[40,75],equip:101,equival:[5,8,19,21,25,31,37,65,106],eras:20,erlang:21,error:[5,6,14,29,30,31,43,63,66,74,75,84,86,93],error_clip:25,error_clip_callback:25,errorclipbyvalu:25,especi:[0,42],essenc:[5,7],essenti:[5,27,30,77],establish:19,estim:[5,24,52],eta:87,etc:[1,8,21,22,26,43,51,52,58,63,66,76,85,86,92,95,106],etcd:[10,14,15,17],etcd_addr:15,eth0:86,etyp:21,eval:[8,26,34,59],eval_program:26,eval_result:26,evalu:[17,36,56,66,98,99],even:[0,5,30,51,57,58,72,93,98,104],evenli:[15,86],event:87,event_:54,event_block:54,event_handl:5,eventkind:54,eventlist:54,eventu:[22,70],everi:[5,10,14,15,17,25,26,35,36,38,40,43,47,48,51,57,63,65,69,72,74,75,77,81,83,88,91,93,99,101],everyon:72,everyth:[22,24,34,103],everywher:0,evid:53,evolv:31,exactli:[20,86],exampl:[0,8,18,20,22,24,26,29,31,32,33,34,35,36,38,40,43,47,48,49,53,54,55,57,58,59,60,61,64,67,68,69,70,72,74,75,76,77,81,82,85,86,87,92,93,95,97,98,99,101,105],except:[17,19,31,35,54,66,70,95],excess:40,exchang:53,exe:[22,82,99],exec:93,execut:[0,9,10,14,18,19,20,21,22,26,27,32,34,36,40,43,51,54,60,71,74,86,97,98],executioncontext:[43,69,75,76,77],executor:[19,22,26,30,31,32,34,50,55,57,69,71,82,97,99],exist:[0,4,5,8,10,20,29,31,49,57,58,60,65,67,70,74,77,86,93,104],exit:[15,29,82,87,93],expand:[2,49,74],expect:[69,98],expected_desc:43,expected_kernel_kei:69,experi:[44,66,95],experienc:72,expert:9,expir:10,explain:[10,19,31,33,35,72,76,82,83,97],explan:[18,19,22,63,69],explicit:[20,54,70,74,76],explicitli:[5,22,27,75,77,104],explod:25,explor:[49,61],expos:[7,16,21,43,44,67,70,86],express:[5,24,26,36,40,75,86],extend:[52,70],extens:[17,24,49,75,103],extent:46,extern:[9,42,45,46,59,66],external_librari:9,extingrad_:42,extinval_:42,extoutgrad_:42,extoutval_:42,extra:[22,61,67,106],extraattr:95,extract:[31,53,66,75,86],extrem:[19,31,98],f120da72:87,f7e3:86,fa0wx:87,fabric:85,face:[9,61],fact:[19,31,51,55,57],factori:45,fail:[10,14,49,75,87,93,95],failur:[10,15,75],fake:34,fake_imag:58,faked_imag:34,fall:[30,56],falloc:28,fals:[6,7,8,31,37,39,41,48,55,56,58,64,71,74,75,80,81,87,93,95,101],false_block:[8,37,55],false_label:58,false_neg:26,false_posit:26,false_read:58,familiar:21,far:[25,70],fashion:22,fast:[14,31,98],faster:[2,10,31,98,101],fastest:31,fastli:72,fat:104,father:7,fault:[1,14,59],favorit:0,fbd1f2bb71f4:87,fc1:[36,74,95],fc1_bia:36,fc1_weight:36,fc2:[36,95],fc3:[36,95],fc4:95,fc8a365:86,fc8a:86,fc_layer:[57,65,74,95],fc_op:65,fc_out:8,fc_output:65,fc_without_b:8,fclayer:74,fcop:33,feasibl:52,featur:[7,22,30,36,51,54,66,72,93],feed:[5,22,35,48,61,82,99],feed_dict:34,feed_list:[82,99],feeder:[22,82,99],feel:72,fetch:[10,13,20,22,56,74,99,101],fetch_list:[22,57,82,99],fetch_op:56,few:[0,9,10,21,22,40,52,58,64,66],fewer:[21,57],fft:66,field:[8,36,38,44,56,57,60,64,65,86,98],fifth:35,figur:[5,9,22,24,34,42,48,54,57,66,74,98,101],file:[0,2,4,5,7,9,10,12,14,15,17,18,20,21,28,29,31,32,36,44,46,58,59,66,67,71,72,74,75,76,77,80,81,82,84,88,93,99,101,103,104,105,106],filelist:66,filenam:[12,57,97],fileoffset:28,filesystem:[17,18,22,28,86],fill:[10,14,47,57,86],fill_zero_grad:59,fill_zeros_like_op:7,filter:[25,43],find:[1,4,8,10,17,21,30,36,43,49,63,69,82,88,98,103,104],find_var:6,findmemori:43,findop:8,findprimit:43,findprimitivedesc:43,findvar:[8,63],fine:[14,33],fingerprint:86,finish:[1,10,14,17,18,27,40,51,65,81,83,84,86,87],finit:74,finnal:2,first:[0,1,5,7,8,10,14,17,18,19,22,27,29,31,34,35,36,43,48,49,55,56,57,59,64,65,66,67,68,70,72,74,75,77,82,86,93,95,98,101,106],first_seq:101,firstli:76,firstseen:87,fit:[30,39,40,44,49,59,82],fit_a_lin:82,five:[55,98],fix:[22,40,45,57,66,72,97],flag:[41,42,54,72,75,78,93],flatten0:36,flatten:[36,55,57,77],flexibl:[5,15,22,31,35,39,48,49,52,58,67,70,101],flist:81,fliud:19,float16:21,float16_t:30,float32:[22,30,33,34,57,58,75,82,99],float_to_half_rn:30,flow:[8,19,21,48,54,62,68],fluid:[7,20,22,24,27,47,54,57,67,68,69,76,97],fluid_cuda_create_tensor:32,fluid_cuda_mult:32,fluid_cuda_read:32,fly:7,fmt:21,fnt03:86,focu:[21,36,97,98],focus:75,folder:[9,12,18,29,82,86],follow:[0,1,2,4,5,6,7,8,9,10,14,18,19,20,21,22,24,27,30,31,32,33,34,35,36,37,38,40,43,47,48,49,51,52,54,55,56,57,58,59,60,61,63,64,65,66,67,68,69,70,72,74,75,77,78,80,82,86,87,88,89,90,95,97,98,99,101,103,104,105,106],footprint:32,forbid:5,forc:[51,57,69],force_cpu:39,force_cudnn:39,force_load:45,forest:8,forev:21,forget:5,fork:72,form:[4,20,21,26,98],formal:69,format:[6,14,20,22,30,31,47,49,66,70,72,74,75,77,80,81,86,93],former:[5,9,31,40,52],formula:[6,40],forth:34,forward:[6,7,8,13,15,25,31,34,41,42,43,44,50,53,55,58,59,60,61,64,74,95,101],forward_infer:43,forward_list:54,forward_op:6,forward_train:43,forwardactiv:74,found:[30,55,61,63,76,82,85,101,105],four:[21,26,31,35,43,47],foward:56,fp16:[30,59,71],fp32:[47,59,69,71],fp64:[47,71],fpga:[47,99],fpgadevicecontext:67,fpgaengin:42,fpgaplac:[47,67],frame:[27,59,66,70],framework:[5,7,8,21,25,26,30,31,36,47,51,52,54,55,59,61,63,65,67,72,74,75,85,97,99,104],free:[32,67,72,106],freememoryop:32,frequenc:[66,98],frequent:[14,58,59,61,67,84,103,104],fresh:17,friend:63,friendli:34,from:[0,2,4,6,7,8,9,10,12,13,14,15,19,20,21,22,24,25,26,29,30,31,33,34,35,36,37,39,40,43,48,49,50,51,53,55,57,58,59,60,63,66,67,68,69,70,72,74,75,76,77,81,82,83,85,86,87,93,95,97,98,101,103,104,106],fromfil:58,front:[36,40],fulfil:98,full:[10,17,21,48,51,52,74,76,101,106],full_matrix_project:101,fulli:[22,24,66,74,82,98,106],fullsiz:13,fullyconnect:[36,57],fullyconnectedlay:74,func:[14,19,32,60],functor:[33,36],fundament:[20,21,24,30,59],further:[20,65,106],futur:[17,22,30,40,48,59,103],fvs:65,fwd_desc:43,fwd_op:60,fwd_primit:43,fwd_primitive_desc:43,fwd_var:25,g_b0:34,g_b1:34,g_b2:34,g_block:34,g_command_config_arg:[41,42],g_h0:34,g_h0_bn:34,g_h0_relu:34,g_h1:34,g_h1_bn:34,g_h1_relu:34,g_h2:34,g_im:34,g_loss:34,g_optim:34,g_program:57,g_state:54,g_step:34,g_w0:34,g_w1:34,g_w2:34,gan:5,gangliao:9,gap:93,gatedrecurrentlay:41,gather:[7,40,51,53,74,75],gaussian_normal_random:34,gcc:[0,1,30,32,45,59,97,103,105],gcc_3:4,gcreators_:65,gemm:41,gemmconvkernel:76,gendrated_id:49,gener:[5,6,7,8,9,10,12,14,15,17,19,22,31,33,38,40,43,47,51,52,55,56,57,58,59,60,64,65,66,67,68,70,72,75,86,88,93,95,98,103,105],generated_id:49,generated_scor:49,generatedinput:101,get:[1,2,4,6,7,8,9,10,14,15,17,18,20,28,31,34,36,39,40,41,42,43,47,48,49,54,57,59,60,63,65,69,70,72,74,75,76,81,84,86,88,97,98,101,102],get_all_op_proto:65,get_block:57,get_config_arg:95,get_data:87,get_dim:6,get_float_el:6,get_grad_op_desc:7,get_input_lay:74,get_numeric_gradi:6,get_numerical_gradi:6,get_output:6,get_program:40,get_pserver_program:82,get_startup_program:82,get_support:4,get_symbol:36,get_tensor:6,get_trainer_program:82,get_vari:8,get_worker_addr:19,getactualkerneltyp:39,getattr:25,getbatchs:74,getdeviceid:76,geteigendevic:77,getengin:43,getenv:[5,18,81],getexpectedkerneltyp:[39,43,69],getinfervartyp:38,getinput:74,getinputgrad:74,getinputvalu:74,getkerneltyp:30,getkerneltypeforvar:69,getlibrari:43,getmat:13,getoptconfig:13,getoutputgrad:74,getoutputvalu:74,getparam:13,getparameterconfig:13,getparameterptr:74,getparameterspars:13,getparametersremot:13,getplac:[43,67,75,76,77],getsiz:74,gettask:14,gettempl:86,gettensor:69,gettranspos:74,getw:74,getweight:74,getwgrad:74,git:[0,1,62,72,78,103,105],github:[0,1,9,34,47,62,72,78,97,99,103,105],give:[0,10,48,57,59,69,72,74,86,98],given:[7,15,17,21,24,25,31,33,34,49,58,61,70,74,93],glibc:[4,103,105],glibc_2:4,glibcxx_3:4,glide:9,global:[0,5,8,9,10,32,36,39,53,54,59,63,65,67,69,86,93,98],global_block:57,globalstat:98,globalstatinfo:98,glog:72,glog_v:72,glog_vmodul:72,gnueabihf:105,go_librari:9,go_test:9,goal:[20,21,24,30,35,51,59,66,98],gob:14,godep:9,godoc:45,goe:[10,31,37,63,99],going:[7,33,52,91,97,106],golang:9,good:[21,34,52,57,58,61,82,97,98,106],googl:[5,54,59,72,85,97,103],googleapi:86,googlenet:42,goroutin:[19,21],got:[39,63],gpg2:86,gpg:86,gprotos_:65,gpu:[0,1,4,6,18,21,26,30,40,47,51,52,53,54,59,61,62,67,69,76,77,80,81,82,91,96,99,106],gpu_id:[93,95],gpu_per_train:22,gpudevic:67,gpugpu_id:92,gpukernel:59,grab:10,grad:[6,7,15,25,42,57,64,93],grad_info_map:7,grad_n:25,grad_nam:25,grad_op:25,grad_op_class:59,grad_op_desc:25,grad_op_maker_:60,grad_op_typ:[59,60],grad_op_type_:60,grad_s_block:7,grad_share_block_num:[92,93],grad_to_var:[7,25],grad_var_nam:6,gradient:[10,14,21,23,25,35,38,50,51,52,53,57,59,64,75,81,82,83,93,97],gradient_flat:6,gradient_machin:46,gradientmachin:[46,53],gradientmachine_:13,gradopdescmak:[38,60],gradopdescmakerbas:60,gradopmak:60,gradual:98,grain:33,gram:66,grant:86,graph:[7,8,9,10,19,21,22,23,24,26,31,34,48,51,52,55,77],great:[24,66,106],greater:[25,52,81],greaterthan:65,greedi:66,green:[19,34],grep:[2,88],groudtruth:101,group:[14,36,43,67,75,106],group_input1:101,group_input2:101,group_input:101,grow:72,grpc:106,gru:[49,66,101],gru_decod:101,gru_decoder_with_attent:101,gru_out:49,gru_step:[49,101],grumemori:101,gserver:[41,42,74],gsizex:98,gtx:40,guarante:[43,57,74],guard:13,guest:[0,4],gui:[97,98],guid:[28,40,59,72,74,82,86,87,98,101,104],gzip:[14,87],h0_bn:34,h_prev:8,had:0,hadoop:5,half:[30,86],half_to_float:30,hand:[40,59,66,67,77,82,83],handi:9,handl:[5,7,18,19,22,36,40,43,47,53,58,63,67,70,76,99],handler:8,hannun:66,happen:[14,65],hard:[0,22,31,49,66,70,86],hardwar:[0,31,32,67,76,98],has:[0,1,5,6,7,8,9,10,14,15,17,20,21,22,24,25,26,30,31,34,36,40,44,47,49,51,54,55,59,64,65,67,69,72,74,86,87,98,99,101,104,106],has_kei:[7,25],has_var_recurs:7,hasdependentvar:56,hash:[47,51],hasn:31,hasnext:20,have:[0,1,2,5,6,7,8,9,10,14,15,17,18,20,21,22,24,25,27,30,31,32,33,34,35,39,40,43,44,47,48,49,51,52,53,54,55,57,58,59,60,63,64,66,67,69,71,72,74,75,76,81,86,93,95,98,101,104,105,106],haven:[0,31],hdf:12,head:[72,75,81,88],header:[15,44,46,59,67,74,76,103,104,105],headip:88,heard:0,heavi:84,height:[8,45,58,74,75],height_:64,held:10,hello:5,help:[0,8,29,31,36,43,49,58,59,70,72,84,97],helper:[22,43,60,70,74],henc:[22,52,57,60,61,63],here:[1,2,4,5,9,10,16,21,24,25,27,29,31,35,36,43,47,48,58,61,65,72,75,76,78,81,82,84,86,87,92,95,101,104,105,106],heterogen:[22,24,54],heurist:[24,49,93],hidden:[50,57,86,101],hidden_out:8,hierarch:[55,57,59,101],hierarchi:59,high:[30,51,66,67,74,85,106],higher:[33,48,70,72],highest:8,highli:[66,70,95,101],him:5,hint:[39,97],histor:[33,76],hl_get_sync_flag:74,hold:[5,7,10,14,16,20,21,30,34,36,38,40,63,65,67,69,77,82,86],holder_:[67,77],home:[2,12,29,86,87,88,97],honor:14,host:[9,18,54,86,87,103,104,105],host_c:[103,104,105],hostfil:88,hostnam:86,hostpath:87,hostport:86,hour:0,hourli:72,hous:80,how:[0,5,8,10,14,19,20,21,22,27,29,31,33,36,39,43,48,49,53,54,61,65,69,76,82,83,86,87,93,97,101,102,105],howev:[6,7,17,21,22,27,31,40,47,52,53,57,58,60,61,64,65,66,67,69,86,92,93,101],howto:81,hpp:[30,45],htod:98,http:[0,1,2,9,18,34,62,72,78,86,87,97,99,103,105,106],hub:62,huge:52,human:[54,66],hundr:76,hyper:[34,74],hyperparamet:61,i1117:98,i386:104,iOS:105,iamfullaccess:86,iamusersshkei:86,icc:32,icml:66,id_rsa:88,idea:[9,21,31,32,52,58,61,82,97],ideal:[22,69],ident:[60,75,86],identifi:[37,47,74],ids:[49,74],idx:[14,20,34,40,74],ies:29,if_els:68,if_else_op:7,ifdef:[47,54],ifels:[8,55],ifelseop:55,ignor:93,iii:66,illustr:[10,15,21,22,33,48,74,98,101],im_siz:34,imag:[0,1,5,22,31,34,35,49,50,55,58,66,72,86,89,90,95,106],image_a:58,image_b:58,image_conv_lay:66,image_fil:58,image_lay:58,image_nam:5,image_path:58,image_reader_cr:58,imagenet:12,imagepullpolici:86,images_reader_cr:58,imagin:35,imgsiz:98,imgsizei:98,imgsizex:98,imikolov:81,immedi:[1,40,43,52,61,86],immutable_paramet:5,imper:19,imperfect:59,implement:[8,14,15,16,17,18,19,21,22,24,31,33,36,37,38,40,43,45,46,47,49,53,56,63,65,66,67,68,69,70,101],implemet:13,impli:9,implicitli:19,imposs:[20,49,106],impractic:69,improv:[23,24,40,59,66,86,97,98],inarg:13,inbound:86,includ:[0,1,4,5,8,9,15,18,20,21,30,31,34,36,40,45,46,48,49,54,55,57,59,65,74,75,85,86,87,93,97,98,101,103,104,105],inclus:49,incom:[19,39],increas:[10,14,30,83,93],increment:[21,26,35,40,93],incupd:74,inde:21,independ:[6,7,15,23,63,67,106],index:[6,7,8,10,14,19,55,57,70,76,86],indic:[7,8,15,27,34,48,55,60,64,67,70,84,86,103],indice_map:70,indices_map:70,individu:[10,51,86],industri:[10,44,106],ineffici:[53,69],infer:[5,7,8,10,26,31,37,38,39,40,41,45,47,56,57,59,64,66,69,80,104,105],infer_shap:57,infer_var_type_:38,inferenc:104,inferer:66,inferfac:38,inferior:17,infernec:105,infershap:[8,57,59,75,77],infershapecontext:[75,77],infervartypefn:38,info:[30,48,74,84,106],inform:[8,18,29,36,39,40,43,44,47,48,51,57,61,63,64,72,74,75,77,86,93,97,98,103],infrastructur:[31,86],ingor:93,ingrad_:42,ingredi:[21,66],inherit:[8,20,50,59,67,75],ininst:5,init:[8,23,34,42,48,49,74,80,81,86,95],init_attr:57,init_model_path:[92,93,95],initi:[7,9,14,19,22,23,24,26,35,48,51,52,57,61,65,70,74,75,80,93,99,101],initialize_op_attr:57,initrd:106,inlin:[67,76,77,86],inner:74,input0:77,input1:77,input:[6,7,8,13,17,19,20,22,23,24,25,26,30,31,32,33,34,35,36,38,39,40,42,43,47,48,49,52,53,56,57,58,59,60,63,65,66,67,69,70,72,74,75,76,77,80,82,88,95,99,101],input_data:74,input_data_target:74,input_hassub_sequence_data:74,input_index:74,input_label:74,input_lay:74,input_nam:5,input_seg:70,input_sequence_data:74,input_sequence_label:74,input_sparse_float_value_data:74,input_sparse_non_value_data:74,input_t:74,input_to_check:6,input_valu:6,input_var:[6,57],inputbuff:13,inputdef:74,inputgradi:60,inputlayers_:74,inputs_to_check:6,inputsizechang:43,insert:[7,25,32,51,56,59,60,72],insid:[2,7,10,22,24,25,26,39,43,53,54,58,59,60,86],inspir:54,instal:[0,1,2,18,42,62,72,78,87,91,97,102],install_android:103,instanc:[6,8,10,12,16,19,20,22,23,25,27,32,37,43,48,49,52,57,59,60,74,77,82,93,98,101],instance_ip:86,instanti:[10,27,99],instead:[0,6,7,9,13,18,19,21,22,30,31,35,36,66,72],instrins:30,instruct:[0,2,8,35,98,103],int16:71,int32:[47,55,70,71,93],int64:[22,28,47,64,69,71],int64_t:54,int8:47,integ:[14,18,19,21,30,45,49,74],integer_value_sequ:[49,66,101],integr:[1,106],intel:[31,47,67,76],intellig:40,inteloptimizedpaddl:42,intend:1,intens:66,inter:22,interact:[2,22,86],interchang:[35,59],interconnect:51,interest:[19,30,51,98],interfac:[1,8,14,18,20,29,36,51,53,59,60,66,67,75,77,86,104,106],intermedi:[0,22,29,32,34,40,50,66,103,105],intern:[30,66,83,84,86,97],internel:42,internet:[9,10,106],interpret:[1,27,31,32,71,98],inth:77,intrins:[19,27,30,105],introduc:[8,10,20,34,41,44,61,63,65,75,81,83,85,87,97],introductori:0,intuit:[17,59],inval_:42,invalid:[58,63],invent:31,invoc:[9,33,59],invok:[0,7,20,22,25,53,57,59,60,65,69,72,86,98],involv:[49,75],ios:104,ios_arch:104,ios_deployment_target:104,ios_development_root:104,ios_enable_bitcod:104,ios_platform:104,ios_sdk_root:104,ios_use_veclib_for_bla:104,ipad:104,iphon:104,ips:86,ipt:[57,65,101],ipx:106,ipython:5,is_cpu_plac:43,is_mkldnn_librari:43,is_seq:101,is_target:56,is_tensor:65,is_test:43,is_traget:56,isinst:25,ismkldnnkernel:43,isn:98,isspars:74,issu:[0,2,4,9,34,66,72,98],istag:62,item:[17,30,58,80,106],iter:[5,10,22,31,32,40,43,52,54,58,66,70],iter_multiple_input_and_param:57,its:[4,5,7,8,10,14,19,20,21,24,25,26,31,32,34,35,36,38,40,44,48,49,51,52,53,56,57,59,60,63,64,65,67,69,74,75,76,77,83,86,93,98],itself:[7,10,17,20,32,43,52,63],ivs:65,java:[8,45,55,59],jeremi:98,job:[2,7,17,19,22,25,59,81,82,83,85,92,93,95],job_desc:22,job_dispatch_packag:84,job_nam:[18,86],job_namespac:86,job_path:86,job_workspac:84,jobdesc:22,jobnam:22,jobpath:86,jobport0:86,jobport1:86,jobport2:86,jobport3:86,jobserv:18,join:10,journei:2,jpg:20,json:[36,66,86,87],juditski:52,jupyt:[2,18],just:[0,1,4,9,14,15,19,22,31,32,34,38,43,52,53,57,58,59,60,61,63,64,72,84,86,95,103],jx4xr:86,jypyt:5,k8s:[19,106],k8s_data:86,k8s_job:5,k8s_token:5,k8s_train:86,k8s_user:5,kafka:12,kcpu:54,kcuda:54,kcudnn:76,kdisabl:54,kebilinearinterpbw:98,kebilinearinterpfw:98,keep:[1,10,21,31,32,35,49,52,57,63,65,72,106],kei:[0,6,7,8,10,12,14,28,30,39,43,59,60,65,66,70,72,75,98],kenlm:66,kept:[40,57],kera:61,kernel0:76,kernel1:76,kernel:[6,21,30,32,39,42,52,54,61,64,66,67,69,75,77,98],kernel_hint:39,kernel_type_for_var:69,kerneltyp:[39,43],key1:93,key2:93,key_pair_nam:86,keyid:86,keymetadata:86,keypair:86,keyserv:86,keystat:86,keyusag:86,keyword:[57,68],kforcecpu:39,kill:[10,86],kind:[2,5,6,10,16,20,22,25,32,35,39,43,50,51,54,67,69,71,76,86,87],kind_:54,kmark:54,kmkldnn:76,kms:86,knchw8c:47,knchw:47,knhwc:47,know:[5,14,19,20,40,44,72,74,81,86,97,98,103],knowledg:66,known:[7,8,21,31,33,48],kplain:76,kpoprang:54,kpushrang:54,kqueue:21,kselectedrow:64,kstate:54,kube_cluster_tl:5,kube_ctrl_start_job:5,kube_get_workers_addr:19,kube_list_containers_in_job_and_return_current_containers_rank:5,kubeconfig:86,kubectl:[84,87,88],kuberent:[10,86],kubernet:[5,10,19,22,59,85,89,90,106],kubernetes_service_host:5,kusecudnn:39,kusemkldnn:39,kwarg:[26,36,57,65],l1_regularization_op:61,l2_regularization_op:61,l93:13,label:[22,26,31,34,35,36,50,55,58,66,69,82,87,99],label_fil:58,label_lay:58,label_path:58,lag:93,lambda:[19,25],lan:91,languag:[19,21,31,35,40,54,59,63,66,68,95],larg:[22,24,25,40,44,52,66,72],larger:[40,68],larger_than:[8,37,55],last:[7,25,40,48,54,55,93,101],last_seq:49,lastseen:87,latenc:[30,66,84,86],later:[1,4,9,59,61,66,67,75,77,86],latest:[1,2,4,8,9,10,17,62,78,87,103,104],latter:[52,70,97],launch:[43,86,93],launcher:5,layer:[7,8,13,19,21,22,24,31,34,35,37,50,52,55,58,59,61,65,66,67,70,73,80,82,92,93,99,101],layer_0:74,layer_attr:[95,101],layer_help:39,layer_num:95,layer_typ:[41,42],layerbas:74,layerconfig:74,layergradutil:74,layerhelp:[39,57],layermap:74,layout:[43,69],layout_:[39,47],layouttyp:39,lazi:[52,61],lead:[20,40,47,98],leaki:34,learing_r:50,learn:[0,2,5,7,15,17,21,22,24,27,34,35,40,42,49,51,52,54,58,59,61,67,74,75,78,82,98,101,104],learning_r:[15,22,82,99],leas:10,least:[4,10,103],leav:[8,86],lectur:40,left:[8,77,104],legaci:2,legal:65,len:[15,19,28,31,57,74,80],length:[15,30,41,44,48,49,59,66,70,87,93,101],leran:40,less:[5,25,82,106],less_equ:68,less_than:[5,40],let02:87,let:[5,8,17,19,21,32,33,35,39,43,47,48,49,50,60,67,69,75,82,86,97],level:[30,33,36,44,48,49,54,67,70,71,72,84,93,103],lgtest:9,lgtest_main:9,lib64:[2,93],lib:[1,2,46,81,97,103,104,105],libapi:9,libari:46,libc:4,libcuda:2,libgcc_:4,libgoogl:97,libiomp5:42,libmkldnn:42,libmklml_intel:42,libnvidia:2,libpaddl:[45,46,59,97],libpaddle_capi:46,libpaddle_gserv:46,libpaddle_math:46,libpython2:1,librari:[1,4,9,16,21,22,42,43,46,51,66,69,75,76,81,82,91,93,104,105],library_:47,library_typ:76,library_type_:69,librarydevicecontext:47,librarytyp:[69,76],libstdc:4,licens:[42,51,72],life:10,lifecycl:[54,106],lifetim:[4,63],lightweight:33,like:[0,1,4,7,8,9,10,13,18,19,21,27,31,32,33,34,35,36,38,43,47,51,52,57,58,59,60,61,63,64,66,68,69,70,72,81,85,86,92,95,97,98,99,101,103,104,105,106],limit:[31,40,44,49,59,61,93,98],linaro:105,line:[0,9,13,18,21,29,35,52,55,57,59,61,72,82,83,84,86,95,96,97,98],linear:[49,80],lineno:97,link1:30,link2:30,link:[4,9,28,29,63,75,86,106],linux:[0,2,4,21,28,72,86,91,105],linux_x86_64:[4,62],lipo:104,list:[0,5,7,8,9,14,18,19,27,29,31,34,47,50,53,54,57,60,63,70,74,75,80,81,82,86,93,95,97,101,105],listdir:81,listen:[10,19,22,81,82,93],listen_and_do:19,listenanddo:19,littl:[15,39,44,93],live:[75,99],live_in:40,live_out:40,load:[1,5,10,22,34,51,57,75,86,93],load_missing_parameter_strategi:[92,93,95],load_mnist:34,load_persist:82,loadsave_parameters_in_pserv:[13,92,93],local:[0,1,2,6,8,10,16,17,21,35,40,48,55,57,59,72,84,87,92,93,97],local_scop:6,localhost:[2,78],localpath:29,locat:[9,31,47,54,67,70,74,81,101,105],lock:[9,10,14,15],lod:[21,44,48,64,70,71],lod_desc:[64,71],lod_expand:49,lod_level:[57,64,71],lod_tensor:[48,64,71],lodtenosr:20,lodtensor:[20,21,38,44,59,71],lodtensordesc:[44,64],log:[4,14,22,29,34,74,81,84,86,87,88,93,105],log_barrier_abstract:93,log_barrier_lowest_nod:[92,93],log_barrier_show_log:[92,93],log_clip:[92,93],log_error_clip:[92,93],log_period:[87,93,95],log_period_serv:[92,93],logic:[17,22,24,25,34,38,50,51,53,63,70,75],login:[4,88],logit:[34,69],longer:[10,22,40],look:[8,18,19,21,31,32,35,52,57,60,61,66,81,82,86,87,92,99],lookahead:66,lookup:[38,49,99],lookup_t:40,loop:[6,8,20,31,40,54,58,63],loop_var:70,loss:[7,22,34,36,50,52,61,66,74],lot:[22,47,49,52,57,61,67,81,92,106],low:[50,51,66,67,70],low_rnn:48,lower:[30,48,49,72,84],lower_level_rnn:48,lowest:93,lpaddle_capi_shar:46,lpaddle_capi_whol:46,lrelu:34,lstm:[87,101],lstmemori:101,lstmlayer:41,luckili:40,mac:[0,46,72,103],machin:[0,1,4,22,24,31,34,40,42,51,52,61,74,86,87,88,92,93,95,105,106],machine_transl:101,maco:[0,4,80,104],macro:[33,47,60,75],made:[10,15,31,101],mai:[1,2,6,8,22,26,30,32,39,40,43,51,54,58,59,63,66,69,77,78,81,86,98,105],main:[4,19,21,25,31,32,36,51,55,59,83,86,97],main_program:[7,26],mainli:[1,16,40,47,67,75,93],mainlin:4,maintain:[8,14,52,57,59,86],majel:9,major:[22,30,69,103],make:[0,1,5,7,8,9,10,14,15,17,21,22,23,30,31,35,48,49,52,53,57,58,59,61,66,70,72,74,75,76,82,83,86,97,98,103,104,105,106],make_chan:21,make_channel:21,make_ddim:77,make_function_oper:33,make_vari:65,maker:[59,60],malloc:[67,74],man:28,manag:[4,10,15,16,19,21,22,29,54,63,67,78,85],mandarin:66,mandatori:104,mani:[0,7,9,14,19,21,31,34,39,40,49,53,54,57,59,60,63,64,65,68,69,70,93],manili:36,manipul:[31,57,60,84,104],manner:[52,61,66,67],mantain:40,manual:[22,50,52,60,84,103,104,106],manufactur:31,manylinux1:4,manylinux1_x86_64:[4,62],manylinux:62,map:[0,5,8,14,25,43,47,57,60,63,65,67,70,93,106],map_fn:70,mapreduc:[5,81],mark:[7,24,34,35,48,49,54,63,97,101,106],marker:54,market:30,master:[5,17,59,62,93,105],mastermind:9,mat:[45,46],mat_cache_row:13,mat_norm:13,mat_normal_shar:13,mat_sparse_row:13,mat_sparse_row_auto_grow:13,mat_sparse_row_id:13,mat_sparse_row_prefetch:13,mat_sparse_row_prefetch_full_s:13,mat_value_shar:13,match:[4,9,30,68,98],matchbox:106,math:[42,45,59,72,74,75,98],mathemat:61,matmul:[8,36,48,70,75],matric:[74,101],matrix:[13,45,46,74,75,92,95,104],matrixptr:74,matrixtyp:46,mattyp:13,matur:85,max:[6,23,25,40,57,93,95,98],max_diff:6,max_length:[49,101],max_relative_error:[6,75],maxim:25,maximum:[8,15,75,93,98,101],maxoutfunctor:67,mayb:[8,43,75],md5:11,mean:[0,2,7,9,22,23,25,36,49,56,58,63,66,69,75,82,86,93,95,97,98,99,101,106],meant:70,measur:[26,98],mechan:[7,16,20,26,43,57,60,76,86,101],mem:[8,18,49],mem_per_pserv:22,mem_per_train:22,member:[5,25,35,36,47,53,57,63,75],memcpi:[53,98],memori:[0,7,8,13,14,18,30,32,42,43,44,47,49,52,54,59,69,72,74,77,87,93,95,98,99,101],memory_optim:40,memory_test:0,memory_threshold_on_load_data:93,memoryalloc:67,memorydesc:43,mention:[0,7,9,14,22,24,31,48,51,52,54],merg:[15,17,23,26,42,48,51,53,72,78,93,104],messag:[8,19,21,27,31,32,35,44,54,55,56,57,59,60,64,71,72,87,93],metaclass:75,metadata:[28,86,87],metal:[82,106],metaphor:35,metaplotlib:5,method:[1,2,4,6,8,17,19,20,22,23,25,30,34,35,36,39,50,51,57,58,59,63,64,69,70,74,75,77,78,93,95,97,98],methodolog:52,metric:[26,54],microarchitectur:30,might:[0,8,9,19,21,31,40,55,66,72,74,86,97,103],mileag:98,million:95,min:[23,25,57,86,95,98],min_block:8,min_count:24,min_desc:8,mind:97,mini:[8,10,21,26,27,31,37,48],mini_batch:58,minibatch:[8,26,35,37,55],minim:[8,22,24,25,31,34,50,59,82,93,99,103,104,105],minimum:[66,82,104],minimun:93,minsizerel:[103,104,105],minu:60,minus_grad:60,minusgradop:60,minusop:60,minusopgradmak:60,minusopprotoandcheckermak:60,minut:[0,2,10,17,86],mirror:[2,9],mislead:15,miss:[34,93],mistak:31,misus:76,mit:86,mix:[54,70,101],mixtur:97,mkdir:[1,29,78,86,88],mkl:[1,2,43,59,67,69,76],mkl_packed_:41,mkldnn:[42,47,69],mkldnn_:42,mkldnnactiv:42,mkldnnbase:42,mkldnnlayer:42,mkldnnmatrix:42,mkldnnstream:42,mkldnntester:42,mklml:42,mklpack:41,mklpackedgatedrecurrentlay:41,mklpackedgemm:41,mklpackedlstmlay:41,mklpackedrecurrentlay:41,mlp:36,mnist:[12,22,34,35,55,58,59,97],mnist_random_image_batch_read:58,mnist_train:58,mnist_train_batch_read:58,mobil:[30,31,40,59,78],mod:81,mode:[30,41,51,72,93],model:[7,8,10,11,19,22,24,25,26,35,40,41,50,51,52,59,61,66,69,70,72,74,78,80,82,83,86,93,96],model_list:[93,95],model_path:95,modelparallel:22,modern:40,modif:[3,66],modifi:[22,30,36,61,74,75,81,84,86,101],modul:[22,33,34,49,66,70,75,97],modular:49,moment:97,momentum:[63,76],momentumop:97,mon:87,monitor:[21,54],month:9,more:[0,1,2,5,6,7,9,10,14,17,18,20,21,22,24,29,30,31,32,33,35,39,40,43,47,48,49,50,54,57,58,59,61,66,67,68,70,74,75,76,77,78,80,81,82,83,87,95,97,98,99,101,105,106],most:[4,5,7,9,17,21,22,32,35,36,47,49,52,54,58,61,66,67,68,74,92,97,98,99,101,106],mostli:[30,106],motiv:59,mount:[2,18,81,86,87],mountpath:[86,87],move:[2,10,14,20,29,31,52,86,98,106],movement:98,movidiu:31,mpi:[21,51,88],mpirun:88,mse:[31,35,50,55],much:[10,31,43,50,58,61,70,98],mul:[33,40,57,74,75],mul_grad:75,mul_op:75,mul_result:57,mulgradkernel:75,mulkernel:75,mulop:[33,75],mulopgrad:75,mulopmak:75,mult:[19,32],multi:[26,51,53,69,74,82,84,92,93,97,106],multigradientmachin:53,multipl:[5,6,14,15,17,19,21,22,24,26,31,32,33,39,51,54,59,66,69,71,74,75,81,83,86,93,95,97,101],multiple_input:57,multiple_param_attr:57,multipli:[19,74],must:[7,15,20,25,40,43,44,47,54,56,57,58,59,65,71,74,75,77,81,82,86,93,95,101,103,105],mutabl:[67,77],mutable_data:[43,67,75,77],mutex:21,mutuable_data:[67,77],mxnet:[8,19,21,31],my_cluster_nam:86,my_external_dns_nam:86,my_lib:81,myerrorclip:25,mypaddl:87,naiv:19,name:[0,2,4,5,6,7,8,10,12,13,15,18,19,20,22,26,30,33,36,39,42,43,44,46,47,49,54,55,57,59,62,64,65,70,71,74,75,76,80,81,82,87,89,90,93,95,98,99,101,104,106],name_:54,name_prefix:12,namespac:[8,37,45,57,74,75,87],nativ:[30,72],natur:[14,17,24,49,70,95],navig:78,ncall:97,nccl1:51,nccl2:51,ncclinit:51,nchw8:69,nchw8c:69,nchw:[42,47],ndarrai:12,ndk:103,nearest:30,nearli:[6,20],necess:70,necessari:[7,8,15,17,25,26,40,44,49,53,57,65,70,74,88],necessarili:[19,74],neck:51,need:[1,2,3,4,5,6,7,9,13,14,15,17,18,20,21,22,24,25,26,29,31,32,33,34,39,40,43,47,49,50,51,52,53,54,56,57,59,60,61,63,64,65,66,67,70,74,75,76,77,78,80,81,82,86,87,91,92,93,95,101,103,104,105,106],neighberhood:51,neon:[30,103,105],nervana:31,nessesari:66,nest:[7,8,54,55,71],net:[8,34,48,63],netop:[8,59],network:[5,6,7,8,10,13,22,24,26,34,36,40,41,42,48,50,52,54,57,58,61,63,65,66,67,69,71,74,75,77,80,81,82,83,93,98,106],network_config:95,networkadministr:86,neural:[5,7,8,10,22,36,40,41,42,48,52,61,63,67,69,71,77,80,82,83,93,98],neuralnetwork:53,neuron:74,never:[40,58,63,86,87],new_block_idx:57,new_op_and_kernel:76,new_op_desc:25,new_scop:69,new_stat:48,newblock:57,newbuff:43,newer:103,newest:15,newli:[30,104,106],newop:8,newopdesc:57,newprogram:57,newscop:69,newvardesc:57,next:[7,10,16,20,21,25,49,51,70,74,75,86,87,93,97,98,101],nextlay:42,nfs4:86,nfs:86,nfsver:86,nic:[92,93],nil:[14,21],nnz:74,no_grad_dict:7,no_grad_set:[6,7,75],no_gradi:7,node1ip:88,node2ip:88,node3ip:88,node:[0,9,17,19,22,24,36,40,49,51,59,74,81,82,84,85,86,87,88,91,93,106],node_0:86,node_1:86,node_2:86,node_id:81,nodeattr:36,nodeentri:36,nodefil:84,nodesep:36,nohup:81,nois:[10,34,83],noisi:34,non:[10,21,30,31,64,74,75,86,93],none:[5,6,7,8,21,25,26,34,36,37,48,49,50,55,57,65,70,82,99,101],noneedtran:43,nonlinear:74,nor:[0,19],norm:[34,47],normal:[52,66,74,87,91,93,101],notat:40,note:[1,2,5,7,8,13,14,18,40,44,47,51,58,59,67,69,75,77,78,81,82,86,93,95,98,104],notebook:[2,18],notest_dist_fit_a_lin:82,noteworthi:31,noth:[0,39,57,63,93],notic:[25,31,51,60,72,74,101],notif:72,notimplementederror:25,notin:69,notingradi:75,notion:70,notori:6,now:[7,9,10,20,21,24,34,44,47,52,59,60,61,63,82,86,93,104],nproc:0,nullptr:[43,54,60,63,74],num:[81,93],num_class:36,num_gradient_serv:[81,92,93],num_hidden:36,num_input:72,num_parameter_serv:5,num_pass:[87,92,93,95],num_pserv:22,num_row:64,num_shard:12,num_step:70,num_train:22,number:[0,8,10,12,24,26,40,52,54,58,59,65,70,74,81,82,85,86,93,97],numdevices_:95,numer:75,numeric_grad:6,numerical_grad:6,numlogicaldevices_:95,numpi:[1,12,30,34,57,58,75],numreal:13,numsampl:98,numtimeout:14,nv_:9,nv_gpu:0,nv_librari:9,nv_test:9,nvcc:[9,30,32],nvidia:[0,2,30,47,51,67,93,98],nvlink:51,nvprof:54,object:[5,13,20,22,25,26,34,36,40,45,50,54,57,59,61,63,77,98],observ:[74,98],obtain:[17,20,52,67],obvious:[9,47,97],occup:40,occupi:[30,54],occur:40,occurr:8,oct:87,off:[0,1,2,46,82,91,103,104,105,106],offer:[8,59,65],offici:[9,72,78,86,103],offlin:[10,12,106],offset:13,often:[13,36,40,47,72,81,97],ograd:74,old:[6,15,17,49,59,93],older:[31,103],omega:61,omp_num_thread:97,ompi_comm_world_rank:81,onc:[10,14,19,22,24,26,31,35,52,68,72,74,78,86],one:[0,2,5,6,7,8,10,13,14,15,17,18,19,20,21,22,25,26,27,30,31,32,33,34,36,38,39,43,44,47,48,49,50,51,52,53,55,56,57,58,59,60,63,64,66,67,68,69,70,72,74,75,76,82,84,86,87,88,93,95,99,106],onehotcrossentropyopkernel:75,ones:[33,34,59,72],onli:[0,1,3,4,5,6,7,9,13,14,15,16,17,18,19,20,22,24,25,26,27,29,30,31,34,35,40,43,48,49,50,51,53,54,57,59,64,65,66,67,68,69,70,74,75,76,77,78,82,86,87,92,93,95,98,101,104,106],onlin:[10,12,40,58],only_cpu:6,onnx:31,onto:[22,24,86,88],op1:[40,69],op1_2_op2:69,op1_to_op2:69,op2:[40,69],op3:40,op_:75,op_check:75,op_class:[59,65],op_desc:[25,38,56],op_info:99,op_kei:43,op_maker_class:[59,65],op_proto:65,op_registri:99,op_siz:25,op_test:75,op_typ:[59,75,76],op_unique_kei:43,opattrcheck:75,opcreat:65,opdesc:[8,25,35,55,56,57,59,60,65,71],opdescbind:[38,60],opdescbuild:8,open:[5,12,31,34,42,58,72,85,86,97],openbla:[1,2,103],openmp:97,openmpi:85,opensourc:51,oper:[6,8,19,21,22,23,24,26,27,30,31,32,34,35,36,38,39,48,49,50,51,54,56,61,63,66,67,69,71,72,74,76,77,86,93,98,99,101,103],operand:30,operator_grad:6,operator_list:54,operatorbas:[8,33,59,60,65,75],operatorwithkernel:[69,75],opinfo:[38,59,60],opinfomak:38,opinfomap:60,opkernel:77,opkernelbas:76,opkernelkei:59,opkerneltyp:[47,69,76],opmak:65,opproto:75,opprotoandcheckermak:[60,75],opprotomak:[65,75],opregist:65,opregistri:65,ops:[6,7,8,9,19,20,32,35,36,52,55,56,57,59,67,75,106],ops_:8,ops_test:9,opt:[1,5,50,56,65],opt_op_list:50,optest:75,optestmeta:75,optim:[6,7,22,23,24,32,34,51,52,53,55,59,61,64,66,74,82,83,97,98,99,103,105],optimis:50,optimize_op:82,optimize_op_attr:57,optimizer_op:82,option:[0,5,9,22,34,39,44,55,56,57,59,64,65,66,71,74,81,82,84,95,97,103,104,106],optmization_op_list:50,opts_np:56,optyp:[38,65],opwithkernel:64,order:[1,7,35,44,54,58,61,70,74,83,86,87,93,97,106],oregon:86,org:[12,28,34],orient:65,origin:[6,30,34,63,70,72,77],other:[0,1,8,10,15,19,29,30,31,32,38,40,43,47,48,52,56,61,63,65,66,67,69,72,82,83,86,87,95,97,99,101,103,104,105,106],otherwis:[5,7,10,15,17,34,38,43,58,66,72,84,95,101],our:[0,1,4,5,7,9,20,21,22,24,34,38,40,47,51,52,63,70,72,74,82,86,87,97,101,103],out:[5,8,9,14,17,20,22,25,31,36,40,43,48,49,57,69,75,77,80,86,87,88,93,97,98,101],out_dir:86,out_mem:101,outgrad_:42,outlin:94,output:[0,1,5,6,7,8,12,17,19,20,24,25,29,32,33,34,35,36,37,38,40,43,44,48,49,52,55,56,57,58,59,60,63,64,65,67,69,70,72,74,75,76,77,81,87,93,95,97,98,101,103],output_:[42,74],output_all_step:48,output_arg_nam:25,output_lay:80,output_mem:101,output_nam:6,output_num:48,output_path:12,output_seg:70,outputbuff:13,outputgradi:60,outsid:[22,63],outupt:70,outv:74,outval_:42,over:[5,31,40,51,52,70,72,74,98],overal:[34,52,54,72,106],overfit:61,overhead:98,overlap:74,overload:[30,39],overrid:[8,10,29,43,67,74,75,77],overview:[14,15,16,67],overwhelm:72,overwrit:[29,81],own:[7,15,17,25,27,36,38,50,51,52,61,65,75,81,84,86,103,104],owner:[0,72],pack:[70,103],packag:[1,2,14,18,19,33,42,62,72,75,86,97],pad:[43,66],paddl:[0,1,2,4,5,8,9,10,12,18,20,22,29,32,33,34,37,41,42,43,44,45,46,48,49,53,55,59,61,62,65,66,67,70,72,74,75,76,78,80,81,82,84,86,87,88,91,93,95,97,98,99,101,103,106],paddle_begin_init_param:15,paddle_dir:75,paddle_element_typ:15,paddle_element_type_float32:15,paddle_element_type_float64:15,paddle_element_type_int32:15,paddle_element_type_int64:15,paddle_element_type_uint32:15,paddle_element_type_uint64:15,paddle_enforc:[8,20,43],paddle_enforce_eq:[75,77],paddle_error:[45,46],paddle_exampl:18,paddle_finish_init_param:15,paddle_get_param:15,paddle_gradi:15,paddle_init_num_gradient_serv:81,paddle_init_param:15,paddle_init_port:81,paddle_init_ports_num:81,paddle_init_ports_num_for_spars:81,paddle_init_pserv:81,paddle_init_trainer_count:81,paddle_init_trainer_id:81,paddle_init_use_gpu:81,paddle_job:18,paddle_manylinux_devel:1,paddle_matrix:[45,46],paddle_matrix_cr:46,paddle_matrix_get_shap:45,paddle_matrix_shap:45,paddle_new_etcd_pserver_cli:15,paddle_new_pserver_cli:15,paddle_on_cloud:18,paddle_output:87,paddle_paramet:15,paddle_pserver2:84,paddle_pserver_cli:15,paddle_pserver_client_releas:15,paddle_save_model:15,paddle_send_grad:15,paddle_train:[46,62,84],paddle_with_cuda:54,paddle_with_mkldnn:47,paddlepaddl:[0,3,9,10,12,15,16,17,18,19,22,28,29,33,34,35,37,39,44,48,49,50,53,54,57,58,59,63,70,71,72,74,75,76,80,81,83,84,85,88,89,90,91,97,98,101,106],paddlepaddle_gpu:4,paddlepaddlebook:2,paddlepaddlehub:2,page:[72,86],pair:[7,8,22,35,50,54,59],pakcag:9,panic:21,paper:[34,66],para:13,paradigm:[19,27,59],paragraph:48,paragraph_data:48,paragraph_out:48,parallel:[0,19,21,22,24,51,54,59,69,81,83,85,86,87,93,95,98],parallel_for:19,parallel_nn:[92,93],paralleldo:23,parallelfor:19,paralleliz:66,param:[6,8,15,53,57,67,77],param_attr:[13,57,101],param_config_proto:15,paramattr:101,paramet:[1,6,7,8,9,11,13,17,19,22,23,25,27,29,31,32,34,35,36,38,44,48,50,51,55,58,63,65,66,70,72,74,75,76,77,80,83,84,95,96,99,104],parameter_block_s:[92,93],parameter_block_size_for_spars:[92,93],parameter_list:[7,50],parameter_nam:5,parameter_serv:5,parameter_valu:13,parameterattribut:13,parameterclient_:13,parametermap:74,parametermutex_:13,parameters_:74,parameters_and_grad:50,parameterserver2:13,parameterset:5,parameterupdat:53,parameterupdater_:13,params_grad:[50,82],paramt:86,paraspars:74,parent:[8,19,55,57,59,74],parent_:[8,63],parent_idx:57,parenthes:59,pars:[0,9,22,36,86,95],part:[7,8,17,22,31,43,44,55,57,66,67,74,81,82,83,97,98,101,106],parti:[0,98,103,104,105],particip:75,particular:[35,44,59,66,69,98],partit:[10,12,22,24,59,83,86],pass:[7,8,10,21,25,26,31,34,40,44,50,52,53,56,57,58,59,61,63,66,68,70,72,74,81,82,84,86,87,93,98],pass_id:[22,82],pass_idx:58,pass_num:82,passtyp:74,password:88,past:[2,5,80,86],patch:28,path:[1,2,10,14,15,18,40,49,58,66,81,86,87,93,95,103,104,105],path_to_paddlepaddle_working_directori:78,pattern:[10,45,52,61,86],paus:[10,17],pcie:51,peer:51,pem:[5,12,86],pend:[10,14],peopl:0,pep425tag:4,pep8:72,per:[10,15,51,52,58,61,75,93],percal:97,perf_test:97,perfectli:66,perfom:[93,95],perform:[0,6,15,21,22,26,30,31,34,40,51,53,54,58,59,61,66,67,69,74,75,82,85,92,96,101,103,104,105],perftool:[54,97],period:[10,17,93],permiss:86,persist:[27,64,66,71,86],persistentvolum:86,persistentvolumeclaim:86,person:[5,39],perspect:[59,98],perturb:[6,74],pex:106,pfs:[12,29],pfsclient:12,pfspath:29,pgp:86,phase:[43,49,51,52,58,60,66,106],philosophi:[52,61],photo:34,physic:[104,106],pick:[86,104],pickl:[81,88],pictur:51,piec:[19,54,77,82],pil:81,pillow:18,ping:72,pip:[1,3,62,72,78,80,97],pipelin:[26,66],pivot:43,pixel:22,place:[7,8,10,17,22,24,27,39,43,51,59,69,74,77,82,98,99],place_:[39,47,67,69],place_typ:76,placehold:[34,67,77],placement:24,plain:[18,44,46,47],plan:[10,19,43,59,66,74,103],platform:[4,8,32,43,47,54,67,69,72,75,76,77,81,85,86,99,103,104,105],pleas:[0,1,2,4,5,10,14,15,16,19,21,32,36,47,48,57,58,59,66,67,71,72,74,75,77,78,82,86,91,97,101,103,104,105],plot:5,plug:[51,52],pne:75,pnpairvalidationlay:93,pnpairvalidationpredict_fil:92,pod:[12,18,19,86,87],pod_nam:86,point:[0,8,10,18,21,30,40,43,51,67,72,75,77,97,98,103,106],pointer:[8,15,36,40,47,57,59,63,67,77],polici:86,poll:21,pollut:17,polyak:52,ponit:36,pool3:74,pool:[23,40,66],pop:[8,27],popul:15,popular:[9,34,36,54],port:[9,19,81,82,86,87,92,93,97],port_num:92,portabl:36,portal:78,ports_num:[81,93],ports_num_for_spars:[13,81,92,93,95],pose:10,possibl:[5,8,14,21,24,40,57,61,71,98],post:[18,28],postpon:61,potenti:[30,98],power:[30,40,51,66,77,106],ppo_workspac:78,pprof:97,practic:[74,101],pre:[1,5,15,39,40,72,86,87,103,105],pre_activ:57,pre_bia:57,pre_stat:[48,70],preambl:57,precis:[1,26,30,52],precompil:27,pred:[36,40],predecessor:40,predetermin:93,predict:[22,61,80,93,101],predict_fil:93,predict_output_dir:[92,93],prefer:[31,39],prefetch:[13,74],prefix:[10,12,49,66,86],pregel:21,pregrad:74,prepand:57,prepar:[6,18,53,66,83,89,101],prepend:57,prepend_oper:57,preprocess:[66,70,87],present:[5,7,8,54,70],preserv:29,prev_batch_st:[92,93],prevent:[5,10,14,17,25,61,97],preview:[59,78],previou:[7,10,24,29,48,49,74,86,93,97],previous:87,previous_memori:8,price:[59,80],prim:43,primari:[31,35],primarili:[52,61],primer:72,primit:[30,42,43,51,70],primitive_desc:43,primitivedesc:43,principl:[5,9,47],print:[4,5,21,22,31,36,57,68,80,88,93,97],print_graphviz:36,printallstatu:98,println:21,printstatu:98,prioriti:59,privat:[8,46,54,57,63,64,65,67,70,72,77],privileg:[0,86],pro:51,prob:80,probabl:[2,49,66,72,101],problem:[0,4,5,6,9,17,20,31,34,35,52,59,61],proc:2,proce:[2,10,58,86],procedur:[8,44,77,104],process:[5,7,8,12,13,14,17,19,20,21,22,26,27,31,32,36,40,42,44,51,61,65,69,72,84,85,86,87,93,95,97,101],processor:[30,98],produc:[10,31,36,58],product:[18,31,74,86,87],productgraph:87,prof:97,profil:[29,54,66],profilerst:54,proflier:[54,98],program:[5,7,12,15,17,22,24,27,35,37,40,50,51,54,58,59,63,68,71,84,93,97,98],programdesc:[19,22,27,31,40,44,56,57,60,71],programm:[22,31,57],progress:[10,14,93],project:[18,46,66,72,74,75,101],promis:49,prompt:[29,31,82],prone:5,pronunc:66,prop_kind:43,propag:[7,31,52,75,93,95],proper:[39,91],properli:[0,39,82],properti:[36,61,93],propos:[8,23,24,49,50,51,52,70],proprietari:42,protect:[20,30,65,74,75],proto:[21,39,44,47,55,59,65,71,75],proto_:65,protobuf:[8,18,19,22,27,31,32,35,36,40,44,55,57,59,60,65],protoc:[103,105],protocol:[93,99,106],provi:81,provid:[2,5,8,15,18,19,26,27,30,31,34,36,38,39,47,51,52,54,57,61,65,66,67,68,70,77,80,81,82,84,85,86,97,98,103,104,106],providermemory_threshold_on_load_data:92,provis:[86,106],prune:8,ps_desir:10,pserver:[13,15,16,18,59,81,82,86,92,93],pserver_addr:15,pserver_cpu:18,pserver_endpoint:82,pserver_id:11,pserver_mem:18,pserver_num_thread:[13,92,93],pserver_prog:82,pserver_startup:82,pserverstart_pserv:92,pseudo:[5,7,18,60,70],pseudocod:70,psize:74,ptr:[46,67],pub:88,publish:103,pull:[9,59,62,72,103],purpos:[10,22,24,39,98],push:[8,27,31,54,62,72],push_back:74,put:[9,10,13,24,40,43,57,67,74,82,87,103],pvc:86,pwd:[0,1,2,78,103],pxe:106,pybind:[8,21,30],pypi:4,python2:97,python3:4,python:[0,1,2,4,5,8,16,20,21,26,27,31,33,34,35,36,39,45,49,53,54,59,62,67,70,72,78,80,81,82,88,99,101],pytorch:[31,54],qualcomm:30,queri:86,question:[5,19,24,65,82,86],queue:[21,24],quick:[36,79,93],quick_start:[18,86,87,89],quick_start_data:87,quickli:[49,57,59],quickstart:87,quit:[49,98],r14b:103,rais:[25,36,81],rajathkmp:34,ran:[24,98],rand:[34,93,95,98],random:[12,21,34,47,53,57,58,75,83,93],random_imag:12,randomli:17,randomnumberse:92,rang:[12,19,22,30,34,40,54,58,65,72,82,93,95],rank0:51,rank1:51,rank:[5,70,77,86],rankdir:36,rapid:60,raspberry_pi:105,raspberrypi:105,raspbian:105,rate:[15,66,74],rather:[7,18,34,70,86],ratio:93,raw:44,rdma:93,rdma_tcp:[92,93],reach:[10,40,51,98],read:[0,2,5,7,10,12,19,20,21,22,24,31,32,58,59,66,70,78,83,86,91,101,103,106],read_from_arrai:40,read_from_realistic_imag:5,read_from_rng:5,read_lock:11,read_minibatch:31,read_mnist_imag:5,read_ranking_model_data:5,readabl:[54,59,97],reader:[12,22,30,34,35,55,66,81,82,97],reader_cr:12,reader_creator_bool:58,reader_creator_random_imag:58,reader_creator_random_image_and_label:58,readi:[10,21,86,87,106],readlockguard:13,readm:46,readnext:20,readwritebuffer_:13,readwritemani:86,real:[13,34,58,81],realist:5,realiti:66,realiz:[8,48],realli:[31,61],reaon:76,reason:[5,6,10,21,31,72,82,87],receiv:[10,18,21,22,24,48,82],recent:[40,52],reciev:93,recognit:66,recommend:[1,2,3,5,72,74,78,84,91,93,101,103],recompil:98,record:[14,43,54,65,86],recordev:54,recordio:[5,12,14,20],recov:[10,70],recover:59,recoveri:14,recurr:[41,48,63,66],recurrent_group:[66,101],recurrent_op:70,recurrentgradientmachin:[46,49,70],recurrentlay:[41,93],recurs:[7,8,9,29,40,59],recv:[19,22,24,51,86],recvparametertyp:13,red:[34,97],reduc:[2,24,30,51,59,72,84,93,95,97],reduce_by_kei:59,reduce_mean:34,refactor:[22,24,35,49,52,53,57,61,70],refer:[0,1,2,8,10,14,15,16,19,30,36,43,47,48,51,55,57,59,61,63,67,70,71,74,75,77,82,87,101,103,104],referenc:14,reflect:14,reformat:72,refrain:75,reg:65,regard:[20,106],region:[63,98],regist:[21,40,47,60,67,69,74,98],register_gpu_profil:98,register_lay:74,register_op:[33,59,60,65,75],register_op_cpu_kernel:[67,75],register_op_cuda_kernel:[67,75],register_op_kernel:76,register_op_without_gradi:[59,75],register_oper:[38,60],register_tim:13,register_timer_info:98,registerop:65,registr:[75,76,99],registri:[18,38,67,87,106],regular:[7,74,86],reinit:20,reiniti:[20,43],reinstal:1,rel:[6,17,61,75,103],relat:[10,17,18,30,47,54,63,69,72,87,97,104,105,106],relationship:[60,67],releas:[62,66,86,103,104,105],relev:75,reli:[6,19,49,50,52,61,75,97],reliabl:[10,61],relu1:36,relu2:36,relu:[34,36,40,74],relwithdebinfo:97,remain:70,rememb:72,remot:[9,13,22,59,72,74,86,93,95],remoteparameterupdat:[13,16,93],remov:[7,22,29,31,49,72,93,103,104],renam:[4,7,29,30],reorder:43,reorder_primit:43,repeat:[8,35,55,56,64,65,71,97],repeatedli:[35,40],replac:[9,14,38,52,60,66],repli:72,replic:22,replicaset:18,repo:[9,72,78,105],report:[14,30,31,54,98],reportdataset:14,repositori:[78,103],repres:[7,8,14,19,22,24,25,31,36,44,47,49,52,57,59,61,64,67,70,71,74,86,101],represent:[15,22,32,34,35,40,47,49,64],reproduc:0,request:[9,10,13,17,19,59,62,72,86,87,106],requir:[0,4,5,7,10,15,17,18,20,22,24,25,29,30,36,40,42,48,52,54,55,56,59,61,64,65,66,71,72,74,75,78,81,86,87,103,105,106],requisit:40,rerun:75,research:[22,31],reserv:29,reserveoutput:74,reset:[10,26],reset_program:26,resetingrad:42,resetinvalu:42,resetoutgrad:42,resetoutvalu:42,resetxxx:42,reshap:[6,58,77],resid:0,residu:21,resiz:[13,67,75,77],resolv:[9,72,87],resourc:[22,27,51,54,67,76,86],respect:[6,20,25,30,34,48,74,93,101],respons:[13,21,22,26,34,51,52,53,61,86,87],rest:[8,18,28,32,69,106],restart:[10,15,86,87,106],restartpolici:[86,87],restor:[6,52],restrict:[61,63,93,97],result:[6,7,14,21,26,34,35,36,40,44,49,50,51,53,75,77,86,93,97,98,99],resum:17,retain:77,retran:86,retriev:[0,8,49,63,74,87,97],retriv:81,reuqest:62,reus:[8,17,49,58,59,74,75],rev:0,revamp:22,reveal:[5,97],revers:[7,101],review:[19,87],reviews_electronics_5:87,rewrit:[9,21,75],rid:[20,31],right:[6,7,8,9,18,26,40,59,61,72,76],ring:51,risk:7,rkt:[0,18],rmsprop:52,rmspropoptim:52,rnn:[8,31,34,49,57,59,63,66,92,96],rnn_bias_attr:101,rnn_layer_attr:101,rnn_out:101,rnn_output:70,rnn_use_batch:[41,92,93],rnnalgorithm:49,rnnstep:70,roadmap:[66,70],rocmplac:47,role:[5,14,15,22,51,82,86],rollback:57,root:[7,51,86,87,104],roughli:66,round:[30,51],routin:[30,42,51],row:[13,21,74],rows_:64,rpc:14,rpcserver:14,rpi:105,rpi_arm_neon:105,rpi_toolchain:105,rsize:86,rtk:106,rule:[7,22,25,31,35,74,86],run:[0,3,4,5,6,7,8,9,10,18,19,20,21,22,23,24,26,30,31,32,33,34,35,36,40,43,47,48,50,51,52,54,55,56,57,59,62,63,64,66,67,68,69,72,74,76,77,78,80,81,82,83,84,85,86,88,89,90,91,93,97,98,103,104,105,106],run_test:1,runinitfunct:98,runnabl:24,running_on_cloud:18,runserv:78,runtim:[2,8,19,21,22,38,48,59,71,76,84,103],runtime_table_:8,s_block:7,s_recurrent_group:101,safe:18,sai:[0,20,32,35,37,40,58,93,95],said:31,sake:74,same:[1,5,6,14,15,17,19,20,21,22,33,34,36,39,40,48,49,51,57,59,60,63,66,69,70,75,76,77,82,84,86,95,101],sampl:[2,26,34,57,65,81,84,93,95],sampler:34,satifi:40,satisfi:[4,9,43,64,86],save:[0,10,12,14,15,18,19,22,35,36,40,44,52,64,71,81,86,87,93,95],save_dir:[87,93,95],save_only_on:[92,93],save_persist:82,saving_period:[92,93],saving_period_by_batch:[92,93,95],scalabl:59,scalar:[7,8,37,68,70],scale:[22,24,52,60,65,66,75,85],scaleop:75,scaleopmak:[59,75],scan:[7,14,40,59],scatter:[7,51],scenario:[49,92],scene:92,schdule:86,schedul:[14,18,24,86],scheme:[13,61,75],scienc:40,scope:[6,19,23,27,32,69,99],score:49,scorer:66,scp:88,script:[0,1,51,75,81,84,85,86,88,103],sdk:104,search:[1,10,63,93,101],second:[5,19,29,31,34,36,48,49,55,56,58,63,65,75,84],secret:86,section:[7,24,31,57,72,74,83,86,97,101],see:[5,7,10,19,21,24,30,31,57,66,72,75,77,82,86,97,98],seed:[93,98],seem:[4,9,21,30,31,66],seen:[61,75],segment:[48,70,77],sel:21,select:[49,86],selected_generation_scor:49,selected_id:49,selected_row:[64,71],selected_rows_desc:[64,71],selected_scor:49,selectedrow:[38,71],selector:87,self:[6,25,26,34,36,40,41,42,44,50,57,70,74,75],self_addr:19,semant:[5,49,62],semaphor:21,semat:5,send:[10,15,19,22,24,39,51,59,65,72,82,83,86,93],send_back_parameter_typ:13,sendbackparameterspars:13,sendbackparametertyp:13,sendparameterrequest:13,sendparameterrespons:13,sens:[52,61,72,97],sent:[5,15,19,22,59,65,71,87],sentenc:[31,48,49,70,101],sentence_input:70,separ:[15,22,33,52,60,61,81,82,93],seper:70,seq_len:70,sequenc:[7,8,19,27,31,35,41,50,55,66,70,72,74],sequenti:[8,19,21,101],seri:[4,20],serial:[8,14,44,53,59],serializ:[59,71],serv:[2,22,30,59,70,81,86,98],server:[1,5,9,13,16,17,22,32,51,59,74,83,84,91,92,106],server_endpoint:82,serverless:10,servic:[81,97,106],sess:[34,36,50],session:[36,50,56,98],set:[0,1,5,7,10,18,20,34,38,40,43,47,48,49,54,56,57,59,60,63,66,67,70,72,74,75,77,78,81,82,84,86,87,92,93,95,96,97,98,101,104,105],set_active_typ:74,set_attr:25,set_drop_r:74,set_float_el:6,set_input:25,set_output:25,set_shap:20,set_siz:74,set_typ:[25,74],setdatatyp:64,setdefault:75,setp:86,setq:0,settup:74,setup:[22,52,62,74,75,106],seven:66,sever:[1,6,13,22,24,34,48,49,51,53,54,57,64,67,70,84,85,86,95],sexstant:106,sgd:[5,10,18,24,52,53,64,82,83,99],sgd_optim:[82,99],sgdasync_count:92,shall:[7,9],shape:[6,7,8,20,22,34,37,47,48,55,57,59,64,66,67,75,77,82,99],shapes_:20,shard:[10,11,12,13,14,15,17,22,24,83,86],share:[0,9,20,34,46,53,57,59,61,66,67,70,75,93,98],shared_librari:9,shared_ptr:[43,45,46,63,67,77],shell:[2,86],shoul:6,should:[5,6,7,8,15,18,21,22,25,26,30,32,33,34,38,39,43,47,48,49,50,52,53,54,55,58,59,60,61,64,65,66,68,69,70,71,75,76,78,80,82,84,86,101,103],should_be_fals:5,should_be_tru:5,show:[0,4,7,8,10,20,21,29,31,37,40,44,48,51,52,55,68,70,77,83,86,87,93],show_check_sparse_distribution_log:[92,93],show_layer_stat:[92,93],show_parameter_stats_period:[87,92,93,95],shown:[5,22,26,51,54,66,74,77,86,98,101],shrink:74,shrunk:25,shuffl:[20,22,82],shuffleread:20,sid:86,side:[22,26,40,53,77,83],sig:86,sigint:84,sigmod:65,sigmod_op:65,sigmod_output:65,sigmoid:[8,65,70,74],sign:[28,44,86],signal:84,signatur:86,signific:[66,98],similar:[8,19,21,22,24,27,31,49,52,54,58,59,61,66,67,69,70,75,86,97,106],similarli:[31,40,75],simpl:[19,24,30,32,35,36,40,48,52,55,61,63,65,66,70,82,93,98],simple_attent:101,simple_gru:101,simple_rnn:101,simpler:53,simplest:86,simpli:[2,5,15,22,80,98,101],simplifi:[5,49,57,65,66,74,87],simul:[31,104],simultan:86,sinc:[10,14,16,17,21,22,23,24,31,38,40,43,47,52,57,58,60,61,70,77,82,86,98,104,106],sincer:72,singl:[1,7,10,20,22,24,26,30,39,51,59,63,66,74,80,83,87,97],singleton:[19,23],sit:22,site:[9,86,97],situat:[7,56,69],size:[2,10,12,13,15,21,22,30,34,40,44,49,52,57,58,64,65,66,67,70,74,75,77,80,82,93,99,101,103,104,105],size_in_byt:43,size_t:[13,20,67,70,74],sizeof:8,skip:[7,58,72,84,86],slice:19,slide:10,slight:31,slightli:34,slow:98,slowli:[0,97],small:[6,19,32,34,42,49,72,74,93],small_messag:[92,93],smaller:[6,10,30,49,72],smart:63,snap:87,snapdragon:30,snapshot:[11,17,86],snippet:[21,33,50,74,86,98,101],sock:18,sock_recv_buf_s:[92,93],sock_send_buf_s:[92,93],socket:93,softmax:[5,8,22,24,31,36,37,49,55,74,101],softmaxoutput:36,softwar:[30,54,98,106],solid:34,solut:[51,106],solv:[5,7,20,40,59],some:[5,7,8,9,13,14,15,17,18,20,21,22,24,25,30,32,33,34,35,39,40,43,47,48,49,50,55,56,57,58,59,60,63,67,69,70,72,74,75,76,77,81,86,92,93,95,98,103,104,105,106],some_c_api_funct:46,some_inst:46,some_op:[38,48,70],some_python_class:45,somecppclass:45,somegotyp:45,someth:[0,7,13,57,72,97],sometim:[0,54,58,98],somewhat:15,somewher:63,soon:10,sophist:74,sort:[70,86,93,97],sort_by_length:70,sortagrad:66,sourc:[0,6,9,29,31,34,42,44,46,49,58,59,85,86,87,97,101,104],source_dict_dim:[49,101],source_dict_s:49,source_language_word:[49,101],space:[0,24,30,57,61,66,98,101],span:54,spars:[13,21,74,77,81,86,93],sparse_remot:13,sparse_upd:13,sparseparam:74,sparseprefetchrowcpumatrix:74,speak:101,spec:[86,87],specfii:93,special:[7,15,22,30,32,38,47,49,50,75],specialvartypeinfer:38,specif:[0,7,9,10,20,22,25,29,32,49,59,63,67,75,82,95,103],specifi:[0,2,5,6,13,14,15,18,19,21,22,23,25,26,27,29,34,44,54,57,63,65,70,72,74,77,78,86,93,97,101,103,104],spectrogram:66,speech:66,speed:[1,30,44,51,52,106],speedup:54,sphinx:[45,78],split:[17,19,20,23,31,37,49,59,70,81,83,86,95],split_count:[81,86],spread:7,squar:36,square_error_cost:[82,99],srand:93,src:[9,43,81],src_backward:101,src_embed:[49,101],src_forward:101,src_primitive_desc:43,src_word_id:[49,101],src_word_vec:49,sreializ:71,ssh:[86,87,88,105],ssh_server:84,sstabl:5,stabil:[6,40,75],stabl:[62,86],stack:[27,59,70,86],stage:[9,16,23,34,40,43,66,71,84,103],stale:10,stamp:98,standalon:103,standard:[0,4,21,31,59,61,66,97],stanford:[6,87],star:9,start:[0,1,2,4,7,9,10,13,14,15,17,18,21,22,23,49,51,53,54,84,88,93,97,98,101,102],start_mpi_train:88,start_op_idx:7,start_pass:[92,93],start_pserv:93,startup:[10,18,31,86],stat:[93,98],state:[8,10,26,27,48,49,54,63,66,70,76,87,93,101],statem:40,statement:[21,31,35,40,74,86],static_cast:[43,77],staticinput:101,statist:[26,54,93],statset:98,statu:[18,49,86,87,98],status:87,std:[9,13,20,36,38,39,43,45,46,54,56,59,60,63,65,67,74,75,77,93],stdbuf:81,stderr:84,stdout:84,step:[1,2,6,8,10,15,21,22,24,26,31,34,35,41,49,52,53,57,59,65,66,70,72,74,80,83,86,87,88,97,98,101,103,105,106],step_gradi:7,step_id:70,step_input:70,step_net:8,step_output:70,step_scop:59,stepnet:[8,48,59,63],still:[4,7,14,17,22,31,40,60,77],stirng:57,stmt1482205552000:86,stmt1482205746000:86,stochast:[10,14,17,52,83],stop:[0,57,84,87,93],stop_gradi:57,storag:[28,30,81,86,87],store:[6,8,9,13,27,36,38,44,47,49,53,55,57,59,60,61,63,70,74,75,77,78,81,83,86,87,88,93,104,105],str:[7,18,70,95],straight:[55,58,64],straightforward:43,strategi:[10,57,93],stream:[22,43,54,67,76],stream_:67,strict:[58,83],stride:[43,47,66],string:[7,8,14,29,36,39,44,54,55,56,57,59,60,63,64,65,71,74,75,86,93],strip:97,strongli:91,struct:[14,15,28,30,38,39,46,47,54,60,65,69,76],structur:[7,8,14,31,34,44,49,55,57,59,64,84,86],sts:86,stuff:72,style:[1,59,65],sub:[5,7,17,19,24,34,40,48,51,53,57,74,101,103],sub_block:7,subclass:57,subcommand:29,subgraph:[24,34],submiss:22,submit:[43,59,78,85,86,92,93],subnet0:86,subnet:[5,86],subobjectpath:87,subsequ:51,subset:74,subtract:6,succ:40,succeed:[14,87],success:[15,86,87],successfulcr:87,successfulli:75,successor:93,sucess:40,sucessor:40,sudo:[0,86],suffer:6,suffici:93,suffix:[4,18,81],suggest:[9,72,98],suit:106,suitabl:[64,67,93],sum:[7,8,11,23,38,57,74,101],sum_op:7,summar:[34,54],summari:54,sumopgradmak:60,supercomput:40,suppli:64,support:[0,1,2,4,6,8,10,17,18,19,21,22,24,31,33,34,40,43,44,47,49,52,53,54,56,58,59,60,61,64,66,69,74,75,77,78,80,83,85,86,93,98,101,103,104,105,106],support_inplac:40,suppos:[9,19,33,64,74],suppress:29,sure:[0,1,74,82,86,97],svs:65,swagger:28,swig:[1,16,45,46,103,104],switchop:8,sychron:51,symbol:[4,8,36,46],symbols_ready_:8,symbolt:[8,59],symlink:72,sync:[10,52,61,93],sync_with_cpp:97,syncflag:74,synchron:[10,14,21,43,51,54,83,86,93],syntax:[19,27,31,49,58],sysroot:103,system:[1,2,4,8,9,10,15,17,21,22,24,28,33,34,40,42,66,75,78,81,85,87,97,103],tab:4,tabl:[8,19,31,38,44,64,71,104],tablelookup:64,tablelookupgrad:64,tablelookupop:64,tag:[2,62,91,101],tail:49,take:[0,1,5,7,8,9,10,17,19,20,21,22,25,27,30,32,34,35,37,38,40,43,47,52,55,56,57,58,59,60,67,69,70,72,74,75,76,81,82,86,87,91,97,98,101],taken:[25,36,40,47,70],talk:[15,32,105],tangl:97,tanh:[34,49,74,101],tar:86,tarbal:86,target:[1,7,8,9,25,27,34,36,50,56,59,75,101,103,104,105],target_block:[7,25],target_dict_dim:101,target_dict_s:49,target_language_word:101,target_link_librari:9,target_word:49,task13:66,task14:66,task:[22,44,49,54,65,82,95,101],task_queu:14,taskentri:14,taskqueu:14,tbd:[16,43,66],tcp:[86,93],tear:98,technic:[7,10,82],techniqu:[40,74,97,101],technolog:[0,31],tee:87,tell:[2,10,14,15,49,65,98,103],templat:[33,43,65,67,75,76,77,87,106],tempor:66,temporari:[7,18,27,40,52,57],tempori:40,ten:0,tensor:[6,9,19,21,23,24,30,31,32,34,36,38,39,43,44,47,48,49,64,70,71,75,99],tensor_arrai:19,tensor_array_read:70,tensor_array_s:70,tensor_array_stack:70,tensor_array_unstack:70,tensor_array_writ:70,tensor_data:44,tensor_in:69,tensor_s:6,tensor_test:9,tensor_to_check:6,tensorarrai:23,tensorarraydesc:70,tensordesc:[44,64],tensorflow:[8,19,21,22,24,31,34,37,61,70,77],term:[10,60,61,66],termin:87,terminolog:40,tessorarrai:70,test1:12,test:[0,5,6,9,36,46,52,58,62,77,80,81,82,88,92,98,99],test_:75,test_all_data_in_one_period:87,test_check_grad_ingore_i:75,test_check_grad_ingore_x:75,test_check_grad_norm:75,test_check_output:75,test_data_dir:81,test_fcgrad:74,test_gpuprofil:98,test_layergrad:74,test_mkldnn:42,test_mklpack:41,test_mul_op:[1,75],test_norm:75,test_pass:[92,93,95],test_period:[92,93,95],test_recurrent_op:72,test_wait:[92,93],testa:5,testb:5,testbilinearfwdbwd:98,testcas:75,testconfig:74,testfcgrad:74,testfclay:74,testlayergrad:74,testmodel_list:92,testmulop:75,testq:5,testsave_dir:92,testutil:74,text1:29,text:[5,44,48,54,66,86],tflop:98,tftp:106,tgz:4,than:[0,1,7,10,18,19,20,25,31,32,33,34,57,59,61,68,70,74,81,82,86,101,103,106],the_step:31,theano:31,thei:[0,5,7,9,10,15,17,19,20,21,24,25,29,31,34,35,39,40,49,50,54,57,59,65,70,71,72,74,75,77,82,83,86,92,98,101],them:[0,5,6,7,9,10,13,18,20,21,24,25,31,32,33,38,39,40,49,57,58,59,60,63,64,65,69,70,71,72,75,78,82,86,92,93,98],themselv:[7,9],theori:[31,98],therefor:[7,40,52],therein:8,theta:34,theta_d:34,theta_g:34,thi:[0,2,4,5,6,7,8,9,10,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,30,31,32,33,34,35,36,39,40,43,47,48,49,50,51,52,53,54,55,57,58,59,60,61,64,65,66,67,69,70,72,74,75,76,77,78,80,81,82,83,86,87,91,93,95,97,98,99,101,103,104,105,106],thin:38,thing:[22,34,59,67,98],think:[5,9,82],third:[10,36,75,97,98,103,104,105],third_parti:[42,103,104,105],thirt:0,those:[8,9,10,33,35,36,37,55,103],though:[70,106],thought:[9,98],thread:[19,21,23,54,74,81,93,95,97,98],thread_count:23,thread_id:54,thread_id_:54,thread_local_rand_use_global_se:[92,93],thread_pool:23,threadid:95,threadloc:98,threadpool:19,three:[6,7,10,21,26,30,31,32,35,43,49,50,53,54,55,58,66,67,93,103],threshold:[10,14,25,72,93],through:[7,9,10,14,16,26,40,50,52,74,75,78,81,98,99,101,104],throughout:27,throughput:[81,98],thrust:59,thu:[17,26,36,40,66,74,86],tier:87,time:[0,1,5,6,9,10,14,17,20,21,22,24,25,31,33,38,40,41,47,48,49,51,54,57,58,59,60,64,65,66,70,71,77,87,93,95,97,98,101,106],timelin:[54,59,98],timeo:86,timeout:[10,14],timestamp:11,timestep:63,tip:[103,104],titan:40,tls:28,tmp:57,todo:[8,10,14,17,49,65,66],togeth:[7,70,82,101],token:[5,66,101],toler:[1,6,75],too:[6,19,21,25,43,69,70],took:106,tool:[0,1,4,54,82,85,86,97,101,103,105],toolchain:[97,103,104],toolkit:66,top:[48,49,66,75],top_k:49,top_level_rnn:48,topic:43,topk_generated_scor:49,topk_id:49,topk_scor:49,toplevel:0,topolog:[5,10,22,36,40,44,53],topoloi:36,torch:[8,31],total:[10,24,26,51,54,58,81,85,87,97,98,106],total_pass:58,tottim:97,toward:31,trace:[8,32,34],track:[10,14,36,57],tradit:[8,30,66],traffic:22,train:[7,8,12,14,15,17,19,20,25,26,27,31,32,34,35,40,41,44,51,52,53,54,55,56,57,59,61,64,66,67,71,74,85,89,90,92,96,98,101,104],train_config_dir:86,train_data:81,train_data_dir:81,train_id:86,train_list:81,train_loop:31,train_read:[22,82],trainabl:[44,57],traindot_period:92,trainer:[5,11,12,13,14,16,22,24,32,41,42,52,53,59,74,82,83,84,93,95],trainer_config:[86,87],trainer_config_help:74,trainer_count:[80,81,86,87,92,93,95],trainer_cpu:18,trainer_cr:18,trainer_gpu:18,trainer_id:[81,86,93],trainer_intern:13,trainer_mem:18,trainer_packag:18,trainer_prog:22,trainerid:17,training_rol:82,trainingjob:22,trainingtest_period:92,trainonebatch:13,tran:[43,74,93],trans_var:69,transact:[10,14],transcript:66,transfer:[40,54],transform:[59,66,74,77,101],translat:40,translation_id:49,translation_scor:49,transpar:[49,84],transpil:[19,82],transport:93,transpos:74,travers:[7,35,40],travi:72,treat:[8,15,40],treatment:[15,30],tree:[8,19,27,31,57,93,99,105],trg_dic_siz:49,trg_embed:[49,101],trick:49,tricki:45,trigger:[17,20,53],trivial:[49,70],true_block:[8,37,55],true_imag:58,true_label:58,true_neg:26,true_posit:26,true_read:58,tune:[66,96,97],tuninglog_barrier_abstract:92,tupl:[7,57,58],turn:[2,57,58],tutori:[1,2,74,75,86,88,89,90,97,98,101,104],twice:[24,34,82],twine:62,two:[1,5,7,15,16,17,18,19,20,21,22,26,29,30,31,32,34,35,38,40,44,47,49,52,54,55,58,59,60,61,63,64,65,66,69,70,71,75,77,84,86,95,98,101,103,105],txt:[9,18,29,41,42,74,78,81,86,88],type:[0,5,7,8,10,13,14,17,18,20,22,28,29,30,32,38,39,43,44,45,46,48,49,55,56,57,58,59,60,61,64,65,66,67,69,71,74,76,77,80,81,83,86,87,91,93,95,101,104],type_nam:65,typedef:[15,30,45,46,47,67,76],typeerror:25,typeid:65,typenam:[33,65,67,75,76,77],typic:[22,98,104],ubuntu:[4,62,80,97],ubyt:58,uci_h:[80,82],uid:87,uint16_t:30,uint32:[28,44],uint32_t:54,uint64:[44,45],uint64_t:45,unawar:15,unbalanc:93,unbias:6,unblock:21,unbound:[40,101],unbuff:21,unclear:17,uncreat:7,under:[1,2,9,14,24,51,69,77,78,81,84,85,86],underli:[20,49],understand:[31,57,66,97,98,106],understand_senti:101,undeterminist:98,uni:66,unidirect:66,unifi:[20,27,36,64,72],uniform:[12,34,57,58,93],uniform_random:57,uniniti:7,uninstal:1,uniqu:[5,8,10,17,18,43,47,57,63,75,82,86,91,93],unique_nam:57,unique_name_gener:57,unique_ptr:[60,63,67,74],unit:[0,1,9,52,54,61,67,77,101],unittest:[46,72,75],unittestcheckgrad_ep:92,unix:21,unk:[64,71],unlik:[49,75],unnecessari:[7,66,72],unordered_map:63,unpack:70,unrol:48,unseen:61,unsign:[15,30],unstack:70,unstack_from:70,unsupervis:34,unsupport:75,until:[10,15,21,23,24,31,40,63,82,86],unzip:103,updat:[4,7,10,14,15,22,28,30,34,48,49,50,51,52,53,63,66,70,74,81,83,93,95,97],update_memori:8,update_op:50,updatecallback:74,updatestack:86,upgrad:[1,4,51],upload:[10,18,21,28,62,83],upon:10,upstream:72,uri:86,url:72,usag:[30,37,40,53,57,68,75,98,104],use:[0,1,2,4,5,6,8,9,10,16,20,21,22,23,24,27,30,34,36,38,39,40,43,47,49,50,51,53,54,57,63,64,65,66,69,70,71,72,74,75,76,78,80,81,83,86,87,93,95,97,98,101,103,104,105],use_cpu:39,use_cudnn:39,use_eigen_bla:103,use_eigen_for_bla:[103,104],use_gpu:[80,81,87,92,93,95],use_mkl_pack:41,use_mkldnn:[39,42],use_old_updat:[13,92,93],use_sparse_remote_updat:13,used:[0,1,4,5,6,8,9,10,16,17,20,21,22,25,27,30,31,34,36,40,48,49,52,53,54,57,58,59,61,63,65,67,69,70,74,75,77,86,92,93,95,97,98,101,103,104,105],useful:[6,30,40,57,63,69,74,95,101,103],usegpu:74,user:[2,5,6,7,8,9,12,14,17,18,19,22,23,24,25,26,27,29,33,34,35,36,38,39,43,47,49,50,51,52,54,57,58,59,60,61,63,65,67,69,70,72,76,78,81,86,92,93,97,103,106],user_nam:12,usercert:12,userkei:12,usernam:12,uses:[0,1,4,6,10,17,19,21,22,30,40,47,48,49,53,54,67,69,72,74,77,78,83,86,93,101,103],using:[1,2,3,4,5,6,7,8,9,10,14,15,17,18,21,22,27,29,30,31,33,34,36,38,40,48,50,52,55,57,58,60,61,63,65,66,67,72,74,75,76,77,78,80,81,86,87,88,91,93,95,98,101,103,105],usr:[1,2,81,86,93],usual:[7,18,40,47,54,55,61,67,72,75,86,93,95,97,98],util:[22,41,42,51,74,75,76,98,101,106],uuid:[11,17],v7a:103,v8a:103,val:7,valid:[58,59,63,75,86,104],valu:[6,7,8,10,19,21,25,26,36,37,40,42,44,48,49,50,52,53,55,59,63,64,65,68,70,71,74,75,76,82,86,93,95,101,103,104],value1:93,value2:93,value_:64,valueerror:36,values_:70,vanilla:101,var_nam:[7,69],var_recurs:25,vardesc:[8,35,55,57,59,64],vardescbuild:8,vari:[86,98],variabl:[5,6,8,19,20,21,22,24,25,26,27,32,34,35,36,37,38,47,48,49,50,52,55,56,60,61,64,65,66,69,70,74,75,76,81,85,86,87,97,99,103,104],variablenamemap:75,varialbl:34,variant:[38,47,67,70,76],varibal:7,varibl:36,varienc:70,varient:70,variou:[8,21,30,40,61,103],varproto:65,vars_:[8,63],vartyp:[64,71],vartypeinfer:38,vec2seq:66,veclib:104,vector:[5,8,13,15,20,36,37,43,48,49,54,57,59,60,64,66,68,70,74,77,101],vectorenable_parallel_vector:92,vendor:9,verbos:[29,72],veri:[9,14,19,24,27,31,33,34,40,43,49,53,58,61,63,66,67,70,84,97,98,101],verifi:[8,74,104],version:[0,1,2,7,9,18,22,25,29,32,34,36,37,44,49,62,66,74,80,82,86,87,91,92,93,97,98,103,104,105],versu:5,via:[3,7,10,47,72,82,86,98,104,106],view:[44,47],vim:2,viriabl:81,virtual:[0,20,25,38,39,60,67,76],virtualenv:0,visibl:[17,63],visit:7,visual:[49,98],vlog:[13,72],vocabulari:66,voila:80,volum:[78,87],volumemount:[86,87],volumn:86,wai:[0,5,7,15,17,21,27,31,39,40,49,52,57,58,61,70,72,74,95,101],wait:[10,15,21,23,76,81,82,83,93],walk:104,wangkuiyi:9,want:[0,1,2,5,18,19,21,26,34,39,47,52,54,56,58,61,63,67,68,69,70,72,74,78,81,93,95,97,103,105],warn:29,warp:98,warp_ctc:66,wast:51,watch:10,wbia:86,web:[78,97],websit:78,weight:[41,44,61,74,93,95,101],weightlist:74,weights_:74,weights_primitive_desc:43,weights_t:74,welcom:[9,66,72],well:[7,18,21,22,24,31,33,34,61,64,66,74,82,86,93],wer:66,were:[9,21,31],west:86,wget:103,wgt:43,what:[9,20,31,34,49,57,65,69,75,97,106],whatev:[0,81],wheel:4,when:[0,1,3,7,8,9,10,13,14,15,18,19,22,24,25,26,27,29,30,31,32,36,49,51,52,53,54,55,57,59,67,70,72,74,75,77,78,81,83,85,86,87,93,95,97,98,101,103,104,106],whenev:[57,66,72],where:[5,7,8,10,17,19,22,31,32,35,47,48,49,52,55,59,61,67,70,74,75,93,95,97,98,99,101],wherea:[8,14,33,37,67,68,71],whether:[1,2,6,7,8,20,27,54,58,64,70,74,75,93,104],which:[0,1,4,5,6,7,8,9,10,12,14,15,17,18,19,20,21,22,23,25,27,30,31,32,33,34,36,38,40,43,44,47,48,49,50,51,53,55,56,57,58,59,60,63,64,65,68,69,70,71,72,74,75,76,77,81,82,83,84,86,93,95,97,98,101,103,104,106],while_grad:40,while_loop:[49,70],while_op:7,whileloop:70,whileop:8,white:66,whl:1,who:[7,33,35,51,57,72],whoever:15,whole:[7,20,34,37,40,45,46,48,51,56,65,66,72,83,86,87,106],wholli:20,whose:[6,7,10,17,25,48,59,60,65,70,101],why:[0,6,46],wide:[4,9,25,34,84,88],width:[13,45,58,74,75],wiki:9,window:[0,2,52,66,103],wirt:36,wise:[24,59,66,77],wish:[1,4,78,81,82],with_avx:[1,2,91,103,104],with_bia:65,with_c_api:[1,103,104,105],with_distribut:82,with_doc:1,with_doubl:[1,74,91],with_dso:1,with_golang:[1,103],with_gpu:[0,1,91,103,104],with_mkl:[1,41,42,103],with_mkldnn:42,with_mklml:42,with_profil:98,with_python:[1,91,103,104],with_rdma:[91,103,104],with_style_check:[1,72],with_swig_pi:[1,103,104],with_test:[1,75],with_tim:[91,98],within:[14,22,31,66,104],without:[7,10,15,20,21,54,57,58,59,66,75,81,88,97],wloop:70,wmt14:101,won:[83,98],word2vec:[18,81],word:[7,24,35,38,40,48,49,59,65,66,70,76,81,95,101],word_dict:[81,88],word_vector_dim:[49,101],wordcount:66,work:[0,2,5,8,9,10,22,27,30,31,39,50,52,54,57,72,74,78,82,83,86,87,93,97,98,101,106],worker:[24,71,86],workercount:86,workflow:[59,86],workspac:[72,81,84,93],world:81,worri:6,worth:99,would:[0,2,8,9,10,17,21,22,23,24,31,33,34,35,43,50,52,53,57,58,64,66,70,72,82,86,97,103,106],wouldn:[31,35],wrap:[20,31,33,34,51,106],wrapper:[9,20,21,33,51,52,60,70,98],write:[0,5,10,17,19,20,21,22,24,30,31,32,33,36,38,43,50,52,57,58,59,60,67,70,73,81,85,86],write_lock:11,write_to_arrai:40,writer:[5,57],written:[1,2,7,8,19,24,27,34,44,52,59,60,64,75,77,84,97],wrong:58,wrote:36,wsize:86,x64:[103,105],x86:104,x86_64:[103,104],x_neg:6,x_po:6,xarg:[2,74,88],xcode:104,xcodebuild:104,xeon:76,xgbe0:93,xgbe1:93,xpu:31,xrang:[6,31,34,54,58,74,80],xx_layer:39,xxx:[5,70],xxxx:11,xxxxxxxxx:86,xxxxxxxxxx:86,xxxxxxxxxxxxx:86,xxxxxxxxxxxxxxxxxxx:86,y_dim:34,y_neg:6,y_po:6,y_predict:[80,82,99],yaml:[9,84,86,88,106],yancey1989:18,yapf:72,year:31,yep:[54,97],yet:[31,66,106],yield:[5,12,20,58],you:[0,1,2,3,4,6,18,22,30,63,72,74,75,76,78,80,81,82,84,86,88,91,93,95,97,98,101,103,104,105,106],your:[0,1,3,4,5,9,13,18,29,59,72,74,78,81,82,84,86,91,95,98,103,104,105,106],your_access_key_id:86,your_secrete_access_kei:86,your_source_root:46,yourself:1,yuang:31,yuyang:97,z_dim:34,z_size:34,zero:[6,7,10,21,34,49,53,57,64,74,86,93],zip:[57,103],zone:86,zxvf:86},titles:["Build using Docker","Build from Sources","Run in Docker Containers","Install and Build","Install Using pip","PaddlePaddle Design Doc","Auto Gradient Check Design","Backward Building","Design Doc: Block and Scope","Required CMake Function","Design Doc: Distributed Training","\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\uff08Checkpointing\uff09","\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1","Alalysis of large model distributed training in Paddle","Design Doc: Master Server","Design Doc: The Client Library of Parameter Server","Design Doc: Remote Parameter Updater for Cluster Train","Design Doc: Save Model","Submit a Distributed Training Job","Design Doc: Concurrent Programming with Fluid","C++ Data Feeding","Design Doc: CSP in PaddlePaddle Fluid","Design Doc: Distributed Training Architecture","Design Doc: Execute the Program with Multi CPU","Design Doc: Parameter Server","Error Clip","Evaluator Design","Executor Design Doc","FileManager\u8bbe\u8ba1\u6587\u6863","PFSClient","Design Doc: float16","Design Doc: PaddlePaddle Fluid","PaddlePaddle Fluid: Towards a Compiled Programming Language","Design Doc: Functions, Operators, and Layers","Design for GAN","Design Doc: Computations as a Graph","Survey on Graph","The IfElse Operator","Design Doc: InferVarType","Problem","Memory Optimization","Intel\u00ae MKL Packed on PaddlePaddle: Design Doc","Intel\u00ae MKL-DNN on PaddlePaddle: Design Doc","Design Doc: Add MKLDNN Kernel in Fluid Operator","Design Doc: Model Format","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Design Doc: The Keys of Operator Kernel Type","RNNOp design","Design: Sequence Decoder Generating LoDTensors","Optimizer Design","Design Doc: NCCL support in Paddle Fluid","Averaging Parameter in PaddlePaddle","Design Doc: The C++ Class Parameters","Introduction","Design Doc: PaddlePaddle Programs","Prune","Design Doc: Python API","Python Data Reader Design Doc","Design Doc: Refactorization Overview","Design Doc: Gradient Operators Registration","Regularization in PaddlePaddle","PaddlePaddle\u53d1\u884c\u89c4\u8303","Design of Scope in Paddle","Design Doc: Selected Rows","Interaction between C++ and Python","DeepSpeech2 on PaddlePaddle: Design Doc","Design Doc: Supporting new Device/Library","Design Doc: Switch","Background","Design for TensorArray","Background","Contribute Code","Development","Write New Layers","How to write a new operator","Add Kernels for a New Device","How to use Eigen in Paddle","Contribute Documentation","GET STARTED","Quick Start","Command-line arguments","Fluid Distributed Training","Distributed Training","Cluster Training Using Fabric","Use different clusters","Distributed PaddlePaddle Training on AWS with Kubernetes","PaddlePaddle On Kubernetes","Cluster Training Using OpenMPI","<no title>","<no title>","Preparations","Argument Outline","Detail Description","Set Command-line Parameters","Use Case","HOW TO","Profiling the Python Code","Tune GPU Performance","PaddlePaddle Fluid Source Code Overview","RNN Models","RNN Configuration","PaddlePaddle Documentation","Build PaddlePaddle for Android","Build PaddlePaddle for iOS","Build PaddlePaddle for Raspberry Pi","Cluster bootstrapping tool survey"],titleterms:{"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":12,"\u4e0d\u4f7f\u7528":45,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":45,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":45,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":45,"\u4ec5\u4ec5\u4f7f\u7528void":45,"\u4ece\u5feb\u7167\u6062\u590d":11,"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":45,"\u4f7f\u7528\u8f6c\u6362\u5e93":12,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":46,"\u5206\u5757\u6587\u4ef6\u4f20\u8f93":28,"\u5206\u652f\u89c4\u8303":62,"\u52a0\u901f\u6267\u884c":11,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":45,"\u52a8\u6001\u6269\u5bb9":11,"\u539f\u56e0":45,"\u539f\u56e0\u5217\u8868":45,"\u53c2\u8003\u6587\u6863":28,"\u53d1\u5e03docker\u955c\u50cf":62,"\u53d1\u5e03wheel\u5305\u5230pypi":62,"\u540d\u8bcd\u89e3\u91ca":28,"\u57fa\u672c\u8981\u6c42":45,"\u5b9e\u73b0":45,"\u5b9e\u73b0\u65b9\u5f0f":46,"\u5bfc\u51fac":45,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":11,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":45,"\u63a8\u6d4b\u6267\u884c":11,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":12,"\u6587\u4ef6\u4f20\u8f93\u4f18\u5316":28,"\u6587\u4ef6\u8bbf\u95ee\u65b9\u5f0f":12,"\u6587\u4ef6\u8bbf\u95ee\u7684\u6743\u9650":12,"\u6587\u4ef6\u9884\u5904\u7406":12,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":46,"\u672f\u8bed":11,"\u67b6\u6784\u56fe":28,"\u6846\u67b6\u751f\u6210":28,"\u6982\u5ff5\u89e3\u91ca":12,"\u6a21\u5757":28,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":11,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":46,"\u6d41\u7a0b\u4ecb\u7ecd":12,"\u751f\u6210sparse\u6587\u4ef6":28,"\u7528\u6237\u4f7f\u7528\u6d41\u7a0b":28,"\u76ee\u5f55\u7ed3\u6784":46,"\u76ee\u6807":28,"\u793a\u4f8b\u7a0b\u5e8f":12,"\u7b26\u53f7":45,"\u7c7b":45,"\u7f16\u8bd1\u9009\u9879":46,"\u7f29\u5bb9":11,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":45,"\u80cc\u666f":45,"\u8986\u76d6\u4e0d\u4e00\u81f4\u7684\u90e8\u5206":28,"\u8bad\u7ec3\u6570\u636e\u5b58\u50a8":12,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":12,"\u8f6c\u6362\u5e93":12,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":45,"\u8fdb\u884c\u8bad\u7ec3":12,"abstract":[22,23,24,51,106],"book\u4e2d\u6240\u6709\u7ae0\u8282":62,"case":[7,95],"class":[34,53,57,74],"filemanager\u8bbe\u8ba1\u6587\u6863":28,"final":39,"function":[9,33,34,57],"new":[67,74,75,76],"paddle\u52a8\u6001\u5e93\u4e2d":45,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":45,"paddlepaddle\u53d1\u884c\u89c4\u8303":62,"paddlepaddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":62,"return":[57,58],"switch":[43,67,68],"tensor\u5230eigentensor\u7684\u8f6c\u6362":77,AWS:86,DNS:86,E2E:82,EFS:86,For:[9,87],KMS:86,Not:0,The:[8,15,19,27,31,34,35,37,38,47,50,53,59,60,68,105],Use:[8,55,78,80,85,87,95],Using:[4,9,15,84,88,103],With:18,about:34,absolut:49,access:86,account:86,action:[41,42],activ:42,actor:21,add:[40,43,76,86],address:86,advanc:67,alalysi:13,algorithm:[6,10,22,48,56],all:[63,70],analog:19,analysi:[22,40],android:103,api:[22,41,42,46,50,52,57,61,65],appendix:106,approach:98,arbitrari:31,architectur:[22,54,101],argument:[29,58,81,92,95,103,105],arrai:6,asset:86,associ:[63,86],assumpt:106,async:93,attent:101,attribut:[40,61],auto:6,averag:52,aws:86,background:[6,24,41,67,68,69,70,71,75,76],backward:[7,31,35,59,75],base:[18,49],basic:[40,67,106],batch:58,batch_siz:58,beam:[49,66],benchmark:[41,42],benefit:[24,59],between:[5,21,57,59,65,67],binari:8,bind:75,bla:1,block:[8,32,34,35,55,57,59],blockdesc:55,book:2,bool:1,bootstrap:106,bottleneck:97,bring:106,bucket:86,build:[0,1,3,7,34,59,78,87,103,104,105],built:98,can:63,capi:46,capi_priv:46,challeng:[7,24,56],chang:49,channel:21,check:[6,74,84],checkpoint:[10,11,17],choic:39,choos:[9,86],client:15,clip:25,close:6,cloudform:86,cluster:[16,82,84,85,86,88,95,106],cmake:[9,41,42,105],code:[18,32,57,72,97,99],command:[81,94,95],commit:87,common:93,commun:93,compar:106,comparis:57,compat:31,compil:[1,8,30,32,55,59,75,99,103,104,105],complet:31,compos:58,comput:[8,35,43,59,61,77],con:106,concept:[57,59,86],concern:42,conclus:[17,36,106],concurr:[19,21],condit:34,config:95,configur:[86,101,104],construct:35,contain:[2,87],content:[41,42,46,66,86,98],context:76,contribut:[72,78],control:[40,59],contruct:40,convert:17,convolut:66,core:[6,57,86],corner:7,cpu:[23,95],creat:[7,21,58,59,63,86,87],createreaderop:20,creation:[14,52,61],creator:58,credenti:86,cross:[103,104,105],csp:21,ctc:66,cuda:[1,30],cudnn:1,current:[30,60],custom:58,data:[10,20,22,58,69,86,87],dataflow:40,dataprovid:93,dataset:[10,14,81],datatyp:47,decod:49,decor:58,decoratedread:20,deep:[8,31],deepspeech2:66,defin:[75,86],definit:71,delet:86,demo:[34,82,86],dens:17,dep:4,depend:[1,4,34,66],deploi:18,deriv:74,describ:[31,50],descript:[29,59,93],design:[5,6,8,10,14,15,16,17,19,21,22,23,24,26,27,30,31,33,34,35,38,41,42,43,44,47,48,49,50,51,53,55,57,58,59,60,63,64,66,67,68,70],destroi:[63,86],detail:[13,66,93],develop:[0,59,73],devic:[67,76,95],devicecontext:67,dictionari:58,differ:[59,67,85,95],directori:86,discrimin:34,discuss:[24,34],dispatch:[10,14],distribut:[5,10,13,18,22,24,82,83,86,93],dnn:42,doc:[5,8,10,14,15,16,17,19,21,22,23,24,27,30,31,33,35,38,41,42,43,44,47,51,53,55,57,58,59,60,64,66,67,68],docker:[0,2,18,87,103],document:[78,102],doe:58,down:86,download:[86,87],dure:[49,58],dylib:46,dynam:[10,70],dynet:36,each:4,ec2:86,eigen:77,elast:86,elect:17,els:8,engin:34,enough:6,entri:58,environ:[18,103],equat:74,error:25,evalu:26,event:[5,54],evolut:31,examin:97,exampl:[5,9,19,21,37,46],execut:[8,23,31,55,59],executor:27,explan:6,extern:86,fabric:84,faq:[3,4],fault:10,feed:20,file:[8,86,87,97],fileread:20,find:86,float16:30,flow:40,fluid:[19,21,31,32,43,51,82,99],format:[8,10,44],forward:[35,75],frame:8,framework:[6,76,77],from:[1,3,5,17,65],functor:67,futur:[31,66],gan:34,gate:101,gener:[0,32,34,49,97,101,106],get:[79,82,87],give:58,global:[55,57],gotcha:0,gpu:[2,93,95,98],grad_op:7,gradient:[6,7,15,42,60,74],graph:[35,36,40,59,61],group:86,gru:93,hand:98,handler:[5,45],happen:17,hardwar:30,have:82,helper:57,hierarchi:8,high:[50,52,61,65],how:[1,6,13,52,58,59,67,75,77,78,96,98],iOS:104,iam:86,identifi:97,ifels:37,ifelseop:8,imag:[2,18,87,103],implement:[6,7,9,13,23,25,26,30,44,48,51,52,57,58,59,60,61,74,75,76,77],infer:103,infershap:[55,64],infervartyp:38,ingredi:5,ingress:28,initi:[15,34,86,95],insid:63,inspect:86,instal:[3,4,80,82,86,103,104,105,106],instanc:86,instead:58,integr:[67,86],intel:[41,42],interact:65,interfac:[6,10,15,16,27,50,58,63],intermedi:59,introduc:[49,70,82],introduct:[54,61,82],isn:58,issu:30,job:[10,18,84,86,87,88],kei:[41,47,86],kernel:[43,47,59,76],kill:84,kube:86,kubectl:86,kubernet:[18,86,87],languag:[8,32],larg:13,launch:[2,84,88],layer:[5,33,41,42,57,74,95],layout:47,learn:[8,31],leval:65,level:[50,52,61,65],libpaddle_capi_shar:46,libpaddle_capi_whol:46,librari:[15,30,47,59,67,103],limit:22,line:[81,94],linux:[84,103],list:[11,58],live:40,load:21,local:[22,63,86,95],lod:49,lodtensor:[48,49,70],lodtensordesc:71,log:72,logic:14,look:97,low:[52,61,65],lstm:93,machin:49,macro:59,main:34,make:40,manag:9,map:[58,59],master:[10,14,18,19],math:67,mathemat:6,matrix:[42,93],member:34,memori:[40,48,67],messag:65,method:49,might:34,migrat:59,mileston:59,mini:58,minibatch:21,mix:95,mkl:[41,42],mkldnn:43,mkldnn_helper:43,mkldnndevicecontext:43,model:[5,13,15,17,21,31,34,44,49,84,95,100,101],modifi:87,modul:[59,67,77],more:34,motiv:[7,21,27,44,51,56],multi:[23,32],multipl:58,mxnet:36,name:[63,86],nativ:32,nccl:51,necess:57,necessari:59,need:[0,58,98],nest:48,network:[59,95,101],neural:101,nlp:93,non:82,norm:61,note:6,numer:6,numpi:6,nvprof:98,nvvp:98,object:10,offset:49,onli:[58,63],op_mak:59,openmpi:88,oper:[20,33,37,40,43,47,52,55,57,59,60,64,70,75],opinfomap:59,opkernel:[59,67,75,76],opproto:65,ops:61,optim:[10,15,35,40,50,57],option:[1,29],opwithkernel:59,order:29,org:78,origin:59,orthogon:63,other:42,outlin:92,output:[84,86],overview:[17,25,27,41,42,59,63,66,99],pack:[41,49],packag:[4,9],paddl:[13,51,58,63,77],paddlejob:18,paddlepaddl:[1,2,4,5,8,21,31,32,41,42,52,55,61,62,66,78,82,86,87,99,102,103,104,105],pair:86,paradigm:31,parallel_nn:95,paramet:[5,10,15,16,18,21,24,42,52,53,57,61,81,82,86,93,94],parameteraverageoptim:52,parent:63,part:35,partit:15,pass:[1,95],path:[17,29],penalti:61,perform:[52,93,97,98],persist:14,pfsclient:[28,29],pfsserver:28,pip:4,place:[40,47,67,76],placement:22,point:[41,86],polici:40,pose:[38,60],potenti:39,prefetch:58,prepar:[81,82,84,86,88,91,104],principl:43,privat:86,pro:106,problem:[26,38,39,40,47,50,60,69],procedur:106,process:[0,10,15,18,50,59],profil:[97,98],program:[2,8,19,21,23,31,32,55,57,81,82],programdesc:[32,55],project:9,propos:[38,60,61],protobuf:64,protomak:75,provid:58,prune:56,pserver:17,pull:2,python:[6,18,22,41,42,48,50,52,57,58,61,65,71,74,75,97],qualiti:59,queue:[10,14],quick:80,randomnumb:93,raspberri:105,reader:[5,20,58],readerbas:20,readerhold:20,readi:82,readop:20,realiz:59,recoveri:10,recurr:101,recv:21,refactor:59,refer:[6,22,24,40,41,42,66,98],region:86,regist:[38,59,65,75,76],registr:[59,60],registri:59,regular:[15,61],rel:49,relat:[20,59,70],remark:75,remot:16,remoteexecutor:22,render:86,represent:[8,59],requir:[9,34],result:[84,87],retri:14,reus:57,review:72,rnn:[48,70,93,100,101],rnnop:[8,48,59],route53:86,row:[64,66],rpc:21,run:[1,2,27,75,87,99],runtim:[4,18],save:17,scale:10,scope:[8,48,59,63],script:[82,87],search:[49,66],secur:86,select:[15,21,64],selectedrow:64,semant:68,send:21,separ:59,sequenc:[49,101],server:[10,14,15,18,21,24,81,82,86,93],servic:86,set:94,setup:[86,103],sextant:106,sgd:93,shape:49,share:[5,7,40,63],should:63,shuffl:58,simpl:[49,101],singl:58,slice:82,solut:[38,39,40,41,47,56,60,69],some:0,sourc:[1,3,99],spars:[15,16,17,64,95],specifi:95,stack:8,standard:72,start:[5,79,80,81,86,87],startup:87,statement:26,step:[3,48],storag:61,store:10,strategi:40,style:72,subcommond:29,submit:18,suffici:58,suitabl:9,sulut:43,summar:[5,19],summari:44,support:[30,51,67,70],survei:[30,36,61,106],synopsi:29,syntax:21,system:[31,86],tabl:[46,66],task:[10,14,66],tear:86,tecton:106,templat:86,tensor:[59,67,77],tensorarrai:[49,70],tensordesc:71,tensorflow:36,test:[1,41,42,43,72,74,75,93,95],theori:6,thi:63,think:34,three:70,time:99,timelin:17,timer:98,tip:98,todo:[11,12,23],togeth:63,toler:10,tool:[9,78,98,106],toolchain:105,topic:67,toward:32,train:[2,5,10,13,16,18,22,50,58,81,82,83,84,86,87,88,93,95],trainer:[10,15,17,18,21,81,86],transform:69,translat:49,transpil:[22,23,24,32,40,51],tune:[93,98],ture:31,two:6,type:[1,21,47,75],uniform:70,unit:[41,42,43,72,74,75,93],unpack:49,updat:[5,16,17,78,82,86],usag:[7,25,48,49,58,77],use:[13,58,77],user:10,using:0,valu:57,vardesc:71,variabl:[7,40,57,59,63,71],vector:93,verifi:86,version:[4,19,30],volum:86,vpc:86,what:[0,13,17,98],when:[17,63],whl:4,why:[30,31,52,58,59,70,98],work:66,worker:19,workflow:72,wrapper:74,write:[72,74,75,76,78],www:78,yaml:87,your:[2,76]}})
\ No newline at end of file
diff --git a/develop/doc_cn/_sources/api/index_cn.rst.txt b/develop/doc_cn/_sources/api/index_cn.rst.txt
deleted file mode 100644
index 84f9097a6cdc2da269bd6a0685796e14e26da37e..0000000000000000000000000000000000000000
--- a/develop/doc_cn/_sources/api/index_cn.rst.txt
+++ /dev/null
@@ -1,10 +0,0 @@
-API
-===
-
-.. toctree::
- :maxdepth: 1
-
- 模型配置
-
- paddle.v2.activation.AbsAbs Activation.
-Forward: \(f(z) = abs(z)\)
-Derivative:
-paddle.v2.activation.ExpExponential Activation.
-paddle.v2.activation.IdentityLinear 的别名
paddle.v2.activation.LinearIdentity Activation.
-Just do nothing for output both forward/backward.
-paddle.v2.activation.LogLogarithm Activation.
-paddle.v2.activation.SquareSquare Activation.
-paddle.v2.activation.SigmoidSigmoid activation.
-paddle.v2.activation.SoftmaxSoftmax activation for simple input
-paddle.v2.activation.SequenceSoftmaxSoftmax activation for one sequence. The dimension of input feature must be -1 and a sequence.
-result = softmax(for each_feature_vector[0] in input_feature)
-for i, each_time_step_output in enumerate(output):
- each_time_step_output = result[i]
-paddle.v2.activation.ReluRelu activation.
-forward. \(y = max(0, z)\)
-derivative:
-paddle.v2.activation.BReluBRelu Activation.
-forward. \(y = min(24, max(0, z))\)
-derivative:
-paddle.v2.activation.SoftReluSoftRelu Activation.
-paddle.v2.activation.TanhTanh activation.
-paddle.v2.activation.STanhScaled Tanh Activation.
-paddle.v2.activation.SoftSignSoftSign Activation.
-
- paddle.v2.attr.ParamParameterAttribute 的别名
paddle.v2.attr.ExtraExtraLayerAttribute 的别名
paddle.v2.attr.HookHookAttribute 的别名
paddle.v2.attr.HookAttrHookAttribute 的别名
paddle.v2.attr.ParamAttrParameterAttribute 的别名
paddle.v2.attr.ExtraAttrExtraLayerAttribute 的别名
paddle.v2.attr.ParameterAttribute(name=None, is_static=False, initial_std=None, initial_mean=None, initial_max=None, initial_min=None, l1_rate=None, l2_rate=None, learning_rate=None, momentum=None, gradient_clipping_threshold=None, sparse_update=False, update_hooks=None, initializer=None)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.
-NOTE: IT IS A HIGH LEVEL USER INTERFACE.
-| 参数: |
|
-
|---|
set_default_parameter_name(name)Set default parameter name. If parameter not set, then will use default -parameter name.
-| 参数: | name (basestring) – default parameter name. | -
|---|
paddle.v2.attr.ExtraLayerAttribute(error_clipping_threshold=None, drop_rate=None, device=None)Some high level layer attributes config. You can set all attributes here, -but some layer doesn’t support all attributes. If you set an attribute to a -layer that not support this attribute, paddle will print an error and core.
-| 参数: |
|
-
|---|
- paddle.v2.evaluator.classification_error(*args, **xargs)Classification Error Evaluator. It will print error rate for classification.
-The classification error is:
-The simple usage is:
-eval = classification_evaluator.error(input=prob,label=lbl)
-| 参数: |
|
-
|---|---|
| 返回: | None. - |
-
paddle.v2.evaluator.auc(*args, **xargs)Auc Evaluator which adapts to binary classification.
-The simple usage:
-eval = evaluator.auc(input, label)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.ctc_error(*args, **xargs)This evaluator is to calculate sequence-to-sequence edit distance.
-The simple usage is :
-eval = ctc_evaluator.error(input=input, label=lbl)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.chunk(*args, **xargs)Chunk evaluator is used to evaluate segment labelling accuracy for a -sequence. It calculates precision, recall and F1 scores for the chunk detection.
-To use chunk evaluator, several concepts need to be clarified firstly.
-We can name a label by combining tag type and chunk type. (ie. B-ORG for begining of an organization name)
-The construction of label dictionary should obey the following rules:
-Scheme Description
-plain Use the same label for the whole chunk.
-IOB Two labels for chunk type X, B-X for chunk begining and I-X for chunk inside.
-IOE Two labels for chunk type X, E-X for chunk ending and I-X for chunk inside.
-IOBES Four labels for chunk type X, B-X for chunk begining, I-X for chunk inside, E-X for chunk end and S-X for single word chunk.
-To make it clear, let’s illustrate by an NER example. -Assuming that there are three named entity types including ORG, PER and LOC which are called ‘chunk type’ here, -if ‘IOB’ scheme were used, the label set will be extended to a set including B-ORG, I-ORG, B-PER, I-PER, B-LOC, I-LOC and O, -in which B-ORG for begining of ORG and I-ORG for inside of ORG. -Prefixes which are called ‘tag type’ here are added to chunk types and there are two tag types including B and I. -Of course, the training data should be labeled accordingly.
-The following table are equations to extract tag type and chunk type from a label.
-tagType = label % numTagType
-chunkType = label / numTagType
-otherChunkType = numChunkTypes
-The following table shows the mapping rule between tagType and tag type in each scheme.
-Scheme Begin Inside End Single
-plain 0 - - -
-IOB 0 1 - -
-IOE - 0 1 -
-IOBES 0 1 2 3
-Continue the NER example, and the label dict should look like this to satify above equations:
-B-ORG 0
-I-ORG 1
-B-PER 2
-I-PER 3
-B-LOC 4
-I-LOC 5
-O 6
-In this example, chunkType has three values: 0 for ORG, 1 for PER, 2 for LOC, because the scheme is -“IOB” so tagType has two values: 0 for B and 1 for I. -Here we will use I-LOC to explain the above mapping rules in detail. -For I-LOC, the label id is 5, so we can get tagType=1 and chunkType=2, which means I-LOC is a part of NER chunk LOC -and the tag is I.
-The simple usage is:
-eval = evaluator.chunk(input, label, chunk_scheme, num_chunk_types)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.precision_recall(*args, **xargs)An Evaluator to calculate precision and recall, F1-score. -It is adapt to the task with multiple labels.
-The simple usage:
-eval = precision_evaluator.recall(input, label)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.pnpair(*args, **xargs)Positive-negative pair rate Evaluator which adapts to rank task like -learning to rank. This evaluator must contain at least three layers.
-The simple usage:
-eval = evaluator.pnpair(input, label, query_id)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.sum(*args, **xargs)An Evaluator to sum the result of input.
-The simple usage:
-eval = evaluator.sum(input)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.column_sum(*args, **xargs)This Evaluator is used to sum the last column of input.
-The simple usage is:
-eval = column_evaluator.sum(input, label)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.classification_error_printer(*args, **xargs)This Evaluator is used to print the classification error of each sample.
-The simple usage is:
-eval = classification_error_evaluator.printer(input)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.gradient_printer(*args, **xargs)This Evaluator is used to print the gradient of input layers. It contains -one or more input layers.
-The simple usage is:
-eval = gradient_evaluator.printer(input)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.maxid_printer(*args, **xargs)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.
-The simple usage is:
-eval = maxid_evaluator.printer(input)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.maxframe_printer(*args, **xargs)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. -It contains one or more input layers.
-注解
-The width of each frame is 1.
-The simple usage is:
-eval = maxframe_evaluator.printer(input)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.seqtext_printer(*args, **xargs)Sequence text printer will print text according to index matrix and a -dictionary. There can be multiple input to this layer:
-1. If there is no id_input, the input must be a matrix containing -the sequence of indices;
-The output format will be:
-id prob space_seperated_tokens_from_dictionary_according_to_seq
-id space_seperated_tokens_from_dictionary_according_to_seq
-id space_seperated_tokens_from_dictionary_according_to_sub_seq
- space_seperated_tokens_from_dictionary_according_to_sub_seq
-...
-Typically SequenceTextPrinter layer takes output of maxid or RecurrentGroup -with maxid (when generating) as an input.
-The simple usage is:
-eval = seqtext_evaluator.printer(input=maxid,
- id_input=sample_id,
- dict_file=dict_file,
- result_file=result_file)
-| 参数: |
|
-
|---|---|
| 返回: | The seq_text_printer that prints the generated sequence to a file. - |
-
| 返回类型: | evaluator - |
-
paddle.v2.evaluator.value_printer(*args, **xargs)This Evaluator is used to print the values of input layers. It contains -one or more input layers.
-The simple usage is:
-eval = value_evaluator.printer(input)
-| 参数: |
|
-
|---|
paddle.v2.evaluator.detection_map(*args, **xargs)Detection mAP Evaluator. It will print mean Average Precision (mAP) for detection.
-The detection mAP Evaluator based on the output of detection_output layer counts -the true positive and the false positive bbox and integral them to get the -mAP.
-The simple usage is:
-eval = detection_evaluator.map(input=det_output,label=lbl)
-| 参数: |
|
-
|---|
- paddle.v2.layer.fcThe fully connected layer.
-The example usage is:
-fc = fc(input=layer,
- size=1024,
- act=paddle.v2.activation.Linear(),
- bias_attr=False)
-which is equal to:
-with mixed(size=1024) as fc:
- fc += full_matrix_projection(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.selective_fcSelectived fully connected layer. Different from fc, the output -of this layer can be sparse. It requires an additional input to indicate -several selected columns for output. If the selected columns is not -specified, selective_fc acts exactly like fc.
-The simple usage is:
-sel_fc = selective_fc(input=input, size=128, act=paddle.v2.activation.Tanh())
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.conv_operatorDifferent from img_conv, conv_op is an Operator, which can be used -in mixed. And conv_op takes two inputs to perform convolution. -The first input is the image and the second is filter kernel. It only -supports GPU mode.
-The example usage is:
-op = conv_operator(img=input1,
- filter=input2,
- filter_size=3,
- num_filters=64,
- num_channels=64)
-| 参数: |
|
-
|---|---|
| 返回: | A ConvOperator Object. - |
-
| 返回类型: | ConvOperator - |
-
paddle.v2.layer.conv_projectionDifferent from img_conv and conv_op, conv_projection is a Projection, -which can be used in mixed and concat. It uses cudnn to implement -convolution and only supports GPU mode.
-The example usage is:
-proj = conv_projection(input=input1,
- filter_size=3,
- num_filters=64,
- num_channels=64)
-| 参数: |
|
-
|---|---|
| 返回: | A Projection Object. - |
-
| 返回类型: | ConvTransProjection | ConvProjection - |
-
paddle.v2.layer.conv_shiftThe example usage is:
-conv_shift = conv_shift(a=layer1, b=layer2)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.img_convConvolution layer for image. Paddle can support both square and non-square -input currently.
-The details of convolution layer, please refer UFLDL’s convolution .
-Convolution Transpose (deconv) layer for image. Paddle can support both square -and non-square input currently.
-The details of convolution transpose layer, -please refer to the following explanation and references therein -<http://datascience.stackexchange.com/questions/6107/ -what-are-deconvolutional-layers/>`_ . -The num_channel means input image’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’s -num_filters.
-There are several groups of filters in PaddlePaddle implementation. -If the groups attribute is greater than 1, for example groups=2, -the input will be splitted into 2 parts along the channel axis, and -the filters will also be splitted into 2 parts. The first half of the filters -is only connected to the first half of the input channels, while the second -half of the filters is only connected to the second half of the input. After -the computation of convolution for each part of input, -the output will be obtained by concatenating the two results.
-The details of grouped convolution, please refer to: -ImageNet Classification with Deep Convolutional Neural Networks
-The example usage is:
-conv = img_conv(input=data, filter_size=1, filter_size_y=1,
- num_channels=8,
- num_filters=16, stride=1,
- bias_attr=False,
- act=paddle.v2.activation.Relu())
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.context_projectionContext Projection.
-It just reorganizes input sequence, combines “context_len” elements of the -sequence to one context from context_start. “context_start” will be set to --(context_len - 1) / 2 by default. When context position is out of sequence -length, padding will be filled as zero if padding_attr = False, otherwise -it is trainable.
-For example, origin sequence is [A B C D E F G], context len is 3, padding_attr -is not set, then after context projection, sequence will -be [ 0AB ABC BCD CDE DEF EFG FG0 ].
-| 参数: |
|
-
|---|---|
| 返回: | Projection object. - |
-
| 返回类型: | Projection - |
-
paddle.v2.layer.row_convThe row convolution is called lookahead convolution. It is firstly -introduced in paper of Deep Speech 2: End-to-End Speech Recognition -in English and Mandarin .
-The bidirectional RNN that learns representation for a sequence by -performing a forward and a backward pass through the entire sequence. -However, unlike unidirectional RNNs, bidirectional RNNs are challenging -to deploy in an online and low-latency setting. The lookahead convolution -incorporates information from future subsequences in a computationally -efficient manner to improve unidirectional RNNs.
-The connection of row convolution is different from the 1D sequence -convolution. Assumed that, the future context-length is k, that is to say, -it can get the output at timestep t by using the the input feature from t-th -timestep to (t+k+1)-th timestep. Assumed that the hidden dim of input -activations are d, the activations r_t for the new layer at time-step t are:
-注解
-The context_len is k + 1. That is to say, the lookahead step -number plus one equals context_len.
-row_conv = row_conv(input=input, context_len=3)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.img_poolImage pooling Layer.
-The details of pooling layer, please refer to ufldl’s pooling .
-The example usage is:
-maxpool = img_pool(input=conv,
- pool_size=3,
- pool_size_y=5,
- num_channels=8,
- stride=1,
- stride_y=2,
- padding=1,
- padding_y=2,
- pool_type=MaxPooling())
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sppA layer performs spatial pyramid pooling.
- -The example usage is:
-spp = spp(input=data,
- pyramid_height=2,
- num_channels=16,
- pool_type=MaxPooling())
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.maxoutSo groups should be larger than 1, and the num of channels should be able -to be devided by groups.
-The simple usage is:
-maxout = maxout(input,
- num_channels=128,
- groups=4)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.roi_poolA layer used by Fast R-CNN to extract feature maps of ROIs from the last -feature map.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.padThis operation pads zeros to the input data according to pad_c,pad_h -and pad_w. pad_c, pad_h, pad_w specify the size in the corresponding -dimension. And the input data shape is NCHW.
-For example, pad_c=[2,3] means padding 2 zeros before the input data -and 3 zeros after the input data in the channel dimension. pad_h means -padding zeros in the height dimension. pad_w means padding zeros in the -width dimension.
-For example,
-input(2,2,2,3) = [
- [ [[1,2,3], [3,4,5]],
- [[2,3,5], [1,6,7]] ],
- [ [[4,3,1], [1,8,7]],
- [[3,8,9], [2,3,5]] ]
- ]
-
-pad_c=[1,1], pad_h=[0,0], pad_w=[0,0]
-
-output(2,4,2,3) = [
- [ [[0,0,0], [0,0,0]],
- [[1,2,3], [3,4,5]],
- [[2,3,5], [1,6,7]],
- [[0,0,0], [0,0,0]] ],
- [ [[0,0,0], [0,0,0]],
- [[4,3,1], [1,8,7]],
- [[3,8,9], [2,3,5]],
- [[0,0,0], [0,0,0]] ]
- ]
-The simply usage is:
-pad = pad(input=ipt,
- pad_c=[4,4],
- pad_h=[0,0],
- pad_w=[2,2])
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.img_cmrnormResponse normalization across feature maps.
- -The example usage is:
-norm = img_cmrnorm(input=net, size=5)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.batch_normBatch Normalization Layer. The notation of this layer is as follows.
-\(x\) is the input features over a mini-batch.
-The example usage is:
-norm = batch_norm(input=net, act=paddle.v2.activation.Relu())
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sum_to_one_normA layer for sum-to-one normalization, -which is used in NEURAL TURING MACHINE.
-where \(in\) is a (batchSize x dataDim) input vector, -and \(out\) is a (batchSize x dataDim) output vector.
-The example usage is:
-sum_to_one_norm = sum_to_one_norm(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cross_channel_normNormalize a layer’s output. This layer is necessary for ssd. This -layer applys normalization across the channels of each sample to -a convolutional layer’s output and scales the output by a group of -trainable factors whose dimensions equal to the channel’s number.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.row_l2_normA layer for L2-normalization in each row.
-where the size of \(in\) is (batchSize x dataDim) , -and the size of \(out\) is a (batchSize x dataDim) .
-The example usage is:
-row_l2_norm = row_l2_norm(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.recurrentSimple recurrent unit layer. It is just a fully connect layer through both -time and neural network.
-For each sequence [start, end] it performs the following computation:
-If reversed is true, the order is reversed:
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.lstmemoryLong Short-term Memory Cell.
-The memory cell was implemented as follow equations.
-NOTE: In PaddlePaddle’s implementation, the multiplications -\(W_{xi}x_{t}\) , \(W_{xf}x_{t}\), -\(W_{xc}x_t\), \(W_{xo}x_{t}\) are not done in the lstmemory layer, -so an additional mixed with full_matrix_projection or a fc must -be included in the configuration file to complete the input-to-hidden -mappings before lstmemory is called.
-NOTE: This is a low level user interface. You can use network.simple_lstm -to config a simple plain lstm layer.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.grumemoryGate Recurrent Unit Layer.
-The memory cell was implemented as follow equations.
-1. update gate \(z\): defines how much of the previous memory to -keep around or the unit updates its activations. The update gate -is computed by:
-2. reset gate \(r\): determines how to combine the new input with the -previous memory. The reset gate is computed similarly to the update gate:
-3. The candidate activation \(\tilde{h_t}\) is computed similarly to -that of the traditional recurrent unit:
-4. The hidden activation \(h_t\) of the GRU at time t is a linear -interpolation between the previous activation \(h_{t-1}\) and the -candidate activation \(\tilde{h_t}\):
-NOTE: In PaddlePaddle’s implementation, the multiplication operations -\(W_{r}x_{t}\), \(W_{z}x_{t}\) and \(W x_t\) are not performed -in gate_recurrent layer. Consequently, an additional mixed with -full_matrix_projection or a fc must be included before grumemory -is called.
- -The simple usage is:
-gru = grumemory(input)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.gated_unitThe gated unit layer implements a simple gating mechanism over the input. -The input \(X\) is first projected into a new space \(X'\), and -it is also used to produce a gate weight \(\sigma\). Element-wise -product between \(X'\) and \(\sigma\) is finally returned.
-The example usage is:
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.memoryThe memory takes a layer’s output at previous time step as its own output.
-If boot_bias, the activation of the bias is the initial value of the memory.
-If boot_with_const_id is set, then the memory’s output at the first time step
-is a IndexSlot, the Arguments.ids()[0] is this cost_id.
If boot is specified, the memory’s output at the first time step will -be the boot’s output.
-In other case, the default memory’s output at the first time step is zero.
-mem = memory(size=256, name='state')
-state = fc(input=mem, size=256, name='state')
-If you do not want to specify the name, you can also use set_input() -to specify the layer to be remembered as the following:
-mem = memory(size=256)
-state = fc(input=mem, size=256)
-mem.set_input(mem)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.recurrent_groupRecurrent 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 useful for attention-based models, or Neural -Turning Machine like models.
-The basic usage (time steps) is:
-def step(input):
- output = fc(input=layer,
- size=1024,
- act=paddle.v2.activation.Linear(),
- bias_attr=False)
- return output
-
-group = recurrent_group(input=layer,
- step=step)
-You can see following configs for further usages:
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.lstm_stepLSTM Step Layer. This function is used only in recurrent_group. -The lstm equations are shown as follows.
-The input of lstm step is \(Wx_t + Wh_{t-1}\), and user should use
-mixed and full_matrix_projection to calculate these
-input vectors.
The state of lstm step is \(c_{t-1}\). And lstm step layer will do
-This layer has two outputs. The default output is \(h_t\). The other
-output is \(o_t\), whose name is ‘state’ and users can use
-get_output to extract this output.
| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.gru_step| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.beam_searchBeam 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.
-The example usage is:
-def rnn_step(input):
- last_time_step_output = memory(name='rnn', size=512)
- with mixed(size=512, name='rnn') as simple_rnn:
- simple_rnn += full_matrix_projection(input)
- simple_rnn += last_time_step_output
- return simple_rnn
-
-generated_word_embedding = GeneratedInput(
- size=target_dictionary_dim,
- embedding_name="target_language_embedding",
- embedding_size=word_vector_dim)
-
-beam_gen = beam_search(name="decoder",
- step=rnn_step,
- input=[StaticInput(encoder_last),
- generated_word_embedding],
- bos_id=0,
- eos_id=1,
- beam_size=5)
-Please see the following demo for more details:
-| 参数: |
|
-
|---|---|
| 返回: | The generated word index. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.get_outputGet layer’s output by name. In PaddlePaddle, a layer might return multiple -values, but returns one layer’s output. If the user wants to use another -output besides the default one, please use get_output first to get -the output from input.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.mixedMixed Layer. A mixed layer will add all inputs together, then activate the sum. -Each input is a projection or operator.
-There are two styles of usages.
-with mixed(size=256) as m:
- m += full_matrix_projection(input=layer1)
- m += identity_projection(input=layer2)
-m = mixed(size=256,
- input=[full_matrix_projection(input=layer1),
- full_matrix_projection(input=layer2)])
-| 参数: |
|
-
|---|---|
| 返回: | MixedLayerType object. - |
-
| 返回类型: | MixedLayerType - |
-
paddle.v2.layer.embeddingDefine a embedding Layer.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.scaling_projectionscaling_projection multiplies the input with a scalar parameter.
-The example usage is:
-proj = scaling_projection(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | ScalingProjection object. - |
-
| 返回类型: | ScalingProjection - |
-
paddle.v2.layer.dotmul_projectionDotMulProjection takes a layer as input and performs -element-wise multiplication with weight.
-where \(.*\) means element-wise multiplication.
-The example usage is:
-proj = dotmul_projection(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | DotMulProjection object. - |
-
| 返回类型: | DotMulProjection - |
-
paddle.v2.layer.dotmul_operatorDotMulOperator takes two inputs and performs element-wise multiplication:
-where \(.*\) means element-wise multiplication, and -scale is a config scalar, its default value is 1.
-The example usage is:
-op = dotmul_operator(a=layer1, b=layer2, scale=0.5)
-| 参数: |
|
-
|---|---|
| 返回: | DotMulOperator object. - |
-
| 返回类型: | DotMulOperator - |
-
paddle.v2.layer.full_matrix_projectionFull Matrix Projection. It performs full matrix multiplication.
-There are two styles of usage.
-with mixed(size=100) as m:
- m += full_matrix_projection(input=layer)
-proj = full_matrix_projection(input=layer,
- size=100,
- param_attr=ParamAttr(name='_proj'))
-| 参数: |
|
-
|---|---|
| 返回: | FullMatrixProjection Object. - |
-
| 返回类型: | FullMatrixProjection - |
-
paddle.v2.layer.identity_projectionThe example usage is:
-proj = identity_projection(input=layer)
-The example usage is:
-proj = identity_projection(input=layer,
- offset=10)
-Note that neither of the projections have trainable parameter.
-| 参数: |
|
-
|---|---|
| 返回: | IdentityProjection or IdentityOffsetProjection object - |
-
| 返回类型: | IdentityProjection | IdentityOffsetProjection - |
-
paddle.v2.layer.slice_projectionslice_projection slices the input value into multiple parts, -then selects and merges some of them into a new output.
-The example usage is:
-proj = slice_projection(input=layer, slices=[(0, 10), (20, 30)])
-Note that slice_projection has no trainable parameter.
-| 参数: |
|
-
|---|---|
| 返回: | SliceProjection object. - |
-
| 返回类型: | SliceProjection - |
-
paddle.v2.layer.table_projectionTable Projection. It selects rows from parameter where row_id -is in input_ids.
-where \(out\) is output, \(table\) is parameter, \(ids\) is input_ids, -and \(i\) is row_id.
-There are two styles of usage.
-with mixed(size=100) as m:
- m += table_projection(input=layer)
-proj = table_projection(input=layer,
- size=100,
- param_attr=ParamAttr(name='_proj'))
-| 参数: |
|
-
|---|---|
| 返回: | TableProjection Object. - |
-
| 返回类型: | TableProjection - |
-
paddle.v2.layer.trans_full_matrix_projectionDifferent from full_matrix_projection, this projection performs matrix -multiplication, using the transpose of weight.
-\(w^\mathrm{T}\) means the transpose of weight. -The simply usage is:
-proj = trans_full_matrix_projection(input=layer,
- size=100,
- param_attr=ParamAttr(
- name='_proj',
- initial_mean=0.0,
- initial_std=0.01))
-| 参数: |
|
-
|---|---|
| 返回: | TransposedFullMatrixProjection Object. - |
-
| 返回类型: | TransposedFullMatrixProjection - |
-
paddle.v2.layer.AggregateLevelPaddlePaddle supports three sequence types:
-SequenceType.NO_SEQUENCE means the sample is not a sequence.SequenceType.SEQUENCE means the sample is a sequence.SequenceType.SUB_SEQUENCE means the sample is a nested sequence,
-each timestep of which is also a sequence.Accordingly, AggregateLevel supports two modes:
-AggregateLevel.TO_NO_SEQUENCE means the aggregation acts on each
-timestep of a sequence, both SUB_SEQUENCE and SEQUENCE will
-be aggregated to NO_SEQUENCE.AggregateLevel.TO_SEQUENCE means the aggregation acts on each
-sequence of a nested sequence, SUB_SEQUENCE will be aggregated to
-SEQUENCE.paddle.v2.layer.poolingPooling layer for sequence inputs, not used for Image.
-If stride > 0, this layer slides a window whose size is determined by stride, -and returns the pooling value of the sequence in the window as the output. Thus, -a long sequence will be shortened. Note that for sequence with sub-sequence, the -default value of stride is -1.
-The example usage is:
-seq_pool = pooling(input=layer,
- pooling_type=AvgPooling(),
- agg_level=AggregateLevel.TO_NO_SEQUENCE)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.last_seqGet Last Timestamp Activation of a sequence.
-If stride > 0, this layer will slide a window whose size is determined by stride, -and return the last value of the sequence in the window as the output. Thus, a -long sequence will be shortened. Note that for sequence with sub-sequence, the -default value of stride is -1.
-The simple usage is:
-seq = last_seq(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.first_seqGet First Timestamp Activation of a sequence.
-If stride > 0, this layer will slide a window whose size is determined by stride, -and return the first value of the sequence in the window as the output. Thus, a -long sequence will be shortened. Note that for sequence with sub-sequence, the -default value of stride is -1.
-The simple usage is:
-seq = first_seq(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sub_seqsub_seq will return sub-sequences from the input sequences. For each -sequence in the input sequence layer, sub_seq will slice it by given -offset and size. Please notice that, number of offset value and size value -both are equal to the number of sequence in the input layer.
-sub_seq = sub_seq(input=input_seq, offsets=offsets, sizes=sizes)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.concatConcatenate all input vectors to one vector. -Inputs can be a list of paddle.v2.config_base.Layer or a list of projection.
-The example usage is:
-concat = concat(input=[layer1, layer2])
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.seq_concatConcatenate sequence a and sequence b.
-Output: [a1, ..., am, b1, ..., bn]
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The example usage is:
-concat = seq_concat(a=layer1, b=layer2)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.seq_sliceseq_slice will return one or several sub-sequences from the -input sequence layer given start and end indices.
----
-- If only start indices are given, and end indices are set to None, -this layer slices the input sequence from the given start indices -to its end.
-- If only end indices are given, and start indices are set to None, -this layer slices the input sequence from its beginning to the -given end indices.
-- If start and end indices are both given, they should have the same -number of elements.
-
If start or end indices contains more than one elements, the input sequence -will be sliced for multiple times.
-seq_silce = seq_slice(input=input_seq,
- starts=start_pos, ends=end_pos)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sub_nested_seqThe sub_nested_seq accepts two inputs: the first one is a nested -sequence; the second one is a set of selceted indices in the nested sequence.
-Then sub_nest_seq trims the first nested sequence input according -to the selected indices to form a new output. This layer is useful in -beam training.
-The example usage is:
-sub_nest_seq = sub_nested_seq(input=data, selected_indices=selected_ids)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.block_expandThe expanding method is the same with ExpandConvLayer, but saved the transposed -value. After expanding, output.sequenceStartPositions will store timeline. -The number of time steps is outputH * outputW and the dimension of each -time step is block_y * block_x * num_channels. This layer can be used after -convolutional neural network, and before recurrent neural network.
-The simple usage is:
-block_expand = block_expand(input=layer,
- num_channels=128,
- stride_x=1,
- stride_y=1,
- block_x=1,
- block_x=3)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.ExpandLevelPlease refer to AggregateLevel first.
-ExpandLevel supports two modes:
-ExpandLevel.FROM_NO_SEQUENCE means the expansion acts on
-NO_SEQUENCE, which will be expanded to
-SEQUENCE or SUB_SEQUENCE.ExpandLevel.FROM_SEQUENCE means the expansion acts on
-SEQUENCE, which will be expanded to
-SUB_SEQUENCE.paddle.v2.layer.expandA layer for expanding dense data or (sequence data where the length of each -sequence is one) to sequence data.
-The example usage is:
-expand = expand(input=layer1,
- expand_as=layer2,
- expand_level=ExpandLevel.FROM_NO_SEQUENCE)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.repeatA layer for repeating the input for num_repeats times.
-If as_row_vector:
-If not as_row_vector:
-The example usage is:
-expand = repeat(input=layer, num_repeats=4)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.rotateA layer for rotating 90 degrees (clock-wise) for each feature channel, -usually used when the input sample is some image or feature map.
-where \(x\) is (M x N x C) input, and \(y\) is (N x M x C) output.
-The example usage is:
-rot = rotate(input=layer,
- height=100,
- width=100)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.seq_reshapeA 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.
-Note that T*M/N must be an integer.
-The example usage is:
-reshape = seq_reshape(input=layer, reshape_size=4)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.addtoAddtoLayer.
-where \(y\) is output, \(x\) is input, \(b\) is bias, -and \(f\) is activation function.
-The example usage is:
-addto = addto(input=[layer1, layer2],
- act=paddle.v2.activation.Relu(),
- bias_attr=False)
-This layer just simply adds all input layers together, then activates the -sum. All inputs should share the same dimension, which is also the dimension -of this layer’s output.
-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.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.linear_combwhere \(0 \le i \le N-1\)
-Or in the matrix notation:
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The simple usage is:
-linear_comb = linear_comb(weights=weight, vectors=vectors,
- size=elem_dim)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.interpolationThis layer performs linear interpolation on two inputs, -which is used in NEURAL TURING MACHINE.
-where \(x_1\) and \(x_2\) are two (batchSize x dataDim) inputs, -\(w\) is (batchSize x 1) weight vector, and \(y\) is -(batchSize x dataDim) output.
-The example usage is:
-interpolation = interpolation(input=[layer1, layer2], weight=layer3)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.bilinear_interpThis layer implements bilinear interpolation on convolutional layer’s output.
-Please refer to Wikipedia: https://en.wikipedia.org/wiki/Bilinear_interpolation
-The simple usage is:
-bilinear = bilinear_interp(input=layer1, out_size_x=64, out_size_y=64)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.dropoutThe example usage is:
-dropout = dropout(input=input, dropout_rate=0.5)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.dot_prodA layer for computing the dot product of two vectors.
-The example usage is:
-dot_prod = dot_prod(input1=vec1, input2=vec2)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.out_prodA layer for computing the outer product of two vectors -The result is a matrix of size(input1) x size(input2)
-The example usage is:
-out_prod = out_prod(input1=vec1, input2=vec2)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.powerThis layer applies a power function to a vector element-wise, -which is used in NEURAL TURING MACHINE.
-where \(x\) is an input vector, \(w\) is a scalar exponent, -and \(y\) is an output vector.
-The example usage is:
-power = power(input=layer1, weight=layer2)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.scalingA layer for multiplying input vector by weight scalar.
-where \(x\) is size=dataDim input, \(w\) is size=1 weight, -and \(y\) is size=dataDim output.
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The example usage is:
-scale = scaling(input=layer1, weight=layer2)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.clipA layer for clipping the input value by the threshold.
-clip = clip(input=input, min=-10, max=10)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.resizeThe resize layer resizes the input matrix with a shape of [Height, Width] -into the output matrix with a shape of [Height x Width / size, size], -where size is the parameter of this layer indicating the output dimension.
-| 参数: |
|
-
|---|---|
| 返回: | A paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.slope_interceptThis layer for applying a slope and an intercept to the input.
-The simple usage is:
-scale = slope_intercept(input=input, slope=-1.0, intercept=1.0)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.tensorThis layer performs tensor operation on two inputs. -For example:
-The simple usage is:
-tensor = tensor(a=layer1, b=layer2, size=1000)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cos_simCosine Similarity Layer. The cosine similarity equation is here.
-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.
-Note that the above computation is for one sample. Multiple samples are -processed in one batch.
-The example usage is:
-cos = cos_sim(a=layer1, b=layer2, size=3)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.l2_distanceThis layer calculates and returns the Euclidean distance between two input -vectors x and y. The equation is as follows:
-The output size of this layer is fixed to be 1. Note that the above -computation is for one sample. Multiple samples are processed in one batch.
-The example usage is:
-l2_sim = l2_distance(x=layer1, y=layer2)
-| 参数: |
|
-
|---|---|
| 返回: | The returned paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.transA layer for transposing a minibatch matrix.
-where \(x\) is (M x N) input, and \(y\) is (N x M) output.
-The example usage is:
-trans = trans(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.scale_shiftA layer applies a linear transformation to each element in each row of -the input matrix. For each element, the layer first re-scales it and then -adds a bias to it.
-This layer is very like the SlopeInterceptLayer, except the scale and -bias are trainable.
-scale_shift = scale_shift(input=input, bias_attr=False)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.factorization_machineThe Factorization Machine models pairwise feature interactions as inner -product of the learned latent vectors corresponding to each input feature. -The Factorization Machine can effectively capture feature interactions -especially when the input is sparse.
-This implementation only consider the 2-order feature interactions using -Factorization Machine with the formula:
-注解
-X is the input vector with size n. V is the factor matrix. Each row of V -is the latent vector corresponding to each input dimesion. The size of -each latent vector is k.
-For details of Factorization Machine, please refer to the paper: -Factorization machines.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.max_idA layer for finding the id which has the maximal value for each sample. -The result is stored in output.ids.
-The example usage is:
-maxid = maxid(input=layer)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sampling_idA layer for sampling id from a multinomial distribution from the input layer. -Sampling one id for one sample.
-The simple usage is:
-samping_id = sampling_id(input=input)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.multiplexThis layer multiplex multiple layers according to the indexes, -which are provided by the first input layer. -inputs[0]: the indexes of the layers to form the output of size batchSize. -inputs[1:N]; the candidate output data. -For each index i from 0 to batchSize - 1, the i-th row of the output is the -the same to the i-th row of the (index[i] + 1)-th layer.
-For each i-th row of output: -.. math:
-y[i][j] = x_{x_{0}[i] + 1}[i][j], j = 0,1, ... , (x_{1}.width - 1)
-where, y is output. \(x_{k}\) is the k-th input layer and -\(k = x_{0}[i] + 1\).
-The example usage is:
-maxid = multiplex(input=layers)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cross_entropy_costA loss layer for multi class entropy.
-The example usage is:
-cost = cross_entropy(input=input,
- label=label)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.cross_entropy_with_selfnorm_costA loss layer for multi class entropy with selfnorm. -Input should be a vector of positive numbers, without normalization.
-The example usage is:
-cost = cross_entropy_with_selfnorm(input=input,
- label=label)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.multi_binary_label_cross_entropy_costA loss layer for multi binary label cross entropy.
-The example usage is:
-cost = multi_binary_label_cross_entropy(input=input,
- label=label)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.huber_regression_costIn statistics, the Huber loss is a loss function used in robust regression, -that is less sensitive to outliers in data than the squared error loss. -Given a prediction f(x), a label y and \(\delta\), the loss function -is defined as:
-The example usage is:
-cost = huber_regression_cost(input=input, label=label)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer. - |
-
paddle.v2.layer.huber_classification_costFor classification purposes, a variant of the Huber loss called modified Huber -is sometimes used. Given a prediction f(x) (a real-valued classifier score) and -a true binary class label \(y\in \{-1, 1 \}\), the modified Huber -loss is defined as:
-The example usage is:
-cost = huber_classification_cost(input=input, label=label)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.lambda_costlambdaCost for lambdaRank LTR approach.
-The example usage is:
-cost = lambda_cost(input=input,
- score=score,
- NDCG_num=8,
- max_sort_size=-1)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.square_error_costsum of square error cost:
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.rank_costA cost Layer for learning to rank using gradient descent.
-The example usage is:
-cost = rank_cost(left=out_left,
- right=out_right,
- label=label)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.sum_costA loss layer which calculates the sum of the input as loss.
-The example usage is:
-cost = sum_cost(input=input)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer. - |
-
paddle.v2.layer.crfA layer for calculating the cost of sequential conditional random -field model.
-The example usage is:
-crf = crf(input=input,
- label=label,
- size=label_dim)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.crf_decodingA layer for calculating the decoding sequence of sequential conditional -random field model. The decoding sequence is stored in output.ids. -If the input ‘label’ is provided, it is treated as the ground-truth label, and -this layer will also calculate error. output.value[i] is 1 for an incorrect -decoding and 0 for the correct.
-The example usage is:
-crf_decoding = crf_decoding(input=input,
- size=label_dim)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.ctcConnectionist Temporal Classification (CTC) is designed for temporal -classication task. e.g. sequence labeling problems where the -alignment between the inputs and the target labels is unknown.
-注解
-Considering the ‘blank’ label needed by CTC, you need to use (num_classes + 1) -as the size of the input, where num_classes is the category number. -And the ‘blank’ is the last category index. So the size of ‘input’ layer (e.g. -fc with softmax activation) should be (num_classes + 1). The size of -ctc should also be (num_classes + 1).
-The example usage is:
-ctc = ctc(input=input,
- label=label,
- size=9055,
- norm_by_times=True)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.warp_ctcA layer intergrating the open-source warp-ctc library, which is used in
-Deep Speech 2: End-toEnd Speech Recognition in English and Mandarin, to compute Connectionist Temporal
-Classification (CTC) loss. Besides, another warp-ctc repository, which is forked from
-the official one, is maintained to enable more compiling options. During the
-building process, PaddlePaddle will clone the source codes, build and
-install it to third_party/install/warpctc directory.
注解
-The example usage is:
-ctc = warp_ctc(input=input,
- label=label,
- size=1001,
- blank=1000,
- norm_by_times=False)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.nceNoise-contrastive estimation.
- -The example usage is:
-cost = nce(input=[layer1, layer2], label=layer2,
- param_attr=[attr1, attr2], weight=layer3,
- num_classes=3, neg_distribution=[0.1,0.3,0.6])
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.hsigmoidOrganize the classes into a binary tree. At each node, a sigmoid function -is used to calculate the probability of belonging to the right branch.
-The example usage is:
-cost = hsigmoid(input=[layer1, layer2],
- label=data)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.smooth_l1_costThis is a L1 loss but more smooth. It requires that the -sizes of input and label are equal. The formula is as follows,
-in which
-The example usage is:
-cost = smooth_l1_cost(input=input,
- label=label)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.multibox_lossCompute the location loss and the confidence loss for ssd.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.detection_outputApply the NMS to the output of network and compute the predict bounding -box location. The output’s shape of this layer could be zero if there is -no valid bounding box.
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.eosA layer for checking EOS for each sample: -- output_id = (input_id == conf.eos_id)
-The result is stored in output_.ids. -It is used by recurrent layer group.
-The example usage is:
-eos = eos(input=layer, eos_id=id)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
paddle.v2.layer.preluThe Parametric Relu activation that actives outputs with a learnable weight.
-The example usage is:
-prelu = prelu(input=layers, partial_sum=1)
-| 参数: |
|
-
|---|---|
| 返回: | paddle.v2.config_base.Layer object. - |
-
| 返回类型: | paddle.v2.config_base.Layer - |
-
- The v2.networks module contains pieces of neural network that combine multiple layers.
-paddle.v2.networks.sequence_conv_pool(*args, **kwargs)Text convolution pooling group.
-Text input => Context Projection => FC Layer => Pooling => Output.
-| 参数: |
|
-
|---|---|
| 返回: | layer’s output. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.text_conv_pool(*args, **kwargs)Text convolution pooling group.
-Text input => Context Projection => FC Layer => Pooling => Output.
-| 参数: |
|
-
|---|---|
| 返回: | layer’s output. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.img_conv_bn_pool(*args, **kwargs)Convolution, batch normalization, pooling group.
-Img input => Conv => BN => Pooling => Output.
-| 参数: |
|
-
|---|---|
| 返回: | layer’s output - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.img_conv_group(*args, **kwargs)Image Convolution Group, Used for vgg net.
-| 参数: |
|
-
|---|---|
| 返回: | layer’s output - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.simple_img_conv_pool(*args, **kwargs)Simple image convolution and pooling group.
-Img input => Conv => Pooling => Output.
-| 参数: |
|
-
|---|---|
| 返回: | layer’s output - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.vgg_16_network(input_image, num_channels, num_classes=1000)Same model from https://gist.github.com/ksimonyan/211839e770f7b538e2d8
-| 参数: |
|
-
|---|---|
| 返回: | layer’s output - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.lstmemory_unit(*args, **kwargs)lstmemory_unit defines the caculation process of a LSTM unit during a -single time step. This function is not a recurrent layer, so it can not be -directly used to process sequence input. This function is always used in -recurrent_group (see layers.py for more details) to implement attention -mechanism.
-Please refer to Generating Sequences With Recurrent Neural Networks -for more details about LSTM. The link goes as follows: -.. _Link: https://arxiv.org/abs/1308.0850
-The example usage is:
-lstm_step = lstmemory_unit(input=[layer1],
- size=256,
- act=TanhActivation(),
- gate_act=SigmoidActivation(),
- state_act=TanhActivation())
-| 参数: |
|
-
|---|---|
| 返回: | The lstmemory unit name. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.lstmemory_group(*args, **kwargs)lstm_group is a recurrent_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 and merely use its outputs, -it is recommended to use the lstmemory, which is relatively faster than -lstmemory_group.
-NOTE: In PaddlePaddle’s implementation, the following input-to-hidden -multiplications: -\(W_{x_i}x_{t}\) , \(W_{x_f}x_{t}\), -\(W_{x_c}x_t\), \(W_{x_o}x_{t}\) 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.
-The example usage is:
-lstm_step = lstmemory_group(input=[layer1],
- size=256,
- act=TanhActivation(),
- gate_act=SigmoidActivation(),
- state_act=TanhActivation())
-| 参数: |
|
-
|---|---|
| 返回: | the lstmemory group. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.simple_lstm(*args, **kwargs)Simple LSTM Cell.
-It just combines a mixed layer with fully_matrix_projection and a lstmemory -layer. The simple lstm cell was implemented with follow equations.
-Please refer to Generating Sequences With Recurrent Neural Networks for more -details about lstm. Link is here.
-| 参数: |
|
-
|---|---|
| 返回: | layer’s output. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.bidirectional_lstm(*args, **kwargs)A bidirectional_lstm is a recurrent unit that iterates over the input -sequence both in forward and backward 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.
-Please refer to Neural Machine Translation by Jointly Learning to Align -and Translate for more details about the bidirectional lstm. -The link goes as follows: -.. _Link: https://arxiv.org/pdf/1409.0473v3.pdf
-The example usage is:
-bi_lstm = bidirectional_lstm(input=[input1], size=512)
-| 参数: |
|
-
|---|---|
| 返回: | LayerOutput object. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.gru_unit(*args, **kwargs)gru_unit defines the calculation process of a gated recurrent unit during a single -time step. This function is not a recurrent layer, so it can not be -directly used to process sequence input. This function is always used in -the recurrent_group (see layers.py for more details) to implement attention -mechanism.
-Please see grumemory in layers.py for the details about the maths.
-| 参数: |
|
-
|---|---|
| 返回: | the gru output layer. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.gru_group(*args, **kwargs)gru_group is a recurrent_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 and merely use the outputs of a GRU, it is recommended -to use the grumemory, which is relatively faster.
-Please see grumemory in layers.py for more detail about the maths.
-The example usage is:
-gru = gru_group(input=[layer1],
- size=256,
- act=TanhActivation(),
- gate_act=SigmoidActivation())
-| 参数: |
|
-
|---|---|
| 返回: | the gru group. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.simple_gru(*args, **kwargs)You may 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 \(W x_t\) 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.
-The computational speed is that, grumemory is relatively better than -gru_group, and gru_group is relatively better than simple_gru.
-The example usage is:
-gru = simple_gru(input=[layer1], size=256)
-| 参数: |
|
-
|---|---|
| 返回: | the gru group. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.simple_gru2(*args, **kwargs)simple_gru2 is the same with simple_gru, but using grumemory instead. -Please refer to grumemory in layers.py for more detail about the math. -simple_gru2 is faster than simple_gru.
-The example usage is:
-gru = simple_gru2(input=[layer1], size=256)
-| 参数: |
|
-
|---|---|
| 返回: | the gru group. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.bidirectional_gru(*args, **kwargs)A bidirectional_gru is a recurrent unit that iterates over the input -sequence both in forward and backward 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.
-The example usage is:
-bi_gru = bidirectional_gru(input=[input1], size=512)
-| 参数: |
|
-
|---|---|
| 返回: | LayerOutput object. - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.simple_attention(*args, **kwargs)Calculate and return a context vector with attention mechanism. -Size of the context vector equals to size of the encoded_sequence.
-where \(h_{j}\) is the jth element of encoded_sequence, -\(U_{a}h_{j}\) is the jth element of encoded_proj -\(s_{i-1}\) is decoder_state -\(f\) is weight_act, and is set to tanh by default.
-Please refer to Neural Machine Translation by Jointly Learning to -Align and Translate for more details. The link is as follows: -https://arxiv.org/abs/1409.0473.
-The example usage is:
-context = simple_attention(encoded_sequence=enc_seq,
- encoded_proj=enc_proj,
- decoder_state=decoder_prev,)
-| 参数: |
|
-
|---|---|
| 返回: | a context vector - |
-
| 返回类型: | LayerOutput - |
-
paddle.v2.networks.dot_product_attention(*args, **kwargs)Calculate and return a context vector with dot-product attention mechanism. -The dimension of the context vector equals to that of the attended_sequence.
-where \(h_{j}\) is the jth element of encoded_sequence, -\(z_{j}\) is the jth element of attended_sequence, -\(s_{i-1}\) is transformed_state.
-The example usage is:
-context = dot_product_attention(encoded_sequence=enc_seq,
- attended_sequence=att_seq,
- transformed_state=state,)
-| 参数: |
|
-
|---|---|
| 返回: | The context vector. - |
-
| 返回类型: | LayerOutput - |
-
- paddle.v2.optimizer.Momentum(momentum=None, sparse=False, **kwargs)Momentum Optimizer.
-When sparse=False, the momentum update formula is as follows:
-where, \(k\) is momentum, \(\lambda\) is decay rate, -\(\gamma_t\) is learning rate at the t’th iteration. -\(w_{t}\) is the weight as the t’th iteration. -And the \(v_{t}\) is the history momentum variable.
-When sparse=True, the update scheme:
-where \(k\) is momentum, \(\lambda\) is decay rate, -\(\gamma_t\) is learning rate at the t’th iteration.
-| 参数: |
|
-
|---|
paddle.v2.optimizer.Adam(beta1=0.9, beta2=0.999, epsilon=1e-08, **kwargs)Adam optimizer. -The details of please refer Adam: A Method for Stochastic Optimization
-| 参数: |
|
-
|---|
paddle.v2.optimizer.Adamax(beta1=0.9, beta2=0.999, **kwargs)Adamax optimizer.
-The details of please refer this Adam: A Method for Stochastic Optimization
-| 参数: |
|
-
|---|
paddle.v2.optimizer.AdaGrad(**kwargs)Adagrad(for ADAptive GRAdient algorithm) optimizer.
-For details please refer this Adaptive Subgradient Methods for -Online Learning and Stochastic Optimization.
-paddle.v2.optimizer.DecayedAdaGrad(rho=0.95, epsilon=1e-06, **kwargs)AdaGrad method with decayed sum gradients. The equations of this method -show as follow.
-| 参数: |
|
-
|---|
paddle.v2.optimizer.AdaDelta(rho=0.95, epsilon=1e-06, **kwargs)AdaDelta method. The details of adadelta please refer to this -ADADELTA: AN ADAPTIVE LEARNING RATE METHOD.
-| 参数: |
|
-
|---|
paddle.v2.optimizer.RMSProp(rho=0.95, epsilon=1e-06, **kwargs)RMSProp(for Root Mean Square Propagation) optimizer. For details please -refer this slide.
-The equations of this method as follows:
-| 参数: |
|
-
|---|
- paddle.v2.pooling.BasePool(name)Base Pooling Type. -Note these pooling types are used for sequence input, not for images. -Each PoolingType contains one parameter:
-| 参数: | name (basestring) – pooling layer type name used by paddle. | -
|---|
paddle.v2.pooling.Avg(strategy='average')Average pooling.
-Return the average values for each dimension in sequence or time steps.
-paddle.v2.pooling.Max(output_max_index=None)Max pooling.
-Return the very large values for each dimension in sequence or time steps.
-| 参数: | output_max_index (bool|None) – True if output sequence max index instead of max -value. None means use default value in proto. | -
|---|
paddle.v2.pooling.SumSum pooling.
-Return the sum values of each dimension in sequence or time steps.
-paddle.v2.pooling.SquareRootNSquare Root Pooling.
-Return the square root values of each dimension in sequence or time steps.
-paddle.v2.pooling.CudnnMaxCudnn max pooling only support GPU. Return the maxinum value in the -pooling window.
-
-
- paddle.v2.data_type.dense_array(dim, seq_type=0)Dense Array. It means the input feature is dense array with float type. -For example, if the input is an image with 28*28 pixels, the input of -Paddle neural network could be a dense vector with dimension 784 or a -numpy array with shape (28, 28).
-For the 2-D convolution operation, each sample in one mini-batch must have -the similarly size in PaddlePaddle now. But, it supports variable-dimension -feature across mini-batch. For the variable-dimension, the param dim is not -used. While the data reader must yield numpy array and the data feeder will -set the data shape correctly.
-| 参数: |
|
-
|---|---|
| 返回: | An input type object. - |
-
| 返回类型: | InputType - |
-
paddle.v2.data_type.dense_vector(dim, seq_type=0)Dense Array. It means the input feature is dense array with float type. -For example, if the input is an image with 28*28 pixels, the input of -Paddle neural network could be a dense vector with dimension 784 or a -numpy array with shape (28, 28).
-For the 2-D convolution operation, each sample in one mini-batch must have -the similarly size in PaddlePaddle now. But, it supports variable-dimension -feature across mini-batch. For the variable-dimension, the param dim is not -used. While the data reader must yield numpy array and the data feeder will -set the data shape correctly.
-| 参数: |
|
-
|---|---|
| 返回: | An input type object. - |
-
| 返回类型: | InputType - |
-
paddle.v2.data_type.dense_vector_sequence(dim)Data type of a sequence of dense vector.
-| 参数: | dim (int) – dimension of dense vector. | -
|---|---|
| 返回: | An input type object | -
| 返回类型: | InputType | -
paddle.v2.data_type.integer_value(value_range, seq_type=0)Data type of integer.
-| 参数: |
|
-
|---|---|
| 返回: | An input type object - |
-
| 返回类型: | InputType - |
-
paddle.v2.data_type.integer_value_sequence(value_range)Data type of a sequence of integer.
-| 参数: | value_range (int) – range of each element. | -
|---|
paddle.v2.data_type.sparse_binary_vector(dim, seq_type=0)Sparse binary vector. It means the input feature is a sparse vector and the -every element in this vector is either zero or one.
-| 参数: |
|
-
|---|---|
| 返回: | An input type object. - |
-
| 返回类型: | InputType - |
-
paddle.v2.data_type.sparse_binary_vector_sequence(dim)| 参数: | dim (int) – dimension of sparse vector. | -
|---|---|
| 返回: | An input type object | -
| 返回类型: | InputType | -
paddle.v2.data_type.sparse_float_vector(dim, seq_type=0)Sparse vector. It means the input feature is a sparse vector. Most of the -elements in this vector are zero, others could be any float value.
-| 参数: |
|
-
|---|---|
| 返回: | An input type object. - |
-
| 返回类型: | InputType - |
-
paddle.v2.data_type.sparse_float_vector_sequence(dim)Data type of a sequence of sparse vector, which most elements are zero, -others could be any float value.
-| 参数: | dim (int) – dimension of sparse vector. | -
|---|---|
| 返回: | An input type object | -
| 返回类型: | InputType | -
paddle.v2.data_type.sparse_non_value_slot(dim, seq_type=0)Sparse binary vector. It means the input feature is a sparse vector and the -every element in this vector is either zero or one.
-| 参数: |
|
-
|---|---|
| 返回: | An input type object. - |
-
| 返回类型: | InputType - |
-
paddle.v2.data_type.sparse_value_slot(dim, seq_type=0)Sparse vector. It means the input feature is a sparse vector. Most of the -elements in this vector are zero, others could be any float value.
-| 参数: |
|
-
|---|---|
| 返回: | An input type object. - |
-
| 返回类型: | InputType - |
-
paddle.v2.data_type.InputType(dim, seq_type, tp)InputType is the base class for paddle input types.
-注解
-this is a base class, and should never be used by user.
-| 参数: |
|
-
|---|
At training and testing time, PaddlePaddle programs need to read data. To ease -the users’ work to write data reading code, we define that
-Indeed, data reader doesn’t have to be a function that reads and yields data
-items. It can be any function with no parameter that creates a iterable
-(anything can be used in for x in iterable):
iterable = data_reader()
-Element produced from the iterable should be a single entry of data, -not a mini batch. That entry of data could be a single item, or a tuple of -items. -Item should be of supported type (e.g., numpy 1d -array of float32, int, list of int)
-An example implementation for single item data reader creator:
-def reader_creator_random_image(width, height):
- def reader():
- while True:
- yield numpy.random.uniform(-1, 1, size=width*height)
-return reader
-An example implementation for multiple item data reader creator:
-def reader_creator_random_image_and_label(width, height, label):
- def reader():
- while True:
- yield numpy.random.uniform(-1, 1, size=width*height), label
-return reader
-TODO(yuyang18): Should we add whole design doc here?
-paddle.v2.reader.map_readers(func, *readers)Creates a data reader that outputs return value of function using -output of each data readers as arguments.
-| 参数: |
|
-
|---|---|
| Type: | callable - |
-
| 返回: | the created data reader. - |
-
| 返回类型: | callable - |
-
paddle.v2.reader.buffered(reader, size)Creates a buffered data reader.
-The buffered data reader will read and save data entries into a -buffer. Reading from the buffered data reader will proceed as long -as the buffer is not empty.
-| 参数: |
|
-
|---|---|
| 返回: | the buffered data reader. - |
-
paddle.v2.reader.compose(*readers, **kwargs)Creates a data reader whose output is the combination of input readers.
-If input readers output following data entries: -(1, 2) 3 (4, 5) -The composed reader will output: -(1, 2, 3, 4, 5)
-| 参数: |
|
-
|---|---|
| 返回: | the new data reader. - |
-
| 引发: | ComposeNotAligned – outputs of readers are not aligned. -Will not raise when check_alignment is set to False. - |
-
paddle.v2.reader.chain(*readers)Creates a data reader whose output is the outputs of input data -readers chained together.
-If input readers output following data entries: -[0, 0, 0] -[1, 1, 1] -[2, 2, 2] -The chained reader will output: -[0, 0, 0, 1, 1, 1, 2, 2, 2]
-| 参数: | readers – input readers. | -
|---|---|
| 返回: | the new data reader. | -
| 返回类型: | callable | -
paddle.v2.reader.shuffle(reader, buf_size)Creates a data reader whose data output is shuffled.
-Output from the iterator that created by original reader will be -buffered into shuffle buffer, and then shuffled. The size of shuffle buffer -is determined by argument buf_size.
-| 参数: |
|
-
|---|---|
| 返回: | the new reader whose output is shuffled. - |
-
| 返回类型: | callable - |
-
paddle.v2.reader.firstn(reader, n)Limit the max number of samples that reader could return.
-| 参数: |
|
-
|---|---|
| 返回: | the decorated reader. - |
-
| 返回类型: | callable - |
-
paddle.v2.reader.xmap_readers(mapper, reader, process_num, buffer_size, order=False)Use multiprocess to map samples from reader by a mapper defined by user. -And this function contains a buffered decorator. -:param mapper: a function to map sample. -:type mapper: callable -:param reader: the data reader to read from -:type reader: callable -:param process_num: process number to handle original sample -:type process_num: int -:param buffer_size: max buffer size -:type buffer_size: int -:param order: keep the order of reader -:type order: bool -:return: the decarated reader -:rtype: callable
-paddle.v2.reader.PipeReader(command, bufsize=8192, file_type='plain')PipeReader read data by stream from a command, take it’s -stdout into a pipe buffer and redirect it to the parser to -parse, then yield data as your desired format.
-You can using standard linux command or call another program -to read data, from HDFS, Ceph, URL, AWS S3 etc:
-An example:
-def example_reader():
- for f in myfiles:
- pr = PipeReader("cat %s"%f)
- for l in pr.get_line():
- sample = l.split(" ")
- yield sample
-get_line(cut_lines=True, line_break='\n')---
-- - - - param cut_lines: - cut buffer to lines -- type cut_lines: bool -- param line_break: - - line break of the file, like -
| type line_break: | |
|---|---|
| string | -|
| return: | one line or a buffer of bytes | -
| rtype: | string | -
Creator package contains some simple reader creator, which could -be used in user program.
-paddle.v2.reader.creator.np_array(x)Creates a reader that yields elements of x, if it is a -numpy vector. Or rows of x, if it is a numpy matrix. -Or any sub-hyperplane indexed by the highest dimension.
-| 参数: | x – the numpy array to create reader from. | -
|---|---|
| 返回: | data reader created from x. | -
paddle.v2.reader.creator.text_file(path)Creates a data reader that outputs text line by line from given text file. -Trailing new line (‘\n’) of each line will be removed.
-| Path: | path of the text file. | -
|---|---|
| 返回: | data reader of text file | -
paddle.v2.reader.creator.cloud_reader(paths, etcd_endpoints, timeout_sec=5, buf_size=64)| Paths: | path of recordio files, can be a string or a string list. | -
|---|---|
| Etcd_endpoints: | the endpoints for etcd cluster | -
| 返回: | data reader of recordio files. | -
paddle.v2.minibatch.batch(reader, batch_size)Create a batched reader.
-| 参数: |
|
-
|---|---|
| 返回: | the batched reader. - |
-
| 返回类型: | callable - |
-
- Dataset package.
-MNIST dataset.
-This module will download dataset from http://yann.lecun.com/exdb/mnist/ and -parse training set and test set into paddle reader creators.
-paddle.v2.dataset.mnist.train()MNIST training set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| 返回: | Training reader creator | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.mnist.test()MNIST test set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| 返回: | Test reader creator. | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.mnist.convert(path)Converts dataset to recordio format
-CIFAR dataset.
-This module will download dataset from -https://www.cs.toronto.edu/~kriz/cifar.html and parse train/test set into -paddle reader creators.
-The CIFAR-10 dataset consists of 60000 32x32 colour images in 10 classes, -with 6000 images per class. There are 50000 training images and 10000 test -images.
-The CIFAR-100 dataset is just like the CIFAR-10, except it has 100 classes -containing 600 images each. There are 500 training images and 100 testing -images per class.
-paddle.v2.dataset.cifar.train100()CIFAR-100 training set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 99].
-| 返回: | Training reader creator | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.cifar.test100()CIFAR-100 test set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| 返回: | Test reader creator. | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.cifar.train10()CIFAR-10 training set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| 返回: | Training reader creator | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.cifar.test10()CIFAR-10 test set creator.
-It returns a reader creator, each sample in the reader is image pixels in -[0, 1] and label in [0, 9].
-| 返回: | Test reader creator. | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.cifar.convert(path)Converts dataset to recordio format
-Conll05 dataset. -Paddle semantic role labeling Book and demo use this dataset as an example. -Because Conll05 is not free in public, the default downloaded URL is test set -of Conll05 (which is public). Users can change URL and MD5 to their Conll -dataset. And a pre-trained word vector model based on Wikipedia corpus is used -to initialize SRL model.
-paddle.v2.dataset.conll05.get_dict()Get the word, verb and label dictionary of Wikipedia corpus.
-paddle.v2.dataset.conll05.get_embedding()Get the trained word vector based on Wikipedia corpus.
-paddle.v2.dataset.conll05.test()Conll05 test set creator.
-Because the training dataset is not free, the test dataset is used for -training. It returns a reader creator, each sample in the reader is nine -features, including sentence sequence, predicate, predicate context, -predicate context flag and tagged sequence.
-| 返回: | Training reader creator | -
|---|---|
| 返回类型: | callable | -
IMDB dataset.
-This module downloads IMDB dataset from -http://ai.stanford.edu/%7Eamaas/data/sentiment/. This dataset contains a set -of 25,000 highly polar movie reviews for training, and 25,000 for testing. -Besides, this module also provides API for building dictionary.
-paddle.v2.dataset.imdb.build_dict(pattern, cutoff)Build a word dictionary from the corpus. Keys of the dictionary are words, -and values are zero-based IDs of these words.
-paddle.v2.dataset.imdb.train(word_idx)IMDB training set creator.
-It returns a reader creator, each sample in the reader is an zero-based ID -sequence and label in [0, 1].
-| 参数: | word_idx (dict) – word dictionary | -
|---|---|
| 返回: | Training reader creator | -
| 返回类型: | callable | -
paddle.v2.dataset.imdb.test(word_idx)IMDB test set creator.
-It returns a reader creator, each sample in the reader is an zero-based ID -sequence and label in [0, 1].
-| 参数: | word_idx (dict) – word dictionary | -
|---|---|
| 返回: | Test reader creator | -
| 返回类型: | callable | -
paddle.v2.dataset.imdb.convert(path)Converts dataset to recordio format
-imikolov’s simple dataset.
-This module will download dataset from -http://www.fit.vutbr.cz/~imikolov/rnnlm/ and parse training set and test set -into paddle reader creators.
-paddle.v2.dataset.imikolov.build_dict(min_word_freq=50)Build a word dictionary from the corpus, Keys of the dictionary are words, -and values are zero-based IDs of these words.
-paddle.v2.dataset.imikolov.train(word_idx, n, data_type=1)imikolov training set creator.
-It returns a reader creator, each sample in the reader is a word ID -tuple.
-| 参数: |
|
-
|---|---|
| 返回: | Training reader creator - |
-
| 返回类型: | callable - |
-
paddle.v2.dataset.imikolov.test(word_idx, n, data_type=1)imikolov test set creator.
-It returns a reader creator, each sample in the reader is a word ID -tuple.
-| 参数: |
|
-
|---|---|
| 返回: | Test reader creator - |
-
| 返回类型: | callable - |
-
paddle.v2.dataset.imikolov.convert(path)Converts dataset to recordio format
-Movielens 1-M dataset.
-Movielens 1-M dataset contains 1 million ratings from 6000 users on 4000 -movies, which was collected by GroupLens Research. This module will download -Movielens 1-M dataset from -http://files.grouplens.org/datasets/movielens/ml-1m.zip and parse training -set and test set into paddle reader creators.
-paddle.v2.dataset.movielens.get_movie_title_dict()Get movie title dictionary.
-paddle.v2.dataset.movielens.max_movie_id()Get the maximum value of movie id.
-paddle.v2.dataset.movielens.max_user_id()Get the maximum value of user id.
-paddle.v2.dataset.movielens.max_job_id()Get the maximum value of job id.
-paddle.v2.dataset.movielens.movie_categories()Get movie categoriges dictionary.
-paddle.v2.dataset.movielens.user_info()Get user info dictionary.
-paddle.v2.dataset.movielens.movie_info()Get movie info dictionary.
-paddle.v2.dataset.movielens.convert(path)Converts dataset to recordio format
-paddle.v2.dataset.movielens.MovieInfo(index, categories, title)Movie id, title and categories information are stored in MovieInfo.
-paddle.v2.dataset.movielens.UserInfo(index, gender, age, job_id)User id, gender, age, and job information are stored in UserInfo.
-The script fetch and preprocess movie_reviews data set that provided by NLTK
-TODO(yuyang18): Complete dataset.
-paddle.v2.dataset.sentiment.get_word_dict()Sorted the words by the frequency of words which occur in sample -:return:
---words_freq_sorted
paddle.v2.dataset.sentiment.train()Default training set reader creator
-paddle.v2.dataset.sentiment.test()Default test set reader creator
-paddle.v2.dataset.sentiment.convert(path)Converts dataset to recordio format
-UCI Housing dataset.
-This module will download dataset from -https://archive.ics.uci.edu/ml/machine-learning-databases/housing/ and -parse training set and test set into paddle reader creators.
-paddle.v2.dataset.uci_housing.train()UCI_HOUSING training set creator.
-It returns a reader creator, each sample in the reader is features after -normalization and price number.
-| 返回: | Training reader creator | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.uci_housing.test()UCI_HOUSING test set creator.
-It returns a reader creator, each sample in the reader is features after -normalization and price number.
-| 返回: | Test reader creator | -
|---|---|
| 返回类型: | callable | -
WMT14 dataset. -The original WMT14 dataset is too large and a small set of data for set is -provided. This module will download dataset from -http://paddlepaddle.cdn.bcebos.com/demo/wmt_shrinked_data/wmt14.tgz and -parse training set and test set into paddle reader creators.
-paddle.v2.dataset.wmt14.train(dict_size)WMT14 training set creator.
-It returns a reader creator, each sample in the reader is source language -word ID sequence, target language word ID sequence and next word ID -sequence.
-| 返回: | Training reader creator | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.wmt14.test(dict_size)WMT14 test set creator.
-It returns a reader creator, each sample in the reader is source language -word ID sequence, target language word ID sequence and next word ID -sequence.
-| 返回: | Test reader creator | -
|---|---|
| 返回类型: | callable | -
paddle.v2.dataset.wmt14.convert(path)Converts dataset to recordio format
-ACL2016 Multimodal Machine Translation. Please see this website for more -details: http://www.statmt.org/wmt16/multimodal-task.html#task1
-If you use the dataset created for your task, please cite the following paper: -Multi30K: Multilingual English-German Image Descriptions.
-}
-paddle.v2.dataset.wmt16.train(src_dict_size, trg_dict_size, src_lang='en')WMT16 train set reader.
-This function returns the reader for train data. Each sample the reader -returns is made up of three fields: the source language word index sequence, -target language word index sequence and next word index sequence.
-NOTE: -The original like for training data is: -http://www.quest.dcs.shef.ac.uk/wmt16_files_mmt/training.tar.gz
-paddle.dataset.wmt16 provides a tokenized version of the original dataset by -using moses’s tokenization script: -https://github.com/moses-smt/mosesdecoder/blob/master/scripts/tokenizer/tokenizer.perl
-| 参数: |
|
-
|---|---|
| 返回: | The train reader. - |
-
| 返回类型: | callable - |
-
paddle.v2.dataset.wmt16.test(src_dict_size, trg_dict_size, src_lang='en')WMT16 test set reader.
-This function returns the reader for test data. Each sample the reader -returns is made up of three fields: the source language word index sequence, -target language word index sequence and next word index sequence.
-NOTE: -The original like for test data is: -http://www.quest.dcs.shef.ac.uk/wmt16_files_mmt/mmt16_task1_test.tar.gz
-paddle.dataset.wmt16 provides a tokenized version of the original dataset by -using moses’s tokenization script: -https://github.com/moses-smt/mosesdecoder/blob/master/scripts/tokenizer/tokenizer.perl
-| 参数: |
|
-
|---|---|
| 返回: | The test reader. - |
-
| 返回类型: | callable - |
-
paddle.v2.dataset.wmt16.validation(src_dict_size, trg_dict_size, src_lang='en')WMT16 validation set reader.
-This function returns the reader for validation data. Each sample the reader -returns is made up of three fields: the source language word index sequence, -target language word index sequence and next word index sequence.
-NOTE: -The original like for validation data is: -http://www.quest.dcs.shef.ac.uk/wmt16_files_mmt/validation.tar.gz
-paddle.dataset.wmt16 provides a tokenized version of the original dataset by -using moses’s tokenization script: -https://github.com/moses-smt/mosesdecoder/blob/master/scripts/tokenizer/tokenizer.perl
-| 参数: |
|
-
|---|---|
| 返回: | The validation reader. - |
-
| 返回类型: | callable - |
-
paddle.v2.dataset.wmt16.get_dict(lang, dict_size, reverse=False)return the word dictionary for the specified language.
-| 参数: |
|
-
|---|---|
| 返回: | The word dictionary for the specific language. - |
-
| 返回类型: | dict - |
-
paddle.v2.dataset.wmt16.fetch()download the entire dataset.
-paddle.v2.dataset.wmt16.convert(path, src_dict_size, trg_dict_size, src_lang)Converts dataset to recordio format.
-
- This file contains some common interfaces for image preprocess. -Many users are confused about the image layout. We introduce -the image layout as follows.
-CHW Layout
-Color format: RGB or BGR
-OpenCV use BGR color format. PIL use RGB color format. Both -formats can be used for training. Noted that, the format should -be keep consistent between the training and inference peroid.
-paddle.v2.image.batch_images_from_tar(data_file, dataset_name, img2label, num_per_batch=1024)¶Read images from tar file and batch them into batch file.
-| 参数: |
|
-
|---|---|
| 返回: | path of list file containing paths of batch file - |
-
| 返回类型: | string - |
-
paddle.v2.image.load_image_bytes(bytes, is_color=True)¶Load an color or gray image from bytes array.
-Example usage:
-with open('cat.jpg') as f:
- im = load_image_bytes(f.read())
-| 参数: |
|
-
|---|
paddle.v2.image.load_image(file, is_color=True)¶Load an color or gray image from the file path.
-Example usage:
-im = load_image('cat.jpg')
-| 参数: |
|
-
|---|
paddle.v2.image.resize_short(im, size)¶Resize an image so that the length of shorter edge is size.
-Example usage:
-im = load_image('cat.jpg')
-im = resize_short(im, 256)
-| 参数: |
|
-
|---|
paddle.v2.image.to_chw(im, order=(2, 0, 1))¶Transpose the input image order. The image layout is HWC format -opened by cv2 or PIL. Transpose the input image to CHW layout -according the order (2,0,1).
-Example usage:
-im = load_image('cat.jpg')
-im = resize_short(im, 256)
-im = to_chw(im)
-| 参数: |
|
-
|---|
paddle.v2.image.center_crop(im, size, is_color=True)¶Crop the center of image with size.
-Example usage:
-im = center_crop(im, 224)
-| 参数: |
|
-
|---|
paddle.v2.image.random_crop(im, size, is_color=True)¶Randomly crop input image with size.
-Example usage:
-im = random_crop(im, 224)
-| 参数: |
|
-
|---|
paddle.v2.image.left_right_flip(im, is_color=True)¶Flip an image along the horizontal direction. -Return the flipped image.
-Example usage:
-im = left_right_flip(im)
-| 参数: |
|
-
|---|
paddle.v2.image.simple_transform(im, resize_size, crop_size, is_train, is_color=True, mean=None)¶Simply data argumentation for training. These operations include -resizing, croping and flipping.
-Example usage:
-im = simple_transform(im, 256, 224, True)
-| 参数: |
|
-
|---|
paddle.v2.image.load_and_transform(filename, resize_size, crop_size, is_train, is_color=True, mean=None)¶Load image from the input file filename and transform image for -data argumentation. Please refer to the simple_transform interface -for the transform operations.
-Example usage:
-im = load_and_transform('cat.jpg', 256, 224, True)
-| 参数: |
|
-
|---|
-
-
- paddle.v2.fluid.evaluator.Accuracy(input, label, k=1, **kwargs)Average Accuracy for multiple mini-batches.
-paddle.v2.fluid.evaluator.ChunkEvaluator(input, label, chunk_scheme, num_chunk_types, excluded_chunk_types=None)Accumulate counter numbers output by chunk_eval from mini-batches and -compute the precision recall and F1-score using the accumulated counter -numbers.
-
- paddle.v2.fluid.executor.Executor(places)paddle.v2.fluid.executor.global_scope()paddle.v2.fluid.executor.scope_guard(*args, **kwds)paddle.v2.fluid.executor.switch_scope(scope)
- paddle.v2.fluid.initializer.ConstantConstantInitializer 的别名
paddle.v2.fluid.initializer.UniformUniformInitializer 的别名
paddle.v2.fluid.initializer.NormalNormalInitializer 的别名
paddle.v2.fluid.initializer.XavierXavierInitializer 的别名
- paddle.v2.fluid.io.save_vars(executor, dirname, main_program=None, vars=None, predicate=None, save_file_name=None)Save variables to directory by executor.
-| 参数: |
|
-
|---|
program which fit predicate. Default default_main_program. -:param predicate: The Predicate describes a callable that returns a variable -as a bool. If it returns true, the corresponding input variable will be saved. -:param vars: variables need to be saved. If vars is specified, program & predicate -will be ignored -:param save_file_name: The name of a single file that all vars are saved to. -If it is None, save variables to separate files.
-| 返回: | None | -
|---|
paddle.v2.fluid.io.save_params(executor, dirname, main_program=None, save_file_name=None)Save all parameters to directory with executor.
-paddle.v2.fluid.io.save_persistables(executor, dirname, main_program=None, save_file_name=None)Save all persistables to directory with executor.
-paddle.v2.fluid.io.load_vars(executor, dirname, main_program=None, vars=None, predicate=None, load_file_name=None)Load variables from directory by executor.
-| 参数: |
|
-
|---|
program which fit predicate. Default default_main_program(). -:param predicate: The Predicate describes a callable that returns a variable -as a bool. If it returns true, the corresponding input variable will be loaded. -:param vars: variables need to be loaded. If vars is specified, program & -predicate will be ignored -:param load_file_name: The name of the single file that all vars are loaded from. -If it is None, load variables from separate files.
-| 返回: | None | -
|---|
paddle.v2.fluid.io.load_params(executor, dirname, main_program=None, load_file_name=None)load all parameters from directory by executor.
-paddle.v2.fluid.io.load_persistables(executor, dirname, main_program=None, load_file_name=None)load all persistables from directory by executor.
-paddle.v2.fluid.io.save_inference_model(dirname, feeded_var_names, target_vars, executor, main_program=None, save_file_name=None)Build a model especially for inference, -and save it to directory by the executor.
-| 参数: |
|
-
|---|
If it is None, save parameters to separate files.
-| 返回: | None | -
|---|
paddle.v2.fluid.io.load_inference_model(dirname, executor, load_file_name=None)Load inference model from a directory
-| 参数: |
|
-
|---|
If it is None, load parameters from separate files.
-| 返回: | [program, feed_target_names, fetch_targets] -program: program especially for inference. -feed_target_names: Names of variables that need to feed data -fetch_targets: Variables from which we can get inference results. | -
|---|
paddle.v2.fluid.io.get_inference_program(target_vars, main_program=None)
- paddle.v2.fluid.layers.split_lod_tensor(input, mask, level=0)split_lod_tensor
-This function takes in an input that contains the complete lod information, -and takes in a mask which is used to mask certain parts of the input. -The output is the true branch and the false branch with the mask applied to -the input at a certain level in the tensor.
-| 参数: |
|
-
|---|---|
| 返回: | The true branch of tensor as per the mask applied to input. -Variable: The false branch of tensor as per the mask applied to input. - |
-
| 返回类型: | Variable - |
-
Examples
-x = layers.data(name='x', shape=[1])
-x.persistable = True
-
-y = layers.data(name='y', shape=[1])
-y.persistable = True
-
-out_true, out_false = layers.split_lod_tensor(
- input=x, mask=y, level=level)
-paddle.v2.fluid.layers.merge_lod_tensor(in_true, in_false, x, mask, level=0)merge_lod_tensor
-This function takes in an input \(x\), the True branch, the False -branch and a binary \(mask\). Using this information, this function -merges the True and False branches of the tensor into a single Output -at a certain lod level indiacted by \(level\).
-| 参数: |
|
-
|---|---|
| 返回: | The merged output tensor. - |
-
| 返回类型: | Variable - |
-
Examples
-x = layers.data(
- name='x', shape=[1], dtype='float32', stop_gradient=False)
-y = layers.data(
- name='y', shape=[1], dtype='bool', stop_gradient=False)
-
-level = 0
-
-out_true, out_false = layers.split_lod_tensor(
- input=x, mask=y, level=level)
-out = layers.merge_lod_tensor(
- in_true=out_true, in_false=out_false, mask=y, x=x, level=level)
-paddle.v2.fluid.layers.BlockGuard(main_program)BlockGuard class.
-BlockGuard class is used to create a sub-block in a program by -using the Python with keyword.
-paddle.v2.fluid.layers.BlockGuardWithCompletion(rnn)BlockGuardWithCompletion class.
-BlockGuardWithCompletion class is used to create an op with a block in a program.
-paddle.v2.fluid.layers.StaticRNNMemoryLink(init, pre_mem, mem=None)StaticRNNMemoryLink class.
-| 参数: |
|
-
|---|
StaticRNNMemoryLink class is used to create a link between two -memory cells of a StaticRNN.
-paddle.v2.fluid.layers.WhileGuard(while_op)paddle.v2.fluid.layers.While(cond, name=None)paddle.v2.fluid.layers.lod_rank_table(x, level=0)LoD Rank Table Operator. Given an input variable x and a level number -of LoD, this layer creates a LodRankTable object. A LoDRankTable object -contains a list of bi-element tuples. Each tuple consists of an index and -a length, both of which are int type. Refering to specified level of LoD, -the index is the sequence index number and the length representes the -sequence length. Please note that the list is ranked in descending order by -the length. The following is an example:
-----x is a LoDTensor: - x.lod = [[0, 2, 3], - [0, 5, 6, 7]] - x.data = [a, b, c, d, e, f, g] - -1. set level to 0: - Create lod rank table: - lod_rank_table_obj = lod_rank_table(x, level=0) - - Get: - lod_rank_table_obj.items() = [(0, 2), (1, 1)] - -2. set level to 1: - Create lod rank table: - lod_rank_table_obj = lod_rank_table(x, level=1) - - Get: - lod_rank_table_obj.items() = [(0, 5), (1, 1), (2, 1)] -
| 参数: |
|
-
|---|---|
| 返回: | The created LoDRankTable object. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10],
- dtype='float32', lod_level=1)
-out = layers.lod_rank_table(x=x, level=0)
-paddle.v2.fluid.layers.max_sequence_len(rank_table)Max Sequence Len Operator. Given a LoDRankTable object, this layer -returns the max length of a batch of sequences. In fact, a LoDRankTable -object contains a list of tuples(<sequence index, sequence length>) and -the list is already sorted by sequence length in descending order, so the -operator just returns the sequence length of the first tuple element.
-| 参数: | rank_table (Variable) – Input variable which is a LoDRankTable object. | -
|---|---|
| 返回: | The max length of sequence. | -
| 返回类型: | Variable | -
Examples
-x = fluid.layers.data(name='x', shape=[10],
- dtype='float32', lod_level=1)
-rank_table = layers.lod_rank_table(x=x, level=0)
-max_seq_len = layers.max_sequence_len(rank_table)
-paddle.v2.fluid.layers.topk(input, k)topk
-This function performs the operation that selects the k entries in the input -vector and outputs their values and indices as vectors. Thus topk_out[j] is -the j-th largest entry in input, and its index is topk_indices[j]
-| 参数: |
|
-
|---|---|
| 返回: |
|
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10])
-k = 5
-array = fluid.layers.topk(x, k)
-paddle.v2.fluid.layers.lod_tensor_to_array(x, table)Convert a LOD_TENSOR to an LOD_TENSOR_ARRAY.
-| 参数: |
|
-
|---|---|
| 返回: |
|
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10])
-table = fluid.layers.lod_rank_table(x, level=0)
-array = fluid.layers.lod_tensor_to_array(x, table)
-paddle.v2.fluid.layers.array_to_lod_tensor(x, table)Convert a LoD_Tensor_Aarry to an LoDTensor.
-| 参数: |
|
-
|---|---|
| 返回: |
|
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10])
-table = fluid.layers.lod_rank_table(x, level=0)
-array = fluid.layers.lod_tensor_to_array(x, table)
-lod_tensor = fluid.layers.array_to_lod_tensor(array, table)
-paddle.v2.fluid.layers.increment(x, value=1.0, in_place=True)This function performs an operation that increments each value in the -input \(x\) by an amount: \(value\) as mentioned in the input -parameter. This operation is performed in-place by default.
-| 参数: |
|
-
|---|---|
| 返回: |
|
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.data(name='data', shape=[32, 32], dtype='float32')
-data = fluid.layers.increment(x=data, value=3.0, in_place=True)
-paddle.v2.fluid.layers.array_write(x, i, array=None)This function writes the given input variable to the specified position -indicating by the arrary index to an output LOD_TENSOR_ARRAY. If the -output LOD_TENSOR_ARRAY is not given(None), a new one will be created and -returned.
-| 参数: |
|
-
|---|---|
| 返回: | The output LOD_TENSOR_ARRAY where the input tensor is written. - |
-
| 返回类型: | Variable - |
-
Examples
-paddle.v2.fluid.layers.create_array(dtype)This function creates an array of type \(LOD_TENSOR_ARRAY\) using the -LayerHelper.
-| 参数: | dtype (int|float) – The data type of the elements in the array. | -
|---|---|
| 返回: | The tensor variable storing the elements of data type. | -
| 返回类型: | Variable | -
Examples
-data = fluid.layers.create_array(dtype='float32')
-paddle.v2.fluid.layers.less_than(x, y, cond=None, **ignored)Less than
-This layer returns the truth value of \(x < y\) elementwise.
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the output of less_than. - |
-
| 返回类型: | Variable - |
-
Examples
-less = fluid.layers.less_than(x=label, y=limit)
-paddle.v2.fluid.layers.array_read(array, i)This function performs the operation to read the data in as an -LOD_TENSOR_ARRAY. -:param array: The input tensor that will be written to an array. -:type array: Variable|list -:param i: The subscript index in tensor array, that points the
---place where data will be written to.
| 返回: | The tensor type variable that has the data written to it. | -
|---|---|
| 返回类型: | Variable | -
Examples
-paddle.v2.fluid.layers.shrink_memory(x, i, table)This function creates an operator to shrink_rnn_memory using the RankTable -as mentioned in the input parameter.
-paddle.v2.fluid.layers.array_length(array)This function performs the operation to find the length of the input -LOD_TENSOR_ARRAY.
-| 参数: | array (LOD_TENSOR_ARRAY) – The input array that will be used -to compute the length. | -
|---|---|
| 返回: | The length of the input LoDTensorArray. | -
| 返回类型: | Variable | -
Examples
-paddle.v2.fluid.layers.IfElse(cond, name=None)paddle.v2.fluid.layers.DynamicRNN(name=None)paddle.v2.fluid.layers.ConditionalBlock(inputs, is_scalar_condition=False, name=None)paddle.v2.fluid.layers.StaticRNN(name=None)StaticRNN class.
-StaticRNN class is used to create a StaticRNN. The RNN will have its -own parameters like inputs, outputs, memories, status and length.
-memory(init=None, shape=None, batch_ref=None, init_value=0.0, init_batch_dim_idx=0, ref_batch_dim_idx=1)| 参数: |
|
-
|---|
paddle.v2.fluid.layers.reorder_lod_tensor_by_rank(x, rank_table)ReorderLoDTensorByRankTable operator.
-Input(X) is a batch of sequences. Input(RankTable) stores new orders of the -input sequence batch. The reorder_lod_tensor_by_rank operator reorders the -Input(X) according to the information provided by Input(RankTable).
-For example:
-If the indices stored in the Input(RankTable) are [3, 0, 2, 1], the -Input(X) will be reordered that the fourth sequence in Input(X) will become the -first one, and then followed by the original first, third, and the second one.
-This is: -X = [Seq0, Seq1, Seq2, Seq3]. The indices in RankTable are [3, 0, 2, 1]. -Out = [Seq3, Seq0, Seq2, Seq1] with a new LoD information.
-If the LoD information of Input(X) is empty, this means Input(X) is not sequence -data. This is also identical to a batch of sequences where each sequence has a -fixed length 1. In this case, the reorder_lod_tensor_by_rank operator reorders -each slice of Input(X) along the first axis according to Input(RankTable).
-This is: -X = [Slice0, Slice1, Slice2, Slice3] and its LoD information is empty. The -indices in RankTable are [3, 0, 2, 1]. -Out = [Slice3, Slice0, Slice2, Slice1] with no LoD information is appended.
-NOTE: This operator sorts Input(X) according to a given LoDRankTable which does -not need to be calculated according to Input(X). It can be calculated according -to another different sequence, and then this operator sorts Input(X) according -to the given LoDRankTable.
-| 参数: |
|
-
|---|---|
| 返回: | (LoDTensor), the reordered lod tensor. - |
-
paddle.v2.fluid.layers.ParallelDo(places, name=None)ParallelDo class.
-ParallelDo class is used to create a ParallelDo.
-paddle.v2.fluid.layers.Print(input, first_n=-1, message=None, summarize=-1, print_tensor_name=True, print_tensor_type=True, print_tensor_shape=True, print_tensor_lod=True, print_phase='both')Print operator
-This creates a print op that will print when a tensor is accessed.
-Wraps the tensor passed in so that whenever that a tensor is accessed, -the message message is printed, along with the current value of the -tensor t.
-| 参数: |
|
-
|---|---|
| 返回: | Output tensor, same data with input tensor. - |
-
| 返回类型: | Variable - |
-
Examples
-
-value = some_layer(...) -Print(value, summarize=10,
---message=”The content of some_layer: ”)
paddle.v2.fluid.layers.get_places(device_count=None, device_type=None)Returns a list of places based on flags. The list will be used for parallel -execution.
-| 参数: |
|
-
|---|---|
| 返回: | vector of Place - |
-
paddle.v2.fluid.layers.data(name, shape, append_batch_size=True, dtype='float32', lod_level=0, type=VarType.LOD_TENSOR, stop_gradient=True)Data Layer
-This function takes in the input and based on whether data has -to be returned back as a minibatch, it creates the global variable by using -the helper functions. The global variables can be accessed by all the -following operators in the graph.
-All the input variables of this function are passed in as local variables -to the LayerHelper constructor.
-| 参数: |
|
-
|---|---|
| 返回: | The global variable that gives access to the data. - |
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.data(name='x', shape=[784], dtype='float32')
-paddle.v2.fluid.layers.BlockGuardServ(server)BlockGuardServ class.
-BlockGuardServ class is used to create an op with a block in a program.
-paddle.v2.fluid.layers.ListenAndServ(endpoint, fan_in=1, optimizer_mode=True)ListenAndServ class.
-ListenAndServ class is used to wrap listen_and_serv op to create a server -which can receive variables from clients and run a block.
-paddle.v2.fluid.layers.Send(endpoints, send_vars, get_vars)Send layer
-| 参数: |
|
-
|---|
Send variables to the server side, and get vars from server -side when server have finished running server side program.
-paddle.v2.fluid.layers.fc(input, size, num_flatten_dims=1, param_attr=None, bias_attr=None, act=None, name=None)Fully Connected Layer
-The fully connected layer can take multiple tensors as its inputs. It -creates a variable (one for each input tensor) called weights for each -input tensor, which represents a fully connected weight matrix from -each input unit to each output unit. The fully connected layer -multiplies each input tensor with its coresponding weight to produce -an output Tensor. If multiple input tensors are given, the results of -multiple multiplications will be sumed up. If bias_attr is not None, -a biases variable will be created and added to the output. Finally, -if activation is not None, it will be applied to the output as well.
-This process can be formulated as follows:
-In the above equation:
-| 参数: |
|
-
|---|---|
| 返回: | The output tensor variable. - |
-
| 返回类型: | Variable - |
-
| Raises: |
|
-
Examples
-data = fluid.layers.data(name="data", shape=[32, 32], dtype="float32")
-fc = fluid.layers.fc(input=data, size=1000, act="tanh")
-paddle.v2.fluid.layers.embedding(input, size, is_sparse=False, padding_idx=None, param_attr=None, dtype='float32')Embedding Layer
-This layer is used to lookup embeddings of IDs, provided by input, in
-a lookup table. The result of this lookup is the embedding of each ID in the
-input.
All the input variables are passed in as local variables to the LayerHelper -constructor.
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the embeddings of the supplied inputs. - |
-
| 返回类型: | Variable - |
-
Examples
-dict_size = len(dataset.ids)
-data = fluid.layers.data(name='ids', shape=[32, 32], dtype='float32')
-fc = fluid.layers.embedding(input=data, size=[dict_size, 16])
-paddle.v2.fluid.layers.dynamic_lstm(input, size, param_attr=None, bias_attr=None, use_peepholes=True, is_reverse=False, gate_activation='sigmoid', cell_activation='tanh', candidate_activation='tanh', dtype='float32', name=None)Dynamic LSTM Layer
-The defalut implementation is diagonal/peephole connection -(https://arxiv.org/pdf/1402.1128.pdf), the formula is as follows:
-where the \(W\) terms denote weight matrices (e.g. \(W_{xi}\) is -the matrix of weights from the input gate to the input), \(W_{ic}, W_{fc}, W_{oc}\) are diagonal weight matrices for peephole connections. In -our implementation, we use vectors to reprenset these diagonal weight -matrices. The \(b\) terms denote bias vectors (\(b_i\) is the input -gate bias vector), \(\sigma\) is the non-linear activations, such as -logistic sigmoid function, and \(i, f, o\) and \(c\) are the input -gate, forget gate, output gate, and cell activation vectors, respectively, -all of which have the same size as the cell output activation vector \(h\).
-The \(\odot\) is the element-wise product of the vectors. \(act_g\) -and \(act_h\) are the cell input and cell output activation functions -and tanh is usually used for them. \(\tilde{c_t}\) is also called -candidate hidden state, which is computed based on the current input and -the previous hidden state.
-Set use_peepholes to False to disable peephole connection. The formula -is omitted here, please refer to the paper -http://www.bioinf.jku.at/publications/older/2604.pdf for details.
-Note that these \(W_{xi}x_{t}, W_{xf}x_{t}, W_{xc}x_{t}, W_{xo}x_{t}\) -operations on the input \(x_{t}\) are NOT included in this operator. -Users can choose to use fully-connect layer before LSTM layer.
-| 参数: |
|
-
|---|---|
| 返回: | The hidden state, and cell state of LSTM. The shape of both is (T x D), and lod is the same with the input. - |
-
| 返回类型: | tuple - |
-
Examples
-hidden_dim = 512
-forward_proj = fluid.layers.fc(input=input_seq, size=hidden_dim * 4,
- act=None, bias_attr=None)
-forward, _ = fluid.layers.dynamic_lstm(
- input=forward_proj, size=hidden_dim * 4, use_peepholes=False)
-paddle.v2.fluid.layers.dynamic_lstmp(input, size, proj_size, param_attr=None, bias_attr=None, use_peepholes=True, is_reverse=False, gate_activation='sigmoid', cell_activation='tanh', candidate_activation='tanh', proj_activation='tanh', dtype='float32', name=None)Dynamic LSTMP Layer
-LSTMP (LSTM with recurrent projection) layer has a separate projection -layer after the LSTM layer, projecting the original hidden state to a -lower-dimensional one, which is proposed to reduce the number of total -parameters and furthermore computational complexity for the LSTM, -espeacially for the case that the size of output units is relative -large (https://research.google.com/pubs/archive/43905.pdf).
-The formula is as follows:
-In the above formula:
-Set use_peepholes to False to disable peephole connection. The formula -is omitted here, please refer to the paper -http://www.bioinf.jku.at/publications/older/2604.pdf for details.
-Note that these \(W_{xi}x_{t}, W_{xf}x_{t}, W_{xc}x_{t}, W_{xo}x_{t}\) -operations on the input \(x_{t}\) are NOT included in this operator. -Users can choose to use fully-connected layer before LSTMP layer.
-| 参数: |
|
-
|---|---|
| 返回: | The projection of hidden state, and cell state of LSTMP. The shape of projection is (T x P), for the cell state which is (T x D), and both LoD is the same with the input. - |
-
| 返回类型: | tuple - |
-
Examples
-hidden_dim, proj_dim = 512, 256
-fc_out = fluid.layers.fc(input=input_seq, size=hidden_dim * 4,
- act=None, bias_attr=None)
-proj_out, _ = fluid.layers.dynamic_lstmp(input=fc_out,
- size=hidden_dim * 4,
- proj_size=proj_dim,
- use_peepholes=False,
- is_reverse=True,
- cell_activation="tanh",
- proj_activation="tanh")
-paddle.v2.fluid.layers.dynamic_gru(input, size, param_attr=None, bias_attr=None, is_reverse=False, gate_activation='sigmoid', candidate_activation='tanh', h_0=None)Dynamic GRU Layer
-Refer to Empirical Evaluation of Gated Recurrent Neural Networks on -Sequence Modeling
-The formula is as follows:
-The \(\odot\) is the element-wise product of the vectors. \(act_g\) -is the update gate and reset gate activation function and \(sigmoid\) -is usually used for it. \(act_c\) is the activation function for -candidate hidden state and \(tanh\) is usually used for it.
-Note that these \(W_{ux}x_{t}, W_{rx}x_{t}, W_{cx}x_{t}\) operations on -the input \(x_{t}\) are NOT included in this operator. Users can choose -to use fully-connect layer before GRU layer.
-| 参数: |
|
-
|---|---|
| 返回: | The hidden state of GRU. The shape is \((T \times D)\), and lod is the same with the input. - |
-
| 返回类型: | Variable - |
-
Examples
-hidden_dim = 512
-x = fluid.layers.fc(input=data, size=hidden_dim * 3)
-hidden = fluid.layers.dynamic_gru(input=x, dim=hidden_dim)
-paddle.v2.fluid.layers.gru_unit(input, hidden, size, weight=None, bias=None, activation='tanh', gate_activation='sigmoid')GRU unit layer. The equation of a gru step is:
----\[ \begin{align}\begin{aligned}u_t & = actGate(xu_{t} + W_u h_{t-1} + b_u)\\r_t & = actGate(xr_{t} + W_r h_{t-1} + b_r)\\m_t & = actNode(xm_t + W_c dot(r_t, h_{t-1}) + b_m)\\h_t & = dot((1-u_t), m_t) + dot(u_t, h_{t-1})\end{aligned}\end{align} \]-
The inputs of gru unit includes \(z_t\), \(h_{t-1}\). In terms -of the equation above, the \(z_t\) is split into 3 parts - -\(xu_t\), \(xr_t\) and \(xm_t\). This means that in order to -implement a full GRU unit operator for an input, a fully -connected layer has to be applied, such that \(z_t = W_{fc}x_t\).
-The terms \(u_t\) and \(r_t\) represent the update and reset gates -of the GRU cell. Unlike LSTM, GRU has one lesser gate. However, there is -an intermediate candidate hidden output, which is denoted by \(m_t\). -This layer has three outputs \(h_t\), \(dot(r_t, h_{t-1})\) -and concatenation of \(u_t\), \(r_t\) and \(m_t\).
-| 参数: |
|
-
|---|---|
| 返回: | The hidden value, reset-hidden value and gate values. - |
-
| 返回类型: | tuple - |
-
Examples
-# assuming we have x_t_data and prev_hidden of size=10
-x_t = fluid.layers.fc(input=x_t_data, size=30)
-hidden_val, r_h_val, gate_val = fluid.layers.gru_unit(input=x_t,
- hidden = prev_hidden)
-paddle.v2.fluid.layers.linear_chain_crf(input, label, param_attr=None)paddle.v2.fluid.layers.crf_decoding(input, param_attr, label=None)paddle.v2.fluid.layers.cos_sim(X, Y, **kwargs)This function performs the cosine similarity between two tensors -X and Y and returns that as the output.
-paddle.v2.fluid.layers.cross_entropy(input, label, **kwargs)Cross Entropy Layer
-This layer computes the cross entropy between input and label. It -supports both standard cross-entropy and soft-label cross-entropy loss -computation.
-soft_label = False, Label[i, 0] indicates the class index for sample i:
-soft_label = True, Label[i, j] indicates the soft label of class j -for sample i:
-Please make sure that in this case the summation of each row of label -equals one.
-As a special case of 2), when each row of ‘label’ has only one -non-zero element which is equal to 1, soft-label cross-entropy degenerates -to a one-hot cross-entropy with one-hot label representation.
-| 参数: |
|
-
|---|---|
| 返回: | A 2-D tensor with shape [N x 1], the cross entropy loss. - |
-
| Raises: | ValueError – 1) the 1st dimension of input and label are not equal. -2) when soft_label == True, and the 2nd dimension of ---
|
-
Examples
-predict = fluid.layers.fc(input=net, size=classdim, act='softmax')
-cost = fluid.layers.cross_entropy(input=predict, label=label)
-paddle.v2.fluid.layers.square_error_cost(input, label, **kwargs)Square error cost layer
-This layer accepts input predictions and target label and returns the -squared error cost.
-For predictions, \(X\), and target labels, \(Y\), the equation is:
-In the above equation:
----
-- \(X\): Input predictions, a tensor.
-- \(Y\): Input labels, a tensor.
-- \(Out\): Output value, same shape with \(X\).
-
| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the element-wise squared error difference of input and label. - |
-
| 返回类型: | Variable - |
-
Examples
-y = layers.data(name='y', shape=[1], dtype='float32')
-y_predict = layers.data(name='y_predict', shape=[1], dtype='float32')
-cost = layers.square_error_cost(input=y_predict, label=y)
-paddle.v2.fluid.layers.accuracy(input, label, k=1, correct=None, total=None, **kwargs)This function computes the accuracy using the input and label. -The output is the top_k inputs and their indices.
-paddle.v2.fluid.layers.chunk_eval(input, label, chunk_scheme, num_chunk_types, excluded_chunk_types=None, **kwargs)This function computes and outputs the precision, recall and -F1-score of chunk detection.
-paddle.v2.fluid.layers.sequence_conv(input, num_filters, filter_size=3, filter_stride=1, padding=None, bias_attr=None, param_attr=None, act=None)This function creates the op for sequence_conv, using the inputs and -other convolutional configurations for the filters and stride as given -in the input parameters to the function.
-paddle.v2.fluid.layers.conv2d(input, num_filters, filter_size, stride=None, padding=None, groups=None, param_attr=None, bias_attr=None, use_cudnn=True, act=None)Convlution2D Layer
-The convolution2D layer calculates the output based on the input, filter -and strides, paddings, dilations, groups parameters. Input(Input) and -Output(Output) are in NCHW format. Where N is batch size, C is the number of -channels, H is the height of the feature, and W is the width of the feature. -The details of convolution layer, please refer UFLDL’s convolution, . -If bias attribution and activation type are provided, bias is added to the -output of the convolution, and the corresponding activation function is -applied to the final result.
-For each input \(X\), the equation is:
-In the above equation:
-Example
-Input:
-Input shape: $(N, C_{in}, H_{in}, W_{in})$
-Filter shape: $(C_{out}, C_{in}, H_f, W_f)$
-Output: -Output shape: $(N, C_{out}, H_{out}, W_{out})$
-Where
-H_{out}&= frac{(H_{in} + 2 * paddings[0] - (dilations[0] * (H_f - 1) + 1))}{strides[0]} + 1 \ -W_{out}&= frac{(W_{in} + 2 * paddings[1] - (dilations[1] * (W_f - 1) + 1))}{strides[1]} + 1
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the convolution and non-linearity activation result. - |
-
| 返回类型: | Variable - |
-
| Raises: |
|
-
Examples
-data = fluid.layers.data(
- name='data', shape=[3, 32, 32], dtype='float32')
-conv2d = fluid.layers.conv2d(
- input=data, num_filters=2, filter_size=3, act="relu")
-paddle.v2.fluid.layers.sequence_pool(input, pool_type, **kwargs)This function add the operator for sequence pooling. -It pools features of all time-steps of each instance, and is applied -on top of the input using pool_type mentioned in the parameters.
-It supports four pool_type:
-x is a 1-level LoDTensor:
- x.lod = [[0, 2, 5, 7]]
- x.data = [1, 3, 2, 4, 6, 5, 1]
- x.dims = [7, 1]
-
-then output is a Tensor:
- out.dim = [3, 1]
- with condition len(x.lod[-1]) - 1 == out.dims[0]
-
-for different pool_type:
- average: out.data = [2, 4, 3], where 2=(1+3)/2, 4=(2+4+6)/3, 3=(5+1)/2
- sum : out.data = [4, 12, 6], where 4=1+3, 12=2+4+6, 6=5+1
- sqrt : out.data = [2.82, 6.93, 4.24], where 2.82=(1+3)/sqrt(2),
- 6.93=(2+4+6)/sqrt(3), 4.24=(5+1)/sqrt(2)
- max : out.data = [3, 6, 5], where 3=max(1,3), 6=max(2,4,6), 5=max(5,1)
-| 参数: |
|
-
|---|---|
| 返回: | The sequence pooling variable which is a Tensor. - |
-
Examples
-x = fluid.layers.data(name='x', shape=[7, 1],
- dtype='float32', lod_level=1)
-avg_x = fluid.layers.sequence_pool(input=x, pool_type='average')
-sum_x = fluid.layers.sequence_pool(input=x, pool_type='sum')
-sqrt_x = fluid.layers.sequence_pool(input=x, pool_type='sqrt')
-max_x = fluid.layers.sequence_pool(input=x, pool_type='max')
-paddle.v2.fluid.layers.pool2d(input, pool_size, pool_type, pool_stride=None, pool_padding=None, global_pooling=False, use_cudnn=True, name=None)This function adds the operator for pooling in 2 dimensions, using the -pooling configurations mentioned in input parameters.
-paddle.v2.fluid.layers.batch_norm(input, act=None, is_test=False, momentum=0.9, epsilon=1e-05, param_attr=None, bias_attr=None, data_layout='NCHW', name=None, moving_mean_name=None, moving_variance_name=None)This function helps create an operator to implement -the BatchNorm layer using the configurations from the input parameters.
-paddle.v2.fluid.layers.layer_norm(input, scale=True, shift=True, begin_norm_axis=1, epsilon=1e-05, param_attr=None, bias_attr=None, act=None, name=None)Layer Normalization
-Assume feature vectors exist on dimensions
-begin_norm_axis ... rank(input) and calculate the moment statistics
-along these dimensions for each feature vector \(a\) with size
-\(H\), then normalize each feature vector using the corresponding
-statistics. After that, apply learnable gain and bias on the normalized
-tensor to scale and shift if scale and shift are set.
Refer to Layer Normalization
-The formula is as follows:
-| 参数: |
|
-
|---|---|
| 返回: | A tensor variable with the same shape as the input. - |
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.data(
- name='data', shape=[3, 32, 32], dtype='float32')
-x = fluid.layers.layer_norm(input=data, begin_norm_axis=1)
-paddle.v2.fluid.layers.beam_search_decode(ids, scores, name=None)paddle.v2.fluid.layers.conv2d_transpose(input, num_filters, output_size=None, filter_size=None, padding=None, stride=None, dilation=None, param_attr=None, use_cudnn=True, name=None)Convlution2D transpose layer
-The convolution2D transpose layer calculates the output based on the input, -filter, and dilations, strides, paddings. Input(Input) and output(Output) -are in NCHW format. Where N is batch size, C is the number of channels, -H is the height of the feature, and W is the width of the feature. -Parameters(dilations, strides, paddings) are two elements. These two elements -represent height and width, respectively. The details of convolution transpose -layer, please refer to the following explanation and references -therein.
-For each input \(X\), the equation is:
-In the above equation:
-Example
-Input:
-Input shape: $(N, C_{in}, H_{in}, W_{in})$
-Filter shape: $(C_{in}, C_{out}, H_f, W_f)$
-Output:
-Output shape: $(N, C_{out}, H_{out}, W_{out})$
-Where
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the convolution transpose result. - |
-
| 返回类型: | Variable - |
-
| Raises: |
|
-
Examples
-data = fluid.layers.data(
- name='data', shape=[3, 32, 32], dtype='float32')
-conv2d_transpose = fluid.layers.conv2d_transpose(
- input=data, num_filters=2, filter_size=3)
-paddle.v2.fluid.layers.sequence_expand(x, y, name=None)Sequence Expand Layer. This layer will expand the input variable x -according to LoD information of y. And the following examples will -explain how sequence_expand works:
-* Case 1
- x is a LoDTensor:
- x.lod = [[0, 2, 3],
- [0, 1, 3, 4]]
- x.data = [a, b, c, d]
- x.dims = [4, 1]
-
- y is a LoDTensor:
- y.lod = [[0, 2, 4],
- [0, 3, 6, 7, 8]]
-
- with condition len(y.lod[-1]) - 1 == x.dims[0]
-
- then output is a 2-level LoDTensor:
- out.lod = [[0, 2, 4],
- [0, 3, 6, 7, 8]]
- out.data = [a, a, a, b, b, b, c, d]
- out.dims = [8, 1]
-
-* Case 2
- x is a Tensor:
- x.data = [a, b, c]
- x.dims = [3, 1]
-
- y is a LoDTensor:
- y.lod = [[0, 2, 3, 6]]
-
- with condition len(y.lod[-1]) - 1 == x.dims[0]
-
- then output is a 1-level LoDTensor:
- out.lod = [[0, 2, 3, 6]]
- out.data = [a, a, b, c, c, c]
- out.dims = [6, 1]
-| 参数: |
|
-
|---|---|
| 返回: | The expanded variable which is a LoDTensor. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[10], dtype='float32')
-y = fluid.layers.data(name='y', shape=[10, 20],
- dtype='float32', lod_level=1)
-out = layers.sequence_expand(x=x, y=y)
-paddle.v2.fluid.layers.lstm_unit(x_t, hidden_t_prev, cell_t_prev, forget_bias=0.0, param_attr=None, bias_attr=None, name=None)Lstm unit layer. The equation of a lstm step is:
----\[ \begin{align}\begin{aligned}i_t & = \sigma(W_{x_i}x_{t} + W_{h_i}h_{t-1} + b_i)\\f_t & = \sigma(W_{x_f}x_{t} + W_{h_f}h_{t-1} + b_f)\\c_t & = f_tc_{t-1} + i_t tanh (W_{x_c}x_t + W_{h_c}h_{t-1} + b_c)\\o_t & = \sigma(W_{x_o}x_{t} + W_{h_o}h_{t-1} + b_o)\\h_t & = o_t tanh(c_t)\end{aligned}\end{align} \]-
The inputs of lstm unit include \(x_t\), \(h_{t-1}\) and -\(c_{t-1}\). The 2nd dimensions of \(h_{t-1}\) and \(c_{t-1}\) -should be same. The implementation separates the linear transformation and -non-linear transformation apart. Here, we take \(i_t\) as an example. -The linear transformation is applied by calling a fc layer and the -equation is:
----\[L_{i_t} = W_{x_i}x_{t} + W_{h_i}h_{t-1} + b_i\]-
The non-linear transformation is applied by calling lstm_unit_op and the -equation is:
----\[i_t = \sigma(L_{i_t})\]-
This layer has two outputs including \(h_t\) and \(o_t\).
-| 参数: |
|
-
|---|---|
| 返回: | The hidden value and cell value of lstm unit. - |
-
| 返回类型: | tuple - |
-
| Raises: |
|
-
Examples
-x_t = fluid.layers.fc(input=x_t_data, size=10)
-prev_hidden = fluid.layers.fc(input=prev_hidden_data, size=30)
-prev_cell = fluid.layers.fc(input=prev_cell_data, size=30)
-hidden_value, cell_value = fluid.layers.lstm_unit(x_t=x_t,
- hidden_t_prev=prev_hidden,
- cell_t_prev=prev_cell)
-paddle.v2.fluid.layers.reduce_sum(input, dim=None, keep_dim=False, name=None)Computes the sum of tensor elements over the given dimension.
-| 参数: |
|
-
|---|---|
| 返回: | The reduced Tensor variable. - |
-
| 返回类型: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_sum(x) # [3.5]
-fluid.layers.reduce_sum(x, dim=0) # [0.3, 0.5, 1.1, 1.6]
-fluid.layers.reduce_sum(x, dim=-1) # [1.9, 1.6]
-fluid.layers.reduce_sum(x, dim=1, keep_dim=True) # [[1.9], [1.6]]
-paddle.v2.fluid.layers.reduce_mean(input, dim=None, keep_dim=False, name=None)Computes the mean of tensor elements over the given dimension.
-| 参数: |
|
-
|---|---|
| 返回: | The reduced Tensor variable. - |
-
| 返回类型: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_mean(x) # [0.4375]
-fluid.layers.reduce_mean(x, dim=0) # [0.15, 0.25, 0.55, 0.8]
-fluid.layers.reduce_mean(x, dim=-1) # [0.475, 0.4]
-fluid.layers.reduce_mean(x, dim=1, keep_dim=True) # [[0.475], [0.4]]
-paddle.v2.fluid.layers.reduce_max(input, dim=None, keep_dim=False, name=None)Computes the maximum of tensor elements over the given dimension.
-| 参数: |
|
-
|---|---|
| 返回: | The reduced Tensor variable. - |
-
| 返回类型: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_max(x) # [0.9]
-fluid.layers.reduce_max(x, dim=0) # [0.2, 0.3, 0.6, 0.9]
-fluid.layers.reduce_max(x, dim=-1) # [0.9, 0.7]
-fluid.layers.reduce_max(x, dim=1, keep_dim=True) # [[0.9], [0.7]]
-paddle.v2.fluid.layers.reduce_min(input, dim=None, keep_dim=False, name=None)Computes the minimum of tensor elements over the given dimension.
-| 参数: |
|
-
|---|---|
| 返回: | The reduced Tensor variable. - |
-
| 返回类型: | Variable - |
-
Examples
-# x is a Tensor variable with following elements:
-# [[0.2, 0.3, 0.5, 0.9]
-# [0.1, 0.2, 0.6, 0.7]]
-# Each example is followed by the correspending output tensor.
-fluid.layers.reduce_min(x) # [0.1]
-fluid.layers.reduce_min(x, dim=0) # [0.1, 0.2, 0.5, 0.7]
-fluid.layers.reduce_min(x, dim=-1) # [0.2, 0.1]
-fluid.layers.reduce_min(x, dim=1, keep_dim=True) # [[0.2], [0.1]]
-paddle.v2.fluid.layers.sequence_first_step(input, **kwargs)This funciton get the first step of sequence.
-x is a 1-level LoDTensor:
- x.lod = [[0, 2, 5, 7]]
- x.data = [1, 3, 2, 4, 6, 5, 1]
- x.dims = [7, 1]
-
-then output is a Tensor:
- out.dim = [3, 1]
- with condition len(x.lod[-1]) - 1 == out.dims[0]
- out.data = [1, 2, 5], where 1=first(1,3), 2=first(2,4,6), 5=first(5,1)
-| 参数: | input (variable) – The input variable which is a LoDTensor. | -
|---|---|
| 返回: | The sequence’s first step variable which is a Tensor. | -
Examples
-x = fluid.layers.data(name='x', shape=[7, 1],
- dtype='float32', lod_level=1)
-x_first_step = fluid.layers.sequence_first_step(input=x)
-paddle.v2.fluid.layers.sequence_last_step(input, **kwargs)This funciton get the last step of sequence.
-x is a 1-level LoDTensor:
- x.lod = [[0, 2, 5, 7]]
- x.data = [1, 3, 2, 4, 6, 5, 1]
- x.dims = [7, 1]
-
-then output is a Tensor:
- out.dim = [3, 1]
- with condition len(x.lod[-1]) - 1 == out.dims[0]
- out.data = [3, 6, 1], where 3=last(1,3), 6=last(2,4,6), 1=last(5,1)
-| 参数: | input (variable) – The input variable which is a LoDTensor. | -
|---|---|
| 返回: | The sequence’s last step variable which is a Tensor. | -
Examples
-x = fluid.layers.data(name='x', shape=[7, 1],
- dtype='float32', lod_level=1)
-x_last_step = fluid.layers.sequence_last_step(input=x)
-paddle.v2.fluid.layers.dropout(x, dropout_prob, is_test=False, seed=None, **kwargs)Computes dropout.
-Drop or keep each element of x independently. Dropout is a regularization -technique for reducing overfitting by preventing neuron co-adaption during -training. The dropout operator randomly set (according to the given dropout -probability) the outputs of some units to zero, while others are remain -unchanged.
-| 参数: |
|
-
|---|---|
| 返回: | A tensor variable. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name="data", shape=[32, 32], dtype="float32")
-droped = fluid.layers.dropout(input=x, dropout_rate=0.5)
-paddle.v2.fluid.layers.split(input, num_or_sections, dim=-1, name=None)Split the input tensor into multiple sub-tensors.
-| 参数: |
|
-
|---|---|
| 返回: | The list of segmented tensor variables. - |
-
| 返回类型: | List - |
-
Examples
-# x is a Tensor variable with shape [3, 9, 5]:
-x0, x1, x2 = fluid.layers.split(x, num_or_sections=3, dim=1)
-x0.shape # [3, 3, 5]
-x1.shape # [3, 3, 5]
-x2.shape # [3, 3, 5]
-x0, x1, x2 = fluid.layers.split(x, num_or_sections=[2, 3, 4], dim=1)
-x0.shape # [3, 2, 5]
-x1.shape # [3, 3, 5]
-x2.shape # [3, 4, 5]
-paddle.v2.fluid.layers.ctc_greedy_decoder(input, blank, name=None)This op is used to decode sequences by greedy policy by below steps: -1. Get the indexes of max value for each row in input. a.k.a.
---numpy.argmax(input, axis=0).
A simple example as below:
-Given:
-
-input.data = [[0.6, 0.1, 0.3, 0.1],
- [0.3, 0.2, 0.4, 0.1],
- [0.1, 0.5, 0.1, 0.3],
- [0.5, 0.1, 0.3, 0.1],
-
- [0.5, 0.1, 0.3, 0.1],
- [0.2, 0.2, 0.2, 0.4],
- [0.2, 0.2, 0.1, 0.5],
- [0.5, 0.1, 0.3, 0.1]]
-
-input.lod = [[0, 4, 8]]
-
-Then:
-
-output.data = [[2],
- [1],
- [3]]
-
-output.lod = [[0, 2, 3]]
-| 参数: |
|
-
|---|---|
| 返回: | CTC greedy decode result. If all the sequences in result were -empty, the result LoDTensor will be [-1] with LoD [[0]] and dims [1, 1]. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[8], dtype='float32')
-
-cost = fluid.layers.ctc_greedy_decoder(input=x, blank=0)
-paddle.v2.fluid.layers.edit_distance(input, label, normalized=False, ignored_tokens=None, name=None)EditDistance operator computes the edit distances between a batch of -hypothesis strings and their references. Edit distance, also called -Levenshtein distance, measures how dissimilar two strings are by counting -the minimum number of operations to transform one string into anthor. -Here the operations include insertion, deletion, and substitution.
-For example, given hypothesis string A = “kitten” and reference -B = “sitting”, the edit distance is 3 for A will be transformed into B -at least after two substitutions and one insertion:
-“kitten” -> “sitten” -> “sittin” -> “sitting”
-Input(Hyps) is a LoDTensor consisting of all the hypothesis strings with -the total number denoted by batch_size, and the separation is specified -by the LoD information. And the batch_size reference strings are arranged -in order in the same way in the LoDTensor Input(Refs).
-Output(Out) contains the batch_size results and each stands for the edit -distance for a pair of strings respectively. If Attr(normalized) is true, -the edit distance will be divided by the length of reference string.
-| 参数: |
|
-
|---|---|
| 返回: | sequence-to-sequence edit distance in shape [batch_size, 1]. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[8], dtype='float32')
-y = fluid.layers.data(name='y', shape=[7], dtype='float32')
-
-cost = fluid.layers.edit_distance(input=x,label=y)
-paddle.v2.fluid.layers.l2_normalize(x, axis, epsilon=1e-12, name=None)L2 normalize Layer
-The l2 normalize layer normalizes x along dimension axis using an L2 -norm. For a 1-D tensor (dim is fixed to 0), this layer computes
-output = x / sqrt(max(sum(x**2), epsilon))
-For x with more dimensions, this layer independently normalizes each 1-D -slice along dimension axis.
-| 参数: |
|
-
|---|---|
| 返回: | The output tensor variable. - |
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.data(name="data",
- shape=(3, 17, 13),
- dtype="float32")
-normed = fluid.layers.l2_normalize(x=data, axis=1)
-paddle.v2.fluid.layers.matmul(x, y, transpose_x=False, transpose_y=False, name=None)Applies matrix multiplication to two tensors.
-Currently, the input tensors’ rank can be any, but when the rank of any -inputs is bigger than 3, this two inputs’ rank should be equal.
-The actual behavior depends on the shapes of \(x\), \(y\) and the
-flag values of transpose_x, transpose_y. Specifically:
Also note that if the raw tensor \(x\) or \(y\) is rank-1 and -nontransposed, the prepended or appended dimension \(1\) will be -removed after matrix multiplication.
-| 参数: |
|
-
|---|---|
| 返回: | The product Tensor variable. - |
-
| 返回类型: | Variable - |
-
Examples
-# Examples to clarify shapes of the inputs and output
-# x: [B, ..., M, K], y: [B, ..., K, N]
-fluid.layers.matmul(x, y) # out: [B, ..., M, N]
-
-# x: [B, M, K], y: [B, K, N]
-fluid.layers.matmul(x, y) # out: [B, M, N]
-
-# x: [B, M, K], y: [K, N]
-fluid.layers.matmul(x, y) # out: [B, M, N]
-
-# x: [M, K], y: [K, N]
-fluid.layers.matmul(x, y) # out: [M, N]
-
-# x: [B, M, K], y: [K]
-fluid.layers.matmul(x, y) # out: [B, M]
-
-# x: [K], y: [K]
-fluid.layers.matmul(x, y) # out: [1]
-
-# x: [M], y: [N]
-fluid.layers.matmul(x, y, True, True) # out: [M, N]
-paddle.v2.fluid.layers.warpctc(input, label, blank=0, norm_by_times=False, **kwargs)An operator integrating the open source Warp-CTC library -(https://github.com/baidu-research/warp-ctc) -to compute Connectionist Temporal Classification (CTC) loss. -It can be aliased as softmax with CTC, since a native softmax activation is -interated to the Warp-CTC library, to to normlize values for each row of the -input tensor.
-| 参数: |
|
-
|---|---|
| 返回: | The Connectionist Temporal Classification (CTC) loss, -which is a 2-D Tensor of the shape [batch_size, 1]. - |
-
| 返回类型: | Variable - |
-
Examples
-paddle.v2.fluid.layers.sequence_reshape(input, new_dim)Sequence Reshape Layer
-This layer will rearrange the input sequences. The new dimension is set by -user. Length of each sequence is computed according to original length, -original dimension and new dimension. The following example will help to -illustrate the function of this layer:
-x is a LoDTensor:
- x.lod = [[0, 2, 6]]
- x.data = [[1, 2], [3, 4],
- [5, 6], [7, 8], [9, 10], [11, 12]]
- x.dims = [6, 2]
-
-set new_dim = 4
-
-then out is a LoDTensor:
- out.lod = [[0, 1, 3]]
- out.data = [[1, 2, 3, 4],
- [5, 6, 7, 8], [9, 10, 11, 12]]
- out.dims = [3, 4]
-Currently, only 1-level LoDTensor is supported and please make sure -(original length * original dimension) can be divided by new dimension with -no remainder for each sequence.
-| 参数: |
|
-
|---|---|
| 返回: | Reshaped LoDTensor according to new dimension. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[5, 20],
- dtype='float32', lod_level=1)
-x_reshaped = layers.sequence_reshape(input=x, new_dim=10)
-paddle.v2.fluid.layers.transpose(x, perm, name=None)transpose Layer
-Permute the dimensions of input according to perm.
-The i-th dimension of the returned tensor will correspond to the -perm[i]-th dimension of input.
-| 参数: |
|
-
|---|---|
| 返回: | A transposed Tensor. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[5, 10, 15], dtype='float32')
-x_transposed = layers.transpose(x, perm=[1, 0, 2])
-paddle.v2.fluid.layers.im2sequence(input, filter_size=1, stride=1, padding=0, name=None)Extracts image patches from the input tensor to form a tensor of shape -{input.batch_size * output_height * output_width, filter_size_H * -filter_size_W * input.channels} which is similar with im2col. -This op use filter / kernel to scan images and convert these images to -sequences. After expanding, the number of time step are -output_height * output_width for an image, in which output_height and -output_width are calculated by below equation:
-And the dimension of each time step is block_y * block_x * input.channels.
-| 参数: |
|
-
|---|---|
| 返回: | The output is a LoDTensor with shape -{input.batch_size * output_height * output_width, -filter_size_H * filter_size_W * input.channels}. -If we regard output as a matrix, each row of this matrix is -a step of a sequence. - |
-
| 返回类型: | output - |
-
Examples:
-As an example:
-----Given: - -x = [[[[ 6. 2. 1.] - [ 8. 3. 5.] - [ 0. 2. 6.]] - - [[ 2. 4. 4.] - [ 6. 3. 0.] - [ 6. 4. 7.]]] - - [[[ 6. 7. 1.] - [ 5. 7. 9.] - [ 2. 4. 8.]] - - [[ 1. 2. 1.] - [ 1. 3. 5.] - [ 9. 0. 8.]]]] - -x.dims = {2, 2, 3, 3} - -And: - -filter = [2, 2] -stride = [1, 1] -padding = [0, 0] - -Then: - -output.data = [[ 6. 2. 8. 3. 2. 4. 6. 3.] - [ 2. 1. 3. 5. 4. 4. 3. 0.] - [ 8. 3. 0. 2. 6. 3. 6. 4.] - [ 3. 5. 2. 6. 3. 0. 4. 7.] - [ 6. 7. 5. 7. 1. 2. 1. 3.] - [ 7. 1. 7. 9. 2. 1. 3. 5.] - [ 5. 7. 2. 4. 1. 3. 9. 0.] - [ 7. 9. 4. 8. 3. 5. 0. 8.]] - -output.dims = {8, 9} - -output.lod = [[0, 4, 8]] -The simple usage is:
---output = fluid.layers.im2sequence( - input=layer, stride=[1, 1], filter_size=[2, 2]) -
paddle.v2.fluid.layers.nce(input, label, num_total_classes, sample_weight=None, param_attr=None, bias_attr=None, num_neg_samples=None)Compute and return the noise-contrastive estimation training loss. -See [Noise-contrastive estimation: A new estimation principle for unnormalized statistical models](http://www.jmlr.org/proceedings/papers/v9/gutmann10a/gutmann10a.pdf). -By default this operator uses a uniform distribution for sampling.
-| 参数: |
|
-
|---|---|
| 返回: | (Tensor) A tensor of shape [batch_size, 1]. Cost of samples. - |
-
paddle.v2.fluid.layers.beam_search(pre_ids, ids, scores, beam_size, end_id, level=0)This function implements the beam search algorithm.
-paddle.v2.fluid.layers.row_conv(input, future_context_size, param_attr=None, act=None)Row Conv Operator. This layer will apply lookahead convolution to -input. The input variable should be a 2D LoDTensor with shape [T, D]. -Parameters with shape [future_context_size + 1, D] will be created. The math -equation of row convolution is as follows:
-In the above equation:
-More details about row_conv please refer to the paper (http://www.cs.cmu.edu/~dyogatam/papers/wang+etal.iclrworkshop2016.pdf) and -the design document (https://github.com/PaddlePaddle/Paddle/issues/2228#issuecomment-303903645).
-| 参数: |
|
-
|---|---|
| 返回: | The output tensor with same shape as input tensor. - |
-
| 返回类型: | Variable - |
-
Examples
-x = fluid.layers.data(name='x', shape=[16],
- dtype='float32', lod_level=1)
-out = fluid.layers.row_conv(input=x, future_context_size=2)
-paddle.v2.fluid.layers.multiplex(inputs, index)Multiplex Layer
-Referring to the given index variable, this layer selects rows from the -input variables to construct a multiplex variable. Assuming that there are -\(m\) input variables and \(I_i\) represents the i-th input -variable and \(i\) is in [0, \(m\)). All input variables are -tensors with same shape [\(d_0\), \(d_1\), ..., \(d_R\)]. -Please note that rank of the input tensor should be at least 2. Each input -variable will be treated as a 2-D matrix with shape [\(M\), \(N\)] -where \(M\) for \(d_0\) and \(N\) for \(d_1\) * \(d_2\) -* ... * \(d_R\). Let \(I_i[j]\) be the j-th row of the i-th input -variable. The given index variable should be a 2-D tensor with shape -[\(M\), 1]. Let ID[i] be the i-th index value of the index variable. -Then the output variable will be a tensor with shape [\(d_0\), -\(d_1\), ..., \(d_R\)]. If we treat the output tensor as a 2-D -matrix with shape [\(M\), \(N\)] and let \(O[i]\) be the i-th -row of the matrix, then O[i] is equal to \(I_{ID[i]}[i]\).
-| 参数: |
|
-
|---|---|
| 返回: | Multiplex variable gathered from input variables. - |
-
| 返回类型: | Variable - |
-
Examples
-x1 = fluid.layers.data(name='x1', shape=[4], dtype='float32')
-x2 = fluid.layers.data(name='x2', shape=[4], dtype='float32')
-index = fluid.layers.data(name='index', shape=[1], dtype='int32')
-out = fluid.layers.multiplex(inputs=[x1, x2], index=index)
-paddle.v2.fluid.layers.mean(**kwargs)Mean Operator.
-Out is a scalar which is the mean of all elements in X.
-| 参数: | x – The input of mean op -Duplicable: False Optional: False | -
|---|---|
| 返回: | The output of mean op | -
paddle.v2.fluid.layers.mul(**kwargs)Mul Operator.
-This operator is used to perform matrix multiplication for input $X$ and $Y$.
-The equation is:
-$$Out = X * Y$$
-Both the input $X$ and $Y$ can carry the LoD (Level of Details) information, -or not. But the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | (Tensor), The output tensor of mul op. - |
-
paddle.v2.fluid.layers.reshape(**kwargs)Reshape Operator.
-Reshape Input(X) into the shape specified by Attr(shape).
-An example: -Given a 2-D tensor X with 2 rows and 2 columns : [[1, 2], [3, 4]]
-and target shape = [1, 4], the reshape operator will transform -the tensor X into a 2-D tensor: [[1, 2, 3, 4]]
-One dimension in the target shape can be set -1, representing that its -size is unknown. In this case, the real dimension will be infered from -the original shape of Input(X) and other dimensions in the target shape.
-| 参数: |
|
-
|---|---|
| 返回: | The output tensor of reshape operator. - |
-
paddle.v2.fluid.layers.scale(**kwargs)Scale operator
-$$Out = scale*X$$
-| 参数: |
|
-
|---|---|
| 返回: | (Tensor) Output tensor of scale operator. - |
-
paddle.v2.fluid.layers.sigmoid_cross_entropy_with_logits(**kwargs)SigmoidCrossEntropyWithLogits Operator.
-This measures the element-wise probability error in classification tasks -in which each class is independent. This can be thought of as predicting labels -for a data-point, where labels are not mutually exclusive. -For example, a news article can be about politics, technology or sports -at the same time or none of these.
-The logistic loss is given as follows:
---$$loss = -Labels * log(sigma(X)) - (1 - Labels) * log(1 - sigma(X))$$
We know that $$sigma(X) = (1 / (1 + exp(-X)))$$. By substituting this we get:
---$$loss = X - X * Labels + log(1 + exp(-X))$$
For stability and to prevent overflow of $$exp(-X)$$ when X < 0, -we reformulate the loss as follows:
---$$loss = max(X, 0) - X * Labels + log(1 + exp(-|X|))$$
Both the input X and Labels can carry the LoD (Level of Details) information. -However the output only shares the LoD with input X.
-| 参数: |
|
-
|---|---|
| 返回: | (Tensor, default Tensor<float>), a 2-D tensor with shape N x D of elementwise logistic losses. - |
-
paddle.v2.fluid.layers.elementwise_add(**kwargs)Limited Elementwise Add Operator.
-The equation is:
-$$Out = X + Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_div(**kwargs)Limited Elementwise Div Operator.
-The equation is:
-$$Out = X / Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_sub(**kwargs)Limited Elementwise Sub Operator.
-The equation is:
-$$Out = X - Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_mul(**kwargs)Limited Elementwise Mul Operator.
-The equation is:
-$$Out = X odotY$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_max(**kwargs)Limited Elementwise Max Operator.
-The equation is:
-$$Out = max(X, Y)$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_min(**kwargs)Limited Elementwise Max Operator.
-The equation is:
-$$Out = min(X, Y)$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.elementwise_pow(**kwargs)Limited Elementwise Pow Operator.
-The equation is:
-$$Out = X ^ Y$$
-$X$ is a tensor of any dimension and the dimensions of tensor $Y$ must be -smaller than or equal to the dimensions of $X$.
-There are two cases for this operator: -1. The shape of $Y$ is same with $X$; -2. The shape of $Y$ is a subset of $X$.
-For case 2: -$Y$ will be broadcasted to match the shape of $X$ and axis should be -set to index of the start dimension to broadcast $Y$ onto $X$.
-shape(X) = (2, 3, 4, 5), shape(Y) = (,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (5,)
-shape(X) = (2, 3, 4, 5), shape(Y) = (4, 5)
-shape(X) = (2, 3, 4, 5), shape(Y) = (3, 4), with axis=1
-shape(X) = (2, 3, 4, 5), shape(Y) = (2), with axis=0
-Either of the inputs $X$ and $Y$ or none can carry the LoD (Level of Details) -information. However, the output only shares the LoD information with input $X$.
-| 参数: |
|
-
|---|---|
| 返回: | The output of elementwise op. - |
-
paddle.v2.fluid.layers.clip(**kwargs)Clip Operator.
-The clip operator limits the value of given input within an interval. The -interval is specified with arguments ‘min’ and ‘max’:
-$$ -Out = min(max(X, min), max) -$$
-| 参数: |
|
-
|---|---|
| 返回: | (Tensor)The output of clip op with shape as input(X) - |
-
paddle.v2.fluid.layers.clip_by_norm(**kwargs)ClipByNorm Operator.
-This operator limits the L2 norm of the input $X$ within $max_norm$. -If the L2 norm of $X$ is less than or equal to $max_norm$, $Out$ will be -the same as $X$. If the L2 norm of $X$ is greater than $max_norm$, $X$ will -be linearly scaled to make the L2 norm of $Out$ equal to $max_norm$, as -shown in the following formula:
-$$ -Out = frac{max_norm * X}{norm(X)}, -$$
-where $norm(X)$ represents the L2 norm of $X$.
-| 参数: |
|
-
|---|---|
| 返回: | (Tensor) The output of clip_by_norm op with shape as input(X) - |
-
paddle.v2.fluid.layers.sequence_softmax(**kwargs)Sequence Softmax Operator.
-SequenceSoftmaxOp computes the softmax activation among all time-steps for each -sequence. The dimension of each time-step should be 1. Thus, the shape of -input Tensor can be either [N, 1] or [N], where N is the sum of the length -of all sequences.
-The algorithm works as follows:
---for i-th sequence in a mini-batch:
$$ -Out(X[lod[i]:lod[i+1]], :) = frac{exp(X[lod[i]:lod[i+1], :])} {sum(exp(X[lod[i]:lod[i+1], :]))} -$$
-For example, for a mini-batch of 3 sequences with variable-length, -each containing 2, 3, 2 time-steps, the lod of which is [0, 2, 5, 7], -then softmax will be computed among X[0:2, :], X[2:5, :], X[5:7, :] -and N turns out to be 7.
-| 参数: | x – (LoDTensor) 1-D or 2-D input LoDTensor with the 2-nd dimension of length 1. -Duplicable: False Optional: False | -
|---|---|
| 返回: | (LoDTensor) 1-D or 2-D output LoDTensor with the 2-nd dimension of length 1. | -
paddle.v2.fluid.layers.sigmoid(**kwargs)Sigmoid Activation Operator
-$$out = frac{1}{1 + e^{-x}}$$
-| 参数: | x – Input of Sigmoid operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Sigmoid operator | -
paddle.v2.fluid.layers.logsigmoid(**kwargs)Logsigmoid Activation Operator
-$$out = log frac{1}{1 + e^{-x}}$$
-| 参数: | x – Input of LogSigmoid operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of LogSigmoid operator | -
paddle.v2.fluid.layers.exp(**kwargs)Exp Activation Operator.
-$out = e^x$
-| 参数: | x – Input of Exp operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Exp operator | -
paddle.v2.fluid.layers.relu(**kwargs)Relu Activation Operator.
-$out = max(x, 0)$
-| 参数: | x – Input of Relu operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Relu operator | -
paddle.v2.fluid.layers.tanh(**kwargs)Tanh Activation Operator.
-$$out = frac{e^{x} - e^{-x}}{e^{x} + e^{-x}}$$
-| 参数: | x – Input of Tanh operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Tanh operator | -
paddle.v2.fluid.layers.tanh_shrink(**kwargs)TanhShrink Activation Operator.
-$$out = x - frac{e^{x} - e^{-x}}{e^{x} + e^{-x}}$$
-| 参数: | x – Input of TanhShrink operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of TanhShrink operator | -
paddle.v2.fluid.layers.softshrink(**kwargs)Softshrink Activation Operator.
-$$ -out = begin{cases}
---x - lambda, text{if } x > lambda \ -x + lambda, text{if } x < -lambda \ -0, text{otherwise} -end{cases}
$$
-| 参数: |
|
-
|---|---|
| 返回: | Output of Softshrink operator - |
-
paddle.v2.fluid.layers.sqrt(**kwargs)Sqrt Activation Operator.
-$out = sqrt{x}$
-| 参数: | x – Input of Sqrt operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Sqrt operator | -
paddle.v2.fluid.layers.abs(**kwargs)Abs Activation Operator.
-$out = |x|$
-| 参数: | x – Input of Abs operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Abs operator | -
paddle.v2.fluid.layers.ceil(**kwargs)Ceil Activation Operator.
-$out = ceil(x)$
-| 参数: | x – Input of Ceil operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Ceil operator | -
paddle.v2.fluid.layers.floor(**kwargs)Floor Activation Operator.
-$out = floor(x)$
-| 参数: | x – Input of Floor operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Floor operator | -
paddle.v2.fluid.layers.round(**kwargs)Round Activation Operator.
-$out = [x]$
-| 参数: | x – Input of Round operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Round operator | -
paddle.v2.fluid.layers.reciprocal(**kwargs)Reciprocal Activation Operator.
-$$out = frac{1}{x}$$
-| 参数: | x – Input of Reciprocal operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Reciprocal operator | -
paddle.v2.fluid.layers.log(**kwargs)Log Activation Operator.
-$out = ln(x)$
-Natural logarithm of x.
-| 参数: | x – Input of Log operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Log operator | -
paddle.v2.fluid.layers.square(**kwargs)Square Activation Operator.
-$out = x^2$
-| 参数: | x – Input of Square operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Square operator | -
paddle.v2.fluid.layers.softplus(**kwargs)Softplus Activation Operator.
-$out = ln(1 + e^{x})$
-| 参数: | x – Input of Softplus operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Softplus operator | -
paddle.v2.fluid.layers.softsign(**kwargs)Softsign Activation Operator.
-$$out = frac{x}{1 + |x|}$$
-| 参数: | x – Input of Softsign operator -Duplicable: False Optional: False | -
|---|---|
| 返回: | Output of Softsign operator | -
paddle.v2.fluid.layers.brelu(**kwargs)BRelu Activation Operator.
-$out = max(min(x, t_{min}), t_{max})$
-| 参数: |
|
-
|---|---|
| 返回: | Output of BRelu operator - |
-
paddle.v2.fluid.layers.leaky_relu(**kwargs)LeakyRelu Activation Operator.
-$out = max(x, alpha * x)$
-| 参数: |
|
-
|---|---|
| 返回: | Output of LeakyRelu operator - |
-
paddle.v2.fluid.layers.soft_relu(**kwargs)SoftRelu Activation Operator.
-$out = ln(1 + exp(max(min(x, threshold), threshold))$
-| 参数: |
|
-
|---|---|
| 返回: | Output of SoftRelu operator - |
-
paddle.v2.fluid.layers.elu(**kwargs)ELU Activation Operator.
-Applies the following element-wise computation on the input according to -https://arxiv.org/abs/1511.07289.
-$out = max(0, x) + min(0, alpha * (e^x - 1))$
-| 参数: |
|
-
|---|---|
| 返回: | Output of ELU operator - |
-
paddle.v2.fluid.layers.relu6(**kwargs)Relu6 Activation Operator.
-$out = min(max(0, x), 6)$
-| 参数: |
|
-
|---|---|
| 返回: | Output of Relu6 operator - |
-
paddle.v2.fluid.layers.pow(**kwargs)Pow Activation Operator.
-$out = x^{factor}$
-| 参数: |
|
-
|---|---|
| 返回: | Output of Pow operator - |
-
paddle.v2.fluid.layers.stanh(**kwargs)STanh Activation Operator.
-$$out = b * frac{e^{a * x} - e^{-a * x}}{e^{a * x} + e^{-a * x}}$$
-| 参数: |
|
-
|---|---|
| 返回: | Output of STanh operator - |
-
paddle.v2.fluid.layers.hard_shrink(**kwargs)HardShrink Activation Operator.
-$$ -out = begin{cases}
---x, text{if } x > lambda \ -x, text{if } x < -lambda \ -0, text{otherwise} -end{cases}
$$
-| 参数: |
|
-
|---|---|
| 返回: | Output of HardShrink operator - |
-
paddle.v2.fluid.layers.thresholded_relu(**kwargs)ThresholdedRelu Activation Operator.
-$$ -out = begin{cases}
---x, text{if } x > threshold \ -0, text{otherwise} -end{cases}
$$
-| 参数: |
|
-
|---|---|
| 返回: | Output of ThresholdedRelu operator - |
-
paddle.v2.fluid.layers.hard_sigmoid(**kwargs)HardSigmoid Activation Operator.
-Segment-wise linear approximation of sigmoid(https://arxiv.org/abs/1603.00391), -which is much faster than sigmoid.
-$out = max(0, min(1, slope * x + shift))$
-The slope should be positive. The offset can be either positive or negative. -The default slope and shift are set according to the above reference. -It is recommended to use the defaults for this activation.
-| 参数: |
|
-
|---|---|
| 返回: | Output of HardSigmoid operator - |
-
paddle.v2.fluid.layers.swish(**kwargs)Swish Activation Operator.
-$$out = frac{x}{1 + e^{- beta x}}$$
-| 参数: |
|
-
|---|---|
| 返回: | Output of Swish operator - |
-
paddle.v2.fluid.layers.create_tensor(dtype, name=None, persistable=False)paddle.v2.fluid.layers.create_parameter(shape, dtype, name=None, attr=None, is_bias=False, default_initializer=None)Create a parameter -:param shape: shape of the parameter -:type shape: list[int] -:param dtype: element type of the parameter -:type dtype: string -:param attr: attributes of the parameter -:type attr: ParamAttr -:param is_bias: This can affect which default initializer is chosen
---when default_initializer is None. If is_bias, -initializer.Constant(0.0) will be used. Otherwise, -Xavier() will be used.
| 参数: | default_initializer (Initializer) – initializer for the parameter | -
|---|---|
| 返回: | the created parameter | -
| 返回类型: | Parameter | -
paddle.v2.fluid.layers.create_global_var(shape, value, dtype, persistable=False, force_cpu=False, name=None)Create a global variable. such as global_step -:param shape: shape of the variable -:type shape: list[int] -:param value: the value of the variable -:type value: float -:param dtype: element type of the parameter -:type dtype: string -:param persistable: if this variable is persistable -:type persistable: bool -:param force_cpu: force this variable to be on CPU -:type force_cpu: bool
-| 返回: | the created Variable | -
|---|---|
| 返回类型: | Variable | -
paddle.v2.fluid.layers.cast(x, dtype)This function takes in the input with input_dtype -and casts it to the output_dtype as the output.
-paddle.v2.fluid.layers.concat(input, axis=0)Concat
-This function concatenates the input along the axis mentioned -and returns that as the output.
-| 参数: |
|
-
|---|---|
| 返回: | Output variable of the concatenation - |
-
| 返回类型: | Variable - |
-
Examples
-paddle.v2.fluid.layers.sums(input, out=None)This function performs the sum operation on the input and returns the -result as the output.
-| 参数: | input (Variable|list) – The input tensor that has the elements -that need to be summed up. | -
|---|---|
| 返回: |
|
-
| 返回类型: | Variable | -
Examples
-paddle.v2.fluid.layers.assign(input, output)Assign
-This function copies the input Variable to the output Variable.
-| 参数: |
|
-
|---|---|
| 返回: | The destination variable that was supplied as the output. - |
-
| 返回类型: | Variable - |
-
Examples
-paddle.v2.fluid.layers.fill_constant_batch_size_like(input, shape, dtype, value, input_dim_idx=0, output_dim_idx=0)fill_constant_batch_size_like
-This function creates a tensor of specified shape, dtype and batch size, -and initializes this with a constant supplied in value. The batch size is -obtained from the input tensor.
-It also sets stop_gradient to True.
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the output - |
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.fill_constant_batch_size_like(
- input=like, shape=[1], value=0, dtype='int64')
-paddle.v2.fluid.layers.fill_constant(shape, dtype, value, force_cpu=False, out=None)fill_constant
-This function creates a tensor with specified shape and dtype, and -initializes it with a constant specifed by value.
-The attribute stop_gradient of the created tensor is set to True.
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the output. - |
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.fill_constant(shape=[1], value=0, dtype='int64')
-paddle.v2.fluid.layers.ones(shape, dtype, force_cpu=False)ones
-This function creates a tensor of specified shape and -dtype, and initializes this with 1.
-It also sets stop_gradient to True.
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the output - |
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.ones(shape=[1], dtype='int64')
-paddle.v2.fluid.layers.zeros(shape, dtype, force_cpu=False)zeros
-This function creates a tensor of specified shape and -dtype, and initializes this with 0.
-It also sets stop_gradient to True.
-| 参数: |
|
-
|---|---|
| 返回: | The tensor variable storing the output - |
-
| 返回类型: | Variable - |
-
Examples
-data = fluid.layers.zeros(shape=[1], dtype='int64')
-
- paddle.v2.fluid.nets.simple_img_conv_pool(input, num_filters, filter_size, pool_size, pool_stride, act, param_attr=None, pool_type='max', use_cudnn=True)paddle.v2.fluid.nets.sequence_conv_pool(input, num_filters, filter_size, param_attr=None, act='sigmoid', pool_type='max')paddle.v2.fluid.nets.glu(input, dim=-1)The gated linear unit composed by split, sigmoid activation and elementwise -multiplication. Specifically, Split the input into two equal sized parts -\(a\) and \(b\) along the given dimension and then compute as -following:
----\[{GLU}(a, b)= a \otimes \sigma(b)\]-
Refer to Language Modeling with Gated Convolutional Networks.
-| 参数: |
|
-
|---|---|
| 返回: | The Tensor variable with half the size of input. - |
-
| 返回类型: | Variable - |
-
Examples
-# x is a Tensor variable with shape [3, 6, 9]
-fluid.nets.glu(input=x, dim=1) # shape of output: [3, 3, 9]
-paddle.v2.fluid.nets.scaled_dot_product_attention(queries, keys, values, num_heads=1, dropout_rate=0.0)The dot-product attention.
-Attention mechanism can be seen as mapping a query and a set of key-value -pairs to an output. The output is computed as a weighted sum of the values, -where the weight assigned to each value is computed by a compatibility -function (dot-product here) of the query with the corresponding key.
-The dot-product attention can be implemented through (batch) matrix -multipication as follows:
----\[Attention(Q, K, V)= softmax(QK^\mathrm{T})V\]-
Refer to Attention Is All You Need.
-| 参数: |
|
-
|---|---|
| 返回: | A 3-D Tensor computed by multi-head scaled dot product attention. - |
-
| 返回类型: | Variable - |
-
| Raises: |
|
-
注解
-1. When num_heads > 1, three linear projections are learned respectively -to map input queries, keys and values into queries’, keys’ and values’. -queries’, keys’ and values’ have the same shapes with queries, keys -and values.
-1. When num_heads == 1, scaled_dot_product_attention has no learnable -parameters.
-Examples
-# Suppose q, k, v are Tensors with the following shape:
-# q: [3, 5, 9], k: [3, 6, 9], v: [3, 6, 10]
-
-contexts = fluid.nets.scaled_dot_product_attention(q, k, v)
-contexts.shape # [3, 5, 10]
-
- paddle.v2.fluid.optimizer.SGDSGDOptimizer 的别名
paddle.v2.fluid.optimizer.MomentumMomentumOptimizer 的别名
paddle.v2.fluid.optimizer.AdagradAdagradOptimizer 的别名
paddle.v2.fluid.optimizer.AdamAdamOptimizer 的别名
paddle.v2.fluid.optimizer.AdamaxAdamaxOptimizer 的别名
paddle.v2.fluid.optimizer.DecayedAdagradDecayedAdagradOptimizer 的别名
- paddle.v2.fluid.param_attr.ParamAttr(name=None, initializer=None, learning_rate=1.0, regularizer=None, trainable=True, gradient_clip=None)paddle.v2.fluid.param_attr.WeightNormParamAttr(dim=None, **kwargs)Used for weight normalization. Any field in ParamAttr can also be set here. -Besides, an extra field dim can be set to indicate the dimension except -which to normalize.
-
- paddle.v2.fluid.profiler.cuda_profiler(*args, **kwds)The CUDA profiler. -This fuctions is used to profile CUDA program by CUDA runtime application -programming interface. The profiling result will be written into -output_file with Key-Value pair format or Comma separated values format. -The user can set the output mode by output_mode argument and set the -counters/options for profiling by config argument. The default config -is [‘gpustarttimestamp’, ‘gpustarttimestamp’, ‘gridsize3d’, -‘threadblocksize’, ‘streamid’, ‘enableonstart 0’, ‘conckerneltrace’].
-| 参数: |
|
-
|---|
paddle.v2.fluid.profiler.reset_profiler()The profiler clear interface. -reset_profiler will clear the previous time record.
-paddle.v2.fluid.profiler.profiler(*args, **kwds)The profiler interface. -Different from cuda_profiler, this profiler can be used to profile both CPU -and GPU program. By defalut, it records the CPU and GPU operator kernels, -if you want to profile other program, you can refer the profiling tutorial -to add more records.
-| 参数: |
|
-
|---|
- paddle.v2.fluid.regularizer.append_regularization_ops(parameters_and_grads, regularization=None)Create and add backward regularization Operators
-Creates and adds backward regularization operators in the BlockDesc. -This will add gradients of the regularizer function to the gradients -of the parameters and return these modified gradients. This is the -same as implementing weight decay in optimizers for regularization.
-| 参数: |
|
-
|---|---|
| 返回: | list of (parameters, gradients) pair with the regularized gradient - |
-
| Raises: |
|
-
paddle.v2.fluid.regularizer.L1DecayL1DecayRegularizer 的别名
paddle.v2.fluid.regularizer.L2DecayL2DecayRegularizer 的别名
-
- paddle.v2.parameters.ParametersParameters manages all the learnable parameters in a neural network. -It stores parameters’ information in an OrderedDict. The key is -the name of a parameter, and value is a parameter’s configuration(in -protobuf format), such as initialization mean and std, its size, whether it -is a static parameter, and so on.
-| 参数: |
|
-
|---|
Basically usage is
-data = paddle.layers.data(...)
-...
-out = paddle.layers.fc(...)
-
-parameters = paddle.parameters.create(out)
-
-parameter_names = parameters.names()
-fc_mat = parameters.get('fc')
-print fc_mat
-keys()keys are the names of each parameter.
-| 返回: | list of parameter name | -
|---|---|
| 返回类型: | list | -
names()names of each parameter.
-| 返回: | list of parameter name | -
|---|---|
| 返回类型: | list | -
has_key(key)has_key return true if there are such parameter name == key
-| 参数: | key (basestring) – Parameter name | -
|---|---|
| 返回: | True if contains such key | -
get_shape(key)get shape of the parameter.
-| 参数: | key (basestring) – parameter name | -
|---|---|
| 返回: | parameter’s shape | -
| 返回类型: | tuple | -
get(parameter_name)Get parameter by parameter name.
-| Note: | It will always copy the parameter from C++ side. | -
|---|---|
| 参数: | parameter_name (basestring) – parameter name | -
| 返回: | The parameter matrix. | -
| 返回类型: | np.ndarray | -
get_grad(key)Get grandient by parameter name.
-| Note: | It will always copy the parameter from C++ side. | -
|---|---|
| 参数: | key (basestring) – parameter name | -
| 返回: | The grandient matrix. | -
| 返回类型: | np.ndarray | -
set(parameter_name, value)Set parameter by parameter name & matrix.
-| 参数: |
|
-
|---|---|
| 返回: | Nothing. - |
-
append_gradient_machine(gradient_machine)append gradient machine to parameters. This method is used internally in -Trainer.train.
-| 参数: | gradient_machine (api.GradientMachine) – PaddlePaddle C++ GradientMachine object. | -
|---|---|
| 返回: | - |
serialize(name, f)| 参数: |
|
-
|---|---|
| 返回: | - | -
deserialize(name, f)| 参数: |
|
-
|---|---|
| 返回: | - | -
to_tar(f)Save parameters to a tar file.
-| 参数: | f (file) – | -
|---|---|
| 返回: | - |
from_tar(f)Create a Parameters object from the given file. And -the Parameters only contains the parameters in this -file. It is adapted the parameters are same in the -defined network and the given file. For example, it -can be used in the inference.
-| 参数: | f (tar file) – the initialized model file. | -
|---|---|
| 返回: | A Parameters object. | -
| 返回类型: | Parameters. | -
init_from_tar(f, exclude_params=[])Different from from_tar, this interface can be used to -init partial network parameters from another saved model.
-| 参数: |
|
-
|---|---|
| 返回: | Nothing. - |
-
Module Trainer
-paddle.v2.trainer.SGD(cost, parameters, update_equation, extra_layers=None, is_local=True, pserver_spec=None, use_etcd=True)Simple SGD Trainer. -SGD Trainer combines data reader, network topolopy and update_equation together -to train/test a neural network.
-| 参数: |
|
-
|---|
train(reader, num_passes=1, event_handler=None, feeding=None)Training method. Will train num_passes of input data.
-| 参数: |
|
-
|---|---|
| 返回: | - | -
test(reader, feeding=None)Testing method. Will test input data.
-| 参数: |
|
-
|---|---|
| 返回: | - | -
Testing and training events.
-There are:
-paddle.v2.event.TestResult(evaluator, cost)Result that trainer.test return.
-paddle.v2.event.BeginPass(pass_id)Event On One Pass Training Start.
-paddle.v2.event.EndPass(pass_id, evaluator, gm)Event On One Pass Training Complete. -To get the output of a specific layer, add “event.gm.getLayerOutputs(‘predict_layer’)” -in your event_handler call back
-paddle.v2.event.BeginIteration(pass_id, batch_id)Event On One Batch Training Start.
-paddle.v2.event.EndForwardBackward(pass_id, batch_id, gm)Event On One Batch ForwardBackward Complete.
-paddle.v2.event.EndIteration(pass_id, batch_id, cost, evaluator, gm)Event On One Batch Training Complete. -To get the output of a specific layer, add “event.gm.getLayerOutputs(‘predict_layer’)” -in your event_handler call back
-paddle.v2.infer(output_layer, parameters, input, feeding=None, field='value')Infer a neural network by given neural network output and parameters. The -user should pass either a batch of input data or reader method.
-Example usage for sinlge output_layer:
-result = paddle.infer(output_layer=prediction,
- parameters=parameters,
- input=SomeData)
-print result
-Example usage for multiple outout_layers and fields:
-result = paddle.infer(output_layer=[prediction1, prediction2],
- parameters=parameters,
- input=SomeData,
- field=[id, value]])
-print result
-| 参数: |
|
-
|---|---|
| 返回: | The prediction result. If there are multiple outout_layers and fields, -the return order is outout_layer1.field1, outout_layer2.field1, ..., -outout_layer1.field2, outout_layer2.field2 ... - |
-
| 返回类型: | numpy.ndarray - |
-
| - |
| - |
| - | - |
|
-
| - | - |
| - |
| - |
pooling 的使用示例如下，详细见 pooling 配置API。
+pooling 的使用示例如下，详细见 api_v2.layer_pooling 配置API。
seq_pool = pooling(input=layer,
pooling_type=pooling.Max(),
agg_level=AggregateLevel.TO_SEQUENCE)
@@ -256,7 +256,7 @@
last_seq 的使用示例如下（ first_seq 类似），详细见 last_seq 配置API。
+last_seq 的使用示例如下（ api_v2.layer_first_seq 类似），详细见 api_v2.layer_last_seq 配置API。
last = last_seq(input=layer,
agg_level=AggregateLevel.TO_SEQUENCE)
expand 的使用示例如下，详细见 ExpandLevel 配置API。
+expand 的使用示例如下，详细见 api_v2.layer_expand 配置API。
ex = expand(input=layer1,
expand_as=layer2,
expand_level=ExpandLevel.FROM_NO_SEQUENCE)
diff --git a/develop/doc_cn/objects.inv b/develop/doc_cn/objects.inv
index 62dc1f11802c1658c781df6e5b848d9625dafcaa..e1a1a249c4782467733fe8c815614892b7fab197 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 3532bace7f3b887553b9c4dd4604bbd1a4ebbd2f..0000000000000000000000000000000000000000
--- a/develop/doc_cn/py-modindex.html
+++ /dev/null
@@ -1,284 +0,0 @@
-
-
-
-
-
-
-
-
-
-
- Python 模块索引 — PaddlePaddle 文档
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
\ No newline at end of file
diff --git a/develop/doc_cn/searchindex.js b/develop/doc_cn/searchindex.js
index be677eb2b5618ebeafe7cfcff472777b13a37629..a6956aea96c0d25b076ee9ff8772f8d14d9b0516 100644
--- a/develop/doc_cn/searchindex.js
+++ b/develop/doc_cn/searchindex.js
@@ -1 +1 @@
-Search.setIndex({docnames:["api/index_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/data/data_reader","api/v2/data/dataset","api/v2/data/image","api/v2/fluid","api/v2/fluid/data_feeder","api/v2/fluid/evaluator","api/v2/fluid/executor","api/v2/fluid/initializer","api/v2/fluid/io","api/v2/fluid/layers","api/v2/fluid/nets","api/v2/fluid/optimizer","api/v2/fluid/param_attr","api/v2/fluid/profiler","api/v2/fluid/regularizer","api/v2/model_configs","api/v2/run_logic","build_and_install/build_cn","build_and_install/build_from_source_cn","build_and_install/docker_install_cn","build_and_install/index_cn","build_and_install/pip_install_cn","design/api","design/auto_gradient_check","design/backward","design/block","design/build_system/README","design/cluster_train/README","design/cluster_train/checkpointing","design/cluster_train/data_dispatch","design/cluster_train/large_model_dist_train","design/cluster_train/master_server","design/cluster_train/pserver_client","design/cluster_train/remote_parameter_updater","design/cluster_train/save_model","design/cluster_train/submit-job","design/concurrent_programming","design/cpp_data_feeding","design/csp","design/dist_refactor/distributed_architecture","design/dist_refactor/multi_cpu","design/dist_refactor/parameter_server","design/error_clip","design/evaluator","design/executor","design/file_manager/README","design/file_manager/pfs/pfsclient","design/float16","design/fluid","design/fluid_compiler","design/functions_operators_layers","design/gan_api","design/graph","design/graph_survey","design/if_else_op","design/infer_var_type","design/kernel_hint_design","design/memory_optimization","design/mkl/mkl_packed","design/mkl/mkldnn","design/mkl/mkldnn_fluid","design/model_format","design/multi_language_interface/00.why_plain_c","design/multi_language_interface/01.inference_implementation","design/operator_kernel_type","design/ops/rnn","design/ops/sequence_decoder","design/optimizer","design/paddle_nccl","design/parameter_average","design/parameters_in_cpp","design/profiler","design/program","design/prune","design/python_api","design/reader/README","design/refactorization","design/register_grad_op","design/regularization","design/releasing_process","design/scope","design/selected_rows","design/simple_op_design","design/speech/deep_speech_2","design/support_new_device","design/switch","design/switch_kernel","design/tensor_array","design/var_desc","dev/contribute_to_paddle_cn","dev/index_cn","dev/new_layer_cn","dev/new_op_cn","dev/use_eigen_cn","dev/write_docs_cn","faq/build_and_install/index_cn","faq/cluster/index_cn","faq/index_cn","faq/local/index_cn","faq/model/index_cn","faq/parameter/index_cn","getstarted/concepts/use_concepts_cn","getstarted/index_cn","getstarted/quickstart_cn","howto/capi/compile_paddle_lib_cn","howto/capi/index_cn","howto/capi/organization_of_the_inputs_cn","howto/capi/workflow_of_capi_cn","howto/cluster/cmd_argument_cn","howto/cluster/index_cn","howto/cluster/multi_cluster/fabric_cn","howto/cluster/multi_cluster/index_cn","howto/cluster/multi_cluster/k8s_aws_cn","howto/cluster/multi_cluster/k8s_cn","howto/cluster/multi_cluster/k8s_distributed_cn","howto/cluster/multi_cluster/openmpi_cn","howto/cluster/multi_cluster/src/k8s_data/README","howto/cluster/multi_cluster/src/k8s_train/README","howto/cluster/preparations_cn","howto/cmd_parameter/arguments_cn","howto/cmd_parameter/detail_introduction_cn","howto/cmd_parameter/index_cn","howto/cmd_parameter/use_case_cn","howto/index_cn","howto/optimization/cpu_profiling_cn","howto/optimization/gpu_profiling_cn","howto/read_source","howto/rnn/hierarchical_layer_cn","howto/rnn/hrnn_rnn_api_compare_cn","howto/rnn/index_cn","howto/rnn/recurrent_group_cn","howto/rnn/rnn_config_cn","index_cn","mobile/cross_compiling_for_android_cn","mobile/cross_compiling_for_ios_cn","mobile/cross_compiling_for_raspberry_cn","survey/cluster_bootstrapping_tools"],envversion:50,filenames:["api/index_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/data/data_reader.rst","api/v2/data/dataset.rst","api/v2/data/image.rst","api/v2/fluid.rst","api/v2/fluid/data_feeder.rst","api/v2/fluid/evaluator.rst","api/v2/fluid/executor.rst","api/v2/fluid/initializer.rst","api/v2/fluid/io.rst","api/v2/fluid/layers.rst","api/v2/fluid/nets.rst","api/v2/fluid/optimizer.rst","api/v2/fluid/param_attr.rst","api/v2/fluid/profiler.rst","api/v2/fluid/regularizer.rst","api/v2/model_configs.rst","api/v2/run_logic.rst","build_and_install/build_cn.md","build_and_install/build_from_source_cn.rst","build_and_install/docker_install_cn.rst","build_and_install/index_cn.rst","build_and_install/pip_install_cn.rst","design/api.md","design/auto_gradient_check.md","design/backward.md","design/block.md","design/build_system/README.md","design/cluster_train/README.md","design/cluster_train/checkpointing.md","design/cluster_train/data_dispatch.md","design/cluster_train/large_model_dist_train.md","design/cluster_train/master_server.md","design/cluster_train/pserver_client.md","design/cluster_train/remote_parameter_updater.md","design/cluster_train/save_model.md","design/cluster_train/submit-job.md","design/concurrent_programming.md","design/cpp_data_feeding.md","design/csp.md","design/dist_refactor/distributed_architecture.md","design/dist_refactor/multi_cpu.md","design/dist_refactor/parameter_server.md","design/error_clip.md","design/evaluator.md","design/executor.md","design/file_manager/README.md","design/file_manager/pfs/pfsclient.md","design/float16.md","design/fluid.md","design/fluid_compiler.md","design/functions_operators_layers.md","design/gan_api.md","design/graph.md","design/graph_survey.md","design/if_else_op.md","design/infer_var_type.md","design/kernel_hint_design.md","design/memory_optimization.md","design/mkl/mkl_packed.md","design/mkl/mkldnn.md","design/mkl/mkldnn_fluid.md","design/model_format.md","design/multi_language_interface/00.why_plain_c.md","design/multi_language_interface/01.inference_implementation.md","design/operator_kernel_type.md","design/ops/rnn.md","design/ops/sequence_decoder.md","design/optimizer.md","design/paddle_nccl.md","design/parameter_average.md","design/parameters_in_cpp.md","design/profiler.md","design/program.md","design/prune.md","design/python_api.md","design/reader/README.md","design/refactorization.md","design/register_grad_op.md","design/regularization.md","design/releasing_process.md","design/scope.md","design/selected_rows.md","design/simple_op_design.md","design/speech/deep_speech_2.md","design/support_new_device.md","design/switch.md","design/switch_kernel.md","design/tensor_array.md","design/var_desc.md","dev/contribute_to_paddle_cn.md","dev/index_cn.rst","dev/new_layer_cn.rst","dev/new_op_cn.md","dev/use_eigen_cn.md","dev/write_docs_cn.rst","faq/build_and_install/index_cn.rst","faq/cluster/index_cn.rst","faq/index_cn.rst","faq/local/index_cn.rst","faq/model/index_cn.rst","faq/parameter/index_cn.rst","getstarted/concepts/use_concepts_cn.rst","getstarted/index_cn.rst","getstarted/quickstart_cn.rst","howto/capi/compile_paddle_lib_cn.md","howto/capi/index_cn.rst","howto/capi/organization_of_the_inputs_cn.md","howto/capi/workflow_of_capi_cn.md","howto/cluster/cmd_argument_cn.md","howto/cluster/index_cn.rst","howto/cluster/multi_cluster/fabric_cn.md","howto/cluster/multi_cluster/index_cn.rst","howto/cluster/multi_cluster/k8s_aws_cn.md","howto/cluster/multi_cluster/k8s_cn.md","howto/cluster/multi_cluster/k8s_distributed_cn.md","howto/cluster/multi_cluster/openmpi_cn.md","howto/cluster/multi_cluster/src/k8s_data/README.md","howto/cluster/multi_cluster/src/k8s_train/README.md","howto/cluster/preparations_cn.md","howto/cmd_parameter/arguments_cn.md","howto/cmd_parameter/detail_introduction_cn.md","howto/cmd_parameter/index_cn.rst","howto/cmd_parameter/use_case_cn.md","howto/index_cn.rst","howto/optimization/cpu_profiling_cn.md","howto/optimization/gpu_profiling_cn.rst","howto/read_source.md","howto/rnn/hierarchical_layer_cn.rst","howto/rnn/hrnn_rnn_api_compare_cn.rst","howto/rnn/index_cn.rst","howto/rnn/recurrent_group_cn.md","howto/rnn/rnn_config_cn.rst","index_cn.rst","mobile/cross_compiling_for_android_cn.md","mobile/cross_compiling_for_ios_cn.md","mobile/cross_compiling_for_raspberry_cn.md","survey/cluster_bootstrapping_tools.md"],objects:{"paddle.v2":{image:[11,0,0,"-"]},"paddle.v2.image":{batch_images_from_tar:[11,1,1,""],center_crop:[11,1,1,""],left_right_flip:[11,1,1,""],load_and_transform:[11,1,1,""],load_image:[11,1,1,""],load_image_bytes:[11,1,1,""],random_crop:[11,1,1,""],resize_short:[11,1,1,""],simple_transform:[11,1,1,""],to_chw:[11,1,1,""]}},objnames:{"0":["py","module","Python \u6a21\u5757"],"1":["py","function","Python \u51fd\u6570"]},objtypes:{"0":"py:module","1":"py:function"},terms:{"00m":134,"01org":104,"03m":134,"0424m":134,"0473v3":5,"04\u4ee5\u4e0a":30,"04\u4ee5\u53camaco":112,"055ee37d":121,"0630u":134,"06u":134,"0810u":134,"0957m":134,"0\u53f7\u8bad\u7ec3\u8282\u70b9\u662f\u4e3b\u8bad\u7ec3\u8282\u70b9":129,"0\u5c42\u5e8f\u5217":136,"0_cudnn5":27,"0_cudnn5_avx_mkl":[28,30],"0_cudnn7_avx_mkl":30,"0ab":4,"0rc1":88,"0rc2":88,"0x10f256d50":62,"0x7ffe4de00110":62,"100gi":121,"100m":107,"10g":44,"1150u":134,"11\u5b9e\u73b0\u4e86c":72,"11e6":122,"124n":134,"12\u4ee5\u4e0a":30,"12\u64cd\u4f5c\u7cfb\u7edf":104,"12gb":66,"13m":122,"1490u":134,"14\u7248\u672c\u4ee5\u4e0a\u7684":144,"14\u8fd9\u79cd\u5199\u6cd5\u5c06\u4f1a\u6d4b\u8bd5\u6a21\u578b":131,"1550u":134,"15\u884c":137,"16\u5b57\u8282\u8868\u793a\u4fdd\u5b58\u7684\u53c2\u6570\u603b\u4e2a\u6570":109,"16u":134,"173n":134,"1770u":134,"18ad":121,"18e457ce3d362ff5f3febf8e7f85ffec852f70f3b629add10aed84f930a68750":122,"197u":134,"1\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2a\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":27,"1\u7684\u5c42\u4e4b\u5916":131,"1\u7a00\u758f\u6570\u636e":100,"1\u8f6e\u5b58\u50a8\u7684\u6240\u6709\u6a21\u578b":131,"1st":18,"210u":134,"211839e770f7b538e2d8":5,"215n":134,"228u":134,"2520u":134,"2680u":134,"26\u884c":137,"279n":134,"27m":134,"285m":134,"2863m":134,"28m":134,"2977m":134,"2\u4e09\u7c7b\u7684\u6bd4\u4f8b\u4e3a":109,"2\u4e2a\u5b50\u5e8f\u5217":115,"2\u5206\u522b\u4ee3\u88683\u4e2a\u8282\u70b9\u7684trainer":123,"2\u610f\u5473\u77400\u53f7\u548c1\u53f7gpu\u5c06\u4f1a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97fc1\u548cfc2\u5c42":131,"2\u8fd9\u51e0\u4e2a\u76ee\u5f55\u8868\u793apaddlepaddle\u8282\u70b9\u4e0etrain":123,"2cbf7385":121,"2nd":18,"302n":134,"30u":134,"328n":134,"32u":134,"32x32":10,"331n":134,"3320u":134,"365e":121,"36u":134,"3710m":134,"3768m":134,"387u":134,"38u":134,"3920u":134,"39u":134,"3\u4ee5\u4e0a\u7684\u7b26\u53f7":30,"3\u53f7gpu":107,"4035m":134,"4090u":134,"4096mb":129,"4279m":134,"43u":134,"448a5b355b84":122,"4560u":134,"4563m":134,"45u":134,"4650u":134,"4726m":134,"473m":122,"4\u4e2a\u5e8f\u5217\u7684\u957f\u5ea6\u5206\u522b\u4e3a":115,"4\u5b57\u8282\u8868\u793apaddlepaddle\u7248\u672c\u4fe1\u606f":109,"4gb":129,"500m":107,"50bd":121,"50gi":121,"514u":134,"525n":134,"526u":134,"536u":134,"5460u":134,"5470u":134,"54u":134,"5690m":134,"573u":134,"578n":134,"5798m":134,"586u":134,"58s":122,"5969m":134,"5\u4f5c\u4e3a\u7f16\u8bd1\u73af\u5883":30,"5\u5373\u5c06\u505c\u6b62\u7ef4\u62a4":30,"5_cudnn5_avx_mkl":30,"5_cudnn5_avx_openbla":[30,112],"6080u":134,"6140u":134,"6305m":134,"639u":134,"64\u5e73\u53f0\u4e3a\u4f8b":142,"64m":70,"655u":134,"6780u":134,"6810u":134,"682u":134,"6970u":134,"6\u4e07\u4ebf\u6b21\u6d6e\u70b9\u8fd0\u7b97\u6bcf\u79d2":134,"6\u4ee5\u4e0a":[30,112],"6\u4f5c\u4e3a\u6807\u51c6\u7f16\u8bd1\u73af\u5883":30,"6ce9":121,"6th":10,"704u":134,"7090u":134,"72u":134,"73u":134,"75u":134,"760u":134,"767u":134,"783n":134,"784u":134,"78m":134,"7\u4ee5\u4e0a":142,"7\u4ee5\u4e0a\u7684\u7b26\u53f7":30,"7\u4ee5\u4e0b":142,"7\u548cpip":104,"7\u7248\u672c\u5f00\u59cb":142,"7\u7cfb\u5217":30,"7eamaa":10,"7kb":122,"8000\u5c31\u53ef\u4ee5\u5728\u7f51\u9875\u4e0a\u751f\u6210\u9700\u8981\u7684\u6587\u6863":103,"8250u":134,"8300u":134,"830n":134,"849m":134,"861u":134,"8661m":134,"892m":134,"8\u5b57\u8282\u8868\u793a\u6bcf\u4e2a\u53c2\u6570\u5360\u7528\u7684\u5b57\u8282\u6570":109,"901n":134,"90u":134,"918u":134,"9247m":134,"924n":134,"9261m":134,"9330m":134,"94u":134,"9530m":134,"983m":134,"988u":134,"997u":134,"99u":134,"9a235":143,"9f18":122,"\u4e00\u4e2a":115,"\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a0\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u4e0d\u5171\u4eab\u7684\u4f8b\u5b50\u662f":101,"\u4e00\u4e2a\u5178\u578b\u7684chunk\u5982\u4e0b\u6240\u793a":54,"\u4e00\u4e2a\u5206\u5e03\u5f0fpaddlepaddle\u8bad\u7ec3\u4efb\u52a1\u4e2d":122,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8fdb\u5165":139,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8fdb\u5165":139,"\u4e00\u4e2a\u53cc\u5c42rnn\u7531\u591a\u4e2a\u5355\u5c42rnn\u7ec4\u6210":139,"\u4e00\u4e2a\u53ef\u8c03\u7528\u7684\u51fd\u6570":139,"\u4e00\u4e2a\u5e38\u7528\u7684cmake\u914d\u7f6e\u5982\u4e0b":144,"\u4e00\u4e2a\u6570\u636e\u96c6\u5927\u90e8\u5206\u5e8f\u5217\u957f\u5ea6\u662f100":107,"\u4e00\u4e2a\u662f\u6d6e\u70b9\u8ba1\u7b97\u91cf":134,"\u4e00\u4e2a\u72ec\u7acb\u7684\u5143\u7d20":136,"\u4e00\u4e2a\u72ec\u7acb\u7684\u8bcd\u8bed":136,"\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u6a21\u578b\u7531\u5927\u91cf\u7684\u53c2\u6570\u7ec4\u6210":37,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad\u90e8\u5206\u628a\u8f93\u5165\u8f6c\u5316\u4e3a\u76f8\u5e94\u7684\u8f93\u51fa":100,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u53c2\u6570\u662f\u5728":100,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684c":100,"\u4e00\u4e2a\u8f93\u51fa\u6570\u636e\u540c\u6837\u88ab\u7ec4\u7ec7\u4e3a\u4e00\u4e2a":115,"\u4e00\u4e2a\u8f93\u51fa\u7ec4\u6210":101,"\u4e00\u4e2a\u91cd\u8981\u7684\u95ee\u9898\u662f\u9009\u62e9\u6b63\u786e\u7684learning_r":109,"\u4e00\u4e2achunk\u7531\u6240\u5728\u7684\u6587\u4ef6\u504f\u79fb":54,"\u4e00\u4e2agpu\u8bbe\u5907\u4e0a\u4e0d\u5141\u8bb8\u914d\u7f6e\u591a\u4e2a\u6a21\u578b":129,"\u4e00\u4e2agradientmachine\u7c7b\u7684\u5bf9\u8c61\u7ba1\u7406\u7740\u4e00\u7ec4\u8ba1\u7b97\u5c42":116,"\u4e00\u4e2alabel":137,"\u4e00\u4e2amemory\u5305\u542b":140,"\u4e00\u4e2aposix\u517c\u5bb9\u7684\u6587\u4ef6\u7cfb\u7edf":54,"\u4e00\u4e9b\u60c5\u51b5\u4e3a\u4e86\u4fbf\u4e8e\u53d1\u5e03":116,"\u4e00\u53e5\u8bdd\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":139,"\u4e00\u53f0\u7535\u8111":26,"\u4e00\u662fbatch":107,"\u4e00\u6837\u7684\u65b9\u5f0f":26,"\u4e00\u7ef4\u6570\u7ec4":[115,116],"\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":115,"\u4e00\u81f4":[136,137],"\u4e00\u822c\u4e0d\u5141\u8bb8\u518d\u4ece":88,"\u4e00\u822c\u4ece":98,"\u4e00\u822c\u5728paddlepaddle\u4e2d":137,"\u4e00\u822c\u662f\u7531\u4e8e\u76f4\u63a5\u4f20\u9012\u5927\u5b57\u5178\u5bfc\u81f4\u7684":109,"\u4e00\u822c\u6765\u8bf4":140,"\u4e00\u822c\u7531mkl":68,"\u4e00\u822c\u8868\u793a":137,"\u4e00\u822c\u8bbe\u7f6e":109,"\u4e00\u8282":116,"\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":110,"\u4e0a":98,"\u4e0a\u4ea4\u53c9\u7f16\u8bd1raspberri":144,"\u4e0a\u4f20\u5230cloud\u6216\u8005\u4e0b\u8f7d\u5230\u672c\u5730\u7684\u65f6\u95f4\u53ef\u80fd\u6bd4\u8f83\u957f":54,"\u4e0a\u4f20\u65b9\u6cd5":88,"\u4e0a\u4f20\u8ba1\u7b97\u5f97\u51fa\u7684\u68af\u5ea6":118,"\u4e0a\u56fe\u4e2d\u7684":115,"\u4e0a\u56fe\u4e2d\u865a\u7ebf\u7684\u8fde\u63a5":137,"\u4e0a\u56fe\u63cf\u8ff0\u4e86\u4e00\u4e2a3\u8282\u70b9\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3\u573a\u666f":123,"\u4e0a\u6ce8\u518c\u4e00\u4e0b":54,"\u4e0a\u7f16\u8bd1\u5f88\u6162":26,"\u4e0a\u8fd0\u884c":142,"\u4e0a\u8ff0\u4ee3\u7801\u5c06bias\u5168\u90e8\u521d\u59cb\u5316\u4e3a1":109,"\u4e0a\u8ff0\u547d\u4ee4\u4e2d":28,"\u4e0a\u8ff0\u547d\u4ee4\u7f16\u8bd1\u51fa\u4e00\u4e2a":26,"\u4e0a\u8ff0\u7684":108,"\u4e0a\u8ff0\u7684\u4ee3\u7801\u7247\u6bb5\u5305\u542b\u4e86\u4e24\u79cd\u65b9\u6cd5":134,"\u4e0a\u8ff0paddlepaddl":88,"\u4e0a\u9762\u7684\u4ee3\u7801\u5728":101,"\u4e0a\u9762\u7684\u4ee3\u7801\u9996\u5148\u5bfc\u5165\u4f9d\u8d56\u7684\u5305":101,"\u4e0b":[101,103],"\u4e0b\u4e00\u4e2awheel\u5305\u9700\u8981\u66f4\u65b0\u7248\u672c\u53f7\u624d\u53ef\u4ee5\u4e0a\u4f20":88,"\u4e0b\u4f1a\u770b\u5230\u5982\u4e0b\u76ee\u5f55\u7ed3\u6784":113,"\u4e0b\u540c":109,"\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":134,"\u4e0b\u56fe\u662f\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u793a\u610f\u56fe":100,"\u4e0b\u56fe\u662fcsr\u5b58\u50a8\u7a00\u758f\u77e9\u9635\u7684\u793a\u610f\u56fe":115,"\u4e0b\u5b58\u653e\u516c\u5171\u6570\u636e\u96c6\u5408":38,"\u4e0b\u627e\u5230":113,"\u4e0b\u62c9\u6846\u4e2d\u627e\u5230\u751f\u6210\u76843\u4e2a\u4e8c\u8fdb\u5236\u6587\u4ef6":88,"\u4e0b\u6587\u4ee5nlp\u4efb\u52a1\u4e3a\u4f8b":139,"\u4e0b\u6587\u4f1a\u8be6\u7ec6\u8fdb\u884c\u4ecb\u7ecd":115,"\u4e0b\u6587\u4f7f\u7528":123,"\u4e0b\u6587\u5c31\u662f\u7528job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u8fdb\u884c\u8bad\u7ec3":122,"\u4e0b\u6587\u8be6\u7ec6\u89e3\u91ca":115,"\u4e0b\u7684":[116,123],"\u4e0b\u8868\u5217\u51fa\u4e86python\u7aef\u8bad\u7ec3\u63a5\u53e3\u66b4\u9732\u7684\u6570\u636e\u7c7b\u578b":115,"\u4e0b\u8f7d":54,"\u4e0b\u8f7d\u5230\u672c\u5730":54,"\u4e0b\u8f7d\u5b8c\u6570\u636e\u540e":122,"\u4e0b\u8f7d\u5f97\u5230":88,"\u4e0b\u8f7d\u6307\u5b9a\u7248\u672c\u7684docker\u955c\u50cf":28,"\u4e0b\u8f7dgpu\u7248\u672c":28,"\u4e0b\u9762":116,"\u4e0b\u9762\u4e3e\u4e2a\u7b80\u5355\u7684\u4f8b\u5b50":134,"\u4e0b\u9762\u4ecb\u7ecd\u4ecb\u7ecd":101,"\u4e0b\u9762\u4ee5":117,"\u4e0b\u9762\u4ee5\u77e9\u9635\u4e58\u64cd\u4f5c":101,"\u4e0b\u9762\u4ee5addop\u4e3a\u4f8b\u8bf4\u660etensor\u7684\u4f7f\u7528\u8fc7\u7a0b":102,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u67d0\u4e00\u7c7b\u6587\u4ef6\u7684\u5b9e\u73b0\u65b9\u5f0f":72,"\u4e0b\u9762\u5217\u51fa\u4e86":140,"\u4e0b\u9762\u5217\u51fa\u4e86\u5168\u8fde\u63a5\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":100,"\u4e0b\u9762\u5c31\u6839\u636e\u8fd9\u51e0\u4e2a\u6b65\u9aa4\u5206\u522b\u4ecb\u7ecd":123,"\u4e0b\u9762\u6211\u4eec\u4f7f\u7528\u8fd9\u4e2a\u955c\u50cf\u6765\u4e0b\u8f7d\u6570\u636e\u5230docker":122,"\u4e0b\u9762\u662f":101,"\u4e0b\u9762\u662f\u5bf9":101,"\u4e0b\u9762\u662fc":116,"\u4e0b\u9762\u7684\u4ee3\u7801\u5c06\u968f\u673a\u751f\u6210\u7684\u77e9\u9635\u8f6c\u5316\u4e3a\u53ef\u4ee5\u88abpaddlepaddle\u52a0\u8f7d\u7684\u6a21\u578b\u53c2\u6570":109,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u4ecegithub\u62c9\u53d6\u6700\u65b0\u4ee3\u7801":113,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u521b\u5efa\u4e86\u4e00\u4e2a\u9ad8\u5ea6\u4e3a1":115,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u521b\u5efa\u4e86\u4e00\u4e2acpu\u4e0a\u7684\u4e8c\u503c\u7a00\u758f\u77e9\u9635":115,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u521b\u5efa\u4e86\u542b\u6709\u4e09\u4e2a\u5143\u7d20":115,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u5728\u521b\u5efa\u4e86\u4e00\u4e2acpu\u4e0a\u7684\u5e26\u5143\u7d20\u503c\u7684\u7a00\u758f\u77e9\u9635":115,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u5b9e\u73b0\u4e86":100,"\u4e0b\u9762\u7684\u70b9\u5b9e\u73b0\u4e86mulop\u7684\u5b9a\u4e49":101,"\u4e0b\u9762\u7ed9\u51fa\u4e86\u4e00\u4e2a\u4f8b\u5b50":100,"\u4e0b\u9762\u7ed9\u51fa\u5728\u4e09\u7ef4\u7a7a\u95f4\u4e2d\u4f7f\u7528\u7ebf\u6027\u56de\u5f52\u62df\u5408\u4e00\u6761\u76f4\u7ebf\u7684\u4f8b\u5b50":110,"\u4e0b\u9762\u8be6\u7ec6\u89e3\u91ca\u4ec0\u4e48\u662f":115,"\u4e0b\u9762\u8fd9\u4e9blayer\u80fd\u591f\u63a5\u53d7\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165":136,"\u4e0d\u4e00\u81f4\u7684\u7531pfsclient\u4e0b\u8f7d\u6216\u8005\u4f20\u8f93chunk\u5b8c\u6210":54,"\u4e0d\u4ec5\u8981\u63d0\u4f9b\u6bcf\u4e00\u4e2a\u5916\u5c42\u5e8f\u5217\u5728\u6574\u4e2a":115,"\u4e0d\u4f1a\u4fdd\u7559\u5728\u78c1\u76d8\u4e0a":26,"\u4e0d\u4f1a\u518d\u4ece":107,"\u4e0d\u4f1a\u865a\u62df\u4efb\u4f55\u786c\u4ef6":26,"\u4e0d\u4f7f\u7528\u9759\u6001\u5e93":71,"\u4e0d\u4f7f\u7528\u989d\u5916\u7a7a\u95f4":100,"\u4e0d\u4f7f\u7528c":71,"\u4e0d\u4f7f\u7528swig":71,"\u4e0d\u5141\u8bb8\u4e00\u4e2a\u6587\u4ef6\u4e2d\u5305\u542b\u591a\u4e2aop":101,"\u4e0d\u5171\u4eab\u5219\u4e0d\u52a0":101,"\u4e0d\u5171\u4eab\u7684\u4f8b\u5b50\u53ef\u4ee5\u53c2\u8003":101,"\u4e0d\u53ef\u4ee5\u66f4\u6539":88,"\u4e0d\u53ef\u518d\u8fdb\u884c\u62c6\u5206":115,"\u4e0d\u540c":68,"\u4e0d\u540c\u4e8e\u4e0a\u8ff0\u4ecb\u7ecd\u7684recurr":108,"\u4e0d\u540c\u4e8eop\u7684\u7f16\u8bd1\u6d4b\u8bd5":101,"\u4e0d\u540c\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u6570\u636e\u5927\u5c0f\u7684\u6700\u5927\u503c\u4e0e\u6700\u5c0f\u503c\u7684\u6bd4\u7387":129,"\u4e0d\u540c\u5e8f\u5217\u53ef\u80fd\u4f1a\u542b\u6709\u4e0d\u540c\u6570\u76ee\u4e2a\u5143\u7d20":115,"\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u662f\u4e0d\u540c\u7684":140,"\u4e0d\u540c\u7248\u672c\u7684\u7f16\u8bd1\u5668\u4e4b\u95f4":71,"\u4e0d\u540c\u7684\u4f18\u5316\u7b97\u6cd5\u9700\u8981\u4f7f\u7528\u4e0d\u540c\u5927\u5c0f\u7684\u5185\u5b58":107,"\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":123,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":110,"\u4e0d\u540c\u8ba1\u7b97\u5c42\u5bf9\u7a7a\u8f93\u5165\u7684\u5904\u7406\u7b56\u7565\u6709\u53ef\u80fd\u4e0d\u540c":115,"\u4e0d\u540c\u8bbe\u5907":101,"\u4e0d\u540c\u8bed\u8a00\u7684\u63a5\u53e3\u9002\u5e94\u4e0d\u540c\u8bed\u8a00\u7684\u7279\u6027":71,"\u4e0d\u540c\u8f93\u5165\u542b\u6709\u7684\u5b50\u53e5":139,"\u4e0d\u540c\u8f93\u5165\u5e8f\u5217\u542b\u6709\u7684\u8bcd\u8bed\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":139,"\u4e0d\u540cdataprovider\u5bf9\u6bd4\u5982\u4e0b":137,"\u4e0d\u540crank\u7684tensor\u662f\u4e0d\u540c\u7c7b\u578b":102,"\u4e0d\u5728":72,"\u4e0d\u5bb9\u6613\u51fa\u9519":54,"\u4e0d\u5d4c\u5165\u5176\u4ed6\u8bed\u8a00\u89e3\u91ca\u5668":71,"\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":71,"\u4e0d\u5e94\u8be5\u88ab\u62c6\u89e3":139,"\u4e0d\u6307\u5b9a\u65f6":139,"\u4e0d\u652f\u6301":115,"\u4e0d\u652f\u6301\u5e8f\u5217\u957f\u5ea6\u4e3a":115,"\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":110,"\u4e0d\u662f\u771f\u6b63\u7684layer":108,"\u4e0d\u662f\u901a\u8fc7\u4e00\u822c\u7684\u65b9\u5f0f\u6765\u5b9e\u73b0\u5bf9\u8f93\u51fa\u7684\u6fc0\u6d3b":108,"\u4e0d\u663e\u793a\u7684\u5199\u6bcf\u4e2a\u7c7b\u5177\u4f53\u5305\u542b\u4ec0\u4e48":71,"\u4e0d\u6ee1\u8db3\u94a9\u5b50\u7684":98,"\u4e0d\u7528mount\u7684\u65b9\u5f0f\u6765\u8bbf\u95ee\u6570\u636e":38,"\u4e0d\u80fd\u4fee\u6539op\u7684\u6210\u5458\u53d8\u91cf":101,"\u4e0d\u80fd\u592a\u968f\u610f":98,"\u4e0d\u80fd\u88ab\u63d0\u4ea4\u5230":98,"\u4e0d\u8981\u5728\u6ce8\u91cd\u6027\u80fd\u7684\u8bad\u7ec3\u573a\u666f\u4e0b\u4f7f\u7528":107,"\u4e0d\u8bba\u5e8f\u5217\u4e2d\u7684\u5143\u7d20\u5728\u5185\u5b58\u4e2d\u5360\u7528\u591a\u5c11\u5b9e\u9645\u5b58\u50a8\u7a7a\u95f4":115,"\u4e0d\u8bba\u6570\u636e\u57df\u662f":115,"\u4e0d\u8bba\u662f\u4e00\u7ef4\u6574\u578b\u6570\u7ec4\u8fd8\u662f\u4e8c\u7ef4\u6d6e\u70b9\u6570\u77e9\u9635":115,"\u4e0d\u8bba\u662f\u5355\u5c42\u5e8f\u5217\u8fd8\u662f\u53cc\u5c42\u5e8f\u5217\u7684\u5e8f\u5217\u4fe1\u606f":115,"\u4e0d\u8fc7\u5b9e\u9645\u4e0a\u662f\u8fd0\u884c\u5728\u4e00\u4e2a":26,"\u4e0d\u9700\u5728\u4f7f\u7528c":116,"\u4e0d\u9700\u8981\u4f9d\u8d56\u5176\u4ed6\u4efb\u4f55\u8f6f\u4ef6\u4e86":26,"\u4e0d\u9700\u8981\u63d0\u4f9b\u5143\u7d20\u503c":115,"\u4e0d\u9700\u8981\u8bbe\u7f6e":142,"\u4e0e":[68,101,123,133],"\u4e0e\u4e4b\u76f8\u5bf9\u7684\u662flocal":54,"\u4e0e\u5176\u4ed6\u7b2c\u4e09\u65b9\u5e93\u4e00\u6837":68,"\u4e0e\u5176\u5b83":116,"\u4e0e\u529f\u80fd\u5206\u652f\u4e0d\u540c\u7684\u662f":88,"\u4e0e\u5355\u5c42rnn\u7684\u914d\u7f6e\u7c7b\u4f3c":137,"\u4e0e\u53ef\u80fd\u6709\u7684":88,"\u4e0e\u540c\u6b65sgd\u76f8\u6bd4":118,"\u4e0e\u5bfb\u627epython\u4ee3\u7801\u7684\u6027\u80fd\u74f6\u9888\u7c7b\u4f3c":133,"\u4e0e\u5f53\u524d\u7684\u8870\u51cf\u56e0\u5b50\u7684\u4e58\u79ef":109,"\u4e0e\u672c\u5730\u8bad\u7ec3\u76f8\u540c":119,"\u4e0e\u6b64\u4e0d\u540c\u7684\u662f":123,"\u4e0e\u8c03\u4f18":133,"\u4e0e\u8f93\u5165\u4e0d\u540c\u7684\u662f":116,"\u4e0e\u8fd9\u4e2a\u8bad\u7ec3\u6570\u636e\u4ea4\u4e92\u7684layer":107,"\u4e0ebatch":67,"\u4e0ejob":123,"\u4e0eoperator\u524d\u5411\u8ba1\u7b97\u7684\u8f93\u51fa\u8fdb\u884c\u5bf9\u6bd4":101,"\u4e0eoperator\u6ce8\u518c\u65f6\u6ce8\u518c\u7684\u7c7b\u578b\u4e00\u81f4":101,"\u4e0epython\u4e0d\u540c":133,"\u4e14\u4e0d\u6392\u9664commit\u4e4b\u95f4\u7684\u4fee\u6539\u5b58\u5728\u76f8\u4e92\u8986\u76d6\u7684\u60c5\u51b5":98,"\u4e14\u4f7f\u7528":113,"\u4e14\u589e\u52a0\u4e00\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00":71,"\u4e14\u5c55\u793a\u6548\u679c\u66f4\u597d":133,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":110,"\u4e14\u6bcf\u4e2a\u53e5\u5b50\u8868\u793a\u4e3a\u5bf9\u5e94\u7684\u8bcd\u8868\u7d22\u5f15\u6570\u7ec4":137,"\u4e14\u8c03\u7528\u65f6\u4e0d\u80fd\u629b\u51fa\u5f02\u5e38\u6216\u51fa\u73b0\u8fd0\u884c\u65f6\u9519\u8bef":72,"\u4e14c99\u652f\u6301bool\u7c7b\u578b\u548c\u5b9a\u957f\u6574\u6570":71,"\u4e14c99\u76f8\u5bf9\u4e8ec11\u4f7f\u7528\u66f4\u52a0\u5e7f\u6cdb":71,"\u4e24\u4e2a\u5b50\u76ee\u5f55\u4e0b":103,"\u4e24\u4e2a\u5d4c\u5957\u7684":139,"\u4e24\u4e2a\u64cd\u4f5c":134,"\u4e24\u4e2a\u8f93\u5165\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u4e5f\u5e76\u4e0d\u76f8\u540c":137,"\u4e24\u4e2a\u90e8\u5206":103,"\u4e24\u4e2a\u9690\u5c42\u7684\u7b80\u5355\u5168\u8fde\u63a5\u7f51\u7edc":116,"\u4e24\u6b21":115,"\u4e24\u79cd\u5e38\u7528\u7684\u6a21\u578b\u52a0\u8f7d\u65b9\u5f0f":116,"\u4e24\u79cd\u65b9\u6cd5\u7684\u533a\u522b":107,"\u4e24\u79cdblas\u5e93":27,"\u4e24\u8005\u90fd\u662f\u5bf9\u68af\u5ea6\u7684\u622a\u65ad":107,"\u4e25\u683c\u7684\u547d\u540d\u89c4\u8303pep":88,"\u4e2a\u5185\u5b58\u6c60\u5b9e\u9645\u4e0a\u51b3\u5b9a\u4e86shuffle\u7684\u7c92\u5ea6":107,"\u4e2a\u6027\u5316\u63a8\u8350":88,"\u4e2a\u6279\u6b21\u7684\u53c2\u6570\u5e73\u5747\u503c\u8fdb\u884c\u6d4b\u8bd5":129,"\u4e2a\u6a21\u578b\u6d4b\u8bd5\u6570\u636e":129,"\u4e2d":[67,68,71,72,100,101,102,107,115,123,133],"\u4e2d\u4e0d\u8981\u6dfb\u52a0\u5927\u6587\u4ef6\u7b49":98,"\u4e2d\u4f1a\u4f7f\u7528\u5230\u7684\u5b57\u5178\u6570\u636e\u6587\u4ef6":117,"\u4e2d\u4f1a\u63d0\u4f9b\u4e00\u4e9b\u5fc5\u8981\u7684\u63a5\u53e3\u548c\u51fd\u6570":68,"\u4e2d\u4f20\u5165\u53c2\u6570":117,"\u4e2d\u4f20\u5165\u7684\u53c2\u6570":117,"\u4e2d\u5143\u7d20\u4e2a\u6570\u603b\u662f\u7b49\u4e8e\u884c\u6570":115,"\u4e2d\u5143\u7d20\u7684\u4e2a\u6570\u7b49\u4e8e\u7f51\u7edc\u4e2d\u8f93\u51fa\u5c42\u7684\u4e2a\u6570":107,"\u4e2d\u5173\u4e8e\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc\u7684\u4ecb\u7ecd":137,"\u4e2d\u5199\u5165json\u5185\u5bb9":37,"\u4e2d\u5305\u542b\u4e00\u4e2araspberri":144,"\u4e2d\u5305\u542b\u6240\u4f9d\u8d56\u7684\u6240\u6709\u7b2c\u4e09\u65b9\u5e93":142,"\u4e2d\u5305\u542b\u82e5\u5e72\u4e2a\u4e0d\u540candroid":142,"\u4e2d\u5305\u542bc":[142,144],"\u4e2d\u5355\u5143\u6d4b\u8bd5\u7684\u4e00\u90e8\u5206":98,"\u4e2d\u5355\u5143\u6d4b\u8bd5\u80fd\u987a\u5229\u901a\u8fc7":98,"\u4e2d\u542b\u6709\u591a\u4e2a\u5e8f\u5217":115,"\u4e2d\u5b8c\u5168\u4e00\u81f4":71,"\u4e2d\u5b9a\u4e49":140,"\u4e2d\u5b9a\u4e49\u548c\u4f7f\u7528":139,"\u4e2d\u5b9e\u73b0\u4e86\u4e00\u4e2amerge\u7684\u65b9\u6cd5":68,"\u4e2d\u5b9e\u73b0\u7684\u7ed3\u6784\u4f53":72,"\u4e2d\u5bf9\u5e94\u7684layer\u5904":67,"\u4e2d\u5f15\u5165\u7684":67,"\u4e2d\u6253\u5370\u5176\u503c":107,"\u4e2d\u6307\u5b9a":129,"\u4e2d\u6307\u5b9a\u7684\u540d\u5b57":131,"\u4e2d\u63d0\u4f9b\u4e00\u4e2a\u4e0emkl\u6709\u5173\u7684\u603b\u5f00\u5173":68,"\u4e2d\u63d0\u4f9b\u4e86\u4e00\u4e9b\u5168\u5c40\u51fd\u6570\u7528\u6765\u5b9e\u73b0paddl":102,"\u4e2d\u641c\u7d22\u8fd9\u51e0\u4e2a\u5e93":27,"\u4e2d\u64cd\u4f5c":115,"\u4e2d\u6587\u6587\u6863":103,"\u4e2d\u6587\u6587\u6863\u76ee\u5f55":103,"\u4e2d\u6587\u7ef4\u57fa\u767e\u79d1\u9875\u9762":137,"\u4e2d\u6839\u636e":67,"\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7ed3\u679c\u77e9\u9635":107,"\u4e2d\u6bcf\u4e2apod\u7684ip\u5730\u5740":123,"\u4e2d\u6bcf\u5c42\u7684\u6570\u503c\u7edf\u8ba1":129,"\u4e2d\u6dfb\u52a0":67,"\u4e2d\u6dfb\u52a0\u4e00\u4e2a":68,"\u4e2d\u6dfb\u52a0\u4e24\u4e2a\u8f93\u5165":101,"\u4e2d\u7528\u4e8e\u5b58\u50a8\u6570\u636e\u7684":116,"\u4e2d\u7684":[102,116],"\u4e2d\u7684\u4e00\u884c":98,"\u4e2d\u7684\u4ee3\u7801\u4f5c\u4e3a\u5b9e\u4f8b":117,"\u4e2d\u7684\u504f\u79fb":115,"\u4e2d\u7684\u5bf9\u5e94\u5206\u652f\u5373\u53ef":98,"\u4e2d\u7684\u7248\u672c\u4fe1\u606f":88,"\u4e2d\u7684\u76f8\u5173\u811a\u672c":116,"\u4e2d\u7684\u8d77\u59cb\u504f\u79fb":115,"\u4e2d\u83b7\u53d6":123,"\u4e2d\u8bbe\u7f6e\u7684\u6240\u6709\u8282\u70b9":119,"\u4e2d\u8be6\u7ec6\u4ecb\u7ecd":100,"\u4e2d\u8c03\u7528":101,"\u4e2d\u8fd0\u884c\u4efb\u52a1\u7684\u89d2\u5ea6":38,"\u4e2d\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":140,"\u4e34\u65f6\u53d8\u91cf\u7b49\u7b49":107,"\u4e3a":[67,68,101,115,140,142,143,144],"\u4e3a\u4e86\u4f7f":101,"\u4e3a\u4e86\u4f7f\u8bc4\u5ba1\u4eba\u5728\u8bc4\u5ba1\u4ee3\u7801\u65f6\u66f4\u597d\u5730\u4e13\u6ce8\u4e8e\u4ee3\u7801\u672c\u8eab":98,"\u4e3a\u4e86\u4fdd\u8bc1\u6548\u7387":100,"\u4e3a\u4e86\u4fdd\u8bc1gpu\u9a71\u52a8\u80fd\u591f\u5728\u955c\u50cf\u91cc\u9762\u6b63\u5e38\u8fd0\u884c":28,"\u4e3a\u4e86\u51cf\u5c11\u751f\u6210\u94fe\u63a5\u5e93\u7684\u5927\u5c0f\u628a":113,"\u4e3a\u4e86\u5c01\u88c5\u80fd\u591f\u6b63\u786e\u5de5\u4f5c":100,"\u4e3a\u4e86\u5c3d\u53ef\u80fd\u5c11\u7684\u5728\u7236\u7c7blayer\u4e2d\u6dfb\u52a0\u53d8\u91cf\u6216\u8005\u51fd\u6570":68,"\u4e3a\u4e86\u5e94\u5bf9\u4ee5\u4e0a\u7684\u95ee\u9898":54,"\u4e3a\u4e86\u5f00\u53d1paddlepaddl":26,"\u4e3a\u4e86\u63cf\u8ff0\u65b9\u4fbf":139,"\u4e3a\u4e86\u65b9\u4fbf\u5927\u5bb6":98,"\u4e3a\u4e86\u66b4\u9732\u7684\u63a5\u53e3\u5c3d\u91cf\u7b80\u5355":72,"\u4e3a\u4e86\u66f4\u597d\u7684\u7b26\u5408paddlepaddle\u7684\u4ee3\u7801\u98ce\u683c":68,"\u4e3a\u4e86\u6700\u5927\u7a0b\u5ea6\u51cf\u5c11\u591a\u6b21\u8c03\u7528":67,"\u4e3a\u4e86\u751f\u6210\u66f4\u53ef\u8bfb\u7684\u6027\u80fd\u5206\u6790\u7ed3\u679c":133,"\u4e3a\u4e86\u7b80\u5316cmake\u914d\u7f6e":142,"\u4e3a\u4e86\u8fbe\u5230\u6027\u80fd\u6700\u4f18":134,"\u4e3a\u4e86\u8fbe\u5230\u6700\u5feb\u7684\u8ba1\u7b97\u901f\u5ea6":[142,143],"\u4e3a\u4e86\u8fdb\u4e00\u6b65\u63d0\u5347paddlepaddle\u5728\u57fa\u672c\u6570\u5b66\u8fd0\u7b97\u7684\u8ba1\u7b97\u901f\u5ea6":68,"\u4e3a\u4ec0\u4e48\u7528":26,"\u4e3a\u4f7f\u7528c":116,"\u4e3a\u4f8b":[101,108],"\u4e3a\u4f8b\u6765\u4ecb\u7ecd\u5982\u4f55\u5199\u5e26kernel\u7684oper":101,"\u4e3a\u53c2\u6570\u77e9\u9635\u7684\u5bbd\u5ea6":109,"\u4e3a\u5b83\u4eec\u9644\u52a0\u4e0a\u5e8f\u5217\u4fe1\u606f\u5c06\u53d8\u6210\u5e8f\u5217\u8f93\u5165":115,"\u4e3a\u5bb9\u5668\u5185\u6267\u884c\u7684\u547d\u4ee4":28,"\u4e3a\u60a8\u505a\u6027\u80fd\u8c03\u4f18\u63d0\u4f9b\u4e86\u65b9\u5411":134,"\u4e3a\u65b9\u4fbf\u4f5c\u4e1a\u542f\u52a8\u63d0\u4f9b\u4e86\u4e24\u4e2a\u72ec\u7279\u7684\u547d\u4ee4\u9009\u9879":119,"\u4e3a\u6b64":122,"\u4e3a\u6bcf\u4e00\u4e2a":[115,116],"\u4e3a\u6bcf\u4e00\u4e2a\u8f93\u5165":[115,116],"\u4e3a\u6bcf\u4e2aop\u521b\u5efa\u5355\u72ec\u7684":101,"\u4e3a\u8f93\u51fa\u5206\u914d\u5185\u5b58":100,"\u4e3aconst\u51fd\u6570":101,"\u4e3aoutput_\u7533\u8bf7\u5185\u5b58":100,"\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":129,"\u4e3b\u8981\u529f\u80fd\u5305\u62ec":54,"\u4e3b\u8981\u5305\u62ec":68,"\u4e3b\u8981\u5305\u62ec\u4e86\u6df1\u5ea6\u5b66\u4e60\u76f8\u5173\u7684\u6570\u5b66\u539f\u8bed\u4e0e\u64cd\u4f5c":68,"\u4e3b\u8981\u5305\u62ec\u56db\u79cd\u7c7b\u578b":110,"\u4e3b\u8981\u539f\u56e0\u5305\u62ec\u4e24\u4e2a\u65b9\u9762":107,"\u4e3b\u8981\u7528\u4e8epython":101,"\u4e3b\u8981\u9488\u5bf9paddlepaddle\u5728\u91cd\u6784\u4e4b\u524d\u7684\u4ee3\u7801\u6846\u67b6\u4ee5\u53cav1\u7684api":68,"\u4e3e\u4e00\u4e2a\u4f8b\u5b50":109,"\u4e3e\u4f8b":107,"\u4e3e\u4f8b\u8bf4\u660e":137,"\u4e4b\u524d":98,"\u4e4b\u540e":[100,110],"\u4e4b\u540e\u4f7f\u7528":100,"\u4e4b\u540e\u4f7f\u7528\u77e9\u9635\u8fd0\u7b97\u51fd\u6570\u6765\u8ba1\u7b97":100,"\u4e4b\u540e\u518d\u7528\u7f51\u9875\u8fde\u5230http":103,"\u4e4b\u540e\u521d\u59cb\u5316\u6240\u6709\u7684\u6743\u91cd\u77e9\u9635":100,"\u4e4b\u540e\u624d\u80fd\u5f00\u59cb\u7f16\u8bd1\u7684\u6b65\u9aa4":27,"\u4e4b\u5916\u7684\u6240\u6709\u5934\u6587\u4ef6":72,"\u4e4b\u7c7b\u7684\u7a0b\u5e8f\u6765\u7f16\u8bd1\u6e90\u7801":26,"\u4e4b\u95f4\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":139,"\u4e58\u4e0a\u8f93\u51fa\u7684\u68af\u5ea6":100,"\u4e58\u6cd5\u548c\u4e58\u6cd5\u68af\u5ea6\u7684\u8ba1\u7b97\u5360\u75282":133,"\u4e58\u9664\u7b49\u65f6\u5019":107,"\u4e5f\u4e0d\u4f7f\u7528\u5176\u4ed6\u52a8\u6001\u5e93":71,"\u4e5f\u4e0d\u5b58\u5728\u4e00\u4e2asubseq\u76f4\u63a5\u751f\u6210\u4e0b\u4e00\u4e2asubseq\u7684\u60c5\u51b5":139,"\u4e5f\u4e0d\u5e94\u8be5\u62a5\u9519":72,"\u4e5f\u4e0d\u751f\u6210":72,"\u4e5f\u4e0d\u80fd\u63a5\u6536\u5e8f\u5217\u6570\u636e\u4f5c\u4e3a\u8f93\u5165":108,"\u4e5f\u4f1a\u5360\u7528\u78c1\u76d8":26,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":98,"\u4e5f\u53ef\u4ee5\u4f7f\u7528\u8fd9\u4e9b\u955c\u50cf":88,"\u4e5f\u53ef\u4ee5\u5229\u7528paddlepaddl":103,"\u4e5f\u53ef\u4ee5\u662f\u4e00\u4e2a\u8bcd\u8bed":139,"\u4e5f\u53ef\u4ee5\u662f\u5728\u4efb\u52a1\u542f\u52a8\u524d\u4e0b\u8f7d\u5230\u672c\u5730\u7684":117,"\u4e5f\u53ef\u4ee5\u76f4\u63a5\u5728\u7f51\u9875\u9884\u89c8\u6587\u6863":103,"\u4e5f\u53ef\u4ee5\u8bf4\u662f\u67d0\u4e9b\u7279\u5b9a\u6307\u4ee4\u7684\u4f7f\u7528\u60c5\u51b5":134,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539":123,"\u4e5f\u53ef\u5199\u6210":101,"\u4e5f\u53ef\u81ea\u884c\u524d\u5f80\u5b98\u7f51\u4e0b\u8f7d":143,"\u4e5f\u53ef\u901a\u8fc7\u4ee5\u4e0b\u547d\u4ee4\u83b7\u53d6":142,"\u4e5f\u5c31\u662f":98,"\u4e5f\u5c31\u662f\u672c\u5730\u7684\u6e90\u7801\u6811\u6839\u76ee\u5f55\u91cc\u7684":26,"\u4e5f\u5c31\u662f\u7a7a\u8f93\u5165":115,"\u4e5f\u5c31\u662f\u81ea\u5df1\u7528\u6237\u540d\u4e0b\u7684":98,"\u4e5f\u5c31\u662f\u8bf4":[115,129,131],"\u4e5f\u5c31\u662f\u8bf4\u8f93\u51fa\u7684\u7ed3\u679c\u4e0d\u4f1a\u5728\u539f\u6765\u7684\u6570\u636e\u4e0a\u7d2f\u52a0":68,"\u4e5f\u5c31\u662fpaddlepaddle\u4e2d\u7684\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":115,"\u4e5f\u63cf\u8ff0\u4e86\u5bb9\u5668\u9700\u8981\u4f7f\u7528\u7684\u5b58\u50a8\u5377\u6302\u8f7d\u7684\u60c5\u51b5":123,"\u4e5f\u652f\u6301cpu\u7684\u6027\u80fd\u5206\u6790":134,"\u4e5f\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":137,"\u4e5f\u662fdecoder\u5faa\u73af\u5c55\u5f00\u7684\u4f9d\u636e":139,"\u4e5f\u6ca1\u7528":104,"\u4e66\u5199":71,"\u4e86":26,"\u4e86\u89e3\u5176\u8c03\u7528\u5173\u7cfb":133,"\u4e86\u89e3\u60a8\u7684\u786c\u4ef6":134,"\u4e86\u89e3\u66f4\u591a\u7ec6\u8282":140,"\u4e86\u89e3\u66f4\u591a\u8be6\u7ec6\u4fe1\u606f":140,"\u4e8c\u7ef4\u6d6e\u70b9\u578b\u77e9\u9635":115,"\u4e8c\u7ef4\u6d6e\u70b9\u6570\u77e9\u9635":115,"\u4e8c\u7ef4\u77e9\u9635":116,"\u4e8c\u7ef4\u77e9\u9635\u53ef\u4ee5\u8868\u793a\u884c\u5411\u91cf\u548c\u5217\u5411\u91cf":115,"\u4e8c\u8005\u8bed\u610f\u4e0a\u5b8c\u5168\u4e00\u81f4":137,"\u4e8e\u662f":115,"\u4e8e\u662f\u6211\u4eec\u53ef\u4ee5\u70b9\u51fb":133,"\u4e8e\u662f\u8fd9\u91cc\u4f7f\u7528":133,"\u4ea4\u4e92\u7684\u65b9\u6cd5":133,"\u4ea4\u53c9\u7f16\u8bd1\u5de5\u5177\u94fe\u4e3a":142,"\u4ea4\u53c9\u7f16\u8bd1android\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":142,"\u4ea4\u53c9\u7f16\u8bd1android\u7248\u672c\u7684paddlepaddle\u5e93\u65f6":142,"\u4ea4\u53c9\u7f16\u8bd1ios\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":143,"\u4ea4\u53c9\u7f16\u8bd1ios\u7248\u672c\u7684paddlepaddle\u5e93\u65f6":143,"\u4ea4\u53c9\u7f16\u8bd1raspberri":144,"\u4ea4\u7531cmake\u7cfb\u7edf\u672c\u8eab\u6765\u5904\u7406":142,"\u4ea6\u53ef\u4ee5\u901a\u8fc7\u624b\u52a8\u8bbe\u7f6e":[142,143],"\u4eab\u53d7\u60a8\u7684\u65c5\u7a0b":28,"\u4eba\u8138\u8bc6\u522b":38,"\u4ec0\u4e48\u662f":26,"\u4ec5\u4ec5\u4f7f\u7528":71,"\u4ec5\u4f1a\u5728\u652f\u6301avx2\u6307\u4ee4\u96c6\u53ca\u4ee5\u4e0a\u7684\u673a\u5668\u624d\u4f7f\u7528mkl":68,"\u4ec5\u5728\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u65f6\u6709\u6548":100,"\u4ec5\u5bf9\u7a00\u758f\u6570\u636e\u6709\u6548":100,"\u4ec5\u652f\u6301\u6574\u578b\u503c":115,"\u4ec5\u7528\u4e8e\u5b58\u50a8\u6574\u578b\u503c":116,"\u4ecb\u7ecd\u4e86\u4e00\u79cd\u901a\u8fc7ssh\u8fdc\u7a0b\u5206\u53d1\u4efb\u52a1":123,"\u4ecb\u7ecd\u4ea4\u53c9\u7f16\u8bd1android\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93\u7684\u65b9\u6cd5\u548c\u6b65\u9aa4":142,"\u4ecb\u7ecd\u4f7f\u7528paddlepaddl":117,"\u4ece":[88,105,134],"\u4ece0\u5230num":129,"\u4ece0\u5f00\u59cb\u7684\u6574\u6570":117,"\u4ece\u4e00\u4e2aword\u751f\u6210\u4e0b\u4e00\u4e2aword":139,"\u4ece\u5185\u6838\u51fd\u6570\u7684\u89d2\u5ea6":134,"\u4ece\u6a21\u578b\u6587\u4ef6\u5c06\u9884\u8bad\u7ec3\u53c2\u6570\u8f7d\u5165":109,"\u4ece\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u770b":137,"\u4ece\u6e90\u7801\u4ea4\u53c9\u7f16\u8bd1ios\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":143,"\u4ece\u6e90\u7801\u4ea4\u53c9\u7f16\u8bd1paddlepaddl":142,"\u4ece\u6e90\u7801\u7f16\u8bd1":29,"\u4ece\u78c1\u76d8\u52a0\u8f7d\u9884\u6d4b\u6a21\u578b":116,"\u4ece\u78c1\u76d8\u6587\u4ef6\u4e2d\u52a0\u8f7duuid\u6587\u4ef6\u540d\u7684\u68c0\u67e5\u70b9\u5feb\u7167\u6587\u4ef6":37,"\u4ece\u800c\u53ef\u4ee5\u505a\u4e00\u4e9b\u4e0e\u8ba1\u7b97\u91cd\u53e0\u7684\u5de5\u4f5c":100,"\u4ece\u800c\u5f15\u53d1\u5176\u4ed6\u8282\u70b9\u65e0\u6cd5\u8fde\u63a5\u5bfc\u81f4":105,"\u4ece\u800c\u907f\u514d\u4e86packing\u5197\u4f59":67,"\u4ece\u8bed\u4e49\u4e0a\u770b":139,"\u4ece\u8d77\u59cb\u7aef\u53e3\u76d1\u542c\u591a\u4e2a\u7aef\u53e3\u7528\u4e8e\u901a\u4fe1":117,"\u4ece\u8f93\u5165\u6570\u636e\u4e0a\u770b":137,"\u4ececmake":142,"\u4eceetcd\u4e2d\u8bfb\u53d6\u8282\u70b9":37,"\u4ecestart":129,"\u4ed3\u5e93\u7684\u8fdc\u7a0b\u4e3b\u673a":98,"\u4ed6\u4e3b\u8981\u5305\u542b\u4e86\u5b9e\u9645\u66b4\u9732\u7684\u7c7b\u578b\u7ed3\u6784":72,"\u4ed6\u4eec\u5206\u522b\u662f":137,"\u4ed6\u4eec\u5728\u81ea\u5df1\u7684":26,"\u4ed6\u4eec\u5728paddle\u7684\u6587\u6863\u548capi\u4e2d\u662f\u4e00\u4e2a\u6982\u5ff5":137,"\u4ed6\u662f\u5c06":72,"\u4ed6\u7684\u76ee\u6807\u662f\u4f7f\u7528c":71,"\u4ee3\u66ff":123,"\u4ee3\u7801\u4e2d9":137,"\u4ee3\u7801\u53c2\u8003":117,"\u4ee3\u7801\u5982\u4e0b":[107,108,109,140],"\u4ee3\u7801\u6ce8\u91ca\u8bf7\u9075\u5b88":98,"\u4ee3\u7801\u7247\u6bb5\u5982\u4e0b":115,"\u4ee3\u7801\u751f\u6210\u7684\u7b26\u53f7\u53ef\u80fd\u4e0d\u4e00\u81f4":71,"\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790":133,"\u4ee3\u7801\u793a\u4f8b\u5982\u4e0b":[101,116],"\u4ee3\u8868\u5bbf\u4e3b\u673a\u76ee\u5f55":123,"\u4ee3\u8868\u8fd9\u4e2alayer\u662f\u7528\u4e8e\u8dd1\u5728mkl":68,"\u4ee3\u8868\u8fd9\u4e2ashard\u7684\u6700\u5927index":38,"\u4ee3\u8868shard\u7684index":38,"\u4ee5":108,"\u4ee5\u4e0a":[98,142],"\u4ee5\u4e0a\u4e24\u79cd\u65b9\u5f0f\u53ea\u9700\u9009\u62e9\u5176\u4e00\u5373\u53ef":116,"\u4ee5\u4e0a\u4ee3\u7801\u7684reader\u8f93\u51fa\u7684data":38,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u4e0b\u751f\u6210100\u4e2a\u6587\u4ef6":38,"\u4ee5\u4e0b":38,"\u4ee5\u4e0b\u4ee3\u7801\u7247\u6bb5\u5b9a\u4e49":140,"\u4ee5\u4e0b\u547d\u4ee4\u542f\u52a8\u4e00\u4e2a":26,"\u4ee5\u4e0b\u6307\u4ee4\u80fd\u68c0\u67e5linux\u7535\u8111\u662f\u5426\u652f\u6301avx":28,"\u4ee5\u4e0b\u64cd\u4f5c\u5747\u5728head\u8282\u70b9\u4e2d\u6267\u884c":124,"\u4ee5\u4e0b\u6559\u7a0b\u5c06\u6307\u5bfc\u60a8\u63d0\u4ea4\u4ee3\u7801":98,"\u4ee5\u4e0b\u7b80\u79f0rnn":67,"\u4ee5\u4ea4\u4e92\u5f0f\u7684\u65b9\u5f0f\u6267\u884c\u6216\u8c03\u8bd5\u60a8\u7684\u4ee3\u7801":28,"\u4ee5\u4f7f\u7528":142,"\u4ee5\u4f7f\u7528adam\u7b97\u6cd5\u4e3a\u4f8b":109,"\u4ee5\u4fbf\u6211\u4eec\u53ef\u4ee5\u628a\u66f4\u591a\u7684\u7cbe\u529b\u653e\u5230\u903b\u8f91\u672c\u8eab\u4e0a":54,"\u4ee5\u4fbf\u83b7\u5f97\u8bad\u7ec3\u6570\u636e\u7684\u4f4d\u7f6e\u548c\u83b7\u53d6\u73af\u5883\u53d8\u91cf\u914d\u7f6e":117,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u7684\u6b63\u786e\u8ba1\u7b97":100,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u8ba1\u7b97\u7684\u6b63\u786e\u6027":100,"\u4ee5\u4fdd\u8bc1\u7f16\u8bd1\u9ad8\u6548":26,"\u4ee5\u53ca":[26,67,100,115],"\u4ee5\u53ca\u4f7f\u7528\u5b50\u5e8f\u5217\u6765\u5b9a\u4e49\u5206\u7ea7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":140,"\u4ee5\u53ca\u5207\u6362\u673a\u5668\u65f6\u9700\u8981\u65b0\u5b89\u88c5\u7684\u8f9b\u82e6":26,"\u4ee5\u53ca\u53cc\u5c42\u5e8f\u5217":136,"\u4ee5\u53ca\u5982\u4f55\u89e3\u6790\u795e\u7ecf\u7f51\u7edc\u524d\u5411\u8ba1\u7b97\u7684\u8f93\u51fa\u7ed3\u679c":115,"\u4ee5\u53ca\u76ee\u6807\u673a\u7248openblas\u5e93":144,"\u4ee5\u53ca\u76f8\u5173\u7684\u5c5e\u6027\u53c2\u6570":101,"\u4ee5\u53ca\u7b2c\u4e09\u65b9\u4f9d\u8d56\u94fe\u63a5\u5e93\u548c\u5934\u6587\u4ef6":113,"\u4ee5\u53ca\u8ba1\u7b97\u903b\u8f91\u5728\u5e8f\u5217\u4e0a\u7684\u5faa\u73af\u5c55\u5f00":139,"\u4ee5\u53ca\u8f93\u5165\u7684\u68af\u5ea6":100,"\u4ee5\u53caandroid":142,"\u4ee5\u53cagcc":27,"\u4ee5\u53canumpi":38,"\u4ee5\u53carelu":100,"\u4ee5\u63d0\u4f9b\u4e00\u4e9b\u9ed8\u8ba4\u7684\u7f16\u8bd1\u5668\u548c\u7f16\u8bd1\u53c2\u6570\u76f8\u5173\u914d\u7f6e":142,"\u4ee5\u63d0\u4f9b\u4e00\u4e9b\u9ed8\u8ba4\u7684\u7f16\u8bd1\u5668\u548c\u7f16\u8bd1\u53c2\u6570\u914d\u7f6e":143,"\u4ee5\u6b64\u8fbe\u5230\u6700\u597d\u7684\u6027\u80fd":68,"\u4ee5\u786e\u4fdd\u6240\u6709\u7684\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93\u548cpaddlepaddle\u4ee3\u7801\u90fd\u662f\u9488\u5bf9\u65b0\u7684cmake\u914d\u7f6e\u91cd\u65b0\u7f16\u8bd1\u7684":[142,143,144],"\u4ee5\u793a\u533a\u5206":[67,68],"\u4ee5\u8f93\u51fa":107,"\u4ee5\u9017\u53f7\u95f4\u9694":129,"\u4ee5\u907f\u514d\u94fe\u63a5\u4e0d\u5fc5\u8981\u7684\u5e93":113,"\u4ee5eigentensor\u4e3a\u4f8b":102,"\u4ee5embedding\u5c42\u4e3a\u4f8b":109,"\u4ee5lstm\u4e3a\u4f8b":108,"\u4efb\u4f55\u65f6\u5019\u5982\u679c\u9700\u8981\u6d6e\u70b9\u578b\u6570\u7ec4":115,"\u4efb\u52a1\u6765\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":119,"\u4efb\u52a1\u88ab\u8c03\u5ea6\u5728\u96c6\u7fa4\u4e2d\u65f6":120,"\u4efb\u610f\u5c06\u4e00\u4e9b\u6570\u636e\u7ec4\u5408\u6210\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":137,"\u4efb\u610f\u65f6\u523b\u53ea\u53ef\u80fd\u540c\u65f6\u6709\u4e00\u53f0\u670d\u52a1\u5668\u6545\u969c":37,"\u4f18\u5316\u524d":67,"\u4f18\u5316\u540e":67,"\u4f18\u5316\u5668\u5219\u7528\u94fe\u5f0f\u6cd5\u5219\u6765\u5bf9\u6bcf\u4e2a\u53c2\u6570\u8ba1\u7b97\u635f\u5931\u51fd\u6570\u7684\u68af\u5ea6":100,"\u4f1a\u4ea7\u751f\u5f53\u524dpython\u4e8c\u8fdb\u5236\u7684\u5b8c\u6574\u8def\u5f84":133,"\u4f1a\u4ee5":[67,68],"\u4f1a\u4f7f\u7528":116,"\u4f1a\u4f7f\u7528\u76f8\u540c\u7684\u539f\u6570\u636e":67,"\u4f1a\u5148\u4e34\u65f6\u4fdd\u5b58\u5728":68,"\u4f1a\u5148\u8fdb\u884c\u53c2\u6570\u7684\u521d\u59cb\u5316\u4e0e\u89e3\u6790":123,"\u4f1a\u5171\u4eab\u53c2\u6570":109,"\u4f1a\u5173\u8054\u53c2\u6570":108,"\u4f1a\u5206\u522b\u4ecb\u7ecd\u96c6\u7fa4\u4f5c\u4e1a\u7684\u542f\u52a8\u548c\u505c\u6b62\u65b9\u6cd5":120,"\u4f1a\u52a0\u8f7d\u4e0a\u4e00\u8f6e\u7684\u53c2\u6570":129,"\u4f1a\u53d8\u6210\u8bcd\u8868\u4e2d\u7684\u4f4d\u7f6e":137,"\u4f1a\u542f\u52a8pserver\u4e0etrainer\u8fdb\u7a0b":123,"\u4f1a\u5728":[68,103],"\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u751f\u6210\u4e24\u4e2a\u5b50\u76ee\u5f55":103,"\u4f1a\u5728\u7f16\u8bd1paddlepaddle\u7684\u65f6\u5019\u4e0b\u8f7d\u5e76\u7f16\u8bd1mkl":68,"\u4f1a\u5927\u4e0d\u76f8\u540c":117,"\u4f1a\u5bf9\u6bcf\u4e00\u4e2a\u6fc0\u6d3b\u6682\u5b58\u4e00\u4e9b\u6570\u636e":107,"\u4f1a\u5bf9\u8bad\u7ec3\u6027\u80fd\u9020\u6210\u5f71\u54cd":107,"\u4f1a\u5bf9\u8fd9\u7c7b\u8f93\u5165\u8fdb\u884c\u62c6\u89e3":139,"\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7248\u672cpython\u5728\u4e00\u4e2a\u8fdb\u7a0b\u91cc\u7684bug":71,"\u4f1a\u5c06\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u62fc\u63a5":139,"\u4f1a\u5c06\u7b2c\u4e00\u4e2a":107,"\u4f1a\u5f15\u5165":68,"\u4f1a\u6210\u4e3astep\u51fd\u6570\u7684\u8f93\u5165":139,"\u4f1a\u6253\u5370\u5230\u6807\u51c6\u8f93\u51fa":133,"\u4f1a\u628a\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\u5206\u522b\u5206\u5272\u6210\u591a\u4e2a\u6587\u4ef6":117,"\u4f1a\u628acpu\u7684buffer\u5bf9\u9f50\u4e3a4096":68,"\u4f1a\u62a5\u5982\u4e0b\u7684\u9519\u8bef":107,"\u4f1a\u62a5\u9519":139,"\u4f1a\u6dfb\u52a0\u76f8\u5e94\u7684\u811a\u672c\u5728":68,"\u4f1a\u6dfb\u52a0\u76f8\u5e94\u7684\u811a\u672c\u7528\u4e8e\u6d4b\u8bd5\u548c\u5bf9\u6bd4\u5728\u4f7f\u7528mkl":67,"\u4f1a\u72ec\u7acb\u62e5\u6709\u4e00\u4efd\u8bad\u7ec3\u597d\u7684\u6a21\u578b":116,"\u4f1a\u751f\u6210\u6027\u80fd\u5206\u6790\u7ed3\u679c\u6587\u4ef6":133,"\u4f1a\u76f4\u63a5\u62a5\u9519\u9000\u51fa":71,"\u4f1a\u76f8\u5e94\u5730\u6539\u53d8\u8f93\u51fa\u7684\u5c3a\u5bf8":100,"\u4f1a\u81ea\u52a8\u4f7f\u7528mklml\u5e93\u4f5c\u4e3apaddlepaddle\u7684cblas\u548clapack\u5e93":68,"\u4f1a\u81ea\u52a8\u5173\u95ed\u5bf9\u5e94\u7684issu":98,"\u4f1a\u81ea\u52a8\u5728\u7f16\u8bd1\u65f6\u4e0b\u8f7d":27,"\u4f1a\u81ea\u52a8\u6839\u636e\u786c\u4ef6\u914d\u7f6e":68,"\u4f1a\u83b7\u53d6\u5f53\u524dnamespace\u4e0b\u7684\u6240\u6709pod":123,"\u4f1a\u88ab":117,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":139,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u975e\u5e8f\u5217":139,"\u4f1a\u88abpickle\u5e8f\u5217\u5316\u6210\u5b57\u7b26\u4e32":38,"\u4f1a\u901a\u8fc7\u5224\u6570\u636e\u662f\u5426\u9644\u5e26\u6709\u5e8f\u5217\u4fe1\u606f\u6765\u5224\u65ad\u4e00\u4e2a\u5411\u91cf":115,"\u4f1a\u9020\u6210\u90ae\u4ef6\u707e\u96be":98,"\u4f20\u5165":38,"\u4f20\u7ed9dataprovider\u7684\u67d0\u4e00\u4e2aargs\u8fc7\u5927":109,"\u4f20\u9012\u7ed9\u914d\u7f6e\u6587\u4ef6\u7684\u53c2\u6570":129,"\u4f46":72,"\u4f46\u4e0d\u66b4\u9732":72,"\u4f46\u4e0d\u7528\u4e8e\u8ba1\u7b97\u68af\u5ea6":100,"\u4f46\u4e0d\u9700\u8981\u63d0\u524d\u521b\u5efa":129,"\u4f46\u4e8e\u53cc\u5c42\u5e8f\u5217\u7684lstm\u6765\u8bf4":137,"\u4f46\u53ef\u4ee5\u83b7\u53d6":107,"\u4f46\u548c\u5355\u5c42rnn\u4e0d\u540c":137,"\u4f46\u5b50\u53e5\u542b\u6709\u7684\u8bcd\u8bed\u6570\u53ef\u4ee5\u4e0d\u76f8\u7b49":139,"\u4f46\u5c3d\u91cf\u8bf7\u4fdd\u6301\u7f16\u8bd1\u548c\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u662f\u540c\u4e00\u4e2a\u7248\u672c":27,"\u4f46\u5e76\u6ca1\u6709\u7ecf\u8fc7\u56de\u5f52\u6d4b\u8bd5":88,"\u4f46\u5e8f\u5217\u8f93\u51fa\u65f6":137,"\u4f46\u622a\u65ad\u65f6\u673a\u4e0d\u540c":107,"\u4f46\u6240\u6709fork\u7684\u7248\u672c\u5e93\u7684\u6240\u6709\u5206\u652f\u90fd\u76f8\u5f53\u4e8e\u7279\u6027\u5206\u652f":88,"\u4f46\u662f":[107,137],"\u4f46\u662f\u53c8\u8fc7\u4e8e\u7410\u788e":72,"\u4f46\u662f\u5728mkl":68,"\u4f46\u662f\u5728paddlepaddle\u4e2d":68,"\u4f46\u662f\u5927\u90e8\u5206\u53c2\u6570\u662f\u4e3a\u5f00\u53d1\u8005\u63d0\u4f9b\u7684":128,"\u4f46\u662f\u5b50\u5e8f\u5217\u7684\u6570\u76ee\u5fc5\u987b\u4e00\u6837":137,"\u4f46\u662f\u5e76\u4e0d\u80fd\u4fdd\u8bc1\u53c2\u6570\u540c\u6b65\u66f4\u65b0":118,"\u4f46\u662f\u652f\u6301avx\u6307\u4ee4\u96c6":98,"\u4f46\u662f\u6574\u4e2a\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4e0d\u9700\u8981\u4efb\u4f55\u8f6c\u6362":68,"\u4f46\u662f\u6bcf\u4e2a\u6837\u672c\u4ec5\u5305\u542b\u51e0\u4e2a\u8bcd":131,"\u4f46\u662f\u6ce8\u610f\u7684\u662f":68,"\u4f46\u662f\u7a81\u7136\u6709\u4e00\u4e2a10000\u957f\u7684\u5e8f\u5217":107,"\u4f46\u662f\u865a\u62df\u7684\u4e0d\u4ec5\u4ec5\u662f":26,"\u4f46\u662f\u89e3\u91ca\u6027\u8bed\u8a00":71,"\u4f46\u662f\u8c03\u8bd5python\u4e2d\u4f7f\u7528\u7684\u52a8\u6001\u94fe\u63a5\u5e93\u4e0e\u76f4\u63a5\u8c03\u8bd5\u539f\u59cb\u4e8c\u8fdb\u5236\u76f8\u6bd4\u589e\u52a0\u4e86\u5f88\u591a\u590d\u6742\u5ea6":133,"\u4f46\u662fbatch":107,"\u4f46\u6709\u503c\u7684\u5730\u65b9\u5fc5\u987b\u4e3a1":110,"\u4f46\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":110,"\u4f46\u7531\u4e8ecuda\u5e93\u901a\u5e38\u9700\u8981cento":30,"\u4f46\u9700\u6ce8\u610f\u53cd\u5411op\u6ca1\u6709":101,"\u4f46eigen":102,"\u4f5c\u4e3a\u4e0b\u4e00\u4e2a\u5b50\u53e5memory\u7684\u521d\u59cb\u72b6\u6001":137,"\u4f5c\u4e3a\u4f8b\u5b50\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":140,"\u4f5c\u4e3a\u53c2\u6570\u5c5e\u6027":101,"\u4f5c\u4e3a\u53c2\u6570\u7684id":109,"\u4f5c\u4e3a\u53e6\u4e00\u4e2a\u7b2c\u4e09\u65b9\u5e93\u96c6\u6210\u8fdbpaddlepaddl":68,"\u4f5c\u4e3a\u5b58\u50a8\u7cfb\u7edf":38,"\u4f5c\u4e3a\u5f53\u524d\u65f6\u523b\u8f93\u5165":139,"\u4f5c\u4e3a\u7c7b\u53e5\u67c4":71,"\u4f5c\u4e3a\u7edf\u8ba1\u7684\u57fa\u672c\u5355\u4f4d":115,"\u4f5c\u4e3a\u7f16\u8bd1\u5de5\u5177":27,"\u4f5c\u4e3a\u8c03\u7528":116,"\u4f5c\u4e3a\u8f93\u5165":115,"\u4f5c\u4e3a\u8f93\u51fa":140,"\u4f5c\u4e3aboot_layer\u4f20\u7ed9\u4e0b\u4e00\u4e2a\u5b50\u53e5\u7684memori":137,"\u4f5c\u7528":136,"\u4f60\u53ef\u4ee5\u5c06\u7f51\u7edc\u914d\u7f6e\u6210\u67d0\u4e9b\u5c42\u4f7f\u7528gpu\u8ba1\u7b97":131,"\u4f60\u8fd8\u53ef\u4ee5\u901a\u8fc7\u8fd0\u884cdjango\u6846\u67b6\u76f4\u63a5\u6fc0\u6d3b\u5de5\u5177\u7684\u670d\u52a1\u5668":103,"\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":100,"\u4f60\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a\u8bbe\u5907\u7684id\u53f7":131,"\u4f60\u9700\u8981\u5728\u914d\u7f6ecmake\u65f6\u5c06":100,"\u4f60\u9700\u8981\u628a\u8be5\u6587\u4ef6\u52a0\u5165":100,"\u4f7f\u4e4b\u53d8\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8f93\u5165":115,"\u4f7f\u4e4b\u53d8\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":115,"\u4f7f\u5f97\u5355\u5143\u6d4b\u8bd5\u6709\u4e00\u4e2a\u5e72\u51c0\u7684\u73af\u5883":104,"\u4f7f\u5f97\u642d\u6a21\u578b\u65f6\u66f4\u65b9\u4fbf":100,"\u4f7f\u68af\u5ea6\u7684\u63d0\u4ea4\u548c\u53c2\u6570\u7684\u66f4\u65b0\u6309\u7167\u987a\u5e8f\u65b9\u5f0f\u6267\u884c":118,"\u4f7f\u7528":[68,72,88,100,107,108,109,113,115,116,129,133,134,137,139,140,142],"\u4f7f\u75280\u53f7\u548c1\u53f7gpu\u8ba1\u7b97fc2\u5c42":131,"\u4f7f\u75280\u53f7gpu\u8ba1\u7b97fc2\u5c42":131,"\u4f7f\u75281\u53f7gpu\u8ba1\u7b97fc3\u5c42":131,"\u4f7f\u75282\u53f7\u548c3\u53f7gpu\u8ba1\u7b97fc3\u5c42":131,"\u4f7f\u7528\u4e00\u4e2a\u5c3a\u5ea6\u4e3a":100,"\u4f7f\u7528\u4e00\u4e2a\u8bcd\u524d\u4e24\u4e2a\u8bcd\u548c\u540e\u4e24\u4e2a\u8bcd":107,"\u4f7f\u7528\u4e0a\u6587\u521b\u5efa\u7684yaml\u6587\u4ef6\u521b\u5efakubernet":122,"\u4f7f\u7528\u4e0b\u9762\u547d\u4ee4":38,"\u4f7f\u7528\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u8fd0\u884c\u5b83":103,"\u4f7f\u7528\u4e86\u540c\u6837\u7684parameter\u548cbia":109,"\u4f7f\u7528\u4ee5\u4e0a\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u8fdb\u884c\u9884\u6d4b":110,"\u4f7f\u7528\u52a8\u6001\u5e93":71,"\u4f7f\u7528\u53c2\u6570":[27,117],"\u4f7f\u7528\u540c\u6837\u7684\u8bad\u7ec3\u6570\u636eblock":37,"\u4f7f\u7528\u57fa\u4e8edocker\u5bb9\u5668\u7684\u7f16\u8bd1\u65b9\u5f0f":142,"\u4f7f\u7528\u591a\u5757\u663e\u5361\u8bad\u7ec3":107,"\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bad\u7ec3":107,"\u4f7f\u7528\u5b66\u4e60\u5b8c\u6210\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u751f\u6210\u5e8f\u5217":140,"\u4f7f\u7528\u5b83\u4f1a\u5f00\u542f\u4e00\u4e2ahttp\u670d\u52a1":133,"\u4f7f\u7528\u5bb9\u5668\u65b9\u5f0f\u8fd0\u884c\u8bad\u7ec3\u4efb\u52a1\u7684kubernet":123,"\u4f7f\u7528\u6211\u4eec\u4e4b\u524d\u6784\u9020\u7684\u955c\u50cf":122,"\u4f7f\u7528\u6570\u503c\u6cd5\u68c0\u6d4b\u68af\u5ea6\u6b63\u786e\u6027\u548c\u7a33\u5b9a\u6027":101,"\u4f7f\u7528\u6587\u6863":101,"\u4f7f\u7528\u663e\u5361\u8bad\u7ec3":107,"\u4f7f\u7528\u667a\u80fd\u6307\u9488\u7684\u539f\u56e0\u662f":72,"\u4f7f\u7528\u6848\u4f8b":130,"\u4f7f\u7528\u73af\u5883\u53d8\u91cf":117,"\u4f7f\u7528\u7684\u53c2\u6570\u4e0epaddlepaddle\u7533\u8bf7\u7684buffer\u5171\u7528\u4e00\u5757\u5185\u5b58":68,"\u4f7f\u7528\u76f8\u5bf9\u8def\u5f84\u7684\u5f15\u7528\u65b9\u5f0f":72,"\u4f7f\u7528\u8005\u4e0d\u9700\u8981\u5173\u5fc3":129,"\u4f7f\u7528\u8005\u53ea\u9700\u8981\u5173\u6ce8\u4e8e\u8bbe\u8ba1rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":139,"\u4f7f\u7528\u8005\u65e0\u9700\u5173\u5fc3\u8fd9\u4e2a\u53c2\u6570":129,"\u4f7f\u7528\u8005\u901a\u5e38\u65e0\u9700\u5173\u5fc3":129,"\u4f7f\u7528\u8be5learning_rate_schedule\u65f6":109,"\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u5b8c\u6210\u5bf9\u65b0\u6570\u636e\u7684\u9884\u6d4b":37,"\u4f7f\u7528\u8fd9\u79cd\u65b9\u5f0f":[116,137],"\u4f7f\u7528\u8fdc\u7a0b\u7a00\u758f\u65b9\u5f0f\u8bad\u7ec3\u65f6":100,"\u4f7f\u7528\u9759\u6001\u5e93\u548c\u52a8\u6001\u5e93\u96be\u5ea6\u5dee\u4e0d\u591a":71,"\u4f7f\u7528c":[72,113],"\u4f7f\u7528c99\u505a\u63a5\u53e3":71,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c11\u7684\u539f\u56e0\u662f":71,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c89":71,"\u4f7f\u7528checkgrad\u6a21\u5f0f\u65f6\u7684\u53c2\u6570\u53d8\u5316\u5927\u5c0f":129,"\u4f7f\u7528cmake\u7684\u8bdd":133,"\u4f7f\u7528cpu\u4e24\u7ebf\u7a0b\u8ba1\u7b97fc4\u5c42":131,"\u4f7f\u7528cpu\u8ba1\u7b97fc4\u5c42":131,"\u4f7f\u7528docker":28,"\u4f7f\u7528docker\u5b89\u88c5\u548c\u8fd0\u884cpaddlepaddle\u53ef\u4ee5\u65e0\u9700\u8003\u8651":28,"\u4f7f\u7528docker\u5b89\u88c5\u8fd0\u884c":29,"\u4f7f\u7528docker\u5c31\u4e0d\u7528\u914d\u7f6e\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883\u4e86":26,"\u4f7f\u7528docker\u6784\u5efapaddlepaddle\u7684\u6587\u6863":103,"\u4f7f\u7528docker\u7684\u60c5\u51b5\u4e0b":27,"\u4f7f\u7528eigen\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":142,"\u4f7f\u7528fabric\u542f\u52a8\u96c6\u7fa4\u8bad\u7ec3":120,"\u4f7f\u7528init":131,"\u4f7f\u7528lstm\u4f5c\u4e3aencod":137,"\u4f7f\u7528memory\u7684rnn\u5b9e\u73b0\u4fbf\u5982\u4e0b\u56fe\u6240\u793a":137,"\u4f7f\u7528model":131,"\u4f7f\u7528openblas\u7684\u955c\u50cf":28,"\u4f7f\u7528openblas\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":142,"\u4f7f\u7528paddlepaddl":[113,116],"\u4f7f\u7528pip\u5b89\u88c5":29,"\u4f7f\u7528rdma\u8fd8\u662ftcp\u4f20\u8f93\u534f\u8bae":129,"\u4f7f\u7528regress":88,"\u4f7f\u7528swig\u53ea\u652f\u6301cpython\u89e3\u91ca\u5668":71,"\u4f7f\u7528swig\u9700\u8981\u591a\u8bed\u8a00\u7ed1\u5b9a\u7684\u5f00\u53d1\u4eba\u5458\u719f\u7ec3\u638c\u63e1swig\u914d\u7f6e":71,"\u4f7f\u7528void":71,"\u4f7f\u8be5\u5c42\u7684\u53c2\u6570\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4fdd\u6301\u4e0d\u53d8":109,"\u4f8b\u5982":[27,38,71,72,88,100,107,108,110,115,123,128,129,131,134,137,140],"\u4f8b\u5982\u4e0b\u56fe\u4e2d":133,"\u4f8b\u5982\u4e0b\u9762\u4ee3\u7801":107,"\u4f8b\u5982\u4e5f\u53ef\u5728\u7a0b\u5e8f\u8fd0\u884c\u8fc7\u7a0b\u4e2d\u518d\u52a0\u8f7d\u53e6\u5916\u4e00\u4e2a\u6a21\u578b":116,"\u4f8b\u5982\u4f7f\u7528":107,"\u4f8b\u5982\u542b\u6709\u591a\u4e2a\u901a\u9053\u7684\u56fe\u7247":115,"\u4f8b\u5982\u5728deepspeech2":67,"\u4f8b\u5982\u5bf9\u4e8ejava\u6216\u8005python":71,"\u4f8b\u5982\u5bf9\u4e8ejava\u6765\u8bf4":71,"\u4f8b\u5982\u5bf9\u4e8epython":71,"\u4f8b\u5982\u5c06\u7b2c\u4e00\u6761\u6570\u636e\u8f6c\u5316\u4e3a":137,"\u4f8b\u5982\u6587\u672c\u5206\u7c7b\u4e2d":137,"\u4f8b\u5982\u672c\u4f8b\u4e2d\u7684\u4e24\u4e2a\u7279\u5f81":137,"\u4f8b\u5982\u673a\u5668\u4e0a\u67094\u5757gpu":107,"\u4f8b\u5982c":71,"\u4f8b\u5982java\u4e0epython\u7684\u9519\u8bef\u5904\u7406\u662f\u76f4\u63a5\u6254\u51fa\u6765except":71,"\u4f8b\u5982output\u76ee\u5f55\u4e0b\u5c31\u5b58\u653e\u4e86\u8f93\u51fa\u7ed3\u679c":123,"\u4f8b\u5982python\u53ef\u4ee5\u4f7f\u7528":71,"\u4f8b\u5982python\u7684":71,"\u4f8b\u5982rnn":67,"\u4f8b\u5982sigmoid":100,"\u4f8b\u5b50\u4e2d\u4e3a3\u4e2a":117,"\u4f8b\u5b50\u4e2d\u662f":100,"\u4f8b\u5b50\u4e2d\u662f0":100,"\u4f8b\u5b50\u4e2d\u662f100":100,"\u4f8b\u5b50\u4e2d\u662f4096":100,"\u4f8b\u5b50\u4e2d\u662f8192":100,"\u4f8b\u5b50\u4e2d\u662ffc":100,"\u4f8b\u5b50\u4e2d\u662fsoftmax":100,"\u4f9bpaddlepaddle\u52a0\u8f7d":129,"\u4f9d\u636e\u662f\u5426\u5305\u542bkernel":101,"\u4f9d\u6b21\u7c7b\u63a8":88,"\u4f9d\u8d56":[27,30],"\u4f9d\u8d56\u73af\u5883\u5373\u53ef\u8fd0\u884c":28,"\u4f9d\u8d56libpython2":27,"\u4fbf\u548c\u5355\u5c42rnn\u914d\u7f6e\u4e2d\u7684":137,"\u4fbf\u662f\u5c06\u9759\u6001\u5e93\u52a0\u5165jvm\u4e2d":71,"\u4fdd\u5b58\u6a21\u578b\u53c2\u6570\u7684\u76ee\u5f55":129,"\u4fdd\u5b58\u7684\u53c2\u6570\u4e5f\u662ffloat\u7c7b\u578b":109,"\u4fdd\u5b58\u7f51\u7edc\u5c42\u8f93\u51fa\u7ed3\u679c\u7684\u76ee\u5f55":129,"\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c\u7684\u6587\u4ef6\u540d":129,"\u4fdd\u6301\u5c3d\u91cf\u5c11\u7684commit":98,"\u4fdd\u8bc1\u4f7f\u7528gpu\u8bad\u7ec3\u65f6\u4e5f\u53ef\u4ee5\u83b7\u5f97":107,"\u4fe1\u53f7\u6765\u81ea\u52a8\u7ec8\u6b62\u5b83\u542f\u52a8\u7684\u6240\u6709\u8fdb\u7a0b":119,"\u4fe1\u606f":115,"\u4fee\u590d\u6240\u6709bug\u540e":88,"\u4fee\u590ddocker\u7f16\u8bd1\u955c\u50cf\u95ee\u9898":88,"\u4fee\u6539":[68,88,122],"\u4fee\u6539\u542f\u52a8\u811a\u672c\u540e":122,"\u4fee\u6539\u6210":88,"\u4fee\u6539\u6210\u66f4\u5feb\u7684\u7248\u672c":134,"\u503c\u5f97\u6ce8\u610f\u7684\u662f":[98,137],"\u503c\u5f97\u6df1\u5165\u5206\u6790":134,"\u503c\u7c7b\u578b":131,"\u5047\u5982\u6211\u4eec\u662f\u4e09\u5206\u7c7b\u95ee\u9898":109,"\u5047\u8bbe":100,"\u5047\u8bbe\u60a8\u5df2\u7ecf\u5728\u5f53\u524d\u76ee\u5f55":28,"\u5047\u8bbe\u635f\u5931\u51fd\u6570\u662f":100,"\u5047\u8bbe\u7b2c\u4e00\u4e2alayer\u7684\u8f93\u51faa\u662f\u4e00\u4e2a":107,"\u504f\u7f6e\u53c2\u6570\u7684\u5927\u5c0f":100,"\u505a\u4e00\u4e2a\u4ecb\u7ecd":102,"\u505a\u53ea\u8bfb\u6302\u8f7d":38,"\u505a\u5982\u4e0b\u51e0\u4e2a\u64cd\u4f5c":88,"\u505a\u63a5\u53e3":71,"\u505a\u68af\u5ea6\u68c0\u6d4b":101,"\u505a\u68c0\u67e5":101,"\u505c\u6b62\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":37,"\u505c\u6b62\u52a0\u8f7d\u6570\u636e":129,"\u5141\u8bb8\u5916\u7f51\u8bbf\u95ee\u8fd9\u4e2ahttp\u670d\u52a1":133,"\u5143\u7d20":136,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u5e8f\u5217\u6240\u643a\u5e26\u7684\u91cd\u8981\u4fe1\u606f":115,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684\u8f93\u5165\u4fe1\u606f":136,"\u5145\u5206\u53d1\u6325\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u4f18\u52bf":67,"\u5145\u5206\u5c55\u73b0\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u4f18\u52bf":68,"\u5148\u4ece\u5355\u7ebf\u7a0b\u5f00\u59cb":133,"\u5148\u5378\u8f7d\u4e4b\u524d\u7684\u7248\u672c":27,"\u5148\u5b8c\u6210\u5bf9\u6743\u91cd\u7684packing\u64cd\u4f5c":67,"\u5148\u5b9e\u73b0\u6a21\u578b\u63a8\u65ad\u7684api":72,"\u5148\u627e\u51fa\u53c2\u6570":108,"\u5148\u67e5\u770b\u4e00\u4e0b\u662f\u5426\u66fe\u7ecf\u5b89\u88c5\u8fc7paddl":104,"\u5148\u68c0\u67e5\u5173\u952e\u8def\u5f84\u7684\u6027\u80fd\u95ee\u9898":133,"\u514b\u9686\u4e0b\u9762":144,"\u5168\u8fde\u63a5\u5c42\u4ee5\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":100,"\u5168\u8fde\u63a5\u5c42\u6ca1\u6709\u7f51\u7edc\u5c42\u914d\u7f6e\u7684\u8d85\u53c2\u6570":100,"\u5168\u8fde\u63a5\u5c42\u7684\u5b9e\u73b0\u4f4d\u4e8e":100,"\u5168\u8fde\u63a5\u5c42\u7684\u6bcf\u4e2a\u8f93\u51fa\u90fd\u8fde\u63a5\u5230\u4e0a\u4e00\u5c42\u7684\u6240\u6709\u7684\u795e\u7ecf\u5143\u4e0a":100,"\u5168\u8fde\u63a5\u5c42python\u5c01\u88c5\u7684\u4f8b\u5b50\u4e2d\u5305\u542b\u4e0b\u9762\u51e0\u6b65":100,"\u516c\u5f0f":28,"\u5171\u4eab\u4e00\u4e2aop\u5b9a\u4e49":101,"\u5171\u4eab\u5185\u5b58":68,"\u5171\u4eab\u540c\u4e00\u4e2a\u6743\u91cd":67,"\u5171\u4eab\u540c\u4e00\u4e2akernel\u65f6":101,"\u5171\u4eab\u5b58\u50a8\u6302\u5728\u7684\u8def\u5f84":123,"\u5173\u4e8e\u4ec0\u4e48\u662f":115,"\u5173\u4e8e\u5728paddlepaddle\u4e2d\u5982\u4f55\u4f7f\u7528eigen\u5e93":101,"\u5173\u4e8e\u65f6\u95f4\u5e8f\u5217":137,"\u5173\u4e8e\u6784\u5efa\u548c\u6d4b\u8bd5\u7684\u66f4\u591a\u4fe1\u606f":98,"\u5173\u4e8eavx":28,"\u5173\u4e8eeigen":102,"\u5173\u4e8elstm":108,"\u5173\u4e8epaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":123,"\u5173\u4e8epaddlepaddle\u7684\u66f4\u591a\u4f7f\u7528\u65b9\u6cd5\u8bf7\u53c2\u8003":110,"\u5173\u4e8eunbound":139,"\u5173\u952e\u8bcd\u5305\u62ec":98,"\u5176\u4e2d":[71,88,100,107,109,110,133,140,142,144],"\u5176\u4e2d\u5305\u542b\u4e86\u7528\u6237\u7684\u8bad\u7ec3\u7a0b\u5e8f":117,"\u5176\u4e2d\u5305\u542b\u6240\u4f9d\u8d56\u7684\u6240\u6709\u7b2c\u4e09\u65b9\u5e93":143,"\u5176\u4e2d\u5305\u542b\u6240\u6709c":143,"\u5176\u4e2d\u5305\u542bpaddlepaddle\u7684c":143,"\u5176\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u53cc\u5c42\u5e8f\u5217\u4e2d\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a":136,"\u5176\u4e2dcheckgrad\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":129,"\u5176\u4e2dmean\u548cstd\u662f\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u53c2\u6570":129,"\u5176\u4e2dx\u8868\u793a\u8f93\u5165\u6570\u636e\u662f\u4e00\u4e2a\u7ef4\u5ea6\u4e3a2\u7684\u7a20\u5bc6\u5411\u91cf":110,"\u5176\u4e3b\u8981\u63a5\u53e3\u5982\u4e0b":102,"\u5176\u4ed6\u4eba\u53ef\u4ee5\u590d\u73b0\u95ee\u9898\u4ee5\u4fbf\u5e2e\u52a9":26,"\u5176\u4ed6\u5185\u5b58\u6742\u9879":107,"\u5176\u4ed6\u5185\u5b58\u6742\u9879\u662f\u6307paddlepaddle\u672c\u8eab\u6240\u7528\u7684\u4e00\u4e9b\u5185\u5b58":107,"\u5176\u4ed6\u51fd\u6570\u5747\u8fd4\u56de":72,"\u5176\u4ed6\u6240\u6709\u5c42\u90fd\u4f1a\u4f7f\u7528gpu\u8ba1\u7b97":131,"\u5176\u4ed6\u7528\u6237\u7684fork\u7248\u672c\u5e93\u5e76\u4e0d\u9700\u8981\u4e25\u683c\u9075\u5b88":88,"\u5176\u4ed6\u7684\u4f9d\u8d56\u8f6f\u4ef6":27,"\u5176\u4ed6\u914d\u7f6e\u53c2\u6570":[142,143],"\u5176\u4ed6\u9ad8\u7ea7\u529f\u80fd\u5305\u62ec\u5b9a\u4e49\u591a\u4e2amemori":140,"\u5176\u4f1a\u81ea\u52a8\u88ab\u52a0\u5165\u7f16\u8bd1\u5217\u8868":100,"\u5176\u547d\u4ee4\u5982\u4e0b":133,"\u5176\u5b83\u53ef\u9009\u7f16\u8bd1\u9009\u9879\u6309\u9700\u8fdb\u884c\u8bbe\u5b9a":113,"\u5176\u5b83layer\u7684\u8f93\u51fa":139,"\u5176\u5b9e\u4e5f\u662f\u548c\u6bcf\u4e2amini":107,"\u5176\u6b21":137,"\u5176\u8bf4\u660e\u5982\u4e0b":137,"\u5176\u8f6c\u6362\u6b21\u6570\u51cf\u5c11\u81f3":67,"\u5176\u8f93\u51fa\u88ab\u7528\u4f5cmemory\u7684\u521d\u59cb\u503c":140,"\u5176name\u7531\u53c2\u6570":108,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":[72,107],"\u5177\u4f53\u505a\u6cd5\u8bf7\u53c2\u8003":26,"\u5177\u4f53\u539f\u56e0\u53c2\u8003":72,"\u5177\u4f53\u53ef\u4ee5\u53c2\u8003":[100,107],"\u5177\u4f53\u53ef\u4ee5\u53c2\u8003mkl":68,"\u5177\u4f53\u53ef\u53c2\u8003\u6587\u6863":139,"\u5177\u4f53\u5b9e\u73b0\u65b9\u5f0f\u6bd4\u5982":[67,68],"\u5177\u4f53\u60c5\u51b5\u56e0\u4eba\u800c\u5f02":134,"\u5177\u4f53\u64cd\u4f5c\u5982\u4e0b":104,"\u5177\u4f53\u6b65\u9aa4\u5982\u4e0b":104,"\u5177\u4f53\u7684\u5b8c\u6210\u72b6\u6001\u53ef\u4ee5\u53c2\u89c1":68,"\u5177\u4f53\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":104,"\u5177\u4f53\u8bf7\u53c2\u8003":[72,98],"\u5177\u4f53\u8bf7\u89c1":98,"\u5177\u6709\u76f8\u540c\u7684\u7ed3\u679c\u4e86":137,"\u5185":140,"\u5185\u5b58":134,"\u5185\u5b58\u4e0d\u8db3":105,"\u5185\u5b58\u5bb9\u9650\u9608\u503c":129,"\u5185\u5bb9":101,"\u5185\u5bb9\u5982\u4e0b":122,"\u5185\u5c42\u5e8f\u5217\u5728":115,"\u5185\u5c42inner_step\u7684recurrent_group\u548c\u5355\u5c42\u5e8f\u5217\u7684\u51e0\u4e4e\u4e00\u6837":137,"\u5185\u5df2\u7ecf\u5305\u542bpaddlepaddle\u7684\u6267\u884c\u7a0b\u5e8f\u4f46\u662f\u8fd8\u6ca1\u4e0a\u8ff0\u529f\u80fd":123,"\u5185\u7f6e\u7684":116,"\u5185\u90e8":[116,123],"\u5185\u90e8\u5b58\u50a8":68,"\u5185\u90e8\u7531":[115,116],"\u5185\u90e8\u9a71\u52a8python\u89e3\u91ca\u5668\u8fdb\u884c\u6a21\u578b\u914d\u7f6e\u89e3\u6790\u548c\u6570\u636e\u8bfb\u53d6":71,"\u518d\u4ee5":101,"\u518d\u505a\u4e00\u5b9a\u7684reshap":108,"\u518d\u5199\u5165\u7f51\u7edc\u53c2\u6570":109,"\u518d\u5728\u6bcf\u4e00\u4e2aapi\u4e2d\u81ea\u5df1\u68c0\u67e5\u7c7b\u578b":71,"\u518d\u57fa\u4e8e":88,"\u518d\u5b89\u88c5":[30,104],"\u518d\u5bf9\u6bcf\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u8fdb\u884c\u5904\u7406":137,"\u518d\u5bf9\u6bcf\u4e00\u53e5\u8bdd\u7684\u7f16\u7801\u5411\u91cf\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u6bb5\u843d\u7684\u5411\u91cf":137,"\u518d\u5bf9\u8fd9\u4e2a\u6bb5\u843d\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":137,"\u518d\u5c06\u66f4\u65b0\u540e\u7684\u53c2\u6570\u4e0b\u53d1\u5230\u6bcf\u4e2a\u8ba1\u7b97\u8282\u70b9":118,"\u518d\u5f00\u542f\u591a\u7ebf\u7a0b":133,"\u518d\u628a\u5df2\u8f6c\u6362\u4e3apacked\u683c\u5f0f\u7684\u6570\u636e\u4f20\u9012\u7ed9\u90a3\u4e9b\u590d\u7528\u540c\u4e00\u6570\u636e\u7684gemm":67,"\u518d\u6307\u5b9a":27,"\u518d\u68c0\u67e5\u5176\u4ed6\u90e8\u5206\u7684\u6027\u80fd\u95ee\u9898":133,"\u518d\u6b21\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":134,"\u518d\u6b21\u8fdb\u884c\u6027\u80fd\u5206\u6790":133,"\u518d\u7528\u8fd9\u4e2a\u68af\u5ea6\u53bb\u548c":100,"\u518d\u901a\u8fc7\u51fd\u6570":123,"\u518d\u91cd\u65b0\u5b89\u88c5":27,"\u5199\u4ee3\u7801":71,"\u5199\u5165\u5feb\u7167\u6570\u636e":37,"\u5199\u5165\u6587\u4ef6\u4e2d":116,"\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":100,"\u5199\u7684":133,"\u51c6\u5907\u60a8\u7684\u8ba1\u7b97\u96c6\u7fa4":127,"\u51c6\u5907\u8bad\u7ec3\u6570\u636e":124,"\u51c6\u5907\u8bad\u7ec3\u6570\u636e\u548c\u9a8c\u8bc1\u6570\u636e\u96c6":117,"\u51c6\u5907\u9884\u6d4b\u6a21\u578b\u548c":116,"\u51c6\u5907\u9884\u6d4b\u6a21\u578b\u90e8\u5206":116,"\u51cf\u5c0f\u5e8f\u5217\u7684\u957f\u5ea6":107,"\u51cf\u5c0f\u8fd9\u4e2a\u5185\u5b58\u6c60\u5373\u53ef\u51cf\u5c0f\u5185\u5b58\u5360\u7528":107,"\u51cf\u5c0fbatch":107,"\u51e0\u53f0\u5230\u51e0\u5343\u53f0\u89c4\u6a21":127,"\u51fa\u73b0":104,"\u51fa\u73b0\u4ee5\u4e0b\u9519\u8bef":109,"\u51fa\u73b0\u8be5\u9519\u8bef\u7684\u539f\u56e0\u4e00\u822c\u662f\u7528\u6237\u5bf9\u4e0d\u540clayer\u7684\u53c2\u6570":108,"\u51fa\u73b0\u8fd9\u4e2a\u95ee\u9898\u7684\u4e3b\u8981\u539f\u56e0\u662f":[30,104],"\u51fd\u6570":[67,68,100,115,133,134,140],"\u51fd\u6570\u4e2d\u64cd\u4f5c\u7684\u91cd\u8981\u53d8\u91cf\u7684\u8be6\u7ec6\u89e3\u91ca":101,"\u51fd\u6570\u5047\u8bbe":140,"\u51fd\u6570\u52a0\u5230\u4ee3\u7801\u4e2d":134,"\u51fd\u6570\u5373\u53ef\u5b8c\u6210\u8f6c\u6362":38,"\u51fd\u6570\u53ea\u5173\u6ce8\u4e8ernn\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8ba1\u7b97":139,"\u51fd\u6570\u540d":133,"\u51fd\u6570\u540d\u4e3a":72,"\u51fd\u6570\u547d\u540d":71,"\u51fd\u6570\u5b9a\u4e49\u8f93\u5165":101,"\u51fd\u6570\u5b9e\u9645\u4f7f\u7528\u7684\u603b\u65f6\u95f4":133,"\u51fd\u6570\u5f97\u5230\u7684\u68af\u5ea6\u53bb\u5bf9\u6bd4":100,"\u51fd\u6570\u5fc5\u987b\u5148\u8c03\u7528\u57fa\u7c7b\u4e2d\u7684\u51fd\u6570":100,"\u51fd\u6570\u5fc5\u987b\u8fd4\u56de\u4e00\u4e2a\u6216\u591a\u4e2alayer\u7684\u8f93\u51fa":139,"\u51fd\u6570\u603b\u65f6\u95f4":133,"\u51fd\u6570\u6307\u51fa\u4e86\u5728\u8bad\u7ec3\u65f6\u9700\u8981\u4ece\u53c2\u6570\u670d\u52a1\u5668\u53d6\u51fa\u7684\u884c":100,"\u51fd\u6570\u6765\u5c06\u4fe1\u606f\u8f93\u51fa\u5230\u754c\u9762\u4e2d":134,"\u51fd\u6570\u7684\u5b9e\u73b0\u662f\u6b63\u786e\u7684":100,"\u51fd\u6570\u7684\u5f00\u5934\u5fc5\u987b\u8c03\u7528":100,"\u51fd\u6570\u7684\u603b\u5171\u8017\u65f6\u5f88\u957f":133,"\u51fd\u6570\u7684\u8c03\u7528\u6b21\u6570":133,"\u51fd\u6570\u80fd\u591f\u5c06\u4f7f\u7528":116,"\u51fd\u6570\u91cc\u5b9e\u73b0":101,"\u5206\u4e3a":116,"\u5206\u522b\u4e3a\u6570\u636e\u8f93\u5165\u6dfb\u52a0\u5916\u5c42\u5e8f\u5217\u548c\u5185\u5c42\u5e8f\u5217\u7684\u5e8f\u5217\u4fe1\u606f":115,"\u5206\u522b\u4ece\u8bcd\u8bed\u548c\u53e5\u5b50\u7ea7\u522b\u7f16\u7801\u8f93\u5165\u6570\u636e":139,"\u5206\u522b\u4ee3\u8868\u8f93\u5165\u6570\u636e":68,"\u5206\u522b\u4f7f\u7528\u5355\u53cc\u5c42rnn\u4f5c\u4e3a\u7f51\u7edc\u914d\u7f6e\u7684\u6a21\u578b":137,"\u5206\u522b\u5b9a\u4e49\u5b50\u53e5\u7ea7\u522b\u548c\u8bcd\u8bed\u7ea7\u522b\u4e0a\u9700\u8981\u5b8c\u6210\u7684\u8fd0\u7b97":139,"\u5206\u522b\u5bf9\u5e94capi":88,"\u5206\u522b\u662f":136,"\u5206\u522b\u662frnn\u72b6\u6001\u548c\u8f93\u5165\u7684\u53d8\u6362\u77e9\u9635":140,"\u5206\u522b\u662fsentences\u548clabel":137,"\u5206\u522b\u662fwords\u548clabel":137,"\u5206\u522b\u8ba1\u7b97\u6bcf\u4e2a\u53c2\u6570\u7684\u68af\u5ea6":100,"\u5206\u522b\u8fdb\u884c\u5e8f\u5217\u64cd\u4f5c":137,"\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":37,"\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u901a\u5e38\u4f1a\u901a\u8fc7api\u6216\u8005\u73af\u5883\u53d8\u91cf\u63d0\u4f9b\u4efb\u52a1\u8fd0\u884c\u9700\u8981\u7684\u53c2\u6570":120,"\u5206\u5e03\u5f0f\u8bad\u7ec3":132,"\u5206\u5e03\u5f0f\u8bad\u7ec3\u67b6\u6784\u5982\u4e0b\u56fe\u6240\u793a":118,"\u5206\u652f":[88,98],"\u5206\u652f\u4e00\u65e6\u5efa\u7acb":88,"\u5206\u652f\u4e0a":98,"\u5206\u652f\u4e0a\u521b\u5efa\u65b0\u5206\u652f":98,"\u5206\u652f\u4e2d":88,"\u5206\u652f\u4e3a\u5f00\u53d1":88,"\u5206\u652f\u4e3a\u6bcf\u4e00\u6b21release\u65f6\u5efa\u7acb\u7684\u4e34\u65f6\u5206\u652f":88,"\u5206\u652f\u4e3a\u7a33\u5b9a":88,"\u5206\u652f\u529f\u80fd\u7684\u5c01\u95ed":88,"\u5206\u652f\u5408\u5165":88,"\u5206\u652f\u5408\u5165master\u5206\u652f":88,"\u5206\u652f\u540c\u6b65\u4e3b\u7248\u672c\u5e93\u7684":88,"\u5206\u652f\u540d":98,"\u5206\u652f\u540d\u4e3a":88,"\u5206\u652f\u5b58\u5728\u7684\u65f6\u5019":88,"\u5206\u652f\u6d3e\u751f\u51fa\u65b0\u7684\u5206\u652f":88,"\u5206\u652f\u7528\u6765\u6d4b\u8bd5\u53ea\u9700\u8981\u8ba1\u7b97\u4e00\u4e2a\u8f93\u5165\u68af\u5ea6\u7684\u60c5\u51b5":101,"\u5206\u652f\u7684\u7248\u672c\u90fd\u662f\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5\u548c\u56de\u5f52\u6d4b\u8bd5\u7684\u7248\u672c":88,"\u5206\u652f\u7684\u7248\u672c\u90fd\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5":88,"\u5206\u652f\u89c4\u8303":98,"\u5206\u6790\u5f97\u5230\u7684\u4fe1\u606f\u7528\u4e8e\u534f\u52a9\u8fdb\u884c\u7a0b\u5e8f\u7684\u4f18\u5316":134,"\u5206\u7247":37,"\u5206\u7c7b\u4efb\u52a1\u4e2d\u7c7b\u522b\u6807\u7b7e":115,"\u5206\u914d\u5230\u5f53\u524d\u6570\u636e\u5757\u6837\u672c\u6570\u7684\u56db\u5206\u4e4b\u4e00":129,"\u5207\u6362\u5230":98,"\u5207\u6362\u5230\u6240\u5efa\u5206\u652f":98,"\u5217\u5143\u7d20\u6392\u5217\u6210\u7684\u77e9\u5f62\u9635\u5217":115,"\u5217\u540d":133,"\u5217\u8868\u5982\u4e0b":110,"\u5219\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":100,"\u5219\u4f1a\u4f7f\u7528openblas\u4f5c\u4e3ablas\u5e93":27,"\u5219\u4f7f\u7528":143,"\u5219\u4f7f\u7528\u540c\u6b65\u8bad\u7ec3":129,"\u5219\u4f7f\u7528\u542f\u52a8\u53c2\u6570\u5b9a\u4e49\u7684\u521d\u59cb\u5316\u65b9\u6cd5\u521d\u59cb\u5316\u53c2\u6570":37,"\u5219\u4f7f\u7528\u8be5\u53c2\u6570\u4f5c\u4e3a\u9ed8\u8ba4\u503c":129,"\u5219\u53ef\u8bbe\u7f6e":[142,143,144],"\u5219\u5e76\u4e0d\u4f1a\u7b49\u5f85\u6240\u6709trainer\u63d0\u4ea4\u68af\u5ea6\u624d\u66f4\u65b0\u53c2\u6570":118,"\u5219\u5ffd\u7565":37,"\u5219\u603b\u4f1a\u663e\u793a\u963b\u9694\u6458\u8981\u4fe1\u606f":129,"\u5219\u628a\u53e6\u4e00\u4e2a\u6162\u901f\u7684kill\u6389":37,"\u5219\u662f\u5e26gui\u7684nvidia\u53ef\u89c6\u5316\u6027\u80fd\u5206\u6790\u5de5\u5177":134,"\u5219\u663e\u793a\u963b\u9694\u6027\u80fd\u7684\u6458\u8981\u4fe1\u606f":129,"\u5219\u76f4\u63a5\u5f15\u5165\u53e6\u4e00\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":72,"\u5219\u8bbe\u7f6e\u6210":[142,144],"\u5219\u9700\u8981\u4f7f\u7528\u7b49\u4e8e\u6743\u91cd\u53c2\u6570\u89c4\u6a21\u5927\u7ea65\u500d\u7684\u5185\u5b58":107,"\u5219\u9700\u8981\u5206\u522b\u7f16\u8bd1\u771f\u673a\u548c\u6a21\u62df\u5668\u7248\u672c":143,"\u5219\u9700\u8981\u56de\u6eda\u5230\u4e0a\u4e00\u4e2a\u68c0\u67e5\u70b9":37,"\u5219\u9700\u8981\u5728\u672c\u673a\u5b89\u88c5\u4e0b\u9762\u7ae0\u8282\u5217\u51fa\u7684":27,"\u5219\u9700\u8981\u624b\u52a8\u62f7\u8d1d\u5c5e\u4e8e\u6bcf\u4e2atrainer\u8282\u70b9\u7684\u8bad\u7ec3\u6570\u636e\u5230\u5bf9\u5e94\u7684\u8282\u70b9\u4e0a":117,"\u521b\u5efa":[68,115,116],"\u521b\u5efa\u4e00\u4e2a":112,"\u521b\u5efa\u4e00\u4e2akubernet":123,"\u521b\u5efa\u5e76\u5207\u6362\u5230\u65b0\u5206\u652f":98,"\u521b\u5efa\u6210\u529f\u540e":123,"\u521b\u5efa\u65e5\u5fd7\u76ee\u5f55":124,"\u521b\u5efa\u7a00\u758f\u77e9\u9635\u65f6\u9700\u8981\u663e\u793a\u5730\u6307\u5b9a\u77e9\u9635\u7684":115,"\u521d\u59cb\u5316\u504f\u7f6e\u5411\u91cf":100,"\u521d\u59cb\u5316\u6743\u91cd\u8868":100,"\u521d\u59cb\u5316\u6a21\u578b\u7684\u8def\u5f84":129,"\u521d\u59cb\u5316\u7236\u7c7b":100,"\u521d\u59cb\u5316biases_":100,"\u521d\u59cb\u72b6\u6001":139,"\u5220\u9664":98,"\u5220\u9664\u78c1\u76d8\u76ee\u5f55\u4e2d\u4e0d\u662f\u5f53\u524duuid\u7684\u5feb\u7167\u6587\u4ef6":37,"\u5224\u65ad\u662f\u5426\u5b89\u88c5\u6210\u529f":143,"\u5229\u7528\u5206\u5e03\u5f0f\u8bad\u7ec3\u9a7e\u9a6d\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":107,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90\u53ef\u4ee5\u5206\u4e3a\u4ee5\u4e0b\u51e0\u4e2a\u65b9\u5f0f\u6765\u8fdb\u884c":107,"\u5229\u7528\u8fd9\u79cd\u7279\u6027":139,"\u522b\u4eba\u5e2e\u4e86\u5fd9":98,"\u522b\u5fd8\u4e86":26,"\u5230":[37,104,140],"\u5230\u6307\u5b9a\u6587\u4ef6\u4e2d":116,"\u5230\u672c\u5730":98,"\u5230\u6b64":101,"\u5230\u7b2c\u4e8c\u6b65":88,"\u5236\u4f5c\u65b0\u955c\u50cf\u6765\u5b8c\u6210\u4ee5\u4e0a\u7684\u5de5\u4f5c":123,"\u5236\u4f5cpaddlepaddle\u955c\u50cf":123,"\u524d\u4e00\u7bc7\u6587\u7ae0\u4ecb\u7ecd\u4e86\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673apaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":123,"\u524d\u540e\u7684\u7f51\u7edc\u6027\u80fd":67,"\u524d\u5411\u4f20\u64ad":100,"\u524d\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u5165":100,"\u524d\u5411\u548c\u540e\u5411":100,"\u524d\u5411\u8ba1\u7b97\u4e4b\u540epaddlepaddle\u5185\u90e8\u5df2\u7ecf\u5206\u914d":116,"\u524d\u5411op\u5b9e\u73b0\u5b8c\u6210":101,"\u524d\u8005\u5728":107,"\u524d\u8005\u5b58\u50a8op\u7684\u8f93\u5165\u8f93\u51fa\u548c\u53c2\u6570\u5c5e\u6027":101,"\u524d\u8005\u622a\u65ad\u53ef\u5b66\u4e60\u53c2\u6570\u7684\u68af\u5ea6":107,"\u524d\u8005op\u7684\u5b9a\u4e49\u7ee7\u627f\u81ea":101,"\u524d\u81ea\u52a8\u68c0\u67e5\u4e00\u4e9b\u57fa\u672c\u4e8b\u5b9c":98,"\u524d\u9700\u8981\u5b89\u88c5":133,"\u524d\u9988":118,"\u529f\u80fd":54,"\u529f\u80fd\u7684\u6b63\u786e\u6027\u5305\u62ec\u9a8c\u8bc1paddlepaddle\u76ee\u524d\u7684":88,"\u52a0\u4e0a\u504f\u7f6e\u5411\u91cf":100,"\u52a0\u5165":134,"\u52a0\u6743\u548c\u7528\u6765\u751f\u6210":140,"\u52a0\u6743\u7f16\u7801\u5411\u91cf":140,"\u52a0\u8f7d\u5177\u4f53\u7f51\u7edc\u53c2\u6570":109,"\u52a0\u8f7d\u6a21\u578b\u53ef\u5176\u5b83\u591a\u79cd\u65b9\u5f0f":116,"\u52a0\u8f7d\u6a21\u578b\u9700\u540c\u65f6\u6307\u5b9a":116,"\u52a0\u8f7d\u9884\u6d4b\u6a21\u578b":116,"\u52a0\u8f7d\u9884\u8bad\u7ec3\u53c2\u6570":109,"\u52a0\u8f7dtest":129,"\u52a0\u901f\u7f16\u8bd1":27,"\u52a0\u901fpaddlepaddle\u8bad\u7ec3\u53ef\u4ee5\u8003\u8651\u4ece\u4ee5\u4e0b\u51e0\u4e2a\u65b9\u9762":107,"\u52a8\u6001\u5e93":[71,113],"\u52a9\u624b":100,"\u5305":133,"\u5305\u542b\u4e86\u67d0\u79cd\u7c7b\u578b\u7684\u7c7b\u578b\u5b9a\u4e49\u548c\u66b4\u9732\u7684\u5168\u90e8\u51fd\u6570":72,"\u5305\u542b\u4f46\u4e0d\u9650\u4e8e":27,"\u5305\u542b\u6d4b\u8bd5\u6570\u636e\u96c6\u7684\u76ee\u5f55":117,"\u5305\u542b\u8bad\u7ec3\u6570\u636e\u7684\u76ee\u5f55":117,"\u5305\u542b\u8fd9\u4e2a\u51fd\u6570\u8c03\u7528\u5176\u4ed6\u51fd\u6570\u7684\u65f6\u95f4":133,"\u5305\u542bkernel\u7684op\u548c\u4e0d\u5305\u542bkernel\u7684op":101,"\u5305\u62ec":[38,67,68,113,120,129],"\u5305\u62ec\u4e86\u7f16\u8bd1\u51fa\u7684paddlepaddle\u5934\u6587\u4ef6\u548c\u94fe\u63a5\u5e93":113,"\u5305\u62ec\u5b57\u7b26\u4e32\u5206\u914d":107,"\u5305\u62ec\u6743\u91cdw\u548c\u504f\u7f6eb":37,"\u5305\u62ec\u751f\u6210cpu":27,"\u5305\u62ec\u795e\u7ecf\u7f51\u7edc\u62d3\u6251\u7ed3\u6784":110,"\u5305\u62ecbool":131,"\u5305\u62eclinux":142,"\u5305\u62ecmkl":68,"\u5305\u7684\u65b9\u6cd5\u662f":104,"\u533a\u522b\u662f\u540c\u65f6\u5904\u7406\u4e86\u4e24\u4e2a\u8f93\u5165":137,"\u533a\u522b\u662frnn\u4f7f\u7528\u4e24\u5c42\u5e8f\u5217\u6a21\u578b":137,"\u534f\u540c\u5b8c\u6210releas":88,"\u5355\u4e2a\u503c":38,"\u5355\u4f4d\u662fmb":129,"\u5355\u5143\u6d4b\u8bd5":[26,102],"\u5355\u5143\u6d4b\u8bd5\u4f1a\u5f15\u7528site":104,"\u5355\u5143\u6d4b\u8bd5\u4f1a\u88ab\u81ea\u52a8\u52a0\u5165\u5de5\u7a0b\u8fdb\u884c\u7f16\u8bd1":101,"\u5355\u5143\u6d4b\u8bd5\u5728\u5185\u7684\u6240\u6709\u5355\u5143\u6d4b\u8bd5":26,"\u5355\u5143\u6d4b\u8bd5checkgrad_ep":128,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u53e5\u5b50\u662f\u4e00\u6837\u7684":137,"\u5355\u53cc\u5c42rnn":138,"\u5355\u5c42":139,"\u5355\u5c42\u4e0d\u7b49\u957frnn":137,"\u5355\u5c42\u548c\u53cc\u5c42\u5e8f\u5217\u7684\u4f7f\u7528\u548c\u793a\u4f8b2\u4e2d\u7684\u793a\u4f8b\u7c7b\u4f3c":137,"\u5355\u5c42\u5e8f\u5217":[115,136],"\u5355\u5c42\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20":136,"\u5355\u5c42\u5e8f\u5217\u7b2ci\u4e2a\u5143\u7d20":136,"\u5355\u5c42\u6216\u53cc\u5c42":136,"\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":137,"\u5355\u5c42rnn":[137,139],"\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":137,"\u5355\u673acpu\u8bad\u7ec3":107,"\u5355\u673agpu\u8bad\u7ec3":107,"\u5355\u6b65\u51fd\u6570":140,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u5728":140,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u90fd\u975e\u5e38\u7b80\u5355":140,"\u5355\u6b65\u51fd\u6570\u7684\u5b9e\u73b0\u5982\u4e0b\u6240\u793a":140,"\u5355\u6d4b\u5305\u62ec\u5bf9\u6bd4\u524d\u5411op\u4e0d\u540c\u8bbe\u5907":101,"\u5355\u70b9\u6545\u969c":37,"\u5355\u7eaf\u7684":133,"\u5355\u8fdb\u5355\u51fa":139,"\u5360\u7528\u4e8617":133,"\u5373":[72,101,103,107,108,123],"\u5373\u4e0a\u8ff0\u4ee3\u7801\u4e2d\u7684\u7b2c19\u884c":137,"\u5373\u4e0b\u8f7d\u5931\u8d25":104,"\u5373\u4e0d\u5141\u8bb8\u5728":101,"\u5373\u4e0d\u9700\u8981\u4f7f\u7528memori":137,"\u5373\u4e3a\u4e00\u4e2a\u65f6\u95f4\u6b65":137,"\u5373\u4e3a\u5355\u5c42rnn\u5e8f\u5217\u7684\u4f7f\u7528\u4ee3\u7801":137,"\u5373\u4e3a\u65f6\u95f4\u5e8f\u5217\u7684\u8f93\u5165":137,"\u5373\u4e3a\u8fd9\u4e2a\u53cc\u5c42rnn\u7684\u7f51\u7edc\u7ed3\u6784":137,"\u5373\u4e8c\u7ef4\u6570\u7ec4":137,"\u5373\u4f7f\u7528":[72,108],"\u5373\u4f7f\u7528\u6237\u76f4\u63a5\u5f15\u7528\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":72,"\u5373\u4f7f\u95f4\u9694\u5f88\u5c0f":129,"\u5373\u4f7fc":72,"\u5373\u4f8b\u5982":72,"\u5373\u4fbf\u662f":26,"\u5373\u4fbf\u8bbe\u7f6e":104,"\u5373\u4fbfpaddl":72,"\u5373\u521d\u59cb\u72b6\u6001\u4e3a0":139,"\u5373\u5355\u65f6\u95f4\u6b65\u6267\u884c\u7684\u51fd\u6570":140,"\u5373\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":137,"\u5373\u53cc\u5c42rnn\u7684\u6bcf\u4e2a\u72b6\u6001":139,"\u5373\u53ef":98,"\u5373\u53ef\u4ee5\u6781\u5927\u7684\u52a0\u901f\u6570\u636e\u8f7d\u5165\u6d41\u7a0b":107,"\u5373\u53ef\u4f7f\u7528\u5f00\u53d1\u955c\u50cf\u6765\u7f16\u8bd1android\u7248paddlepaddl":142,"\u5373\u53ef\u5728":144,"\u5373\u53ef\u5f00\u59cb\u4e0b\u8f7d":30,"\u5373\u53ef\u5f00\u59cb\u4e0b\u9762\u7684\u6b65\u9aa4":28,"\u5373\u53ef\u663e\u793a\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c":133,"\u5373\u53ef\u68c0\u67e5\u6211\u4eec\u8c03\u4f18\u540e\u7684\u4fee\u6b63\u662f\u5426\u80fd\u591f\u6539\u5584\u7a0b\u5e8f\u7684\u6027\u80fd":133,"\u5373\u5728\u53cc\u5c42\u5e8f\u5217\u7684\u539f\u59cb\u6570\u636e\u4e2d":137,"\u5373\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":107,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":110,"\u5373\u5b8c\u6210\u67d0\u4e00\u4e2a\u4efb\u52a1\u7684\u6700\u5c11\u51fd\u6570":72,"\u5373\u5c06\u4e00\u6bb5\u8bdd\u8fdb\u884c\u5206\u7c7b":137,"\u5373\u5c06nchw\u8f6c\u6362\u6210nhwc":108,"\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":137,"\u5373\u6211\u4eec\u53ef\u4ee5\u5148\u5b9a\u4e49\u4e00\u4e2atensor":102,"\u5373\u628a\u5355\u5c42rnn\u751f\u6210\u540e\u7684subseq\u7ed9\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42seq":139,"\u5373\u6574\u4e2a\u53cc\u5c42group\u662f\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":137,"\u5373\u6574\u4e2a\u8f93\u5165\u5e8f\u5217":136,"\u5373\u6574\u6570\u6570\u7ec4":137,"\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":137,"\u5373\u662f\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684\u7f51\u7edc\u8fde\u63a5":137,"\u5373\u66b4\u9732":72,"\u5373\u7279\u5f81\u7684\u6570\u7ec4":137,"\u5373\u7f51\u5361\u540d":123,"\u5373\u8868\u793a\u4e0d\u9700\u8981\u8f6c\u6362":68,"\u5373\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0nan\u6216\u8005inf":107,"\u5373\u8bbe\u7f6e":107,"\u5373\u8fd0\u884c\u8bad\u7ec3\u7a0b\u5e8f":28,"\u5373\u8fd9\u4e2a\u52a8\u6001\u5e93\u662f\u4e0d\u4f9d\u8d56\u4e8e\u5176\u4ed6\u4efb\u4f55\u6587\u4ef6\u7684":71,"\u5373define_py_data_sources2\u5e94\u6539\u4e3a":109,"\u5373input":139,"\u5373rnn\u4e4b\u95f4\u6709\u4e00\u6b21\u5d4c\u5957\u5173\u7cfb":137,"\u5378\u8f7dpaddlepaddle\u5305":104,"\u538b\u6241\u6210\u4e3aeigen\u7684\u4e00\u7ef4tensor":102,"\u538b\u7f29\u6210\u4e00\u4e2a\u5411\u91cf":137,"\u539f\u56e0":[98,104],"\u539f\u56e0\u5728\u4e8e\u6ca1\u6709\u628a\u673a\u5668\u4e0acuda\u76f8\u5173\u7684\u9a71\u52a8\u548c\u5e93\u6620\u5c04\u5230\u5bb9\u5668\u5185\u90e8":104,"\u539f\u56e0\u662f\u6bcf\u4e2a\u56de\u590d\u90fd\u4f1a\u53d1\u9001\u4e00\u5c01\u90ae\u4ef6":98,"\u539f\u6765\u7684\u65b9\u6848":68,"\u53c2\u6570":[2,3,4,5,6,7,9,10,11,17,18,19,22,23,25,26,71,100,107,116,123,128],"\u53c2\u6570\u4e3a":101,"\u53c2\u6570\u5171\u4eab\u7684\u914d\u7f6e\u793a\u4f8b\u4e3a":109,"\u53c2\u6570\u548c\u73af\u5883\u53d8\u91cf":117,"\u53c2\u6570\u670d\u52a1\u5668":[118,128],"\u53c2\u6570\u670d\u52a1\u5668\u4e4b\u95f4\u4e0d\u76f8\u4e92\u4f9d\u8d56":118,"\u53c2\u6570\u670d\u52a1\u5668\u4e5f\u4e0d\u4f1a\u7b49\u5f85\u8ba1\u7b97\u8282\u70b9\u5168\u90e8\u90fd\u63d0\u4ea4\u68af\u5ea6\u4e4b\u540e\u624d\u5f00\u59cb\u4e0b\u4e00\u6b65":118,"\u53c2\u6570\u670d\u52a1\u5668\u63a5\u6536\u4ece\u8ba1\u7b97\u8282\u70b9\u4e0a\u4f20\u7684\u68af\u5ea6":118,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":129,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u76d1\u542c\u7aef\u53e3":129,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u7f51\u7edc\u8bbe\u5907\u540d\u79f0":129,"\u53c2\u6570\u670d\u52a1\u5668\u7684ip\u5730\u5740":129,"\u53c2\u6570\u670d\u52a1\u5668\u7a00\u758f\u66f4\u65b0\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":129,"\u53c2\u6570\u6765\u63a7\u5236\u7f13\u5b58\u65b9\u6cd5":107,"\u53c2\u6570\u6982\u8ff0":130,"\u53c2\u6570\u7684\u4e2a\u6570\u548c\u53c2\u6570\u5217\u8868":116,"\u53c2\u6570\u7684\u89e3\u6790":123,"\u53c2\u6570\u8bbe\u7f6e":106,"\u53c2\u6570\u8bbe\u7f6e\u4e86\u5916\u5c42":137,"\u53c2\u6570\u8bf4\u660e":117,"\u53c2\u6570\u8bf4\u660e\u5bb9\u5668\u5df2\u4ea4\u4e92\u5f0f\u8fd0\u884c":28,"\u53c2\u6570\u8f93\u5165":107,"\u53c2\u6570\u9700\u8981\u5b9e\u73b0":140,"\u53c2\u7167\u4e0a\u8ff0\u6b65\u9aa4\u66f4\u65b0":98,"\u53c2\u8003":[28,54,71],"\u53c2\u8003\u4e0b\u56fe":88,"\u53c2\u8003\u5f3a\u8c03\u90e8\u5206":134,"\u53c2\u8003\u65f6\u95f4\u5e8f\u5217":137,"\u53c2\u8003\u6837\u4f8b\u6570\u636e\u51c6\u5907\u811a\u672c":117,"\u53c2\u8003\u955c\u50cf\u7684":123,"\u53c8\u53ef\u4ee5\u907f\u514d\u4e0d\u5fc5\u8981\u7684\u8f6c\u6362":68,"\u53c8\u662f\u4e00\u4e2a\u5355\u5c42\u7684\u5e8f\u5217":136,"\u53c8\u8981\u4fdd\u8bc1\u6570\u636e\u662f\u968f\u673a\u7684":107,"\u53ca":100,"\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u9690\u85cf\u72b6\u6001":140,"\u53cc\u5411\u9a8c\u8bc1":54,"\u53cc\u5c42":139,"\u53cc\u5c42\u4e0d\u7b49\u957frnn":137,"\u53cc\u5c42\u5e8f\u5217":[115,136],"\u53cc\u5c42\u5e8f\u5217\u5728\u5904\u7406\u957f\u5e8f\u5217\u7684\u4efb\u52a1\u6216\u662f\u6784\u5efa\u5c42\u7ea7\u6a21\u578b\u65f6\u4f1a\u53d1\u6325\u4f5c\u7528":115,"\u53cc\u5c42\u5e8f\u5217\u6216\u5355\u5c42\u5e8f\u5217":136,"\u53cc\u5c42\u5e8f\u5217\u6570\u636e\u4e00\u5171\u67094\u4e2a\u6837\u672c":137,"\u53cc\u5c42\u5e8f\u5217\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u5e8f\u5217":136,"\u53cc\u5c42\u5e8f\u5217\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":139,"\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u4e2d\u6bcf\u4e2a\u5143\u7d20":136,"\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"\u53cc\u5c42\u5e8f\u5217\u9700\u8981\u8bbe\u7f6e\u5206\u522b\u4e3a\u5916\u5c42\u5e8f\u5217\u548c\u5185\u5c42\u5e8f\u5217\u5206\u522b\u8bbe\u7f6e":115,"\u53cc\u5c42\u6216\u8005\u5355\u5c42":136,"\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u7684\u4ee3\u7801":137,"\u53cc\u5c42rnn":139,"\u53cc\u5c42rnn\u6570\u636e\u968f\u610f\u52a0\u4e86\u4e00\u4e9b\u9694\u65ad":137,"\u53cc\u5c42rnn\u987e\u540d\u601d\u4e49":137,"\u53cc\u8fdb\u5355\u51fa":139,"\u53cc\u8fdb\u53cc\u51fa":139,"\u53cd\u5411\u4f20\u64ad":100,"\u53cd\u5411\u4f20\u64ad\u6839\u636e\u8f93\u51fa\u7684\u68af\u5ea6":100,"\u53cd\u5411\u8ba1\u7b97\u5df2\u7ecf\u81ea\u52a8\u96c6\u6210\u8fdb\u6d4b\u8bd5\u6846\u67b6":101,"\u53cd\u5411op\u7684\u68af\u5ea6\u6d4b\u8bd5":101,"\u53cd\u5411op\u7c7b":101,"\u53cd\u5411op\u7c7b\u7684\u5b9a\u4e49":101,"\u53cd\u5411opkernel\u7684\u5b9a\u4e49\u4e0e\u524d\u5411op\u7c7b\u4f3c":101,"\u53d1\u578b\u7248":88,"\u53d1\u5e03\u5230dockerhub":88,"\u53d1\u5e03docker\u955c\u50cf\u53ea\u9700\u8981\u5bf9\u81ea\u52a8push\u7684\u955c\u50cf\u6253\u4e0a":88,"\u53d1\u6563\u5230\u4e86\u4e00\u4e2a\u6570\u503c\u7279\u522b\u5927\u7684\u5730\u65b9":107,"\u53d1\u884c\u548c\u7ef4\u62a4":98,"\u53d1\u9001\u53c2\u6570\u7684\u7aef\u53e3\u53f7":129,"\u53d6\u503c\u76f8\u540c\u7684layer":108,"\u53d6\u5176\u4e2d\u4e00\u4e2a\u6a21\u578bparams_pass_90":110,"\u53d6\u51b3\u4e8e":101,"\u53d8\u6362\u77e9\u9635":100,"\u53d8\u91cf\u6765\u533a\u5206layer\u7684\u5c5e\u6027":68,"\u53e5\u5b50\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":115,"\u53e6\u4e00\u4e2a\u65b9\u6cd5\u662f\u4ea4\u53c9\u7f16\u8bd1":144,"\u53e6\u4e00\u4e2a\u662f\u5185\u5b58\u64cd\u4f5c\u91cf":134,"\u53e6\u4e00\u4e2a\u662f\u6bcf\u6761\u5e8f\u5217":107,"\u53e6\u4e00\u79cd\u65b9\u5f0f\u662f\u5c06\u7f51\u7edc\u5c42\u5212\u5206\u5230\u4e0d\u540c\u7684gpu\u4e0a\u53bb\u8ba1\u7b97":131,"\u53e6\u5916":[26,137],"\u53e6\u5916\u6700\u65b0\u7684pip\u5b98\u65b9\u6e90\u4e2d\u7684\u5b89\u88c5\u5305\u9ed8\u8ba4\u662fmanylinux1\u6807\u51c6":30,"\u53ea\u4f5c\u4e3aread":139,"\u53ea\u4fdd\u5b58\u6700\u540e\u4e00\u8f6e\u7684\u53c2\u6570":129,"\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":27,"\u53ea\u5bf9\u7279\u6b8a\u5728\u7ebf\u7cfb\u7edf\u8003\u8651\u4e24\u53f0\u4ee5\u4e0a\u540c\u65f6\u6545\u969c\u7684\u5bb9\u707e":37,"\u53ea\u5bf9\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u8fdb\u884c\u5e8f\u5217\u5316":116,"\u53ea\u5c06\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u8fdb\u884c\u5e8f\u5217\u5316":116,"\u53ea\u662f\u53cc\u5c42\u5e8f\u5217\u5c06\u5176\u53c8\u505a\u4e86\u5b50\u5e8f\u5217\u5212\u5206":137,"\u53ea\u66b4\u9732\u6982\u5ff5\u7684\u63a5\u53e3":72,"\u53ea\u6709\u5728\u9047\u5230\u9700\u8981":29,"\u53ea\u6709\u5f53\u8bbe\u7f6e\u4e86spars":129,"\u53ea\u7528\u4e8e\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d\u6307\u5b9a\u8f93\u5165\u6570\u636e":139,"\u53ea\u7559\u4e0b\u6838\u5fc3\u8ba1\u7b97\u5c42":116,"\u53ea\u80fd\u5728recurrent_group\u4e2d\u4f5c\u4e3astep":108,"\u53ea\u80fd\u6309\u884c\u8ba1\u7b97":108,"\u53ea\u80fd\u6d4b\u8bd5\u5355\u4e2a\u6a21\u578b":131,"\u53ea\u80fd\u8bbf\u95ee\u5b83\u4eec\u7684\u8f93\u51fa\u503c":108,"\u53ea\u80fd\u8c03\u7528paddle\u7684\u52a8\u6001\u5e93":71,"\u53ea\u8981\u4e00\u7cfb\u5217\u7279\u5f81\u6570\u636e\u4e2d\u7684":137,"\u53ea\u8981\u51fa\u73b0\u6d6e\u70b9\u6570\u5f02\u5e38":107,"\u53ea\u8bfbmemory\u8f93\u5165":139,"\u53ea\u9488\u5bf9\u5185\u5b58":107,"\u53ea\u9700\u4e2d\u65ad":119,"\u53ea\u9700\u5728\u7f16\u8bd1\u65f6\u9700\u914d\u5236\u4e0b\u9762\u8fd9\u4e9b\u7f16\u8bd1\u9009\u9879":113,"\u53ea\u9700\u7528\u60a8\u5b9a\u4e49\u7684\u76ee\u5f55\u4fee\u6539":119,"\u53ea\u9700\u8981":140,"\u53ea\u9700\u8981\u6062\u590d\u8fd9\u53f0\u8282\u70b9":37,"\u53ea\u9700\u8981\u8bbe\u7f6e\u884c\u504f\u79fb":115,"\u53ea\u9700\u8981\u94fe\u63a5":113,"\u53ea\u9700\u8fdb\u884c\u524d\u5411\u8ba1\u7b97\u800c\u65e0\u9700\u8c03\u7528\u53cd\u5411\u8ba1\u7b97":116,"\u53ef\u4ee5":[28,88,98,103],"\u53ef\u4ee5\u4ece":28,"\u53ef\u4ee5\u4ece\u6211\u4eec\u7684ci\u7cfb\u7edf\u4e2d\u4e0b\u8f7d\u6700\u65b0\u7684whl\u5b89\u88c5\u5305\u548cc":30,"\u53ef\u4ee5\u4f30\u8ba1\u51fa\u5982\u679c\u6a21\u578b\u91c7\u7528\u4e0d\u53d8\u7684\u8f93\u51fa\u6700\u5c0f\u7684cost0\u662f\u591a\u5c11":109,"\u53ef\u4ee5\u4f7f\u7528":[109,116,120],"\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u547d\u4ee4\u66f4\u65b0\u60a8\u7684pip":30,"\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u4ee3\u7801":109,"\u53ef\u4ee5\u4f7f\u7528\u76f8\u5e94\u6570\u636e\u7c7b\u578b\u7684":109,"\u53ef\u4ee5\u4f7f\u7528\u8be5\u53c2\u6570":129,"\u53ef\u4ee5\u4f7f\u7528kubernetes\u7684\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":123,"\u53ef\u4ee5\u5148\u4f7f\u7528":108,"\u53ef\u4ee5\u51cf\u5c0f\u7cfb\u7edf\u590d\u6742\u6027":37,"\u53ef\u4ee5\u51cf\u5c11\u7f13\u5b58\u6c60\u7684\u5927\u5c0f":107,"\u53ef\u4ee5\u521b\u5efa\u4e00\u4e2a":122,"\u53ef\u4ee5\u521b\u5efa\u975e":101,"\u53ef\u4ee5\u52a0\u901fpaddlepaddle\u7684\u8ba1\u7b97":28,"\u53ef\u4ee5\u53c2\u8003":[26,28,98,137,140],"\u53ef\u4ee5\u53c2\u8003\u4e0b\u9762\u7684\u6b65\u9aa4\u6392\u67e5":105,"\u53ef\u4ee5\u53c2\u8003paddlepaddl":110,"\u53ef\u4ee5\u540c\u65f6\u5728cpu":102,"\u53ef\u4ee5\u542b\u6709\u4e00\u6761\u6216\u591a\u6761\u6837\u672c":115,"\u53ef\u4ee5\u544a\u8bc9\u60a8\u67d0\u4e2a\u64cd\u4f5c\u5230\u5e95\u82b1\u4e86\u591a\u957f\u65f6\u95f4":134,"\u53ef\u4ee5\u5728":[27,119],"\u53ef\u4ee5\u5728\u4efb\u4f55\u673a\u5668\u4e0a\u6267\u884c\u7684":71,"\u53ef\u4ee5\u5728\u5171\u4eab\u5b58\u50a8\u4e0a\u67e5\u770b\u8f93\u51fa\u7684\u65e5\u5fd7\u548c\u6a21\u578b":123,"\u53ef\u4ee5\u5728\u6b64\u9875\u9762\u7684":88,"\u53ef\u4ee5\u5728\u8fd9\u4e2a":98,"\u53ef\u4ee5\u5728event_handler\u4e2d":107,"\u53ef\u4ee5\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684sgd\u65b9\u6cd5\u7684\u8bad\u7ec3":118,"\u53ef\u4ee5\u5b9e\u73b0\u4ecepaddl":102,"\u53ef\u4ee5\u5c06cmake":133,"\u53ef\u4ee5\u5c06memory\u7406\u89e3\u4e3a\u4e00\u4e2a\u65f6\u5ef6\u64cd\u4f5c":139,"\u53ef\u4ee5\u5c06op\u5206\u4e3a\u4e24\u79cd":101,"\u53ef\u4ee5\u5c1d\u8bd5\u4ee5\u4e0b\u7684\u65b9\u6cd5":28,"\u53ef\u4ee5\u5e2e\u60a8\u63d0\u4f9b\u4e00\u4e9b\u5b9a\u4f4d\u6027\u80fd\u74f6\u9888\u7684\u5efa\u8bae":134,"\u53ef\u4ee5\u5e76\u884c\u7f16\u8bd1\u5417":26,"\u53ef\u4ee5\u5feb\u901f\u5728\u672c\u5730\u542f\u52a8\u4e00\u4e2a\u5305\u542b\u4e86paddlepaddle\u5b98\u65b9book\u6559\u7a0b\u7684jupyt":28,"\u53ef\u4ee5\u6267\u884c":[30,104],"\u53ef\u4ee5\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u7f16\u8bd1\u751f\u6210\u6587\u6863":103,"\u53ef\u4ee5\u6267\u884cctest\u547d\u4ee4\u5373\u53ef":27,"\u53ef\u4ee5\u628a\u5b83\u60f3\u8c61\u4e3a\u4e00\u4e2a\u7c7b\u4f3c":26,"\u53ef\u4ee5\u628a\u672c\u5730\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5b58\u50a8\u96c6\u7fa4\u4e2d":38,"\u53ef\u4ee5\u6307\u5b9a\u540c\u65f6\u6267\u884cgpu\u4e0a\u7684\u5355\u5143\u6d4b\u8bd5":27,"\u53ef\u4ee5\u6307\u5b9a\u54ea\u4e00\u4e2a\u8f93\u5165\u548c\u8f93\u51fa\u5e8f\u5217\u4fe1\u606f\u4e00\u81f4":137,"\u53ef\u4ee5\u6307\u5b9a\u5f00\u542f\u81ea\u52a8\u68c0\u6d4bsm\u67b6\u6784":27,"\u53ef\u4ee5\u6309\u7167\u4e0b\u9762\u7684\u65b9\u6cd5":27,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":[136,139],"\u53ef\u4ee5\u662f\u4e00\u4e2a\u975e\u5e8f\u5217":139,"\u53ef\u4ee5\u662f\u4ece\u5206\u5e03\u5f0f\u5b58\u50a8\u6302\u8f7d\u8fc7\u6765\u7684":117,"\u53ef\u4ee5\u662f\u4ee5\u4e0b\u51e0\u79cd":100,"\u53ef\u4ee5\u662f\u6574\u578b":115,"\u53ef\u4ee5\u663e\u793a\u5730\u6307\u5b9a\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7528\u4e8e\u521d\u59cb\u5316memori":139,"\u53ef\u4ee5\u66f4\u6709\u6b21\u5e8f\u7684\u5b8c\u6210\u6027\u80fd\u7684\u4f18\u5316":133,"\u53ef\u4ee5\u6709\u4ee5\u4e0b\u4e24\u79cd":139,"\u53ef\u4ee5\u6709\u6548\u51cf\u5c0f\u7f51\u7edc\u7684\u963b\u585e":129,"\u53ef\u4ee5\u6709\u6548\u7684\u907f\u514dparamet":37,"\u53ef\u4ee5\u67e5\u770b":123,"\u53ef\u4ee5\u67e5\u770b\u6b64pod\u8fd0\u884c\u7684\u5bbf\u4e3b\u673a":122,"\u53ef\u4ee5\u6d4b\u8bd5\u591a\u4e2a\u6a21\u578b":131,"\u53ef\u4ee5\u7528":[26,54],"\u53ef\u4ee5\u7528\u4ee5\u4e0b\u6307\u4ee4":38,"\u53ef\u4ee5\u7528\u5982\u4e0b\u547d\u4ee4":98,"\u53ef\u4ee5\u7528\u6765\u8ba1\u7b97cpu\u51fd\u6570\u6216cuda\u5185\u6838\u7684\u65f6\u95f4\u6d88\u8017":134,"\u53ef\u4ee5\u76f4\u63a5\u8fd0\u884c":116,"\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u975e\u5e8f\u5217\u8f93\u5165":136,"\u53ef\u4ee5\u770b\u51fa":118,"\u53ef\u4ee5\u770b\u5230\u6700\u8017\u65f6\u7684\u51fd\u6570\u662fc":133,"\u53ef\u4ee5\u7cbe\u786e\u8bf4\u660e\u4e00\u4e2a\u957f\u8017\u65f6\u64cd\u4f5c\u7684\u5177\u4f53\u539f\u56e0":134,"\u53ef\u4ee5\u7ee7\u7eed\u5728\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f\u63d0\u4ea4\u4ee3\u7801":88,"\u53ef\u4ee5\u8003\u8651\u4f7f\u7528\u4e00\u4e9b\u4f18\u5316\u7b97\u6cd5":107,"\u53ef\u4ee5\u8054\u7cfbop":105,"\u53ef\u4ee5\u8054\u7cfbop\u662f\u5426\u53ef\u4ee5\u66f4\u6362\u96c6\u7fa4\u6216\u5347\u7ea7\u5f53\u524d\u96c6\u7fa4":105,"\u53ef\u4ee5\u83b7\u53d6\u7f51\u7edc\u4e2d\u5b9a\u4e49\u7684\u4efb\u610f\u591a\u4e2a":116,"\u53ef\u4ee5\u88c5\u7684\u662f":26,"\u53ef\u4ee5\u8bbe\u7f6e":[133,142,143,144],"\u53ef\u4ee5\u8bbf\u95ee\u7531recurr":108,"\u53ef\u4ee5\u8c03\u7528resize\u63a5\u53e3\u8fdb\u884c\u6539\u53d8":102,"\u53ef\u4ee5\u8f7b\u677e\u5730\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":110,"\u53ef\u4ee5\u8fd0\u884c":117,"\u53ef\u4ee5\u9009\u5728\u5728\u5f53\u524d\u673a\u5668\u5b89\u88c5\u4e5f\u53ef\u4ee5\u62f7\u8d1d\u5230\u76ee\u6807\u673a\u5668\u5b89\u88c5":27,"\u53ef\u4ee5\u9009\u62e9\u662f\u5426\u4f7f\u7528\u53c2\u6570":131,"\u53ef\u4ee5\u901a\u8fc7":98,"\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539\u8fd9\u4e24\u4e2a\u51fd\u6570\u6765\u5b9e\u73b0\u590d\u6742\u7684\u7f51\u7edc\u914d\u7f6e":140,"\u53ef\u4ee5\u901a\u8fc7\u5728":107,"\u53ef\u4ee5\u901a\u8fc7\u7f51\u9875\u6d4f\u89c8":28,"\u53ef\u4ee5\u901a\u8fc7\u8fd9\u4e2a\u8f93\u51fa\u6765\u5b8c\u6210\u81ea\u5b9a\u4e49\u7684\u8bc4\u4f30\u6307\u6807\u8ba1\u7b97\u7b49\u529f\u80fd":107,"\u53ef\u4ee5\u901a\u8fc7\u9636\u6bb5\u6027\u7684\u4fdd\u5b58\u6bcf\u4e2aparamet":37,"\u53ef\u4ee5\u91c7\u53d6\u4e0b\u9762\u51e0\u70b9\u63aa\u65bd":133,"\u53ef\u4ee5\u91cd\u547d\u540d\u8fd9\u4e2awhl\u5305\u4e3a":[30,104],"\u53ef\u53c2\u8003":116,"\u53ef\u5728":113,"\u53ef\u5728\u547d\u4ee4\u884c\u6267\u884c":143,"\u53ef\u663e\u5f0f\u6307\u5b9a\u4e3a":143,"\u53ef\u7528\u4e8e\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u89e3\u6790\u8fd9\u4e9b\u53c2\u6570":131,"\u53ef\u76f4\u63a5\u8fd0\u884c":116,"\u53ef\u80fd\u4f1a\u5bfc\u81f4\u51fa\u9519":123,"\u53ef\u80fd\u4f1a\u9020\u6210\u7f51\u7edc\u62e5\u585e":37,"\u53ef\u80fd\u7684\u4ee3\u7801\u4e3a":107,"\u53ef\u80fd\u7684\u539f\u56e0\u662f":109,"\u53ef\u80fd\u7684\u60c5\u51b5\u4e0b":134,"\u53ef\u80fd\u9700\u8981\u6ce8\u610f\u7ed9\u8fd9\u4e2a\u865a\u62df\u673a\u591a\u5206\u914d\u4e00\u4e9b":26,"\u53ef\u89c1\u8be5\u8ba1\u7b97\u7531\u4e24\u4e2a\u8f93\u5165":101,"\u53ef\u8bbe\u7f6e":[142,143],"\u53ef\u8bbe\u7f6e\u4e3a":143,"\u53ef\u8bbe\u7f6e\u7684\u76ee\u6807\u67b6\u6784\u5982\u4e0b\u8868\u6240\u793a":143,"\u53ef\u9009":[27,100,116,117],"\u53ef\u9009\u7684\u4e0d\u540c\u7f16\u8bd1\u73af\u5883docker\u955c\u50cf":27,"\u53ef\u9009\u914d\u7f6e\u9009\u9879":113,"\u53ef\u914d\u7f6e\u4e3a":113,"\u53ef\u91c7\u7528\u7b2c\u4e8c\u79cd\u65b9\u5f0f":108,"\u53f3\u4fa7\u7684":88,"\u5404\u6b21\u524d\u5411\u4e4b\u95f4\u4e5f\u90fd\u4f7f\u7528\u4e86\u76f8\u540c\u7684\u6743\u91cd":67,"\u5404\u9879\u66f4\u52a0\u5177\u4f53\u7684\u5355\u5143\u6d4b\u8bd5\u5728":101,"\u5408\u5e76\u5165\u4e00\u4e2a\u6587\u4ef6":116,"\u5408\u5e76\u6a21\u578b\u6587\u4ef6":116,"\u540c\u4e00\u6b21\u524d\u5411":67,"\u540c\u65f6":[67,68,104,107,134],"\u540c\u65f6\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":27,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u52a0\u901f\u5f00\u59cb\u8bad\u7ec3\u524d\u6570\u636e\u8f7d\u5165\u7684\u8fc7\u7a0b":107,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u901a\u8fc7":98,"\u540c\u65f6\u4e5f\u80fd\u591f\u5f15\u5165\u66f4\u52a0\u590d\u6742\u7684\u8bb0\u5fc6\u673a\u5236":139,"\u540c\u65f6\u4f1a\u5f00\u542fintel":68,"\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":134,"\u540c\u65f6\u518d\u5c06":88,"\u540c\u65f6\u53c8\u5c3d\u53ef\u80fd\u5c11\u7684\u727a\u7272mkl":68,"\u540c\u65f6\u5728\u5185\u5b58\u91cc\u76f4\u63a5\u968f\u5373\u9009\u53d6\u6570\u636e\u6765\u505ashuffl":107,"\u540c\u65f6\u5c06\u53c2\u6570\u521d\u59cb\u5316\u4e3a":109,"\u540c\u65f6\u628a\u5f53\u524d\u76ee\u5f55":26,"\u540c\u65f6\u63d0\u8d77":88,"\u540c\u65f6\u6570\u636e\u683c\u5f0f\u5c31\u662f":68,"\u540c\u65f6\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":131,"\u540c\u65f6\u8bbe\u7f6e\u5185\u5b58\u7f13\u5b58\u529f\u80fd":107,"\u540c\u65f6\u8f93\u51fa\u5e8f\u5217\u5c42\u548c\u975e\u5e8f\u5217\u5c42":107,"\u540c\u6837":110,"\u540c\u6837\u4e5f\u53ef\u4ee5\u5728\u6d4b\u8bd5\u6a21\u5f0f\u4e2d\u6307\u5b9a\u6a21\u578b\u8def\u5f84":129,"\u540c\u6837\u53ef\u4ee5\u6269\u5c55\u5230\u53cc\u5c42\u5e8f\u5217\u7684\u5904\u7406\u4e0a":139,"\u540c\u6837\u53ef\u83b7\u53d6\u5230\u8f93\u5165\u8f93\u51fa\u548c\u5c5e\u6027\u53c2\u6570":101,"\u540c\u6b65\u6267\u884c\u64cd\u4f5c\u7684\u7ebf\u7a0b\u6570":129,"\u540c\u7406":101,"\u540d\u5b57\u4fee\u9970":71,"\u540e":[27,98,109,123,142,143,144],"\u540e\u5411":67,"\u540e\u5411\u4f20\u64ad":100,"\u540e\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u51fa\u7684\u68af\u5ea6":100,"\u540e\u5411\u65f6\u590d\u7528\u5df2\u7ecf\u8f6c\u6362\u8fc7\u7684\u6743\u91cd":67,"\u540e\u7f00\u4e3a":117,"\u540e\u8005\u5728\u6fc0\u6d3b\u51fd\u6570\u53cd\u5411\u8ba1\u7b97\u65f6\u88ab\u8c03\u7528":107,"\u540e\u8005\u622a\u65ad\u56de\u4f20\u7ed9\u524d\u5c42\u7684\u68af\u5ea6":107,"\u540e\u8005\u7528\u4e8e\u68c0\u67e5\u53c2\u6570\u5c5e\u6027\u7684\u5408\u6cd5\u6027":101,"\u540e\u8005\u7ee7\u627f\u81ea":101,"\u540e\u9988":118,"\u5411\u6307\u5b9a\u7684\u76ee\u5f55\u4e2d\u4e00\u4e2a\u65b0\u7684\u6587\u4ef6":37,"\u5411\u91cf":115,"\u5411\u91cfenable_parallel_vector":128,"\u5411paddlepaddle\u7684\u4e3b\u7248\u672c\u5e93\u63d0\u4ea4":88,"\u5417":26,"\u5426\u5219":[101,142,144],"\u5426\u5219\u4f1a\u628a":98,"\u5426\u5219\u4f7f\u7528\u591a\u673a\u8bad\u7ec3":129,"\u5426\u5219\u4f7f\u7528cpu\u6a21\u5f0f":129,"\u5426\u5219\u4f7f\u7528gpu":131,"\u5426\u5219\u5b83\u4ee5\u4e00\u4e2a\u5e8f\u5217\u8f93\u5165":140,"\u5426\u5219\u5f97\u628apaddle\u9759\u6001\u5e93\u94fe\u63a5\u5230\u89e3\u91ca\u5668\u91cc":71,"\u5426\u5219\u9891\u7e41\u7684\u591a\u8282\u70b9\u5de5\u4f5c\u7a7a\u95f4\u90e8\u7f72\u53ef\u80fd\u4f1a\u5f88\u9ebb\u70e6":119,"\u542b\u4e49":133,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u548c\u5b50\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":100,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u6574\u6570":100,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":100,"\u542f\u52a8\u4e00\u4e2a\u65b0\u7684\u7ebf\u7a0b\u5f00\u59cb\u4fdd\u5b58\u68c0\u67e5\u70b9":37,"\u542f\u52a8\u4e00\u4e2a\u6d4b\u8bd5\u96c6\u7fa4":119,"\u542f\u52a8\u53c2\u6570\u8bf4\u660e":118,"\u542f\u52a8\u5bb9\u5668\u5f00\u59cb\u8bad\u7ec3":123,"\u542f\u52a8\u5e76\u884c\u5411\u91cf\u7684\u9608\u503c":129,"\u542f\u52a8\u5feb\u901f\u5e94\u7b54":129,"\u542f\u52a8\u8bad\u7ec3\u4efb\u52a1":124,"\u542f\u7528\u68af\u5ea6\u53c2\u6570\u7684\u9608\u503c":129,"\u547d\u4ee4\u4e3a":[104,122],"\u547d\u4ee4\u521b\u5efa\u65b0\u955c\u50cf":122,"\u547d\u4ee4\u5220\u9664":[142,143,144],"\u547d\u4ee4\u53ef\u4ee5\u8bbe\u7f6e":27,"\u547d\u4ee4\u65f6":142,"\u547d\u4ee4\u6709\u65f6\u5019\u4f1a\u4ea7\u751f\u4e00\u4e9b\u4e2d\u95f4\u7ed3\u679c":26,"\u547d\u4ee4\u770b\u5230\u505c\u6b62\u540e\u4f46\u662f\u6ca1\u6709\u5220\u9664\u7684":26,"\u547d\u4ee4\u7f16\u8bd1\u6e90\u7801\u5373\u53ef":26,"\u547d\u4ee4\u884c\u4e2d\u7684":133,"\u547d\u4ee4\u884c\u53c2\u6570\u8bbe\u7f6e":132,"\u547d\u4ee4\u8bbe\u7f6e\u8be5\u7c7b\u7f16\u8bd1\u9009\u9879":27,"\u547d\u4ee4\u9009\u9879\u5e76\u4e14":119,"\u547d\u4ee4\u91cc\u90fd\u7528\u4e86":26,"\u547d\u540d\u4e3a":98,"\u547d\u540d\u89c4\u8303":101,"\u547d\u540d\u8bf7\u9075\u5b88":101,"\u548c":[26,27,38,67,68,71,72,88,98,100,101,102,103,107,108,109,113,115,117,118,119,131,133,134,137,140,142,144],"\u548c\u4e00\u4e2a\u5df2\u7ecf\u5206\u8bcd\u540e\u7684\u53e5\u5b50":137,"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":110,"\u548c\u4e0b\u9762\u5c06\u8981\u4ecb\u7ecd\u7684\u6ce8\u518c\u51fd\u6570\u4e00\u8d77\u653e\u5728":101,"\u548c\u4e2d\u6587\u6587\u6863":103,"\u548c\u4e4b\u524d\u51cf\u5c0f\u901a\u8fc7\u51cf\u5c0f\u7f13\u5b58\u6c60\u6765\u51cf\u5c0f\u5185\u5b58\u5360\u7528\u7684\u539f\u7406\u4e00\u81f4":107,"\u548c\u504f\u7f6e\u5411\u91cf":100,"\u548c\u5185\u5b58":26,"\u548c\u5217\u53f7":115,"\u548c\u53cc\u5c42\u5e8f\u5217\u542b\u6709subseq":136,"\u548c\u5bf9\u5e94\u884c\u7684\u4ee3\u7801":133,"\u548c\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":136,"\u548c\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":118,"\u548c\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165":140,"\u548c\u64cd\u4f5c\u7cfb\u7edf\u4e0a\u76f4\u63a5\u8fd0\u884c\u7684":26,"\u548c\u672a\u6765\u53ef\u80fd\u8fd8\u4f1a\u7528\u5230":68,"\u548c\u793a\u4f8b2\u4e2d\u7684\u914d\u7f6e\u7c7b\u4f3c":137,"\u548c\u79bb\u7ebf\u6570\u636e\u7684\u65b9\u5f0f":38,"\u548c\u90e8\u5206layer":139,"\u548cpool":136,"\u548cpserver\u4e4b\u95f4\u7528\u4e8e\u7a00\u758f\u7c7b\u578b\u53c2\u6570\u901a\u4fe1\u7684\u7aef\u53e3\u4e2a\u6570":117,"\u54ea\u4e2atrainer\u5148\u5b8c\u6210block\u7684\u8bad\u7ec3":37,"\u54ea\u4e9b\u4e0d\u662f":137,"\u56db\u79cd\u6570\u636e\u7c7b\u578b":110,"\u56e0\u4e3a\u5168\u8fde\u63a5\u5c42\u7684\u6fc0\u6d3b\u53ef\u4ee5\u662fsoftmax":100,"\u56e0\u4e3a\u53c2\u6570":131,"\u56e0\u4e3a\u5b83\u4eec\u7684\u8ba1\u7b97\u6548\u7387\u6bd4":140,"\u56e0\u4e3a\u5b83\u6bd4":140,"\u56e0\u4e3a\u5b98\u65b9\u955c\u50cf":123,"\u56e0\u4e3a\u6211\u4eec\u4f1a\u628a\u6240\u6709\u7f16\u8bd1\u5de5\u5177\u90fd\u5b89\u88c5\u8fdb\u4e00\u4e2a":26,"\u56e0\u4e3a\u6e90\u7801\u5c31\u5728\u672c\u673a\u4e0a":26,"\u56e0\u4e3a\u8fd9\u4e2a\u5de5\u5177\u5177\u6709web\u670d\u52a1\u754c\u9762":133,"\u56e0\u4e3a\u8fd9\u6837\u505a\u4e5f\u6ca1\u6cd5\u4fdd\u8bc1\u6d88\u9664\u968f\u673a\u6027":37,"\u56e0\u4e3ac":133,"\u56e0\u4e3apython\u7684\u641c\u7d22\u8def\u5f84\u662f\u4f18\u5148\u5df2\u7ecf\u5b89\u88c5\u7684python\u5305":104,"\u56e0\u4e3aswig\u5728\u7b2c\u4e09\u65b9\u8bed\u8a00\u4e2d\u66b4\u9732\u7684\u51fd\u6570\u540d":71,"\u56e0\u6b64":[67,100,137,139,142],"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":137,"\u56e0\u6b64\u53ef\u80fd\u6d4b\u8bd5\u4e0d\u591f\u5b8c\u5907":102,"\u56e0\u6b64\u5728\u8f6c\u6362\u65f6\u9700\u8981\u663e\u793a\u7684\u6307\u5b9a":102,"\u56e0\u6b64\u5b83\u662finteger_value_sub_sequ":137,"\u56e0\u6b64\u5f53":142,"\u56e0\u6b64\u6211\u4eec\u91c7\u7528\u8f93\u51fa\u7684\u52a0\u6743\u548c":100,"\u56e0\u6b64\u7528\u6237\u5e76\u4e0d\u9700\u8981\u5173\u5fc3\u5b83\u4eec":128,"\u56e0\u6b64\u9519\u8bef\u7684\u4f7f\u7528\u4e8c\u8fdb\u5236\u53d1\u884c\u7248\u53ef\u80fd\u4f1a\u5bfc\u81f4\u8fd9\u79cd\u9519\u8bef":104,"\u56fd\u5185\u7528\u6237\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u955c\u50cf\u6e90\u6765\u52a0\u901f\u8bbf\u95ee":28,"\u56fe1":[115,116],"\u56fe2":115,"\u56fe\u50cf\u5206\u7c7b":88,"\u56fe\u8868":28,"\u5728":[26,67,68,72,88,98,101,115,116,117,133,136,140,143],"\u5728\u4e00\u4e2a\u4e0d\u53ef\u4e2d\u65ad\u5e76\u7f3a\u5c11\u5907\u4efd\u7684\u8bad\u7ec3\u4efb\u52a1\u4e2d":37,"\u5728\u4e00\u4e2a\u529f\u80fd\u9f50\u5168\u7684kubernetes\u673a\u7fa4\u91cc":122,"\u5728\u4e00\u4e2a\u53c2\u6570\u7684\u68af\u5ea6\u88ab\u66f4\u65b0\u540e":100,"\u5728\u4e00\u4e9b\u5206\u5e03\u5f0f\u7cfb\u7edf\u4e2d":117,"\u5728\u4e00\u6b21\u8bad\u7ec3\u4e2d":133,"\u5728\u4e00\u8f6e\u4e2d\u6bcfsave":129,"\u5728\u4e0a\u56fe\u4e2d\u663e\u793a\u4e86\u5728\u4e00\u4e2a\u5b9e\u9645\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u5e94\u7528":38,"\u5728\u4e0a\u9762\u4ee3\u7801\u4e2d":137,"\u5728\u4e0a\u9762\u7684\u4ee3\u7801\u4e2d":101,"\u5728\u4e0b\u4e00\u7bc7\u4e2d":122,"\u5728\u4e0d\u540c\u96c6\u7fa4\u4e2d\u8fd0\u884c":118,"\u5728\u4e4b\u540e\u7684":107,"\u5728\u4e86\u89e3docker\u7684\u57fa\u672c\u4f7f\u7528\u65b9\u6cd5\u4e4b\u540e":28,"\u5728\u4efb\u610f\u65f6\u95f4\u67d0\u4e00\u53f0\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u4fdd\u5b58\u7684\u53c2\u6570\u53ef\u80fd\u6bd4\u53e6\u4e00\u53f0\u8981\u66f4\u65b0":118,"\u5728\u4f7f\u7528\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u65f6":117,"\u5728\u4f7f\u7528\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u8fdb\u884c\u8bad\u7ec3\u65f6":120,"\u5728\u4f7f\u7528\u540c\u6b65sgd\u8bad\u7ec3\u795e\u7ecf\u7f51\u7edc\u65f6":118,"\u5728\u4f7f\u7528\u65f6":108,"\u5728\u4f7f\u7528\u8be5\u6587\u6863\u4e4b\u524d":110,"\u5728\u4f7f\u7528c":113,"\u5728\u4f7f\u7528paddlepaddl":113,"\u5728\u4f7f\u7528twine\u4e0a\u4f20\u4e4b\u524d":88,"\u5728\u5168\u8fde\u63a5\u5c42\u4e2d":100,"\u5728\u5177\u4f53\u7684\u8ba1\u7b97\u4e2d":102,"\u5728\u51c6\u5907\u53d1\u8d77":98,"\u5728\u51fa\u73b0\u5355\u70b9\u6545\u969c\u65f6":37,"\u5728\u51fd\u6570":123,"\u5728\u5206\u5e03\u5f0f\u73af\u5883\u4e2d\u6d4b\u8bd5":129,"\u5728\u5206\u5e03\u5f0f\u8bad\u7ec3\u4e2d":129,"\u5728\u521b\u5efaparameters\u540e":109,"\u5728\u5355\u5c42\u6570\u636e\u7684\u57fa\u7840\u4e0a":137,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u52a0\u8f7d\u548c\u4fdd\u5b58\u53c2\u6570":129,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u7ec8\u7aef\u6bcflog":129,"\u5728\u53cc\u5c42rnn\u4e2d\u7684\u7ecf\u5178\u60c5\u51b5\u662f\u5c06\u5185\u5c42\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u6570\u636e":137,"\u5728\u53cd\u5411\u4f20\u9012\u7684\u65f6\u5019":107,"\u5728\u53d8\u6362\u65f6\u9700\u8981\u5c06\u8f93\u5165\u5e8f\u5217\u4f20\u5165":137,"\u5728\u542f\u52a8job\u4e4b\u524d":123,"\u5728\u547d\u4ee4\u884c\u663e\u793a\u5206\u6790\u7ed3\u679c":133,"\u5728\u56de\u590d\u8bc4\u5ba1\u4eba\u610f\u89c1\u65f6":98,"\u5728\u56fe\u50cf\u4efb\u52a1\u4e2d":108,"\u5728\u591acpu\u8bad\u7ec3\u65f6\u5171\u4eab\u8be5\u53c2\u6570":129,"\u5728\u5b8c\u6210\u4e00\u5b9a\u91cf\u6570\u636e\u7684\u8bad\u7ec3\u540e":118,"\u5728\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684\u642d\u5efa\u4e4b\u540e":110,"\u5728\u5b9a\u4e49\u8f93\u5165layer\u4e4b\u540e":110,"\u5728\u5b9e\u73b0\u6bcf\u4e2a\u5b50\u7c7b\u7684\u65f6\u5019\u5c31\u4e0d\u9700\u8981\u5173\u5fc3\u5206\u652f\u7684\u4e8b\u60c5\u4e86":68,"\u5728\u5b9e\u73b0\u8fc7\u7a0b\u4e2d":72,"\u5728\u5b9e\u9645\u5e94\u7528\u4e2d":108,"\u5728\u5bb9\u5668\u4e2d\u7f16\u8f91\u4ee3\u7801":28,"\u5728\u5bb9\u5668\u521b\u5efa\u540e":123,"\u5728\u5bf9\u5bb9\u5668\u7684\u63cf\u8ff0":123,"\u5728\u5bf9\u5e94\u7684":67,"\u5728\u5c42\u4e2d\u6307\u5b9a":131,"\u5728\u5c42\u521d\u59cb\u5316\u7684\u65f6\u5019":67,"\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d":139,"\u5728\u5f00\u59cb\u8bad\u7ec3\u4e4b\u524d":38,"\u5728\u5f02\u6784\u96c6\u7fa4\u4e2d":37,"\u5728\u5f02\u6b65sgd\u4e2d":118,"\u5728\u5f15\u5165\u5176\u4ed6\u7c7b\u578b\u7684\u5934\u6587\u4ef6\u65f6":72,"\u5728\u5f53\u524d":107,"\u5728\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":100,"\u5728\u5f97\u5230":123,"\u5728\u5feb\u7167\u5199\u5165\u5b8c\u6210\u540e":37,"\u5728\u60a8\u7684\u5b9e\u9645\u73af\u5883\u4e2d":37,"\u5728\u6211\u4eec\u7684\u4f8b\u5b50\u4e2d":140,"\u5728\u6267\u884c\u65f6":102,"\u5728\u63d0\u4ea4":98,"\u5728\u642d\u5efa\u795e\u7ecf\u7f51\u7edc\u7684\u8fc7\u7a0b\u4e2d":110,"\u5728\u65e0\u7279\u6b8a\u9700\u6c42\u60c5\u51b5\u4e0b":113,"\u5728\u6709\u666e\u901a\u7684cpu":68,"\u5728\u672c\u4f8b\u4e2d":[98,131,137],"\u5728\u672c\u6559\u7a0b\u4e2d":140,"\u5728\u672c\u6587\u6863\u4e2d":54,"\u5728\u672c\u793a\u4f8b\u4e2d":137,"\u5728\u672c\u8282\u4e2d":140,"\u5728\u673a\u7fa4\u4e0a\u8fd0\u884c\u8f6c\u6362\u7a0b\u5e8f":38,"\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b":107,"\u5728\u6811\u7684\u6bcf\u4e00\u5c42\u4e0a":129,"\u5728\u6837\u4f8b\u4e2d":72,"\u5728\u6b64":[128,131],"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":140,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u76f8\u7b49":137,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u957f":140,"\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":123,"\u5728\u6d4b\u8bd5\u9636\u6bb5":129,"\u5728\u6e90\u7801\u76ee\u5f55\u6811\u7684\u6839\u76ee\u5f55\u4e2d\u8fd0\u884c":98,"\u5728\u751f\u6210\u65f6":140,"\u5728\u7528\u6237\u4f7f\u7528c":72,"\u5728\u76f8\u5e94\u7684\u4f18\u5316\u7b97\u6cd5\u91cc\u8bbe\u7f6elearning_rate_schedule\u53ca\u76f8\u5173\u53c2\u6570":109,"\u5728\u76f8\u5e94layer\u7684":108,"\u5728\u795e\u7ecf\u7f51\u7edc\u4e2d\u7b49\u4e8e\u4e00\u6b21\u9884\u6d4b\u5904\u7406\u7684\u6837\u672c\u6570":115,"\u5728\u7a0b\u5e8f\u5b9e\u73b0\u4e2d\u90fd\u4f1a\u8f6c\u5316\u4e3a\u4e8c\u7ef4\u77e9\u9635":115,"\u5728\u7b2c\u4e8c\u4e2atab":88,"\u5728\u7ebf\u4e0a\u7cfb\u7edf\u4e2d":117,"\u5728\u7ebf\u6a21\u578b\u9884\u6d4b\u670d\u52a1":38,"\u5728\u7ec4\u5408\u65f6":110,"\u5728\u7ec4\u7ec7\u795e\u7ecf\u7f51\u7edc\u8f93\u5165":116,"\u5728\u7ec4\u7ec7\u795e\u7ecf\u7f51\u7edc\u8f93\u5165\u65f6":115,"\u5728\u7ec8\u7aef\u6267\u884c":116,"\u5728\u7f16\u8bd1\u5bbf\u4e3b\u673a\u7248protoc\u53ef\u6267\u884c\u6587\u4ef6\u548c\u76ee\u6807\u673a\u7248openblas\u5e93\u65f6\u9700\u8981\u7528\u5230":[142,144],"\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":100,"\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d":136,"\u5728\u8bad\u7ec3\u4e2d":140,"\u5728\u8bad\u7ec3\u4e4b\u524d":123,"\u5728\u8bad\u7ec3\u65f6":122,"\u5728\u8bad\u7ec3\u7ed3\u675f\u7684\u65f6\u5019\u518d\u4fdd\u5b58\u4e3apaddlepaddle\u7684\u683c\u5f0f":68,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":123,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6bcfshow":129,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u53c2\u6570\u7684\u6743\u91cd\u548c\u68af\u5ea6":107,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u67d0\u4e00\u4e2alayer\u7684output":107,"\u5728\u8bbe\u7f6e":[142,143,144],"\u5728\u8bc4\u5ba1\u8fc7\u7a0b\u4e2d":88,"\u5728\u8be5\u793a\u4f8b\u4e2d":109,"\u5728\u8be5\u914d\u7f6e\u76847":137,"\u5728\u8c03\u7528":116,"\u5728\u8c03\u7528c":116,"\u5728\u8f6f\u4ef6\u5de5\u7a0b\u7684\u8303\u7574\u91cc":134,"\u5728\u8f93\u51fa\u7684\u8fc7\u7a0b\u4e2d":139,"\u5728\u8fd0\u884c\u5b8c\u6027\u80fd\u5206\u6790\u540e":133,"\u5728\u8fd0\u884c\u65f6\u5c06\u795e\u7ecf\u7f51\u7edc\u7684\u591a\u4e2a\u53ef\u5b66\u4e60\u53c2\u6570\u653e\u5728\u540c\u4e00\u4e2a\u76ee\u5f55\u4e2d":116,"\u5728\u8fd0\u884c\u795e\u7ecf\u7f51\u7edc\u8ba1\u7b97\u56fe\u65f6":102,"\u5728\u8fd9\u4e2a":88,"\u5728\u8fd9\u4e2a\u4f8b\u5b50\u91cc":[100,122],"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":137,"\u5728\u8fd9\u4e2a\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":71,"\u5728\u8fd9\u4e2a\u6559\u7a0b\u4e2d":134,"\u5728\u8fd9\u4e2a\u6a21\u578b\u4e2d":140,"\u5728\u8fd9\u4e2a\u9636\u6bb5\u7684\u4ee3\u7801\u6b63\u5728\u7ecf\u5386\u56de\u5f52\u6d4b\u8bd5":88,"\u5728\u8fd9\u4e9b\u5934\u6587\u4ef6\u4e2d":72,"\u5728\u8fd9\u4e9b\u6587\u4ef6\u4e2d":72,"\u5728\u8fd9\u4e9blayer\u4e2d":137,"\u5728\u8fd9\u65f6\u771f\u6b63\u7684\u5206\u914d\u5185\u5b58":102,"\u5728\u8fd9\u6bb5\u4ee3\u7801\u4e2d":102,"\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b":[100,140],"\u5728\u8fd9\u79cd\u7ed3\u6784\u4e2d":139,"\u5728\u8fd9\u7bc7\u6587\u6863\u91cc":122,"\u5728\u8fd9\u7bc7\u6587\u7ae0\u91cc":123,"\u5728\u8fd9\u91cc":139,"\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":100,"\u5728\u8fd9\u91cc\u7528eigenvector\u6765\u8868\u793a":102,"\u5728\u8fd9\u91cc\u9700\u8981\u6ce8\u610f\u7684\u662f":102,"\u5728\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u65f6":117,"\u5728\u8fdb\u884c\u7f51\u7edc\u914d\u7f6e\u4e4b\u524d":110,"\u5728\u91c7\u7528sgd":109,"\u5728\u91cd\u6784\u524d\u7684paddlepaddle\u4e2d":68,"\u5728\u95ee\u9898\u672c\u8eab\u7684\u8ba1\u7b97\u91cf\u6bd4\u8f83\u5c0f\u7684\u65f6\u5019":67,"\u5728\u96c6\u7fa4\u4e0a\u8bad\u7ec3\u4e00\u4e2a\u7a00\u758f\u6a21\u578b\u9700\u8981\u52a0\u4e0a\u4e0b\u9762\u7684\u53c2\u6570":131,"\u5728\u975e\u5e8f\u5217\u8f93\u5165\u65f6":107,"\u5728android\u5e73\u53f0\u4e0a\u4e0d\u652f\u6301\u901a\u8fc7swig\u8c03\u7528\u6765\u8bad\u7ec3\u6216\u8005\u9884\u6d4b":142,"\u5728android\u5e73\u53f0\u4e0a\u53ea\u652f\u6301\u4f7f\u7528c":142,"\u5728batch":67,"\u5728build\u76ee\u5f55\u4e0b\u6267\u884c":104,"\u5728c":[71,115],"\u5728c\u7684\u5934\u6587\u4ef6":71,"\u5728cmake\u53c2\u6570\u914d\u7f6e\u4e0a":[142,143],"\u5728cmake\u7684\u547d\u4ee4\u884c\u4e2d":27,"\u5728eigen\u4e2d":102,"\u5728ios\u5e73\u53f0\u4e0a\u4e0d\u652f\u6301\u901a\u8fc7swig\u8c03\u7528\u6765\u8bad\u7ec3\u6216\u8005\u9884\u6d4b":143,"\u5728ios\u5e73\u53f0\u4e0a\u53ea\u652f\u6301\u4f7f\u7528c":143,"\u5728main":133,"\u5728openmpi\u96c6\u7fa4\u4e2d\u63d0\u4ea4\u8bad\u7ec3\u4f5c\u4e1a":120,"\u5728packing\u4e0a\u7684\u8017\u65f6":67,"\u5728paddl":123,"\u5728paddle\u4e2d":131,"\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":38,"\u5728paddlepaddl":115,"\u5728paddlepaddle\u4e2d":[110,139],"\u5728paddlepaddle\u4e2d\u4f7f\u7528dropout\u6709\u4e24\u79cd\u65b9\u5f0f":108,"\u5728paddlepaddle\u4e2d\u5305\u542b\u4ee5\u4e0b":108,"\u5728paddlepaddle\u5185\u90e8":[115,116],"\u5728paddlepaddle\u7684\u6587\u6863\u4e2d":137,"\u5728paramet":37,"\u5728python\u811a\u672c\u4e2d\u5b9e\u73b0\u4e0e\u524d\u5411operator\u76f8\u540c\u7684\u8ba1\u7b97\u903b\u8f91":101,"\u5728rnn\u7684\u60c5\u51b5\u4e0b":67,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49":139,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49memori":139,"\u5728trainer":131,"\u5728trainer\u4e2d\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u53d6\u6a21\u7684\u65b9\u6cd5\u4e3a\u6bcf\u4e2atrainer\u5206\u914d\u8bad\u7ec3\u6570\u636e\u6587\u4ef6":117,"\u5747\u4f1a\u5b58\u653e\u4e8e":113,"\u5747\u4f1a\u88ab\u5b89\u88c5\u5230includ":72,"\u5747\u5300\u5206\u5e03":109,"\u5747\u5300\u5206\u5e03\u7684\u8303\u56f4\u662f":129,"\u5747\u662f\u5728":72,"\u5747\u6709\u4e09\u4e2a\u5b50\u5e8f\u5217":137,"\u5747\u6709\u4e24\u7ec4\u7279\u5f81":137,"\u57fa\u4e8e\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":139,"\u57fa\u4e8e\u7c98\u6027\u4f1a\u8bdd\u7684\u8d1f\u8f7d\u5747\u8861\u529f\u80fd":54,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":[111,116],"\u57fa\u7840\u4e0a":115,"\u57fa\u7c7b":101,"\u586b\u5199":98,"\u589e\u52a0":101,"\u589e\u52a0\u4e86\u4e00\u6761cd\u547d\u4ee4":122,"\u589e\u52a0\u4e86\u8bbe\u5907\u7c7b\u578b":101,"\u589e\u52a0\u5982\u4e0b\u53c2\u6570":131,"\u589e\u52a0\u68af\u5ea6\u68c0\u6d4b\u7684\u5355\u5143\u6d4b\u8bd5":100,"\u5904\u7406\u5668\u6709\u4e24\u4e2a\u5173\u952e\u6027\u80fd\u9650\u5236":134,"\u5904\u7406\u7684\u8f93\u5165\u5e8f\u5217\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u4e09\u79cd\u7c7b\u578b":139,"\u5907\u6ce8":134,"\u590d\u6742\u5ea6\u6216\u65f6\u95f4\u590d\u6742\u5ea6":134,"\u5916\u5c42\u5e8f\u5217\u5728":115,"\u5916\u5c42memory\u662f\u4e00\u4e2a\u5143\u7d20":137,"\u5916\u5c42outer_step\u4e2d":137,"\u5916\u90e8\u5b58\u50a8":68,"\u591a\u4e2a":116,"\u591a\u4e2a\u503c":38,"\u591a\u4e2a\u5c42\u7684\u8f93\u51fa\u77e9\u9635\u7684\u9ad8\u5ea6\u4e0d\u4e00\u81f4\u5bfc\u81f4\u62fc\u63a5\u5931\u8d25":107,"\u591a\u4e2a\u6392\u6210\u4e00\u5217\u7684\u5143\u7d20":115,"\u591a\u4e2a\u8f93\u51fa\u5c42\u5904\u7406\u591a\u4e2a\u4e0d\u540c\u957f\u5ea6\u7684\u5e8f\u5217":107,"\u591a\u4e2aip\u4f7f\u7528":117,"\u591a\u4e2aparamet":37,"\u591a\u53e5\u8bdd\u8fdb\u4e00\u6b65\u6784\u6210\u4e86\u6bb5\u843d":139,"\u591a\u673a\u8bad\u7ec3":107,"\u591a\u6b21\u8c03\u7528":67,"\u591a\u7528\u4e8e\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1":116,"\u591a\u8f6e\u5bf9\u8bdd\u7b49\u66f4\u4e3a\u590d\u6742\u7684\u8bed\u8a00\u6570\u636e":139,"\u5927\u4e8e\u7b49\u4e8e\u4e00\u4e2a":116,"\u5927\u591a\u6570\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u51fd\u6570":100,"\u5927\u591a\u6570\u5c42\u9700\u8981\u8bbe\u7f6e\u4e3a":100,"\u5927\u591a\u6570\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u652f\u6301\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":100,"\u5927\u591a\u6570\u8bed\u8a00\u90fd\u652f\u6301\u4f7f\u7528c\u8bed\u8a00api":71,"\u5927\u5bb6\u53ef\u4ee5\u7528\u628a\u5f00\u53d1\u5de5\u5177\u5b89\u88c5\u8fdb\u5165":26,"\u5927\u5bb6\u53ef\u4ee5\u901a\u8fc7\u5b83\u9605\u8bfb\u6559\u7a0b":28,"\u5927\u5c0f\u4e0d\u4e00\u6837\u65f6":107,"\u5927\u6982\u82b1\u5341\u5206\u949f\u770b\u4e00\u4e0b":26,"\u5927\u90e8\u5206":133,"\u5934\u4fe1\u606f\u4e2d":109,"\u5934\u6587\u4ef6\u4e2d\u628a\u53c2\u6570\u5b9a\u4e49\u4e3a\u7c7b\u7684\u6210\u5458\u53d8\u91cf":100,"\u5934\u6587\u4ef6\u5982\u4e0b":100,"\u5982":[98,101,131,140],"\u5982\u4e0a\u4e00\u5c0f\u8282\u6240\u793a":102,"\u5982\u4e0b":117,"\u5982\u4e0b\u56fe\u6240\u793a":[134,137],"\u5982\u4e0b\u6240\u793a":131,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":99,"\u5982\u4f55\u8d21\u732e\u6587\u6863":99,"\u5982\u56fe\u4e2dtrainer":37,"\u5982\u6709":101,"\u5982\u679c\u4e00\u4e2a\u7f51\u7edc\u5c42\u9700\u8981\u914d\u7f6e\u7684\u8bdd":100,"\u5982\u679c\u4e0a\u9762\u4e24\u6b65\u51fa\u73b0\u9519\u8bef":37,"\u5982\u679c\u4e0d\u4e3a0":129,"\u5982\u679c\u4e0d\u4f7f\u7528\u5206\u5e03\u5f0f\u5b58\u50a8":117,"\u5982\u679c\u4e0d\u4f7f\u7528docker":27,"\u5982\u679c\u4e0d\u4f7f\u7528docker\u7f16\u8bd1\u73af\u5883":27,"\u5982\u679c\u4e0d\u60f3\u4f7f\u7528":103,"\u5982\u679c\u4e0d\u6307\u5b9a\u8fd9\u4e2a\u6587\u4ef6":133,"\u5982\u679c\u4e0d\u6536\u655b":109,"\u5982\u679c\u4e0d\u9700\u8981\u5916\u90e8\u5b58\u50a8\u7528\u4e8e\u8f6c\u6362":68,"\u5982\u679c\u4e3a0":129,"\u5982\u679c\u4e3a\u5426\u5219\u662f\u7528openbla":27,"\u5982\u679c\u4e3afals":129,"\u5982\u679c\u4e3atrue":129,"\u5982\u679c\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe\u7f6e":27,"\u5982\u679c\u4ec5\u4ec5\u4fee\u6539\u4e00\u4e2a\u6587\u4ef6\u4f46\u63d0\u4ea4\u4e86\u5341\u51e0\u4e2acommit":98,"\u5982\u679c\u4ecd\u7136\u5b58\u5728\u95ee\u9898":30,"\u5982\u679c\u4ed4\u7ec6\u8bbe\u7f6e\u7684\u8bdd":129,"\u5982\u679c\u4f60\u53ea\u9700\u8981\u4f7f\u7528\u7b80\u5355\u7684rnn":140,"\u5982\u679c\u4f60\u60f3\u4f7f\u7528\u8fd9\u4e9b\u7279\u6027":131,"\u5982\u679c\u4f60\u60f3\u8981\u4fdd\u5b58\u67d0\u4e9b\u5c42\u7684\u7279\u5f81\u56fe":129,"\u5982\u679c\u4f60\u66fe\u5728\u6e90\u7801\u76ee\u5f55\u4e0b\u7f16\u8bd1\u8fc7\u5176\u4ed6\u5e73\u53f0\u7684paddlepaddle\u5e93":143,"\u5982\u679c\u4f60\u66fe\u7ecf\u5728\u6e90\u7801\u76ee\u5f55\u4e0b\u7f16\u8bd1\u8fc7\u5176\u4ed6\u5e73\u53f0\u7684paddlepaddle\u5e93":[142,144],"\u5982\u679c\u4f60\u6b63\u5728\u5904\u7406\u5e8f\u5217\u6807\u8bb0\u4efb\u52a1":140,"\u5982\u679c\u4f60\u8981\u4e3a\u4e86\u6d4b\u8bd5\u800c\u589e\u52a0\u65b0\u7684\u6587\u4ef6":100,"\u5982\u679c\u4f7f\u7528":116,"\u5982\u679c\u4f7f\u7528docker\u7f16\u8bd1\u73af\u5883":27,"\u5982\u679c\u4f7f\u7528mkl\u5e76\u4e14\u673a\u5668\u542b\u6709avx2\u6307\u4ee4\u96c6":27,"\u5982\u679c\u4f7f\u7528swig\u6211\u4eec\u9700\u8981\u5c06\u5728interface\u6587\u4ef6\u91cc":71,"\u5982\u679c\u5173\u95edmkl":27,"\u5982\u679c\u51fa\u73b0\u4ee5\u4e0bpython\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86\u7684\u60c5\u51b5":104,"\u5982\u679c\u53c2\u6570\u4fdd\u5b58\u4e0b\u6765\u7684\u6a21\u578b\u76ee\u5f55":107,"\u5982\u679c\u53d1\u73b0\u6700\u65e9\u7684\u62a5\u9519\u5c31\u662f\u7f51\u7edc\u901a\u4fe1\u7684\u95ee\u9898":105,"\u5982\u679c\u540c\u65f6\u4f7f\u7528":117,"\u5982\u679c\u5728\u4f7f\u7528mkl":68,"\u5982\u679c\u5728\u70b9\u51fb\u4e0b\u9762\u94fe\u63a5\u65f6\u51fa\u73b0\u5982\u4e0b\u767b\u9646\u754c\u9762":30,"\u5982\u679c\u5728\u7f16\u8bd1":113,"\u5982\u679c\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u672a\u8bbe\u7f6easync":129,"\u5982\u679c\u5728\u8bad\u7ec3\u671f\u95f4\u540c\u65f6\u53d1\u8d77\u53e6\u5916\u4e00\u4e2a\u8fdb\u7a0b\u8fdb\u884c\u6d4b\u8bd5":129,"\u5982\u679c\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6ebatch":129,"\u5982\u679c\u5728\u8bad\u7ec3nlp\u76f8\u5173\u6a21\u578b\u65f6":109,"\u5982\u679c\u591a\u4e2aop\u4f9d\u8d56\u4e00\u4e9b\u5171\u7528\u7684\u51fd\u6570":101,"\u5982\u679c\u5931\u8d25":88,"\u5982\u679c\u5b58\u5728\u6570\u636e\u6392\u5217\u683c\u5f0f\u4e0d\u4e00\u6837\u7684\u60c5\u51b5\u65f6":68,"\u5982\u679c\u5b58\u5728\u67d0\u4e9btrainer\u6267\u884c\u901f\u5ea6\u8fc7\u6162\u4f1a\u5f71\u54cd\u6574\u4f53\u96c6\u7fa4\u7684\u901f\u5ea6":37,"\u5982\u679c\u5c06\u8fd9\u4e2a\u5185\u5b58\u6c60\u51cf\u5c0f":107,"\u5982\u679c\u5c0f\u4e8e75m":104,"\u5982\u679c\u5df2\u7ecf\u6709pod\u8fd0\u884c":123,"\u5982\u679c\u5df2\u7ecf\u6b63\u5728\u6267\u884c\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":37,"\u5982\u679c\u5e0c\u671b\u53ef\u4ee5\u5728\u540e\u53f0\u8fd0\u884cpserver\u7a0b\u5e8f":117,"\u5982\u679c\u5f53\u524dmpi\u96c6\u7fa4\u5e76\u4e0d\u652f\u6301\u4efb\u52a1\u72ec\u5360\u6a21\u5f0f":105,"\u5982\u679c\u60a8\u5728\u4f7f\u7528window":28,"\u5982\u679c\u60a8\u60f3\u8981\u66f4\u6df1\u5165\u4e86\u89e3deep":28,"\u5982\u679c\u60a8\u671f\u671b\u5728\u7f16\u8bd1\u5b8c\u6210\u540e\u7acb\u5373\u6267\u884c\u6240\u6709\u7684\u5355\u5143\u6d4b\u8bd5":27,"\u5982\u679c\u60a8\u6ca1\u6709\u542c\u8bf4":26,"\u5982\u679c\u60a8\u7684\u7535\u8111\u4e0d\u652f\u6301avx":28,"\u5982\u679c\u60a8\u7684gpu\u7406\u8bba\u53ef\u4ee5\u8fbe\u52306":134,"\u5982\u679c\u60a8\u9009\u62e9\u4e0d\u4f7f\u7528docker\u955c\u50cf":27,"\u5982\u679c\u60f3\u4f7f\u7528\u53ef\u89c6\u5316\u7684\u5206\u6790\u5668":134,"\u5982\u679c\u60f3\u5f88\u597d\u7684\u7406\u89e3\u7a0b\u5e8f\u7684\u884c\u4e3a":134,"\u5982\u679c\u60f3\u6539\u53d8\u539f\u6709tensor\u7684shape\u4fe1\u606f":102,"\u5982\u679c\u60f3\u8981\u4e86\u89e3\u53cc\u5c42rnn\u5728\u5177\u4f53\u95ee\u9898\u4e2d\u7684\u4f7f\u7528":137,"\u5982\u679c\u60f3\u8981\u542f\u7528paddlepaddle\u7684\u5185\u7f6e\u5b9a\u65f6\u5668":134,"\u5982\u679c\u6211\u4eec\u53ea\u9700\u8981\u7f16\u8bd1\u4e00\u4e2a\u53ea\u652f\u6301":26,"\u5982\u679c\u6211\u77e5\u9053\u5185\u6838\u82b1\u4e8610ms\u6765\u79fb\u52a81gb\u6570\u636e":134,"\u5982\u679c\u6307\u5b9a\u4e862\u4e2alayer\u4f5c\u4e3a\u8f93\u51fa\u5c42":107,"\u5982\u679c\u63d0\u793a\u6b63\u786e":103,"\u5982\u679c\u652f\u6301\u589e\u52a0\u6b64\u53c2\u6570\u63d0\u4ea4":105,"\u5982\u679c\u662f\u4e00\u4e2a\u5e8f\u5217\u8f93\u5165":115,"\u5982\u679c\u662f\u5176\u5b83\u7c7b\u578b":38,"\u5982\u679c\u662f\u7528\u7f16\u8bd1\u65f6\u6307\u5b9acpu\u7248\u672c":113,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165":139,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":139,"\u5982\u679c\u6709\u9700\u8981\u4fee\u6539\u7684\u5730\u65b9":98,"\u5982\u679c\u6709bugfix\u7684\u884c\u4e3a":88,"\u5982\u679c\u672a\u8bbe\u7f6e":129,"\u5982\u679c\u672a\u8bbe\u7f6egpu":131,"\u5982\u679c\u673a\u5668\u4e2d\u5df2\u7ecf\u5b89\u88c5\u8fc7paddlepaddl":27,"\u5982\u679c\u67d0\u4e00\u4e2a\u7c7b\u578b\u9700\u8981\u5f15\u7528\u53e6\u4e00\u4e2a\u7c7b\u578b":72,"\u5982\u679c\u67d0\u4e00\u4e2apaddl":72,"\u5982\u679c\u67d0\u4e00\u4e2apaddle\u6982\u5ff5\u5fc5\u987b\u8981\u66b4\u9732":72,"\u5982\u679c\u67d0\u4e00\u5757\u6839\u672c\u5c31\u4e0d\u600e\u4e48\u8017\u65f6":134,"\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":129,"\u5982\u679c\u6ca1\u6709\u5b89\u88c5nvidia":28,"\u5982\u679c\u6ca1\u6709\u5b9a\u4e49memori":139,"\u5982\u679c\u6ca1\u8fc7":98,"\u5982\u679c\u6d88\u606f\u6570\u636e\u592a\u5c0f":129,"\u5982\u679c\u6ee1\u8db3\u6761\u4ef6":37,"\u5982\u679c\u7528\u516c\u7528\u7684\u7535\u8111\u5f00\u53d1":26,"\u5982\u679c\u7528\u6237\u4e0d\u9700\u8981\u8bbf\u95eelstm\u7684\u4e2d\u95f4\u53d8\u91cf":108,"\u5982\u679c\u7528\u6237\u60f3\u8981\u81ea\u5b9a\u4e49\u521d\u59cb\u5316\u65b9\u5f0f":109,"\u5982\u679c\u7528\u6237\u8981\u628apaddle\u7684\u9759\u6001\u5e93":71,"\u5982\u679c\u7528\u81ea\u5df1\u7684\u7535\u8111\u5f00\u53d1":26,"\u5982\u679c\u771f\u60f3\u6316\u6398\u5185\u6838\u6df1\u5904\u7684\u67d0\u4e2a\u79d8\u5bc6":134,"\u5982\u679c\u795e\u7ecf\u7f51\u7edc\u6709\u591a\u4e2a\u8f93\u5165\u6216\u8005\u591a\u4e2a\u8f93\u51fa":[115,116],"\u5982\u679c\u7a0b\u5e8f\u5d29\u6e83\u4f60\u4e5f\u53ef\u4ee5\u624b\u52a8\u7ec8\u6b62":119,"\u5982\u679c\u7cfb\u7edf\u5b89\u88c5\u4e86\u591a\u4e2apython\u7248\u672c":104,"\u5982\u679c\u7cfb\u7edf\u652f\u6301":[30,104],"\u5982\u679c\u7cfb\u7edf\u652f\u6301\u7684\u662f":[30,104],"\u5982\u679c\u7f16\u8bd1\u65f6\u6307\u5b9a\u7f16\u8bd1cpu\u7248\u672c":113,"\u5982\u679c\u7f16\u8bd1\u65f6\u6307\u5b9a\u7f16\u8bd1gpu\u7248\u672c":113,"\u5982\u679c\u7f16\u8bd1\u7684\u65f6\u5019\u6211\u4eec\u7528\u4e86":26,"\u5982\u679c\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":100,"\u5982\u679c\u7f51\u7edc\u67b6\u6784\u7b80\u5355":140,"\u5982\u679c\u8981\u4e0a\u4f20gpu\u7248\u672c\u7684\u5305":88,"\u5982\u679c\u8981\u542f\u7528gpu":127,"\u5982\u679c\u8981\u8fd0\u884c\u6240\u6709\u7684\u5355\u5143\u6d4b\u8bd5":98,"\u5982\u679c\u89e3\u51b3\u4e86\u67d0\u4e2aissue\u7684\u95ee\u9898":98,"\u5982\u679c\u8bad\u7ec3\u4e00\u4e2apass":109,"\u5982\u679c\u8bad\u7ec3\u8fc7\u7a0b\u7684\u7684cost\u660e\u663e\u9ad8\u4e8e\u8fd9\u4e2a\u5e38\u6570\u8f93\u51fa\u7684cost":109,"\u5982\u679c\u8bbe\u7f6e\u8be5\u53c2\u6570":129,"\u5982\u679c\u8bc4\u5ba1\u610f\u89c1\u6bd4\u8f83\u591a":98,"\u5982\u679c\u8c03\u7528\u9759\u6001\u5e93\u53ea\u80fd\u5c06\u9759\u6001\u5e93\u4e0e\u89e3\u91ca\u5668\u94fe\u63a5":71,"\u5982\u679c\u8f93\u5165\u662f\u5e8f\u5217\u6570\u636e":115,"\u5982\u679c\u8f93\u51fa\u662f\u4e00\u4e2a\u5e8f\u5217":115,"\u5982\u679c\u8f93\u51fa\u662fno":28,"\u5982\u679c\u8fd0\u884c":104,"\u5982\u679c\u8fd8\u4e0d\u884c":104,"\u5982\u679c\u9700\u8981":113,"\u5982\u679c\u9700\u8981\u5728c\u7ef4\u5ea6\u8ba1\u7b97softmax":108,"\u5982\u679c\u9700\u8981\u5b89\u88c5\u652f\u6301gpu\u7684\u7248\u672c":[30,112],"\u5982\u679c\u9700\u8981\u624b\u52a8\u7f16\u8bd1":88,"\u5982\u679c\u9700\u8981\u6269\u5927\u77e9\u9635":100,"\u5982\u679c\u9700\u8981\u7f29\u51cf\u77e9\u9635":100,"\u5982\u679c\u9700\u8981\u83b7\u53d6":30,"\u5982\u679ccuda":101,"\u5982\u679clearning_rate\u592a\u5927":109,"\u5982\u679clearning_rate\u592a\u5c0f":109,"\u5982\u679cmkl":68,"\u5982\u679cop\u6ca1\u6709\u5b9e\u73b0cuda":101,"\u5982\u679cop\u7684\u67d0\u4e2a\u8f93\u5165\u4e0d\u53c2\u4e0e\u53cd\u5411\u68af\u5ea6\u7684\u8ba1\u7b97":101,"\u5982\u679cpaddlepaddle\u5305\u5df2\u7ecf\u5728python\u7684sit":104,"\u5982\u679cpaddlepaddle\u5e93\u9700\u8981\u540c\u65f6\u652f\u6301\u771f\u673a\u548c\u6a21\u62df\u5668":143,"\u5982\u679cparamet":37,"\u5982\u6bcf\u4e2a\u6587\u4ef6\u53ea\u6709\u4e00\u4e2a":98,"\u5982\u795e\u7ecf\u5143\u6fc0\u6d3b\u503c\u7b49":107,"\u5982\u8981build\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf":98,"\u5982\u9ad8\u4eae\u90e8\u5206":134,"\u5982train":117,"\u5b50\u53e5":139,"\u5b50\u53e5\u7684\u5355\u8bcd\u6570\u548c\u6307\u5b9a\u7684\u4e00\u4e2a\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":139,"\u5b50\u76ee\u5f55":26,"\u5b50\u7c7b\u53ea\u9700\u8981\u4f7f\u7528\u5b9a\u4e49\u597d\u7684\u63a5\u53e3":68,"\u5b57\u6bb5\u4e2d":123,"\u5b57\u6bb5\u4e3a\u4f8b":107,"\u5b57\u6bb5\u7684\u53d6\u503c":115,"\u5b57\u6bb5\u8868\u793a\u5bb9\u5668\u7684\u73af\u5883\u53d8\u91cf":123,"\u5b57\u6bb5\u8868\u793a\u8fd9\u4e2ajob\u4f1a\u540c\u65f6\u5f00\u542f3\u4e2apaddlepaddle\u8282\u70b9":123,"\u5b57\u6bb5\u8bbe\u4e3a":88,"\u5b57\u7b26\u4e32":38,"\u5b58\u50a8":[38,115,116],"\u5b58\u50a8\u6d6e\u70b9\u7c7b\u578b\u8f93\u5165":116,"\u5b58\u50a8\u7684\u538b\u7f29\u6587\u4ef6":116,"\u5b58\u6570\u6570\u636e":115,"\u5b66\u4e60":26,"\u5b66\u4e60\u6210\u672c\u9ad8":71,"\u5b66\u4e60\u7387\u4e3a":109,"\u5b83\u4eec\u4e3b\u8981\u662f\u7528\u4e8e":68,"\u5b83\u4eec\u7684\u6587\u4ef6\u540d\u662f":38,"\u5b83\u5305\u542b\u4ee5\u4e0b\u51e0\u6b65":100,"\u5b83\u5305\u542b\u4ee5\u4e0b\u53c2\u6570":100,"\u5b83\u53ea\u4f1a\u5305\u62ec\u751f\u6210\u597d\u7684\u52a8\u6001\u5e93\u548c\u5934\u6587\u4ef6":68,"\u5b83\u53eb\u505a":140,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u51cf\u5c11\u5206\u53d1\u5ef6\u8fdf":119,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u6211\u4eec\u683c\u5f0f\u5316\u6e90\u4ee3\u7801":98,"\u5b83\u53ef\u4ee5\u6307\u6d4b\u91cf\u4e00\u4e2a\u7a0b\u5e8f\u7684\u7a7a\u95f4":134,"\u5b83\u53ef\u80fd\u6709\u4e0d\u6b62\u4e00\u4e2a\u6743\u91cd":100,"\u5b83\u5b9a\u4e49\u4e86":140,"\u5b83\u5b9a\u4e49\u89e3\u7801\u7f51\u7edc\u7684":140,"\u5b83\u5c06\u88ab\u5206\u53d1\u5230":119,"\u5b83\u5e76\u4e0d\u662f\u4e00\u4e2a\u5b8c\u6574\u7684recurr":108,"\u5b83\u5e94\u8be5\u6253\u5370\u51fa\u9884\u6d4b\u4f4f\u623f\u6570\u636e\u7684\u6e05\u5355":112,"\u5b83\u652f\u6301\u591a\u7ebf\u7a0b\u66f4\u65b0":100,"\u5b83\u662finteger_value\u7c7b\u578b\u7684":137,"\u5b83\u662finteger_value_sequence\u7c7b\u578b\u7684":137,"\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":119,"\u5b83\u7684":140,"\u5b83\u7684\u529f\u80fd\u662f":101,"\u5b83\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20":136,"\u5b83\u7684\u8f93\u5165\u4e0e\u7ecf\u8fc7\u5b66\u4e60\u7684\u53c2\u6570\u505a\u5185\u79ef\u5e76\u52a0\u4e0a\u504f\u7f6e":100,"\u5b83\u8868\u793a":26,"\u5b83\u8d1f\u8d23\u51b3\u5b9a\u7f16\u8bd1\u65f6\u662f\u5426\u4f7f\u7528mklml\u548cmkl":68,"\u5b83\u9996\u5148\u8c03\u7528\u57fa\u6784\u9020\u51fd\u6570":100,"\u5b89\u88c5":133,"\u5b89\u88c5\u4e0e\u7f16\u8bd1":141,"\u5b89\u88c5\u4e0e\u7f16\u8bd1c":114,"\u5b89\u88c5\u540e":28,"\u5b89\u88c5\u540e\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":72,"\u5b89\u88c5\u597ddocker\u4e4b\u540e\u53ca\u53ef\u7528\u4ee5\u4e0b\u547d\u4ee4\u542f\u52a8\u5de5\u5177":103,"\u5b89\u88c5\u597ddocker\u4e4b\u540e\u53ef\u4ee5\u4f7f\u7528\u6e90\u7801\u76ee\u5f55\u4e0b\u7684\u811a\u672c\u6784\u5efa\u6587\u6863":103,"\u5b89\u88c5\u5b8c\u6210\u4e4b\u540e":[127,143],"\u5b89\u88c5\u5b8c\u6bd5\u540e":133,"\u5b89\u88c5\u5f00\u53d1\u5de5\u5177\u5230":26,"\u5b89\u88c5\u6587\u6863":110,"\u5b89\u88c5\u65b9\u5f0f\u6765\u5feb\u901f\u5b89\u88c5paddlepaddl":127,"\u5b89\u88c5\u6d41\u7a0b":112,"\u5b8c\u6210":98,"\u5b8c\u6210\u4e00\u4e2a\u4f20\u8f93\u52a8\u4f5c\u5b8c\u6210\u7684\u65f6\u95f4\u4e5f\u6bd4\u8f83\u77ed":54,"\u5b8c\u6210\u4e0a\u8ff0\u51c6\u5907\u4e4b\u540e":116,"\u5b8c\u6210\u4efb\u610f\u7684\u8fd0\u7b97\u903b\u8f91":139,"\u5b8c\u6210\u540evolume\u4e2d\u7684\u6587\u4ef6\u5185\u5bb9\u5927\u81f4\u5982\u4e0b":123,"\u5b8c\u6210\u5728windows\u4e0a\u5b89\u88c5\u548c\u4f7f\u7528dock":28,"\u5b8c\u6210\u5b89\u88c5":30,"\u5b8c\u6210\u5e38\u7528layer\u7684mkl":68,"\u5b8c\u6210\u5e38\u89c1\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edcvgg":68,"\u5b8c\u6210\u6570\u636e\u7684\u9884\u5904\u7406":38,"\u5b8c\u6210\u76f8\u5e94\u7684\u8ba1\u7b97":136,"\u5b8c\u6210\u81ea\u52a8\u5316\u4e8c\u8fdb\u5236\u7f16\u8bd1":88,"\u5b8c\u6210paddlepaddle\u7684\u5b89\u88c5":110,"\u5b8c\u6574\u4ee3\u7801\u53ef\u4ee5\u53c2\u8003\u793a\u4f8b":107,"\u5b8c\u6574\u4ee3\u7801\u53ef\u4ee5\u67e5\u770b":116,"\u5b8c\u6574\u6e90\u7801\u53ef\u53c2\u8003":109,"\u5b8c\u6574\u7684\u53c2\u6570\u77e9\u9635\u88ab\u5206\u5e03\u5728\u4e0d\u540c\u7684\u53c2\u6570\u670d\u52a1\u5668\u4e0a":100,"\u5b8c\u6574\u7684\u914d\u7f6e\u6587\u4ef6\u5728":140,"\u5b98\u65b9\u6587\u6863":27,"\u5b9a\u4e49":68,"\u5b9a\u4e49\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185rnn\u5355\u5143\u5b8c\u6210\u7684\u8ba1\u7b97":139,"\u5b9a\u4e49\u4e00\u4e9b\u9664\u4e86layer\u548cmemory\u76f8\u5173\u7684\u7c7b\u548c\u51fd\u6570":68,"\u5b9a\u4e49\u4e86\u4e00\u4e2a\u53ea\u8bfb\u7684memori":139,"\u5b9a\u4e49\u4e86lstm\u5355\u5143\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u7684\u8ba1\u7b97\u8fc7\u7a0b":108,"\u5b9a\u4e49\u4f4d\u7f6e":101,"\u5b9a\u4e49\u5728\u5916\u5c42":139,"\u5b9a\u4e49\u5f02\u6b65\u8bad\u7ec3\u7684\u957f\u5ea6":129,"\u5b9a\u4e49\u6e90\u8bed\u53e5\u7684\u6570\u636e\u5c42":140,"\u5b9a\u4e49\u7684\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784":116,"\u5b9a\u4e49\u7c7b\u578b":101,"\u5b9a\u4e49\u89e3\u7801\u5668\u7684memori":140,"\u5b9a\u4e49\u8f93\u5165":101,"\u5b9a\u4e49\u8f93\u51fa":101,"\u5b9a\u4e49\u8f93\u51fa\u51fd\u6570":140,"\u5b9a\u4e49\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5355\u6b65\u51fd\u6570":140,"\u5b9d\u5854\u7684\u5e95\u7aef\u9700\u8981\u575a\u5b9e\u7684\u57fa\u5ea7\u6765\u652f\u6491":110,"\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":137,"\u5b9e\u73b0\u5177\u4f53\u7684\u51fd\u6570\u529f\u80fd\u5373\u53ef":68,"\u5b9e\u73b0\u524d\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":100,"\u5b9e\u73b0\u5355\u6b65\u51fd\u6570":140,"\u5b9e\u73b0\u540e\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":100,"\u5b9e\u73b0\u5728":101,"\u5b9e\u73b0\u5bf9":102,"\u5b9e\u73b0\u65b0\u7684op\u90fd\u6dfb\u52a0\u81f3\u76ee\u5f55":101,"\u5b9e\u73b0\u6784\u9020\u51fd\u6570":100,"\u5b9e\u73b0\u7684":108,"\u5b9e\u73b0\u7b80\u5355":71,"\u5b9e\u73b0\u7ec6\u8282":100,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad":100,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u540e\u5411\u4f20\u64ad":100,"\u5b9e\u73b0\u8bcd\u8bed\u548c\u53e5\u5b50\u4e24\u4e2a\u7ea7\u522b\u7684\u53cc\u5c42rnn\u7ed3\u6784":139,"\u5b9e\u73b0\u8be5\u5c42\u7684c":100,"\u5b9e\u9645\u4e0a\u4f7f\u7528\u4e86":108,"\u5b9e\u9645\u4e0a\u9700\u8981\u7684\u8f93\u51fa\u7ed3\u679c\u662f\u4e24\u4e2a\u77e9\u9635":107,"\u5bb9\u5668\u8fd0\u884c\u90fd\u8fd0\u884c":123,"\u5bb9\u5668\u9ed8\u8ba4\u6267\u884c":142,"\u5bbd\u5ea6":115,"\u5bbd\u5ea6\u4e3a":115,"\u5bbd\u5ea6\u7b49\u4e8e\u914d\u7f6e\u4e2dlayer\u7684s":107,"\u5bbf\u4e3b\u673a\u7684c":[142,143,144],"\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":134,"\u5bf9":116,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f01\u5411\u91cf":110,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f\u6d6e\u70b9\u5411\u91cf":110,"\u5bf9\u4e8e":140,"\u5bf9\u4e8e\u4e0d\u540c\u7684\u8bad\u7ec3\u4efb\u52a1":117,"\u5bf9\u4e8e\u4e0d\u540c\u7684\u96c6\u7fa4\u5e73\u53f0":120,"\u5bf9\u4e8e\u4e0d\u540c\u8bed\u8a00":71,"\u5bf9\u4e8e\u4e24\u79cd\u4e0d\u540c\u7684\u8f93\u5165\u6570\u636e\u7c7b\u578b":137,"\u5bf9\u4e8e\u4e60\u60ef\u4f7f\u7528windows\u548cmacos\u7684\u5f00\u53d1\u8005\u6765\u8bf4":26,"\u5bf9\u4e8e\u5355\u5c42rnn":137,"\u5bf9\u4e8e\u5355\u5c42rnn\u7684\u6570\u636e\u4e00\u5171\u6709\u4e24\u4e2a\u6837\u672c":137,"\u5bf9\u4e8e\u53cc\u5c42rnn":137,"\u5bf9\u4e8e\u540c\u4e00\u6bb5c":71,"\u5bf9\u4e8e\u540c\u6837\u7684\u6570\u636e":137,"\u5bf9\u4e8e\u540c\u6837\u8bbe\u7f6e\u7684\u7f51\u7edc\u6a21\u578b":67,"\u5bf9\u4e8e\u56fd\u5185\u7528\u6237":28,"\u5bf9\u4e8e\u591a\u8bed\u8a00\u63a5\u53e3":71,"\u5bf9\u4e8e\u5927\u591a\u6570\u8bed\u8a00":71,"\u5bf9\u4e8e\u5e8f\u5217\u957f\u5ea6":67,"\u5bf9\u4e8e\u6027\u80fd\u7684\u5173\u952e\u8def\u5f84\u90fd\u505a\u51fa\u4e86\u7ea2\u8272\u6807\u8bb0":133,"\u5bf9\u4e8e\u6211\u4eec\u652f\u6301\u7684\u5168\u90e8\u77e9\u9635\u64cd\u4f5c":100,"\u5bf9\u4e8e\u6709\u53c2\u6570\u7684\u5c42":68,"\u5bf9\u4e8e\u672c\u6837\u4f8b\u4ee3\u7801":117,"\u5bf9\u4e8e\u6bb5\u843d\u7684\u6587\u672c\u5206\u7c7b":137,"\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5355\u5c42rnn\u7684\u6570\u636e":137,"\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u65b0\u52a0\u7684rnn":67,"\u5bf9\u4e8e\u6bcf\u79cd\u7c7b\u578b":72,"\u5bf9\u4e8e\u6bcf\u79cdc":72,"\u5bf9\u4e8e\u8fd9\u6837\u7684\u9700\u6c42":116,"\u5bf9\u4e8e\u914d\u5907\u6709\u6ce8\u610f\u529b\u673a\u5236\u7684\u89e3\u7801\u5668":140,"\u5bf9\u4e8enchw":108,"\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":134,"\u5bf9\u4f7f\u7528\u7684\u4e2d\u95f4\u53d8\u91cf\u548c\u8d44\u6e90\u8fdb\u884c\u6e05\u7406\u548c\u91ca\u653e":116,"\u5bf9\u5168\u8fde\u63a5\u5c42\u6765\u8bf4":100,"\u5bf9\u52a0\u8f7d\u9884\u8bad\u7ec3\u53c2\u6570\u7684\u5c42":109,"\u5bf9\u53cc\u5c42\u5e8f\u5217\u6765\u8bb2":115,"\u5bf9\u5df2\u7ecfpush\u5230\u8fdc\u7a0b\u4ed3\u5e93\u7684\u591a\u4e2acommit":98,"\u5bf9\u5e94":143,"\u5bf9\u5e94\u4e00\u4e2a\u5b50\u53e5":139,"\u5bf9\u5e94\u4e00\u4e2a\u8bcd":139,"\u5bf9\u5e94\u4e8e\u8c03\u7528c":115,"\u5bf9\u5e94\u7684\u68af\u5ea6op\u8ba1\u7b97\u4e4b\u4e2d":101,"\u5bf9\u5e94\u7740\u4e0a\u6587\u63d0\u5230\u7684\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":115,"\u5bf9\u5e94\u7740\u4e0a\u6587\u63d0\u5230\u7684\u4e8c\u7ef4\u6d6e\u70b9\u578b\u77e9\u9635":115,"\u5bf9\u63a8\u8350\u914d\u7f6e\u4e2d\u7684\u9009\u9879\u5efa\u8bae\u6309\u7167\u8bbe\u7f6e":113,"\u5bf9\u65b0\u7684\u6743\u91cd\u8fdb\u884c\u8f6c\u6362\u7528\u4e8e\u4e0b\u6b21\u8fed\u4ee3":67,"\u5bf9\u6bcf\u4e2a\u8f93\u5165":100,"\u5bf9\u6bcf\u4e2a\u8f93\u5165\u4e58\u4e0a\u53d8\u6362\u77e9\u9635":100,"\u5bf9\u6bd4":71,"\u5bf9\u6bd4\u4f18\u5316\u540elayer\u4e0e\u76f8\u5bf9\u5e94\u7684paddlepaddle\u539f\u6709lay":67,"\u5bf9\u6bd4\u4f18\u5316\u540elayer\u81ea\u8eab":67,"\u5bf9\u6bd4\u53cd\u5411op\u4e0d\u540c\u8bbe\u5907":101,"\u5bf9\u6fc0\u6d3b\u6c42\u5bfc":100,"\u5bf9\u795e\u7ecf\u7f51\u7edc\u6765\u8bf4":115,"\u5bf9\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u8fdb\u884c\u5e8f\u5217\u5316":116,"\u5bf9\u8bc4\u5ba1\u610f\u89c1\u4e0d\u540c\u610f\u7684":98,"\u5bf9\u8bc4\u5ba1\u610f\u89c1\u540c\u610f\u4e14\u6309\u5176\u4fee\u6539\u5b8c\u7684":98,"\u5bf9\u8c61":109,"\u5bf9\u8c61\u5206\u914d\u7a7a\u95f4":116,"\u5bf9\u8f93\u5165\u53c2\u6570\u7684\u5b89\u5168\u6027\u8fdb\u884c\u4e86\u5fc5\u8981\u7684\u5224\u65ad":72,"\u5bf9\u8f93\u51fa\u7684\u5408\u5e76":139,"\u5bf9\u8fd9\u4e2a\u7248\u672c\u7684\u63d0\u4ea4":88,"\u5bf9sparse_binary_vector\u548csparse_float_vector":110,"\u5bfb\u627e\u6709\u6ca1\u6709\u5176\u4ed6\u53ef\u4ee5\u4f18\u5316\u7684\u53ef\u80fd":68,"\u5bfb\u627epython\u4e0ec":133,"\u5bfc\u51fa\u8fd9\u4e9b\u63a5\u53e3":72,"\u5bfc\u81f4\u4e86\u6d6e\u70b9\u6570\u6ea2\u51fa":107,"\u5bfc\u81f4\u53c2\u6570\u6536\u655b\u5230\u4e86\u4e00\u4e9b\u5947\u5f02\u7684\u60c5\u51b5":107,"\u5bfc\u81f4\u53c2\u6570\u7d2f\u52a0":107,"\u5bfc\u81f4\u7f16\u8bd1paddlepaddle\u5931\u8d25":104,"\u5bfc\u81f4\u8bad\u7ec3\u65f6\u95f4\u8fc7\u957f":109,"\u5bfc\u81f4mklml\u5e93\u4e0b\u8f7d\u4e0d\u6210\u529f":104,"\u5c01\u88c5\u4e86":134,"\u5c01\u88c5\u8be5\u5c42\u7684python\u63a5\u53e3":100,"\u5c06":[88,109,134],"\u5c06\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u62c6\u5206\u6210\u591a\u4efd":37,"\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":140,"\u5c06\u4f1a\u4f18\u5148\u4f7f\u7528":117,"\u5c06\u4f1a\u59cb\u7ec8\u4f7f\u7528":142,"\u5c06\u4f1a\u5c06\u7528\u6237\u4f20\u8fdb\u6765\u7684\u914d\u7f6e\u53c2\u6570\u4f20\u9012cmake\u7cfb\u7edf":142,"\u5c06\u4f1a\u81ea\u52a8\u8ba1\u7b97\u51fa\u4e00\u4e2a\u5408\u9002\u7684\u503c":129,"\u5c06\u5176\u8bbe\u7f6e\u6210":107,"\u5c06\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e\u5148\u53d8\u6362\u6210\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e":137,"\u5c06\u542b\u6709\u5b50\u53e5":139,"\u5c06\u542b\u6709\u8bcd\u8bed\u7684\u53e5\u5b50\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":139,"\u5c06\u56fe\u7247\u5206\u7c7b\u5230":116,"\u5c06\u591a\u53e5\u8bdd\u770b\u6210\u4e00\u4e2a\u6574\u4f53\u540c\u65f6\u4f7f\u7528encoder\u538b\u7f29":137,"\u5c06\u591a\u53f0\u673a\u5668\u7684\u6d4b\u8bd5\u7ed3\u679c\u5408\u5e76":129,"\u5c06\u5927\u91cf\u7684":71,"\u5c06\u5b57\u5178\u7684\u5730\u5740\u4f5c\u4e3aargs\u4f20\u7ed9dataprovid":109,"\u5c06\u5b83\u4eec\u653e\u5728\u540c\u4e00\u76ee\u5f55\u4e2d":116,"\u5c06\u5bf9\u5e94\u6570\u636e\u5c42\u7684\u7ef4\u6570\u8bbe\u7f6e\u6210\u4e00\u4e2a\u5927\u4e8e\u8f93\u5165\u6570\u636e\u7ef4\u6570\u7684\u503c\u7528\u4e8e\u5360\u4f4d\u5373\u53ef":108,"\u5c06\u5e8f\u5217\u5316\u7ed3\u679c\u5199\u5165\u4e00\u4e2a\u6587\u4ef6\u5185":116,"\u5c06\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u524d\u5411\u548c\u53cd\u5411\u90e8\u5206\u6df7\u5408\u5728\u4e00\u8d77":140,"\u5c06\u6027\u80fd\u5206\u6790\u7ed3\u679c\u4ee5\u7f51\u9875\u7684\u5f62\u5f0f\u5c55\u793a\u51fa\u6765":133,"\u5c06\u6027\u80fd\u5206\u6790\u7ed3\u679c\u6309\u7167tottime\u6392\u5e8f":133,"\u5c06\u6570\u636e\u5207\u5206\u6210\u591a\u4efd":117,"\u5c06\u65b0\u5206\u652f\u7684\u7248\u672c\u6253\u4e0atag":88,"\u5c06\u65b0\u5efa\u7684\u6743\u91cd\u52a0\u5165\u6743\u91cd\u8868":100,"\u5c06\u660e\u6587\u53c2\u6570\u8f6c\u5316\u4e3apaddlepaddle\u53ef\u52a0\u8f7d\u7684\u6a21\u578b\u53c2\u6570\u65f6":109,"\u5c06\u672c\u5730\u7684\u4fee\u6539\u63a8\u9001\u5230":98,"\u5c06\u6b64\u76ee\u5f55\u6302\u8f7d\u4e3a\u5bb9\u5668\u7684":123,"\u5c06\u73af\u5883\u53d8\u91cf\u8f6c\u6362\u6210paddle\u7684\u547d\u4ee4\u884c\u53c2\u6570":123,"\u5c06\u7528\u4e8epython":101,"\u5c06\u7ed3\u679c\u4fdd\u5b58\u5230\u6b64\u76ee\u5f55\u91cc":123,"\u5c06\u7f51\u7edc\u7ed3\u6784\u5b9a\u4e49\u548c\u8bad\u7ec3\u7ed3\u675f\u5b58\u50a8\u4e0b\u6765\u7684\u6a21\u578b\u53c2\u6570\u6587\u4ef6":116,"\u5c06\u8bad\u7ec3\u6587\u4ef6\u4e0e\u5207\u5206\u597d\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5171\u4eab\u5b58\u50a8":123,"\u5c06\u8df3\u8fc7\u5206\u53d1\u9636\u6bb5\u76f4\u63a5\u542f\u52a8\u6240\u6709\u8282\u70b9\u7684\u96c6\u7fa4\u4f5c\u4e1a":119,"\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":137,"\u5c06\u8fdc\u7a0b\u4ed3\u5e93":98,"\u5c06\u900f\u660e":119,"\u5c06\u9700\u8981\u8f93\u51fa\u7684\u5c42\u4f5c\u4e3a":107,"\u5c06cuda\u5e93\u548clinux\u8bbe\u5907\u6302\u8f7d\u5230docker\u5bb9\u5668\u5185":28,"\u5c06ip\u6392\u5e8f\u751f\u6210\u7684\u5e8f\u53f7\u4f5c\u4e3atrain":123,"\u5c06master\u5206\u652f\u7684\u5408\u5165commit\u6253\u4e0atag":88,"\u5c06node\u8282\u70b9\u7684ip\u5730\u5740\u4fdd\u5b58\u5230machines\u6587\u4ef6\u4e2d":124,"\u5c06paddlepaddle\u4fdd\u5b58\u7684\u6a21\u578b\u53c2\u6570\u8fd8\u539f\u56de\u660e\u6587\u65f6":109,"\u5c06recurr":108,"\u5c0f\u4e8e\u67d0\u4e2a\u6bd4\u8f83\u5c0f\u7684\u9608\u503c\u8ba4\u4e3a\u901a\u8fc7":68,"\u5c31\u4f1a\u5728\u5b8c\u6210\u7f16\u8bd1\u4e4b\u540e":27,"\u5c31\u53ef\u4ee5\u4e86\u89e3\u5230\u95ee\u9898\u4ee3\u7801\u5728\u54ea\u91cc":133,"\u5c31\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u547d\u4ee4\u5f00\u59cb\u6267\u884c\u8bad\u7ec3":28,"\u5c31\u53ef\u4ee5\u6309":26,"\u5c31\u5c06\u8fd9\u4e9b\u5c42\u52a0\u5165\u4e00\u4e2apython":116,"\u5c31\u5f88\u5bb9\u6613\u5bfc\u81f4\u5185\u5b58\u8d85\u9650":107,"\u5c31\u662f\u7528\u4e8e\u5c55\u793a\u4e0a\u8ff0\u5206\u6790\u5de5\u5177\u7684\u7528\u6cd5":134,"\u5c31\u662fpaddlepaddle\u4e2d\u6240\u6307\u7684":115,"\u5c31\u8fd9\u4e48\u7b80\u5355":28,"\u5c31\u901a\u5e38\u7684gpu\u6027\u80fd\u5206\u6790\u6765\u8bf4":134,"\u5c31\u9700\u8981\u5bf9\u8fd9\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00\u589e\u52a0\u4e00\u4e9b\u5b9a\u4e49":71,"\u5c31\u9700\u8981\u9009\u62e9\u4f7f\u7528no":28,"\u5c3d\u65e9\u62a5\u9519":101,"\u5c42\u524d\u5411\u8ba1\u7b97\u7684\u7ed3\u679c":116,"\u5c42\u548c\u8f93\u5165\u7684\u914d\u7f6e":100,"\u5c42\u6b21\u5316\u7684rnn":139,"\u5c42\u7684\u540d\u79f0\u4e0e":140,"\u5c42\u7684\u5927\u5c0f":100,"\u5c42\u7684\u7c7b\u578b":100,"\u5c42\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684":140,"\u5c55\u793a\u4e86\u4e00\u4e2a\u542b\u67094\u4e2a\u5e8f\u5217\u7684":115,"\u5c55\u793a\u7684\u8c03\u7528\u56fe\u4e5f\u53ef\u4ee5\u5e2e\u52a9\u6211\u4eec\u53d1\u73b0\u6027\u80fd\u4e2d\u7684\u95ee\u9898":133,"\u5c5e\u4e8e\u8fd9\u4e00\u7c7b\u7684\u5b9e\u73b0":108,"\u5c5e\u6027":101,"\u5de5\u4f5c\u6a21\u5f0f":129,"\u5de5\u4f5c\u7a7a\u95f4\u4e2d\u7684":119,"\u5de5\u4f5c\u7a7a\u95f4\u5e94\u5982\u4e0b\u6240\u793a":117,"\u5de5\u5177\u4e0a\u4f20\u5373\u53ef":88,"\u5de5\u5177\u5408\u5e76fat\u5e93":143,"\u5de5\u5177\u670d\u52a1\u5668\u5c06\u8bfb\u53d6\u73af\u5883\u53d8\u91cf":103,"\u5de5\u5177\u6765\u7ba1\u7406":98,"\u5de5\u5177\u6765\u7f16\u8bd1\u6587\u6863":103,"\u5de5\u5177\u94fe":142,"\u5de5\u5177\u94fe\u7684android":142,"\u5de6\u53f3\u7684\u8ba1\u7b97\u65f6\u95f4":133,"\u5df2\u6253\u5f00":98,"\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":129,"\u5e02\u9762\u4e0a\u5df2\u7ecf\u6709nvidia\u6216\u7b2c\u4e09\u65b9\u63d0\u4f9b\u7684\u4f17\u591a\u5de5\u5177":134,"\u5e0c\u671b\u80fd\u591f\u5c06\u5e8f\u5217\u5316\u540e\u7684\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u53c2\u6570\u6253\u5305\u8fdb\u4e00\u4e2a\u6587\u4ef6":116,"\u5e26\u6709\u4e0b\u9762\u4e24\u4e2a\u6a21\u677f\u53c2\u6570":101,"\u5e2e\u52a9\u6211\u4eec\u5b8c\u6210\u5bf9\u8f93\u5165\u5e8f\u5217\u7684\u62c6\u5206":139,"\u5e2e\u52a9\u6211\u4eec\u66f4\u597d\u5730\u63cf\u8ff0\u6bb5\u843d":139,"\u5e2e\u52a9\u6211\u4eec\u6784\u9020\u4e00\u4e9b\u590d\u6742\u7684\u8f93\u5165\u4fe1\u606f":136,"\u5e38\u5e38\u51fa\u73b0":104,"\u5e38\u7528\u4e8e\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1":115,"\u5e38\u7528\u7684cmake\u914d\u7f6e\u5982\u4e0b":[142,143],"\u5e38\u89c1\u7684\u5305\u62ec":133,"\u5e73\u5747\u6545\u969c\u4fee\u590d\u65f6\u95f4":37,"\u5e73\u5747\u6545\u969c\u7387":37,"\u5e76\u4e0d\u4fdd\u8bc1":100,"\u5e76\u4e0d\u662f\u4f7f\u7528\u53cc\u5c42rnn\u89e3\u51b3\u5b9e\u9645\u7684\u95ee\u9898":137,"\u5e76\u4e0d\u662fkubernetes\u4e2d\u7684node\u6982\u5ff5":123,"\u5e76\u4e0d\u771f\u6b63":[115,116],"\u5e76\u4e0d\u771f\u6b63\u7684\u548c":137,"\u5e76\u4e0d\u96be":26,"\u5e76\u4e14":140,"\u5e76\u4e14\u4e5f\u53ef\u4ee5\u5728windows\u7684docker\u4e2d\u8fd0\u884c":28,"\u5e76\u4e14\u4e66\u5199\u4e00\u4efd\u4ee3\u7801":102,"\u5e76\u4e14\u4f1a\u5199\u597d":68,"\u5e76\u4e14\u4f1a\u6839\u636e":142,"\u5e76\u4e14\u4f7f\u7528":72,"\u5e76\u4e14\u5185\u5c42\u7684\u5e8f\u5217\u64cd\u4f5c\u4e4b\u95f4\u72ec\u7acb\u65e0\u4f9d\u8d56":137,"\u5e76\u4e14\u52a0\u4e0a\u4e0b\u9762\u7684\u547d\u4ee4\u884c\u53c2\u6570":131,"\u5e76\u4e14\u5305\u62ecunit":98,"\u5e76\u4e14\u53ea\u9700\u8981\u5728\u5fc5\u8981\u7684\u65f6\u5019\u8f6c\u6362\u8fd9\u79cd\u683c\u5f0f":68,"\u5e76\u4e14\u53ef\u80fd\u4f1a\u52a0\u901f\u8bad\u7ec3\u8fc7\u7a0b":107,"\u5e76\u4e14\u542f\u52a8\u8bad\u7ec3":123,"\u5e76\u4e14\u5728\u5e38\u89c1\u7684\u5e73\u53f0\u4e0a":71,"\u5e76\u4e14\u5c55\u793a\u4e86\u5982\u4f55\u5229\u7528paddlepaddle\u6765\u89e3\u51b3\u4e00\u4e2a\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u95ee\u9898":110,"\u5e76\u4e14\u5f3a\u5236\u8bbe\u7f6e\u4e00\u4e9bpaddlepaddle\u53c2\u6570\u7684\u503c":143,"\u5e76\u4e14\u5f53\u7f16\u8bd1\u65f6":67,"\u5e76\u4e14\u628a\u5404\u79cd\u5f00\u53d1\u5de5\u5177\u5b89\u88c5\u8fdb\u53bb":26,"\u5e76\u4e14\u628a\u7cfb\u7edf\u751f\u6210\u7684ca":54,"\u5e76\u4e14\u628a\u7ed3\u679c\u8fd4\u56depfsclient\u7aef":54,"\u5e76\u4e14\u67e5\u8be2paddlepaddle\u5355\u5143\u6d4b\u8bd5\u7684\u65e5\u5fd7":104,"\u5e76\u4e14\u7f16\u8bd1\u65f6\u9700\u8981\u6253\u5f00":101,"\u5e76\u4e14\u7f16\u8bd1\u80fd\u901a\u8fc7\u4ee3\u7801\u6837\u5f0f\u68c0\u67e5":98,"\u5e76\u4e14\u8ba9\u63a5\u53e3\u8131\u79bb\u5b9e\u73b0\u7ec6\u8282":71,"\u5e76\u4e14\u8bbe\u7f6e\u9ed8\u8ba4\u503c\u4e3a1":101,"\u5e76\u4e14\u8f93\u5165\u8f93\u51fa\u90fd\u662f\u5171\u7528\u4e00\u5757\u5185\u5b58":68,"\u5e76\u4e14\u8f93\u51fa\u4e00\u4e2a":98,"\u5e76\u4e14\u8fd0\u884c":26,"\u5e76\u4e14\u9700\u8981\u91cd\u5199\u57fa\u7c7b\u4e2d\u7684\u4ee5\u4e0b\u51e0\u4e2a\u865a\u51fd\u6570":100,"\u5e76\u4e14cpu":101,"\u5e76\u4e14softmax\u5c42\u7684\u4e24\u4e2a\u8f93\u5165\u4e5f\u4f7f\u7528\u4e86\u540c\u6837\u7684\u53c2\u6570":109,"\u5e76\u4f7f\u7528":119,"\u5e76\u4fdd\u5b58\u8f93\u51fa\u5230\u4e00\u4e2a\u65e5\u5fd7\u6587\u4ef6":117,"\u5e76\u521b\u5efa\u4e86\u4e00\u4e2a\u65b0\u6587\u4ef6":98,"\u5e76\u521b\u5efaoptim":110,"\u5e76\u521d\u59cb\u5316":101,"\u5e76\u5220\u9664":88,"\u5e76\u5220\u9664\u66f4\u65e9\u7684\u5feb\u7167":37,"\u5e76\u52a0\u8f7d\u5176\u4e2d\u7684\u53c2\u6570":37,"\u5e76\u53d1\u5e03\u5230pypi":88,"\u5e76\u53ef\u4ee5\u5728\u5927\u591a\u6570\u4e3b\u6d41\u7684linux\u64cd\u4f5c\u7cfb\u7edf\u4ee5\u53camacos\u4e0a\u6267\u884c":30,"\u5e76\u548c\u53c2\u6570\u670d\u52a1\u5668\u901a\u4fe1":118,"\u5e76\u5728\u4e58\u79ef\u7ed3\u679c\u4e0a\u518d\u52a0\u4e0a\u7ef4\u5ea6\u4e3a":100,"\u5e76\u5728\u6700\u5f00\u59cb\u521d\u59cb\u5316\u4e3a\u8d77\u59cb\u8bcd":140,"\u5e76\u5728\u6bcf\u6b21\u6743\u91cd\u66f4\u65b0\u540e":67,"\u5e76\u5728\u7c7b\u6784\u5efa\u51fd\u6570\u4e2d\u628a\u5b83\u653e\u5165\u4e00\u4e2a\u7c7b\u6210\u5458\u53d8\u91cf\u91cc":100,"\u5e76\u5728\u8be5layer\u91cc\u91c7\u7528\u7b2c\u4e00\u79cd\u65b9\u5f0f\u8bbe\u7f6e":108,"\u5e76\u5728\u96c6\u7fa4\u4e2d\u8fd0\u884c\u591a\u4e2a\u5206\u5e03\u5f0f\u6570\u636e\u5904\u7406\u4efb\u52a1":38,"\u5e76\u5728python\u811a\u672c\u4e2d\u5b8c\u6210\u4e0eoperator\u540c\u6837\u7684\u8ba1\u7b97\u903b\u8f91":101,"\u5e76\u5904\u7406\u4e0e\u4e4b\u76f8\u5173\u7684\u6240\u6709\u7ec6\u8282":116,"\u5e76\u5b89\u88c5\u4e86python":104,"\u5e76\u5b89\u88c5\u6700\u65b0":30,"\u5e76\u5b89\u88c5\u6709python2":112,"\u5e76\u5b8c\u6210\u53c2\u6570\u4f18\u5316\u66f4\u65b0":118,"\u5e76\u5bf9\u6bd4\u662f\u5426\u548c\u6b63\u5728\u5b89\u88c5\u7684\u540e\u7f00\u4e00\u81f4":104,"\u5e76\u5bf9\u76f8\u5e94\u7684\u53c2\u6570\u8c03\u7528":100,"\u5e76\u5c06":88,"\u5e76\u5c06\u5176\u6295\u5c04\u5230":140,"\u5e76\u5c06\u8be5layer\u4e0a\u4e00\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u4f5c\u4e3a\u81ea\u8eab\u5f53\u524d\u65f6\u95f4\u6b65\u7684\u8f93\u51fa":108,"\u5e76\u5c06c":72,"\u5e76\u624b\u52a8\u751f\u6210download\u6210\u529f\u6807\u7b7e":104,"\u5e76\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4":28,"\u5e76\u628a\u5feb\u7167\u4fdd\u5b58\u5230\u8fd9\u4e2a\u76ee\u5f55\u4e0b":37,"\u5e76\u628a\u7ed3\u679c\u653e\u5230\u5f53\u524d\u5c42\u7684":68,"\u5e76\u628a\u8fd9\u4e2a\u5305\u542b\u4e86\u8bad\u7ec3\u6570\u636e\u7684container\u4fdd\u5b58\u4e3a\u4e00\u4e2a\u65b0\u7684\u955c\u50cf":122,"\u5e76\u66f4\u6362job":105,"\u5e76\u6839\u636e\u5206\u5e03\u5f0f\u8bad\u7ec3\u5e76\u53d1\u6570":117,"\u5e76\u68c0\u67e5\u548c\u9700\u5b89\u88c5\u7684\u5305\u662f\u5426\u5339\u914d":30,"\u5e76\u6ca1\u6709paddle\u7279\u522b\u9700\u8981\u7684\u7279\u6027":71,"\u5e76\u6dfb\u52a0\u5934\u6587\u4ef6":67,"\u5e76\u6dfb\u52a0\u6ce8\u91ca":101,"\u5e76\u7279\u5316\u6a21\u677f\u53c2\u6570\u4e3a":101,"\u5e76\u7c98\u8d34\u6b64python\u4ee3\u7801":112,"\u5e76\u81ea\u52a8\u4e0b\u8f7d\u5b89\u88c5\u4f9d\u8d56\u8f6f\u4ef6":30,"\u5e76\u81ea\u52a8\u7f16\u8bd1\u5bbf\u4e3b\u673a\u7248protoc\u53ef\u6267\u884c\u6587\u4ef6":144,"\u5e76\u81ea\u52a8\u7f16\u8bd1paddlepaddle\u6240\u9700\u7684\u6240\u6709\u7b2c\u4e09\u65b9\u5e93":142,"\u5e76\u884c\u5730\u6267\u884c\u6a21\u578b\u7684\u8bad\u7ec3":118,"\u5e76\u884c\u5730\u63a5\u6536\u68af\u5ea6\u548c\u66f4\u65b0\u53c2\u6570":118,"\u5e76\u88ab\u5b58\u50a8\u5728\u8bf8\u5982hadoop":38,"\u5e76\u89c2\u5bdf\u7ed3\u679c":134,"\u5e76\u89e3\u91ca\u4e86\u5404\u81ea\u542b\u4e49":101,"\u5e76\u8bb0\u5f55\u5b83\u7684\u7f16\u53f7":98,"\u5e76\u8fdb\u884c\u521d\u59cb\u5316\u64cd\u4f5c":110,"\u5e76\u9002\u5e94github\u7684\u7279\u6027\u505a\u4e86\u4e00\u4e9b\u533a\u522b":88,"\u5e76\u91cd\u65b0\u6253\u5305wheel\u5305":88,"\u5e76\u94fe\u63a5\u5230\u751f\u6210\u7684lib\u5e93\u4e2d":101,"\u5e78\u800cpython\u7684\u4e00\u4e2a\u7b2c\u4e09\u65b9\u5e93":133,"\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":136,"\u5e8f\u5217\u4e2d\u7684\u4e00\u4e2a\u5143\u7d20":115,"\u5e8f\u5217\u4e2d\u7684\u5143\u7d20\u662f\u8bcd\u8bed":115,"\u5e8f\u5217\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":115,"\u5e8f\u5217\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u662f\u975e\u5e8f\u5217":115,"\u5e8f\u5217\u4fe1\u606f":115,"\u5e8f\u5217\u5316\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u914d\u7f6e":116,"\u5e8f\u5217\u5316\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u65f6":116,"\u5e8f\u5217\u5316\u7ed3\u679c\u4f1a\u5199\u5165\u5f53\u524d\u8fd0\u884c\u76ee\u5f55\u4e0b\u7684":116,"\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":139,"\u5e8f\u5217\u662f\u4e00\u79cd\u5e38\u89c1\u7684\u6570\u636e\u7c7b\u578b":136,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u5927\u591a\u9075\u5faaencod":139,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u7684\u8f93\u5165":139,"\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u539f\u6765\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u5143\u7d20\u7684\u5e73\u5747\u503c":136,"\u5e8f\u5217\u8f93\u5165":115,"\u5e8f\u5217\u8f93\u5165\u65f6\u7b49\u4e8e":107,"\u5e8f\u5217\u8f93\u5165\u793a\u610f\u56fe":115,"\u5e93\u6709\u81ea\u5df1\u72ec\u7acb\u7684\u52a8\u6001\u5e93\u6587\u4ef6":113,"\u5e94\u7528\u524d\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":140,"\u5e94\u7528\u53cd\u5411\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":140,"\u5e94\u80fd\u53cd\u6620\u5f53\u524dcommit\u7684\u5185\u5bb9":98,"\u5e94\u8be5\u4e0e\u5b83\u7684memory\u540d\u5b57\u76f8\u540c":140,"\u5e94\u8be5\u8bf4\u8c22\u8c22":98,"\u5e94\u8be5\u8bfb\u53d6\u5f53\u524d\u76ee\u5f55\u4e0b\u7684":26,"\u5e94\u8be5\u964d\u4f4e\u5b66\u4e60\u7387":109,"\u5e95\u5c42\u8fdb\u7a0b":119,"\u5efa\u7acb\u4e00\u4e2a":98,"\u5efa\u8bae":[29,88,98],"\u5efa\u8bae\u5c06\u8be5\u53c2\u6570\u8bbe\u4e3atrue":129,"\u5f00\u53d1\u4e86\u6a21\u578b\u9884\u6d4b\u7684\u6837\u4f8b\u4ee3\u7801":72,"\u5f00\u53d1\u4eba\u5458\u4f7f\u7528":98,"\u5f00\u53d1\u5206\u652f":30,"\u5f00\u53d1\u6807\u51c6":141,"\u5f00\u53d1\u8005\u4f7f\u7528":26,"\u5f00\u53d1\u8005\u4fee\u6539\u81ea\u5df1\u7684\u4ee3\u7801":88,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4e2d":88,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4f7f\u7528":88,"\u5f00\u53d1\u955c\u50cf":98,"\u5f00\u53d1\u9884\u6d4b\u5e8f":116,"\u5f00\u53d1\u9884\u6d4b\u7a0b\u5e8f\u94fe\u63a5":113,"\u5f00\u542f":27,"\u5f00\u5934":[67,68],"\u5f00\u5934\u7684\u90e8\u5206":117,"\u5f00\u5934\u90e8\u5206\u6307\u5b9a":117,"\u5f00\u59cb\u63d0\u4f9b\u670d\u52a1":37,"\u5f00\u59cb\u6807\u8bb0":140,"\u5f00\u59cb\u795e\u7ecf\u7f51\u7edc\u7684":118,"\u5f00\u59cb\u9636\u6bb5":134,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":128,"\u5f15\u5165\u4e86\u4ee5\u4e0b\u56db\u4e2aapi":67,"\u5f15\u5165\u4e86\u7c7b\u578b\u7684\u5934\u6587\u4ef6":72,"\u5f15\u53d1":9,"\u5f15\u5bfc\u5c42":140,"\u5f15\u7528memory\u5f97\u5230\u8fd9layer\u4e0a\u4e00\u65f6\u523b\u8f93\u51fa":139,"\u5f39\u51fa\u4e0b\u9762\u7684\u9009\u62e9\u6846":88,"\u5f52\u4e00\u5316\u6982\u7387\u5411\u91cf":140,"\u5f53":131,"\u5f53\u4e00\u4e2a":115,"\u5f53\u4e0a\u8ff0\u63a5\u53e3\u7b2c4\u4e2a\u53c2\u6570":115,"\u5f53\u4f60\u6267\u884c\u547d\u4ee4":100,"\u5f53\u4fdd\u5b58\u7684\u7f51\u7edc\u53c2\u6570\u4e3afloat\u7c7b\u578b\u65f6\u4e3a4":109,"\u5f53\u524d\u65f6\u95f4\u6b65\u5904\u7684memory\u7684\u8f93\u51fa\u4f5c\u4e3a\u4e0b\u4e00\u65f6\u95f4\u6b65memory\u7684\u8f93\u5165":140,"\u5f53\u524d\u7684\u5b66\u4e60\u7387\u4e3a\u6240\u8bbe\u7f6e":109,"\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":100,"\u5f53\u524d\u7684\u8f93\u5165y\u548c\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51farnn_state\u505a\u4e86\u4e00\u4e2a\u5168\u94fe\u63a5":137,"\u5f53\u524d\u8bad\u7ec3\u4efb\u52a1\u542f\u52a8\u7684pserver\u7684ip\u5217\u8868":117,"\u5f53\u524d\u8bad\u7ec3\u4efb\u52a1pserver\u603b\u6570":117,"\u5f53\u524d\u8bad\u7ec3\u4efb\u52a1trainer\u603b\u6570":117,"\u5f53\u524dtrainer\u7684\u7ebf\u7a0b\u6570\u76ee":117,"\u5f53\u529f\u80fd\u5206\u652f\u5f00\u53d1\u5b8c\u6bd5\u540e":88,"\u5f53\u53ea\u505a\u63a8\u65ad":67,"\u5f53\u5728\u7f51\u7edc\u5c42\u914d\u7f6e\u4e2d\u8bbe\u7f6e":129,"\u5f53\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":129,"\u5f53\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5927\u4e8e1000\u5c0f\u4e8e\u7b49\u4e8e2000\u65f6":109,"\u5f53\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5927\u4e8e2000\u65f6":109,"\u5f53\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5c0f\u4e8e\u7b49\u4e8e1000\u65f6":109,"\u5f53\u5df2\u8bad\u7ec3pass\u6570\u5927\u4e8e1\u5c0f\u4e8e\u7b49\u4e8e2\u65f6":109,"\u5f53\u5df2\u8bad\u7ec3pass\u6570\u5927\u4e8e2\u65f6":109,"\u5f53\u5df2\u8bad\u7ec3pass\u6570\u5c0f\u4e8e\u7b49\u4e8e1\u65f6":109,"\u5f53\u5f00\u542f":68,"\u5f53\u6211\u4eec\u505a\u51fa\u6027\u80fd\u4fee\u6b63\u540e":133,"\u5f53\u6240\u6709pod\u90fd\u5904\u4e8erunning\u72b6\u6001":123,"\u5f53\u6253\u5f00":68,"\u5f53\u6570\u636e\u683c\u5f0f\u4e0epaddlepaddle\u9ed8\u8ba4\u7684":68,"\u5f53\u6a21\u578b\u53c2\u6570\u4e0d\u5b58\u5728\u65f6":129,"\u5f53\u6a21\u5f0f\u4e3a":129,"\u5f53\u7136":[28,134],"\u5f53\u7136\u53ef\u4ee5":26,"\u5f53\u7136\u8fd9\u4e24\u8005\u4e5f\u53ef\u4ee5\u76f8\u7b49":68,"\u5f53\u7528\u6237\u4f7f\u7528\u5b8c\u8fd9\u4e2a\u53c2\u6570\u540e":72,"\u5f53\u7528\u6237\u6ca1\u6709\u663e\u5f0f\u8bbe\u5b9a\u65f6":108,"\u5f53\u7f51\u7edc\u51fa\u73b0\u5206\u652f\u4e14\u5728":68,"\u5f53\u7f51\u7edc\u5c42\u7528\u4e00\u4e2a\u6279\u6b21\u505a\u8bad\u7ec3\u65f6":100,"\u5f53\u89e3\u8bfb\u6bcf\u4e00\u4e2a":140,"\u5f53\u8d85\u8fc7\u8be5\u9608\u503c\u65f6":129,"\u5f53\u8f93\u5165\u662f\u7ef4\u5ea6\u5f88\u9ad8\u7684\u7a00\u758f\u6570\u636e\u65f6":131,"\u5f53\u9700\u8981\u5b8c\u6210\u8ba1\u7b97\u65f6":102,"\u5f53\u975e\u5e8f\u5217\u8f93\u5165\u65f6":115,"\u5f53destination\u6587\u4ef6\u4e0d\u5b58\u5728\u6216\u8005\u5927\u5c0f\u548csource\u6587\u4ef6\u4e0d\u4e00\u81f4\u65f6":54,"\u5f53n1":107,"\u5f62\u6210recurr":139,"\u5f62\u6210recurrent\u8fde\u63a5":139,"\u5f88\u591a":26,"\u5f88\u6709\u53ef\u80fd\u5b9e\u9645\u5e94\u7528\u5c31\u662f\u6ca1\u6709\u6309\u7167\u60a8\u7684\u9884\u671f\u60c5\u51b5\u8fd0\u884c":134,"\u5f88\u6709\u53ef\u80fd\u662f\u975e\u72ec\u5360\u65b9\u5f0f\u6267\u884c\u5bfc\u81f4\u7684\u7aef\u53e3\u51b2\u7a81":105,"\u5f88\u96be\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":71,"\u5f97\u4f7f\u7528":71,"\u5f97\u5230\u8f93\u51fa\u503c":101,"\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e2a\u65f6\u95f4\u6b65\u603b\u662f\u80fd\u591f\u5f15\u7528\u6240\u6709\u8f93\u5165":139,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4e2d":140,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4f5c\u4e3a\u4f7f\u7528":140,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u548c":140,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u9aa4\u987a\u5e8f\u5730\u5904\u7406\u5e8f\u5217":140,"\u5faa\u73af\u7f51\u7edc\u4ece":140,"\u5fc5\u8981":72,"\u5fc5\u9009":117,"\u5fc5\u987b":100,"\u5fc5\u987b\u5206\u522b\u4e0e":68,"\u5fc5\u987b\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":137,"\u5fc5\u987b\u6307\u5411\u4e00\u4e2apaddlepaddle\u5b9a\u4e49\u7684lay":139,"\u5fc5\u987b\u6307\u5b9a\u4e3a":116,"\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":139,"\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":136,"\u5fc5\u987b\u7531\u53ea\u8bfbmemory\u7684":140,"\u5fc5\u987b\u8bbe\u7f6e\u4e3a":[142,143],"\u5fc5\u987b\u8bbe\u7f6e\u4e3aon":143,"\u5fc5\u987b\u914d\u7f6e\u4e3a":[113,144],"\u5fc5\u987b\u914d\u7f6e\u9009\u9879":113,"\u5feb\u901f\u5f00\u59cb":111,"\u6027\u80fd\u4f18\u5316\u7684\u8fc7\u7a0b\u901a\u5e38\u662f\u4e0d\u65ad\u91cd\u590d\u5730":133,"\u6027\u80fd\u5206\u6790":134,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u662f\u7528\u4e8e\u7ed9\u5e94\u7528\u7a0b\u5e8f\u7684\u6027\u80fd\u505a\u5b9a\u91cf\u5206\u6790\u7684":134,"\u6027\u80fd\u5206\u6790\u662f\u6027\u80fd\u4f18\u5316\u7684\u5173\u952e\u4e00\u6b65":134,"\u6027\u80fd\u548c\u628a\u7f16\u8bd1\u5de5\u5177\u5b89\u88c5\u5728\u672c\u673a\u8fd0\u884c\u4e00\u6837":26,"\u6027\u80fd\u8c03\u4f18":128,"\u6027\u80fdtip":[142,143],"\u60a8\u4e5f\u53ef\u4ee5\u8fdb\u5165\u5230docker\u5bb9\u5668\u4e2d":28,"\u60a8\u4f1a\u5728\u63a5\u4e0b\u6765\u7684\u90e8\u5206\u4e2d\u83b7\u5f97\u66f4\u591a\u7684\u7ec6\u8282\u4ecb\u7ecd":134,"\u60a8\u53ef\u4ee5\u4ece\u4e0b\u9762\u7684\u8868\u683c\u4e2d\u627e\u5230\u9700\u8981\u7684\u7248\u672c":30,"\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":134,"\u60a8\u53ef\u4ee5\u5728":28,"\u60a8\u53ef\u4ee5\u5728\u5bb9\u5668\u4e2d\u6267\u884c":28,"\u60a8\u53ef\u4ee5\u5bfc\u5165":134,"\u60a8\u53ef\u4ee5\u6309\u7167\u4e0b\u9762\u7684\u6b65\u9aa4\u5728openmpi\u96c6\u7fa4\u4e2d\u63d0\u4ea4paddle\u8bad\u7ec3\u4efb\u52a1":124,"\u60a8\u53ef\u4ee5\u91c7\u7528\u4e0b\u9762\u4e94\u4e2a\u6b65\u9aa4":134,"\u60a8\u53ef\u80fd\u9700\u8981\u4fee\u6539":117,"\u60a8\u5c06\u4e86\u89e3\u5982\u4f55":140,"\u60a8\u5c31\u80fd\u83b7\u5f97\u5982\u4e0b\u7684\u5206\u6790\u7ed3\u679c":134,"\u60a8\u6309\u5982\u4e0b\u6b65\u9aa4\u64cd\u4f5c\u5373\u53ef":134,"\u60a8\u6700\u597d\u5148\u786e\u8ba4":134,"\u60a8\u9996\u5148\u9700\u8981\u5728\u76f8\u5173\u4ee3\u7801\u6bb5\u4e2d\u52a0\u5165":134,"\u60c5\u611f\u5206\u6790":88,"\u60f3\u4e86\u89e3\u66f4\u591apaddlepaddl":103,"\u610f\u5473\u7740\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u90fd\u662f\u76f8\u540c\u7684\u503c":140,"\u610f\u601d\u662f\u4e0d\u4f7f\u7528\u5e73\u5747\u53c2\u6570\u6267\u884c\u6d4b\u8bd5":129,"\u610f\u601d\u662f\u4e0d\u4fdd\u5b58\u7ed3\u679c":129,"\u610f\u601d\u662f\u4f7f\u7528\u7b2ctest":129,"\u610f\u601d\u662f\u5728gpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2agpu":129,"\u6210\u529f\u7f16\u8bd1\u540e":113,"\u6210\u529f\u8bad\u7ec3\u4e14\u9000\u51fa\u7684pod\u6570\u76ee\u4e3a3\u65f6":123,"\u6210\u5458":101,"\u6210\u719f\u7684\u9ad8\u6027\u80fd\u5e76\u884c\u8ba1\u7b97\u6846\u67b6":120,"\u6211\u4eec\u4e0d\u80fd\u901a\u8fc7\u5e38\u89c4\u7684\u68af\u5ea6\u68c0\u67e5\u7684\u65b9\u5f0f\u6765\u8ba1\u7b97\u68af\u5ea6":100,"\u6211\u4eec\u4e3b\u8981\u4f1a\u4ecb\u7ecdnvprof\u548cnvvp":134,"\u6211\u4eec\u4e5f\u53ef\u4ee5\u786e\u5b9a\u6bcf\u4e00\u4e2a\u53c2\u6570\u7684\u7c7b\u578b":72,"\u6211\u4eec\u4e5f\u5c06mklml\u5373":68,"\u6211\u4eec\u4ec5\u4ec5\u5bf9\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165\u8fdb\u884c\u4e86\u63cf\u8ff0":110,"\u6211\u4eec\u4ec5\u6709\u4e00\u4e2a\u8f93\u5165":100,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728":122,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884c\u5206\u5e03\u5f0fpaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":123,"\u6211\u4eec\u4ee5\u624b\u5199\u6570\u5b57\u8bc6\u522b\u4efb\u52a1\u4e3a\u4f8b\u8fdb\u884c\u4ecb\u7ecd":116,"\u6211\u4eec\u4f1a\u4fdd\u8bc1":68,"\u6211\u4eec\u4f1a\u5728\u7f51\u7edc\u8bad\u7ec3\u4e4b\u524d\u628a\u683c\u5f0f\u8f6c\u6362\u4e3amkl":68,"\u6211\u4eec\u4f1a\u5bf9\u6bcf\u4e2a\u8bad\u7ec3\u4efb\u52a1\u90fd\u4f1a\u5728\u6bcf\u4e2a\u8282\u70b9\u4e0a\u521b\u5efa\u4e00\u4e2a\u5de5\u4f5c\u7a7a\u95f4":117,"\u6211\u4eec\u4f1a\u5bf9\u6bd4\u5982\u4e0b2\u4e2a\u65b9\u9762":67,"\u6211\u4eec\u4f1a\u628amkl":68,"\u6211\u4eec\u4f1a\u6dfb\u52a0":[67,68],"\u6211\u4eec\u4f1a\u7ee7\u7eed\u4f7f\u7528\u73b0\u6709\u7684\u5185\u5b58\u5757":100,"\u6211\u4eec\u4f1a\u91cd\u65b0\u5206\u914d\u5185\u5b58":100,"\u6211\u4eec\u4f7f\u7528":100,"\u6211\u4eec\u4f7f\u7528\u4e0d\u540c\u7684layer\u8fdb\u884c\u7ec4\u5408":110,"\u6211\u4eec\u4f7f\u7528\u4e86":137,"\u6211\u4eec\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":71,"\u6211\u4eec\u4f7f\u7528paddl":117,"\u6211\u4eec\u5047\u8bbe\u4e00\u53f0\u673a\u5668\u4e0a\u67094\u4e2agpu":131,"\u6211\u4eec\u5148\u8c03\u7528\u6bcf\u4e2a":102,"\u6211\u4eec\u51b3\u5b9a\u4f7f\u7528\u5df2\u6709\u7684":68,"\u6211\u4eec\u5373\u53ef\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684\u642d\u5efa":110,"\u6211\u4eec\u53ea\u6f14\u793a\u4e00\u4e2a\u5355\u673a\u4f5c\u4e1a":122,"\u6211\u4eec\u53ea\u9700\u8981":26,"\u6211\u4eec\u53ea\u9700\u8981\u4f7f\u7528lstm":137,"\u6211\u4eec\u53ea\u9700\u8981\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u628a\u7f16\u8bd1\u597d\u7684paddlepaddle\u6253\u5305\u6210\u4e00\u4e2a":98,"\u6211\u4eec\u53ea\u9700\u8981\u914d\u7f6e":26,"\u6211\u4eec\u53ef\u4ee5":26,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528":133,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5176\u4ed6layer\u8fdb\u884c\u7ec4\u5408":110,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5b83\u6765\u751f\u6210\u5e8f\u5217":140,"\u6211\u4eec\u53ef\u4ee5\u5148\u5b8c\u6210\u5bf9\u539f\u6570\u636e\u7684packing\u64cd\u4f5c":67,"\u6211\u4eec\u53ef\u4ee5\u521b\u5efatrainer\u6765\u5bf9\u7f51\u7edc\u8fdb\u884c\u8bad\u7ec3":110,"\u6211\u4eec\u53ef\u4ee5\u53c2\u8003tensorflow\u7684":102,"\u6211\u4eec\u53ef\u4ee5\u5728":98,"\u6211\u4eec\u53ef\u4ee5\u5728\u547d\u4ee4\u884c\u4e2d\u7b80\u5355\u7684\u770b\u4e00\u4e0b\u751f\u6210\u6548\u679c":133,"\u6211\u4eec\u53ef\u4ee5\u5b9a\u4e49\u5982\u4e0b\u7684layer\u7ec4\u5408":110,"\u6211\u4eec\u53ef\u4ee5\u5b9a\u4e49\u5982\u4e0blayer\u6765\u63cf\u8ff0\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165":110,"\u6211\u4eec\u53ef\u4ee5\u6309\u7167\u5982\u4e0b\u5c42\u6b21\u5b9a\u4e49\u975e\u5e8f\u5217":136,"\u6211\u4eec\u53ef\u4ee5\u67e5\u770b\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c":133,"\u6211\u4eec\u53ef\u4ee5\u8bbe\u8ba1\u642d\u5efa\u4e00\u4e2a\u7075\u6d3b\u7684":139,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7":133,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u65e5\u5fd7\u67e5\u770b\u5bb9\u5668\u8bad\u7ec3\u7684\u60c5\u51b5":123,"\u6211\u4eec\u5728":102,"\u6211\u4eec\u5728\u51fd\u6570\u7684\u7ed3\u5c3e\u8fd4\u56de":140,"\u6211\u4eec\u5bf9\u6a21\u578b\u8fdb\u884c\u4e86\u4ee5\u4e0b\u66f4\u6539":140,"\u6211\u4eec\u5c06":123,"\u6211\u4eec\u5c06\u4e00\u6bb5\u8bdd\u770b\u6210\u53e5\u5b50\u7684\u6570\u7ec4":137,"\u6211\u4eec\u5c06\u4ecb\u7ecd\u5982\u4f55\u542f\u52a8\u5206\u5e03\u5f0f\u8bad\u7ec3\u4f5c\u4e1a":122,"\u6211\u4eec\u5c06\u4f7f\u7528":140,"\u6211\u4eec\u5c06\u4f7f\u7528\u7b80\u5355\u7684":140,"\u6211\u4eec\u5c06\u539f\u59cb\u6570\u636e\u7684\u6bcf\u4e00\u7ec4":137,"\u6211\u4eec\u5c06\u5b83\u4eec\u5212\u5206\u4e3a\u4e0d\u540c\u7684\u7c7b\u522b":128,"\u6211\u4eec\u5c06\u795e\u7ecf\u7f51\u7edc\u4e00\u6b21\u8ba1\u7b97\u63a5\u53d7\u7684\u6240\u6709\u8f93\u5165\u6837\u672c\u79f0\u4e4b\u4e3a\u4e00\u4e2a":115,"\u6211\u4eec\u5c06\u8bad\u7ec3\u7ed3\u675f\u540e\u5b58\u50a8\u4e0b\u6765\u7684\u6a21\u578b\u8f6c\u6362\u6210\u9884\u6d4b\u6a21\u578b":116,"\u6211\u4eec\u5c31\u5b8c\u6210\u4e86\u4e00\u6b21\u4ee3\u7801\u8d21\u732e\u7684\u8fc7\u7a0b":98,"\u6211\u4eec\u5df2\u7ecf\u5b9e\u73b0\u4e86\u5927\u591a\u6570\u5e38\u7528\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":140,"\u6211\u4eec\u5efa\u8bae\u4f60\u4e3a\u4f60\u7684python\u5c01\u88c5\u5b9e\u73b0\u4e00\u4e2a":100,"\u6211\u4eec\u5efa\u8bae\u4f60\u5728\u5199\u65b0\u7f51\u7edc\u5c42\u65f6\u628a\u6d4b\u8bd5\u4ee3\u7801\u653e\u5165\u65b0\u7684\u6587\u4ef6\u4e2d":100,"\u6211\u4eec\u5efa\u8bae\u4f7f\u7528\u7b2c\u4e8c\u7c7b\u5b9e\u73b0":108,"\u6211\u4eec\u603b\u7ed3\u51fa\u4e00\u4e9b\u7279\u522b\u9700\u8981\u6ce8\u610f\u7684\u70b9":68,"\u6211\u4eec\u628apaddlepaddle\u7684\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883\u6253\u5305\u6210\u4e00\u4e2a\u955c\u50cf":142,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528":[28,127],"\u6211\u4eec\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":27,"\u6211\u4eec\u63a8\u8350\u60a8\u4f7f\u7528paddlepaddl":27,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u8f6c\u6362\u65b9\u5f0f":38,"\u6211\u4eec\u63d0\u4f9b\u4e86\u52a0\u901f\u8bbf\u95ee\u7684\u955c\u50cf\u6e90":28,"\u6211\u4eec\u63d0\u4f9b\u53ef\u4ee5\u76f4\u63a5\u8fd0\u884cpaddlepaddl":28,"\u6211\u4eec\u63d0\u51fa\u4e86chunk\u7684\u6982\u5ff5":54,"\u6211\u4eec\u662f\u5bf9\u6bcf\u4e00\u4e2a\u5b50\u5e8f\u5217\u53d6\u6700\u540e\u4e00\u4e2a\u5143\u7d20":137,"\u6211\u4eec\u6700\u7ec8\u7684\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165python\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u8bed\u8a00\u7684\u89e3\u91ca\u5668":71,"\u6211\u4eec\u6709\u4e00\u4e2a\u5e8f\u5217\u4f5c\u4e3a\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u72b6\u6001":140,"\u6211\u4eec\u7684":26,"\u6211\u4eec\u7684\u6807\u51c6\u5f00\u53d1\u6d41\u7a0b\u662f\u628a\u8fd9\u4e9b\u5de5\u5177\u90fd\u88c5\u8fdb\u4e00\u4e2adocker":98,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u5c42rnn\u7684\u914d\u7f6e":137,"\u6211\u4eec\u770b\u4e00\u4e0b\u8bed\u4e49\u76f8\u540c\u7684\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":137,"\u6211\u4eec\u771f\u8bda\u5730\u611f\u8c22\u60a8\u7684\u8d21\u732e":98,"\u6211\u4eec\u79f0\u4e4b\u4e3a\u4e00\u4e2a0\u5c42\u7684\u5e8f\u5217":136,"\u6211\u4eec\u8ba1\u5212\u5c06":67,"\u6211\u4eec\u8ba1\u5212\u5c06\u82f1\u7279\u5c14\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u6570\u5b66\u5e93":68,"\u6211\u4eec\u8bbe\u8ba1\u8bf4\u660e\u4e86\u540d\u4e3afilemanager\u7cfb\u7edf":54,"\u6211\u4eec\u8c03\u7528\u4e86eigenvector\u7684flatten\u63a5\u53e3":102,"\u6211\u4eec\u8fd8\u53ef\u4ee5\u767b\u5f55\u5230\u5bbf\u4e3b\u673a\u4e0a\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":122,"\u6211\u4eec\u8fd8\u5c06\u7f16\u7801\u5411\u91cf\u6295\u5c04\u5230":140,"\u6211\u4eec\u9009\u53d6\u5355\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e\u4e2d\u7684\u4e0d\u540c\u90e8\u5206":137,"\u6211\u4eec\u9009\u62e9":38,"\u6211\u4eec\u901a\u5e38\u501f\u52a9":101,"\u6211\u4eec\u901a\u5e38\u5c06\u4e00\u53e5\u8bdd\u7406\u89e3\u6210\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":137,"\u6211\u4eec\u901a\u8fc7\u4f7f\u7528\u65b0\u5f15\u5165\u7684gemm":67,"\u6211\u4eec\u901a\u8fc7\u8bfb\u53d6":123,"\u6211\u4eec\u90fd\u63d0\u4f9bpython\u7684\u8f6c\u6362\u5e93":38,"\u6211\u4eec\u9700\u8981":26,"\u6211\u4eec\u9700\u8981\u5148\u628a\u8f93\u5165tensor\u548c\u8f93\u51fatensor\u8f6c\u6362\u4e3aeigen\u652f\u6301\u7684\u683c\u5f0f":102,"\u6211\u4eec\u9700\u8981\u5236\u4f5c\u4e00\u4e2a\u5305\u542b\u8bad\u7ec3\u6570\u636e\u7684paddlepaddle\u955c\u50cf":122,"\u6211\u4eec\u9700\u8981\u5728\u96c6\u7fa4\u7684\u6240\u6709\u8282\u70b9\u4e0a\u5b89\u88c5":127,"\u6211\u4eec\u9700\u8981\u7b49\u5f0f\u5de6\u8fb9\u7684eigentensor\u8c03\u7528device\u63a5\u53e3":102,"\u6211\u4eec\u9700\u8981\u8ba1\u7b97":100,"\u6211\u4eec\u9996\u5148\u9700\u8981\u6839\u636e\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u6765\u521b\u5efa\u6240\u9700\u8981\u4f18\u5316\u7684paramet":110,"\u6211\u5220\u9664\u4e86":98,"\u6211\u53ef\u4ee5\u7528":26,"\u6211\u53ef\u4ee5\u9009\u62e9\u4e0d\u7528docker\u5417":26,"\u6216":[115,134,143],"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"\u6216\u4e00\u4e2a\u5411\u91cf":139,"\u6216\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"\u6216\u662f\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d\u8868\u793a\u8bcd\u8bed\u5728\u5b57\u5178\u4e2d\u7684\u5e8f\u53f7":115,"\u6216\u6700\u5927\u503c":136,"\u6216\u79f0\u4f5cweight":107,"\u6216\u7b2c\u4e00\u4e2a":136,"\u6216\u7b2c\u4e00\u4e2a\u5143\u7d20":136,"\u6216\u7f16\u5199\u7a0b\u5e8f\u65f6":117,"\u6216\u8005":[26,68,71,72,98,101,107,115,133,134,136,137],"\u6216\u8005\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":139,"\u6216\u8005\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":[136,139],"\u6216\u8005\u4ece\u5de5\u5177\u7684\u754c\u9762\u91cc\u8fd0\u884c\u60a8\u7684\u5e94\u7528":134,"\u6216\u8005\u5236\u4f5c\u548c\u5206\u4eab\u5e26\u6709\u4ee3\u7801":28,"\u6216\u8005\u53cd\u5411\u5730\u4ece":140,"\u6216\u8005\u53ef\u88abdns\u89e3\u6790\u7684\u4e3b\u673a\u540d":127,"\u6216\u8005\u5728cpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2a\u7ebf\u7a0b":129,"\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":37,"\u6216\u8005\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":128,"\u6216\u8005\u5f15\u8d77\u884c\u65f6\u9519\u8bef":115,"\u6216\u8005\u6570\u7ec4\u7684\u6570\u7ec4\u8fd9\u4e2a\u6982\u5ff5":137,"\u6216\u8005\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":136,"\u6216\u8005\u662f\u51fd\u6570\u8c03\u7528\u7684\u9891\u7387\u548c\u8017\u65f6\u7b49":134,"\u6216\u8005\u66f4\u65e9":109,"\u6216\u8005\u6bcf\u4e00\u4e2a\u7cfb\u5217\u91cc\u7684\u7279\u5f81\u6570\u636e":137,"\u6216\u8005\u7528tuple\u8868\u793a\u7684\u591a\u4e2a\u503c":38,"\u6216\u8005\u7531\u5b83\u4eec\u7ec4\u6210\u7684list":38,"\u6216\u8005\u76f4\u63a5\u6254\u6389\u975e\u5e38\u957f\u7684\u5e8f\u5217":107,"\u6216\u8005\u76f8\u5bf9\u4e8e\u6784\u5efa\u76ee\u5f55\u7684\u76f8\u5bf9\u8def\u5f84":[142,144],"\u6216\u8005\u8f93\u5165\u6570\u636e\u5c3a\u5ea6\u8fc7\u5927":107,"\u6216\u8005\u8fd0\u884c":104,"\u6216\u8005\u91c7\u7528\u5e76\u884c\u8ba1\u7b97\u6765\u52a0\u901f\u67d0\u4e9b\u5c42\u7684\u66f4\u65b0":131,"\u6216activ":68,"\u6216gpu":129,"\u622a\u65ad\u5bf9\u8c61\u4e0d\u540c":107,"\u6240\u4ee5":[28,68,88,107,133],"\u6240\u4ee5\u4e00\u4e2a\u7248\u672c\u53f7\u7684wheel\u5305\u53d1\u5e03\u4e4b\u540e":88,"\u6240\u4ee5\u4e0d\u5b58\u5728\u8fd9\u4e2a\u95ee\u9898":68,"\u6240\u4ee5\u4e0d\u80fd\u91c7\u7528\u7b2c\u4e00\u79cd\u65b9\u5f0f\u5728\u8fd9\u51e0\u4e2alayer\u91cc\u8bbe\u7f6e":108,"\u6240\u4ee5\u505a\u6cd5\u53ef\u4ee5\u6709\u4e24\u79cd":107,"\u6240\u4ee5\u53ef\u4ee5\u7b80\u5316\u5bf9\u73af\u5883\u7684\u8981\u6c42":122,"\u6240\u4ee5\u5728":68,"\u6240\u4ee5\u5728\u5199\u5165\u5feb\u7167\u7684\u8fc7\u7a0b\u4e2d":37,"\u6240\u4ee5\u5916\u5c42\u8f93\u51fa\u7684\u5e8f\u5217\u5f62\u72b6":137,"\u6240\u4ee5\u5bf9":137,"\u6240\u4ee5\u5f00\u53d1\u8005\u9700\u8981\u6839\u636e\u81ea\u5df1\u8bad\u7ec3\u4efb\u52a1\u7684\u5b9e\u9645\u573a\u666f\u5b8c\u6210\u8bad\u7ec3\u6570\u636e\u7684\u5206\u5272\u548c":117,"\u6240\u4ee5\u6027\u80fd\u4e5f\u5c31\u9010\u6b65\u53d8\u6210\u4e86\u6df1\u5ea6\u5b66\u4e60\u9886\u57df\u6700\u91cd\u8981\u7684\u6307\u6807":134,"\u6240\u4ee5\u6211\u4eec\u53ef\u4ee5\u5728\u8fd9\u4e2a\u57fa\u7840\u4e0a":123,"\u6240\u4ee5\u6211\u4eec\u5b9a\u4e49\u4e86\u4e00\u4e2a":68,"\u6240\u4ee5\u6211\u4eec\u786e\u4fdd\u53d1\u5e03\u7684\u4e8c\u8fdb\u5236\u5305\u53ef\u4ee5\u652f\u6301\u4e3b\u6d41\u7684linux\u64cd\u4f5c\u7cfb\u7edf":30,"\u6240\u4ee5\u6211\u4eec\u9700\u8981\u5c06\u8f93\u5165\u6570\u636e\u6807\u8bb0\u6210":137,"\u6240\u4ee5\u6211\u4eec\u9ed8\u8ba4\u4f7f\u7528cento":30,"\u6240\u4ee5\u6574\u4f53\u4e0a":68,"\u6240\u4ee5\u6dfb\u52a0\u4e86\u5bf9\u5e94\u7684":68,"\u6240\u4ee5\u7528\u6237\u9700\u8981\u9996\u5148\u5728":54,"\u6240\u4ee5\u76f8\u6bd4\u4e8erecurr":108,"\u6240\u4ee5\u8fd9\u4e00\u6b65\u662f\u5fc5\u8981\u7684":100,"\u6240\u4ee5\u9700\u8981\u5f15\u5165\u4e00\u4e2a\u8f6c\u6362\u65b9\u6cd5":68,"\u6240\u4f7f\u7528":143,"\u6240\u4f9d\u8d56\u7684\u7b2c\u4e09\u65b9\u5e93\u540c\u65f6\u4e5f\u88ab\u5b89\u88c5\u5230":142,"\u6240\u5bf9\u5e94\u7684\u8bcd\u8868index\u6570\u7ec4":137,"\u6240\u6709\u4e0e\u7c7b\u578b\u76f8\u5173\u7684\u51fd\u6570":72,"\u6240\u6709\u4ee3\u7801\u5fc5\u987b\u5177\u6709\u5355\u5143\u6d4b\u8bd5":98,"\u6240\u6709\u53c2\u6570\u7f6e\u4e3a\u96f6":129,"\u6240\u6709\u547d\u4ee4\u884c\u9009\u9879\u53ef\u4ee5\u8bbe\u7f6e\u4e3a":119,"\u6240\u6709\u5916\u90e8\u7684\u8f6c\u6362\u5de5\u4f5c\u90fd\u4f1a\u5728reset\u7cfb\u5217\u51fd\u6570\u4e2d\u90fd\u51c6\u5907\u597d":68,"\u6240\u6709\u67b6\u6784":142,"\u6240\u6709\u7684":[67,98,100],"\u6240\u6709\u7684\u5355\u6d4b\u90fd\u4f1a\u88ab\u6267\u884c\u4e00\u6b21":100,"\u6240\u6709\u7684\u63a5\u53e3\u5747\u4e3ac\u63a5\u53e3":72,"\u6240\u6709\u7684\u64cd\u4f5c\u90fd\u662f\u9488\u5bf9\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u8fdb\u884c\u7684":137,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u4f7f\u7528":100,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u5728":100,"\u6240\u6709\u76f8\u5173\u7684":67,"\u6240\u6709\u7c7b\u578b\u540d\u4e3a":72,"\u6240\u6709\u7f51\u7edc\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u90fd\u4f4d\u4e8e":100,"\u6240\u6709\u8f93\u5165\u5e8f\u5217\u5e94\u8be5\u6709\u76f8\u540c\u7684\u957f\u5ea6":140,"\u6240\u6709mkl":68,"\u6240\u9700\u652f\u6301\u7684\u6700\u4f4eandroid":142,"\u6240\u9700\u7684\u5f00\u53d1\u5de5\u5177\u548c\u7b2c\u4e09\u65b9\u5e93\u53ef\u4ee5\u53c2\u8003":144,"\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":71,"\u624b\u5199\u6570\u5b57\u8bc6\u522b":116,"\u624b\u5199\u6570\u5b57\u8bc6\u522b\u4efb\u52a1\u5b9a\u4e49\u4e86\u4e00\u4e2a\u542b\u6709":116,"\u624b\u52a8\u4e0b\u8f7d\u4e14\u89e3\u538b\u7f29":104,"\u624b\u52a8\u4e0b\u8f7d\u5e76\u5b89\u88c5":104,"\u624d\u53ef\u4ee5\u5b89\u88c5":30,"\u624d\u80fd\u4fdd\u8bc1\u548c\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":137,"\u624d\u80fd\u53d1\u6325\u5176\u5168\u90e8\u80fd\u529b":134,"\u624d\u80fd\u66f4\u597d\u7684\u53d1\u6325mkl":68,"\u6253\u5f00":134,"\u6253\u5f00\u6d4f\u89c8\u5668\u8bbf\u95ee\u5bf9\u5e94\u76ee\u5f55\u4e0b\u7684index":103,"\u6253\u5f00\u8fd9\u4e2a\u7f16\u8bd1\u9009\u9879":72,"\u6267\u884c":[88,112,113,119],"\u6267\u884c\u4e0a\u8ff0":142,"\u6267\u884c\u4e0a\u8ff0\u4ee3\u7801\u751f\u6210makefile\u6587\u4ef6\u540e":113,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4":27,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u4ee5\u542f\u52a83\u4e2a\u8282\u70b9\u7684openmpi\u96c6\u7fa4\u548c\u4e00\u4e2a":124,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5373\u53ef\u5728\u5f53\u524d\u673a\u5668\u4e0a\u5b89\u88c5paddlepaddle\u7684\u8fd0\u884c\u65f6\u73af\u5883":30,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u53ef\u4ee5\u67e5\u770b\u5df2\u7ecf\u5b89\u88c5\u7684\u7248\u672c":127,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5b8c\u6210\u5feb\u901f\u5b89\u88c5":112,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5":101,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u7f16\u8bd1cpu":27,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u83b7\u53d6\u6700\u65b0\u7684paddlepaddl":28,"\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4":[142,143,144],"\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u542f\u52a8\u4f7f\u7528python\u7f16\u5199\u7684trainer\u7a0b\u5e8f":117,"\u6267\u884c\u4ee5\u4e0b\u64cd\u4f5c":140,"\u6267\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4\u542f\u52a8\u4e00\u4e2a\u53c2\u6570\u670d\u52a1\u5668\u5e76\u7b49\u5f85\u548c\u8ba1\u7b97\u8282\u70b9\u7684\u6570\u636e\u4ea4\u4e92":117,"\u6267\u884c\u5b8c\u5b89\u88c5\u547d\u4ee4\u540e":[142,143,144],"\u6267\u884c\u60a8\u7684\u4ee3\u7801":134,"\u627e\u5230":[27,120,140],"\u627e\u5230\u6700\u65e9\u62a5\u9519\u7684\u5730\u65b9":105,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u539f\u56e0":134,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u90e8\u5206":134,"\u628a":[38,100],"\u628a\u4e4b\u524d\u793a\u4f8b\u4e2d\u8f6c\u6362\u5b8c\u6bd5\u7684random":38,"\u628a\u4efb\u610f\u7ef4\u5ea6\u7684tensor\u8f6c\u4e3a\u4e86\u4e00\u7ef4\u7684eigenvector":102,"\u628a\u5de5\u5177\u548c\u914d\u7f6e\u90fd\u5b89\u88c5\u5728\u4e00\u4e2a":26,"\u628a\u8bad\u7ec3\u6570\u636e\u76f4\u63a5\u653e\u5728":122,"\u628a\u8fd9\u4e9b\u5de5\u5177\u5b89\u88c5\u5230\u672c\u673a":26,"\u6295\u5c04\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b\u5230":140,"\u6295\u5c04\u7f16\u7801\u5411\u91cf\u5230":140,"\u62c6\u6210\u4ee5\u4e0a\u4e24\u4e2a\u9759\u6001\u94fe\u63a5\u5e93":113,"\u62c6\u89e3":139,"\u62c6\u89e3\u6210\u7684\u6bcf\u4e00\u53e5\u8bdd\u518d\u901a\u8fc7\u4e00\u4e2alstm\u7f51\u7edc":137,"\u62f7\u8d1d\u5230numpi":107,"\u62f7\u8d1d\u5fc5\u8981\u7684\u6587\u4ef6\u5230head\u8282\u70b9":124,"\u62f7\u8d1d\u8bad\u7ec3\u6570\u636e\u5230\u5404\u81ea\u7684\u8282\u70b9":124,"\u62f7\u8d1d\u8bad\u7ec3\u6587\u4ef6\u5230\u5bb9\u5668\u5185":123,"\u62f7\u8d1d\u8bad\u7ec3\u7a0b\u5e8f\u548c\u5b57\u5178\u6587\u4ef6\u5230\u6bcf\u53f0mpi\u8282\u70b9":124,"\u62fc\u63a5":107,"\u6302\u8f7d\u5230\u5bb9\u5668\u5185\u90e8\u7684":28,"\u6302\u8f7d\u6216\u4e0b\u8f7d\u7684\u8bad\u7ec3\u6570\u636e\u5206\u7247":117,"\u6307\u53d1\u73b0\u6027\u80fd\u74f6\u9888":133,"\u6307\u5411\u4e00\u4e2alayer":139,"\u6307\u5b9a":[107,108,139,140],"\u6307\u5b9a\u4e00\u53f0\u673a\u5668\u4e0a\u4f7f\u7528\u7684\u7ebf\u7a0b\u6570":129,"\u6307\u5b9a\u4e3a":115,"\u6307\u5b9a\u4f7f\u75282":107,"\u6307\u5b9a\u524d\u5411\u7f51\u7edc\u6700\u7ec8\u7684\u8f93\u51fa\u76ee\u6807\u53d8\u91cf":101,"\u6307\u5b9a\u52a0\u8f7d\u7684\u65b9\u5f0f":129,"\u6307\u5b9a\u5728\u751f\u6210\u6027\u80fd\u5206\u6790\u6587\u4ef6\u4e4b\u540e":133,"\u6307\u5b9a\u5bf9\u8f93\u5165\u53d8\u91cf":101,"\u6307\u5b9a\u5c06\u5f53\u524d\u8def\u5f84":28,"\u6307\u5b9a\u6267\u884c\u5176\u4e2d\u4e00\u4e2a\u5355\u5143\u6d4b\u8bd5":27,"\u6307\u5b9a\u68c0\u6d4b\u68af\u5ea6\u65f6\u80fd\u5bb9\u5fcd\u7684\u6700\u5927\u9519\u8bef\u503c":101,"\u6307\u5b9a\u7684\u5185\u5bb9\u5b58\u50a8\u5e93\u8fd0\u884c\u547d\u4ee4":103,"\u6307\u5b9a\u7684\u8f93\u5165\u4e0d\u4f1a\u88ab":139,"\u6307\u5b9a\u8981\u8f93\u51fa\u7684\u5b57\u6bb5\u8fdb\u884c\u8f93\u51fa":107,"\u6307\u5b9a\u9700\u8981\u4f7f\u7528\u7684\u5bb9\u5668":28,"\u6307\u5b9acudnn\u7684\u6700\u5927\u5de5\u4f5c\u7a7a\u95f4\u5bb9\u9650":129,"\u6307\u5bf9\u4e8e\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u8f93\u5165\u6570\u636e":137,"\u6307\u5f00\u542fhttp\u670d\u52a1":133,"\u6307\u6d88\u9664\u74f6\u9888":133,"\u6307\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u4e4b\u540e\u5f97\u5230\u7684\u6240\u6709\u53c2\u6570":37,"\u6307\u793a\u4f7f\u7528\u54ea\u4e2agpu\u6838":129,"\u6307\u793a\u5728\u7b80\u5355\u7684recurrentlayer\u5c42\u7684\u8ba1\u7b97\u4e2d\u662f\u5426\u4f7f\u7528\u6279\u5904\u7406\u65b9\u6cd5":129,"\u6307\u793a\u5f53\u6307\u5b9a\u8f6e\u7684\u6d4b\u8bd5\u6a21\u578b\u4e0d\u5b58\u5728\u65f6":129,"\u6307\u793a\u662f\u5426\u4f7f\u7528\u591a\u7ebf\u7a0b\u6765\u8ba1\u7b97\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc":129,"\u6307\u793a\u662f\u5426\u5f00\u542f\u53c2\u6570\u670d\u52a1\u5668":129,"\u6307\u793a\u662f\u5426\u663e\u793a\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u65e5\u5fd7\u7ec6\u8282":129,"\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":129,"\u6309\u542f\u53d1\u5f0f\u635f\u5931\u7684\u5927\u5c0f\u9012\u589e\u6392\u5e8f":129,"\u6309\u7167\u4e0b\u9762\u6b65\u9aa4\u5373\u53ef":123,"\u6309\u7167\u5176\u5185\u5bb9\u521b\u5efa\u4e00\u4e2a\u540d\u4e3a":26,"\u6309\u7167\u5177\u4f53\u5b9e\u73b0\u65b9\u5f0f\u53ef\u4ee5\u5f52\u7eb3\u4e3a2\u7c7b":108,"\u6309\u7167\u57fa\u672c\u6570\u636e\u7c7b\u578b\u5728paddlepaddle\u5185\u90e8\u7684\u5b9a\u4e49\u548c\u5b9e\u73b0":115,"\u6309\u94ae":[88,98],"\u635f\u5931\u51fd\u6570\u5c42":116,"\u6392\u6210\u4e00\u5217\u7684\u591a\u4e2a\u5143\u7d20":136,"\u63a5\u4e0a\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42":110,"\u63a5\u4e0a\u5e73\u65b9\u8bef\u5dee\u5c42":110,"\u63a5\u4e0b\u6765":101,"\u63a5\u4e0b\u6765\u53ef\u4ee5\u8003\u8651\u4e0b\u65f6\u95f4\u7ebf\u7684\u5206\u6790":134,"\u63a5\u4e0b\u6765\u5c31\u53ef\u4ee5\u4f7f\u7528":134,"\u63a5\u4e0b\u6765\u6211\u4eec\u521b\u5efa\u4e00\u4e2a\u539f\u59cb":98,"\u63a5\u4e0b\u6765\u6211\u4eec\u53d6\u6d88\u5bf9":98,"\u63a5\u4e0b\u6765\u7b49\u5f85":98,"\u63a5\u53d7\u4e00\u4e2a\u8f93\u5165\u53c2\u6570":101,"\u63a5\u53e3":[71,72,101,102,116],"\u63a5\u53e3\u4f1a\u88ab\u8c03\u7528":102,"\u63a5\u53e3\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684\u524d\u5411\u8ba1\u7b97":116,"\u63a5\u53e3\u5bf9\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u53c2\u6570\u8fdb\u884c\u5e8f\u5217\u5316":116,"\u63a5\u53e3\u5c42\u505a\u8fc7\u591a\u5c01\u88c5":72,"\u63a5\u53e3\u662f":38,"\u63a5\u53e3\u6700\u7ec8\u4f1a\u8c03\u7528\u5bf9\u5e94":102,"\u63a5\u53e3\u6709\u4e00\u4e2a":107,"\u63a5\u53e3\u7684":107,"\u63a5\u53e3\u8bf4\u660e\u8bf7\u67e5\u770b":115,"\u63a5\u6536\u5904\u7406pfsclient\u7aef\u7684\u6587\u4ef6\u7ba1\u7406\u8bf7\u6c42":54,"\u63a5\u7740\u7f16\u8bd1\u5373\u53ef":104,"\u63a7\u5236":129,"\u63a7\u5236\u662f\u5426\u4f7f\u7528mkl":68,"\u63a7\u5236\u662f\u5426\u4f7f\u7528mklml\u5e93":68,"\u63a7\u5236\u7528\u6237\u6743\u9650":38,"\u63a8\u5bfc\u8be5\u5c42\u524d\u5411\u548c\u540e\u5411\u4f20\u9012\u7684\u65b9\u7a0b":100,"\u63a8\u8350\u4f7f\u7528\u6b64\u65b9\u5f0f":113,"\u63a8\u8350\u4f7f\u7528centos\u7684devtools2":27,"\u63a8\u8350\u6e05\u7406\u6574\u4e2a\u7f16\u8bd1\u76ee\u5f55":27,"\u63a8\u8350\u8bbe\u7f6e\u4e3a":113,"\u63a8\u8350\u914d\u7f6e\u4e3a":113,"\u63a8\u8350\u914d\u7f6e\u9009\u9879":113,"\u63a8\u9001\u5230\u8fdc\u7a0b\u4ed3\u5e93":98,"\u63cf\u8ff0\u7684\u9ed8\u8ba4\u5165\u53e3\u7a0b\u5e8f":26,"\u63cf\u8ff0\u8be5op\u7684\u8f93\u5165":101,"\u63cf\u8ff0\u95ee\u9898":98,"\u63d0\u4ea4\u65b9\u5f0f\u53c2\u89c1":103,"\u63d0\u4ea4pull":98,"\u63d0\u4f9b":119,"\u63d0\u4f9b\u4e03\u5c42\u534f\u8bae\u7684\u53cd\u5411\u4ee3\u7406":54,"\u63d0\u4f9b\u4e86\u4e00\u4e2a\u542f\u52a8\u811a\u672c":123,"\u63d0\u4f9b\u4e86\u547d\u4ee4\u6837\u4f8b\u6765\u8fd0\u884c":119,"\u63d0\u4f9b\u4e86\u65b9\u4fbf\u7684\u548c":133,"\u63d0\u4f9b\u4e86\u81ea\u52a8\u5316\u811a\u672c\u6765\u542f\u52a8\u4e0d\u540c\u8282\u70b9\u4e2d\u7684\u6240\u6709":119,"\u63d0\u4f9b\u51e0\u4e4e\u6240\u6709\u8bad\u7ec3\u7684\u5185\u90e8\u8f93\u51fa\u65e5\u5fd7":119,"\u63d0\u4f9b\u5e38\u7528\u7684\u547d\u4ee4\u884c\u7ba1\u7406\u547d\u4ee4\u7ba1\u7406\u6587\u4ef6\u548c\u76ee\u5f55":54,"\u63d0\u4f9b\u6269\u5c55\u7684\u957f\u5ea6\u4fe1\u606f":136,"\u63d0\u4f9b\u7528\u6237\u7ba1\u7406\u6587\u4ef6\u7684\u547d\u4ee4":54,"\u63d0\u4f9b\u7ed9paddle\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e":38,"\u63d0\u4f9b\u8bad\u7ec3\u8fc7\u7a0b\u7684":119,"\u63d0\u793a":104,"\u641c\u7d22\u4ee3\u7801\u5e93":103,"\u642d\u5efa\u795e\u7ecf\u7f51\u7edc\u5c31\u50cf\u4f7f\u7528\u79ef\u6728\u642d\u5efa\u5b9d\u5854\u4e00\u6837":110,"\u64cd\u4f5c":137,"\u64cd\u4f5c\u7cfb\u7edf":[26,30],"\u652f\u6301\u4e24\u79cd\u5e8f\u5217\u7c7b\u578b":115,"\u652f\u6301\u4ea4\u53c9\u7f16\u8bd1":144,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":[138,139],"\u652f\u6301\u5927\u6587\u4ef6\u7684\u65ad\u70b9\u4e0a\u4f20":54,"\u652f\u6301\u5927\u89c4\u6a21\u96c6\u7fa4\u751f\u4ea7\u73af\u5883\u7684\u5b8c\u6574\u96c6\u7fa4\u65b9\u6848":120,"\u652f\u6301\u7684\u6700\u5c0f\u7684android":142,"\u652f\u6301\u7684\u6700\u5c0fandroid":142,"\u652f\u6301\u7ef4\u6570\u53ef\u53d8\u7684\u6570\u636e\u8f93\u5165":108,"\u652f\u6301\u7f16\u8bd1\u5668":142,"\u6539\u53d8\u7ef4\u5ea6\u987a\u5e8f":108,"\u653e\u5728\u8fd9\u4e2a\u76ee\u5f55\u91cc\u7684\u6587\u4ef6\u5176\u5b9e\u662f\u4fdd\u5b58\u5230\u4e86mfs\u4e0a":123,"\u6545\u800c\u662f\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":137,"\u6548\u679c\u5982\u4e0b":133,"\u6559\u7a0b":28,"\u6570":139,"\u6570\u5b66\u5e93":113,"\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":139,"\u6570\u636e":[54,115,116],"\u6570\u636e\u4e2d0":109,"\u6570\u636e\u5206\u7247":118,"\u6570\u636e\u63d0\u4f9b\u5668":128,"\u6570\u636e\u8bbf\u95ee":0,"\u6570\u636e\u8bfb\u53d6\u5747\u4ea4\u7531\u5176\u4ed6\u8bed\u8a00\u5b8c\u6210":71,"\u6570\u636e\u8f93\u5165":[115,139],"\u6570\u636e\u957f\u5ea6\u53ca\u6821\u9a8c\u503c\u7ec4\u6210":54,"\u6570\u636e\u96c6":116,"\u6570\u636e\u96c6\u9700\u8981\u9884\u5148\u88ab\u8f6c\u6362\u6210paddlepaddle\u5206\u5e03\u5f0f\u8bad\u7ec3\u4f7f\u7528\u7684\u5b58\u50a8\u683c":38,"\u6570\u636e\u9884\u5904\u7406\u4efb\u52a1":38,"\u6570\u76ee":131,"\u6574\u4f53\u4f7f\u7528\u6d41\u7a0b":116,"\u6574\u4f53\u6570\u636e\u548c\u539f\u59cb\u6570\u636e\u5b8c\u5168\u4e00\u6837":137,"\u6574\u4f53\u7684\u7ed3\u6784\u56fe\u5982\u4e0b":123,"\u6574\u578b\u6570\u7ec4":115,"\u6574\u6570":100,"\u6574\u6570\u6807\u7b7e":110,"\u6587\u4ef6":[26,71,98,101,115,122],"\u6587\u4ef6\u4e2d":[101,116,123],"\u6587\u4ef6\u4e2d\u6ce8\u518c\u524d\u5411":101,"\u6587\u4ef6\u4e2d\u6ce8\u518c\u8be5op\u548ckernel":101,"\u6587\u4ef6\u4e2d\u6ce8\u518ccuda":101,"\u6587\u4ef6\u4e3a":107,"\u6587\u4ef6\u4e4b\u5916":98,"\u6587\u4ef6\u4f20\u8f93\u7684\u7684\u5173\u952e\u5728\u4e8e\u9700\u8981pfsclient\u7aef\u5bf9\u6bd4source\u548cdestination\u7684\u6587\u4ef6chunks\u7684checksum\u662f\u5426\u4fdd\u6301\u4e00\u81f4":54,"\u6587\u4ef6\u5185\u5bb9\u4e3a":71,"\u6587\u4ef6\u540d":133,"\u6587\u4ef6\u540d\u4e3a\u4efb\u610f\u6587\u4ef6\u540d":117,"\u6587\u4ef6\u540d\u4e3a\u6b64uuid":37,"\u6587\u4ef6\u547d\u540d\u4ee5":101,"\u6587\u4ef6\u59390":123,"\u6587\u4ef6\u5bf9\u5e94\u7684data":38,"\u6587\u4ef6\u5de5\u5177\u662f\u4f7f\u7528docker":103,"\u6587\u4ef6\u7684\u4e0a\u4f20\u548c\u4e0b\u8f7d\u90fd\u662f\u901a\u8fc7\u5bf9chunk\u7684\u64cd\u4f5c\u6765\u5b9e\u73b0\u7684":54,"\u6587\u4ef6\u7684\u6539\u53d8":98,"\u6587\u4ef6\u7684\u8def\u5f84\u6765\u52a0\u8f7d\u9884\u6d4b\u6a21\u578b":116,"\u6587\u4ef6model":131,"\u6587\u5b57\u7684\u4ea4\u4e92\u5f0f\u6587\u6863":28,"\u6587\u6863":104,"\u6587\u68631":102,"\u6587\u68632":102,"\u6587\u6863\u8f83\u5c11":102,"\u6587\u6863\u90fd\u662f\u901a\u8fc7":103,"\u6587\u7ae0":123,"\u65b0\u5efa\u4e00\u4e2a\u6743\u91cd":100,"\u65b0\u624b\u5165\u95e8":141,"\u65b0\u624b\u5165\u95e8\u7ae0\u8282":88,"\u65b0\u7248\u672c":68,"\u65b9\u4fbf\u5feb\u901f\u5b89\u88c5":29,"\u65b9\u4fbf\u6392\u67e5\u4ee5\u53ca\u5feb\u901f\u5b9a\u4f4d\u95ee\u9898":107,"\u65b9\u4fbf\u6d4b\u8bd5\u4eba\u5458\u6d4b\u8bd5paddlepaddle\u7684\u884c\u4e3a":88,"\u65b9\u4fbf\u7528\u6237\u4e0a\u4f20\u81ea\u5df1\u7684\u8bad\u7ec3\u6570\u636e\u4ee5\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":54,"\u65b9\u4fbf\u7528\u6237\u5728python\u7aef\u9009\u62e9\u662f\u5426\u542f\u7528\u8fd9\u4e2a\u529f\u80fd":67,"\u65b9\u4fbf\u7528\u6237\u9009\u62e9\u4f7f\u7528mkl":68,"\u65b9\u5f0f1":107,"\u65b9\u5f0f2":107,"\u65b9\u5f0f\u5c06\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u53c2\u6570\u5e8f\u5217\u5316\u5230\u4e00\u4e2a\u6587\u4ef6":116,"\u65b9\u5f0f\u7c7b\u4f3c\u4e8e":68,"\u65b9\u6cd5\u4e00":131,"\u65b9\u6cd5\u4e09":131,"\u65b9\u6cd5\u4e8c":131,"\u65e0\u5ef6\u8fdf":129,"\u65e0\u6cd5\u505a\u5230\u5bf9\u4e8e\u5404\u79cd\u8bed\u8a00\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u7684\u9002\u914d":71,"\u65e0\u8bba\u5728\u672c\u5730\u8fd8\u662f\u5728\u4e91\u7aef":38,"\u65e0\u8bba\u662f\u4ece":38,"\u65e0\u8bba\u662f\u5728\u672c\u5730\u6216\u662f\u4e91\u7aef\u8f6c\u6362":38,"\u65e0\u8bba\u662f\u91cd\u6784\u524d\u7684layer\u8fd8\u662f\u91cd\u6784\u540e\u7684op":68,"\u65e0\u9700\u5173\u5fc3\u548c\u5904\u7406\u5e8f\u5217\u4fe1\u606f":115,"\u65e0\u9700\u63d0\u4f9b\u975e\u96f6\u5143\u7684\u503c":115,"\u65e0\u9700\u9644\u52a0\u5e8f\u5217\u4fe1\u606f":115,"\u65e0\u9ed8\u8ba4\u503c":[142,144],"\u65e5\u5fd7\u62a5\u9519\u4e3a\u7f51\u7edc\u901a\u4fe1\u7c7b\u9519\u8bef":105,"\u65f6":[37,67,68,100,107,109,113,115,123,129,136,140,142],"\u65f6\u4e00\u8d77\u66f4\u65b0":68,"\u65f6\u4f7f\u7528openblas\u6570\u5b66\u5e93":113,"\u65f6\u5982\u4f55\u7ec4\u7ec7\u8f93\u5165\u6570\u636e":115,"\u65f6\u6709\u6548":143,"\u65f6\u88ab\u8bad\u7ec3\u7684":100,"\u65f6\u8bbe\u5907id\u53f7\u7684\u5206\u914d":131,"\u65f6\u95f4":137,"\u65f6\u95f4\u6b65\u7684\u6982\u5ff5":137,"\u65f6\u987b\u4ece\u7b2c17\u5b57\u8282\u5f00\u59cb":109,"\u6620\u5c04\u4e3a":26,"\u6620\u5c04\u5230\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":100,"\u662f":[30,54,68,104],"\u662f\u4e00\u4e2a\u51681\u7684\u5411\u91cf":100,"\u662f\u4e00\u4e2a\u5185\u7f6e\u7684\u5b9a\u65f6\u5668\u5c01\u88c5":134,"\u662f\u4e00\u4e2a\u52a8\u6001\u7a0b\u5e8f\u5206\u6790\u7684\u672f\u8bed":134,"\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u662f\u4e00\u4e2a\u53cc\u5c42\u7684\u5e8f\u5217":136,"\u662f\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3\u7684\u4ee3\u7801\u751f\u6210\u5668":71,"\u662f\u4e00\u4e2a\u5c01\u88c5\u5bf9\u8c61":134,"\u662f\u4e00\u4e2a\u5f88\u6709\u7528\u7684\u53c2\u6570":131,"\u662f\u4e00\u4e2a\u7c7b\u578b\u7684\u6807\u5fd7":72,"\u662f\u4e00\u4e2a\u975e\u7ebf\u6027\u7684":100,"\u662f\u4e00\u4e2apython\u7684\u7b2c\u4e09\u65b9\u5e93":133,"\u662f\u4e00\u4e2aunbound":139,"\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":110,"\u662f\u4e00\u6b21\u9884\u6d4b\u63a5\u53d7\u7684\u6837\u672c\u6570\u76ee":115,"\u662f\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":139,"\u662f\u4e0d\u5305\u62ec\u6e90\u7801\u7684":122,"\u662f\u4e0d\u5e38\u89c1\u7684\u505a\u6cd5":71,"\u662f\u4f7f\u5f97\u8981\u5171\u4eab\u7684\u53c2\u6570\u4f7f\u7528\u540c\u6837\u7684":109,"\u662f\u4f7f\u7528mkl\u6570\u5b66\u5e93":113,"\u662f\u504f\u5dee":140,"\u662f\u5404\u4e2a\u5b9e\u73b0\u4e2d\u5171\u4eab\u7684\u5934\u6587\u4ef6":72,"\u662f\u5411\u91cf":100,"\u662f\u5426\u4ec5\u7f16\u8bd1capi":27,"\u662f\u5426\u4ee5\u9006\u5e8f\u5904\u7406\u8f93\u5165\u5e8f\u5217":139,"\u662f\u5426\u4f7f\u7528":143,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":27,"\u662f\u5426\u4f7f\u7528\u65e7\u7684remoteparameterupdat":129,"\u662f\u5426\u4f7f\u7528\u6743\u91cd":100,"\u662f\u5426\u4f7f\u7528arm\u6a21\u5f0f":142,"\u662f\u5426\u4f7f\u7528eigen\u5e93\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":[142,143],"\u662f\u5426\u4f7f\u7528mkl\u6570\u5b66\u5e93":27,"\u662f\u5426\u4f7f\u7528neon\u6307\u4ee4":[142,144],"\u662f\u5426\u4f7f\u80fd":143,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":27,"\u662f\u5426\u5219\u5171\u4eab\u540c\u4e00\u4e2a":101,"\u662f\u5426\u542f\u7528gpu\u8bad\u7ec3":117,"\u662f\u5426\u5c06\u5168\u5c40\u79cd\u5b50\u5e94\u7528\u4e8e\u672c\u5730\u7ebf\u7a0b\u7684\u968f\u673a\u6570":129,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":27,"\u662f\u5426\u6253\u5370\u7248\u672c\u4fe1\u606f":129,"\u662f\u5426\u6253\u5f00":67,"\u662f\u5426\u652f\u6301gpu":27,"\u662f\u5426\u663e\u793a":129,"\u662f\u5426\u7a00\u758f":100,"\u662f\u5426\u7f16\u8bd1\u4e2d\u82f1\u6587\u6587\u6863":27,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236\u6587\u4ef6":27,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":27,"\u662f\u5426\u7f16\u8bd1c":143,"\u662f\u5426\u7f16\u8bd1go\u8bed\u8a00\u7684\u53ef\u5bb9\u9519paramet":27,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":27,"\u662f\u5426\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":27,"\u662f\u5426\u9700\u8981\u7b49\u5f85\u8be5\u8f6e\u6a21\u578b\u53c2\u6570":129,"\u662f\u56e0\u4e3a\u8fd9\u4e2a\u6d41\u7a0b\u6bd4\u5176\u4ed6\u65b9\u6cd5\u90fd\u66f4\u7b80\u4fbf":26,"\u662f\u56e0\u4e3ac99\u652f\u6301":71,"\u662f\u5728paddlepaddle\u4e2d\u6784\u9020\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u6700\u91cd\u8981\u7684\u6982\u5ff5":140,"\u662f\u5b58\u6709\u4e00\u7cfb\u5217\u53d8\u6362\u77e9\u9635\u7684\u6743\u91cd":100,"\u662f\u5b58\u6709\u504f\u7f6e\u5411\u91cf\u7684\u6743\u91cd":100,"\u662f\u5bf9\u7528\u6237\u6587\u4ef6\u5b58\u50a8\u7a7a\u95f4\u7684\u62bd\u8c61":54,"\u662f\u5bfb\u627e\u74f6\u9888\u7684\u5173\u952e\u6307\u6807":133,"\u662f\u5f00\u542favx\u7f16\u8bd1\u7684":28,"\u662f\u5f85\u6269\u5c55\u7684\u6570\u636e":136,"\u662f\u6211\u4eec":98,"\u662f\u6211\u4eec\u8981\u5206\u6790\u7684\u7a0b\u5e8f":133,"\u662f\u6307":72,"\u662f\u6307\u4e00\u7cfb\u5217\u7684\u7279\u5f81\u6570\u636e":137,"\u662f\u6307recurrent_group\u7684\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":137,"\u662f\u6570\u636e\u8f93\u5165":140,"\u662f\u6709\u610f\u4e49\u7684":137,"\u662f\u6784\u5efa\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u6700\u91cd\u8981\u7684\u5de5\u5177":140,"\u662f\u6ca1\u6709\u540d\u5b57\u7684":26,"\u662f\u7528\u6237\u4f7f\u7528c":72,"\u662f\u7684":26,"\u662f\u77e9\u9635":100,"\u662f\u795e\u7ecf\u7f51\u7edc\u5b9a\u4e49\u65f6":115,"\u662f\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u6807\u8bc6\u7b26":100,"\u662f\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26":100,"\u662f\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":100,"\u662f\u7f51\u7edc\u5c42\u8f93\u51fa\u7684\u5927\u5c0f":100,"\u662f\u8be5\u5c42\u7684\u6807\u8bc6\u7b26":100,"\u662f\u8be5\u5c42\u7684\u7c7b\u540d":100,"\u662f\u8be5\u7f51\u7edc\u5c42\u7684":100,"\u662f\u8f93\u5165":140,"\u662f\u8fd9\u4e00\u7c7b\u7684":108,"\u662f\u8fdb\u884c\u8ba1\u7b97\u7684\u57fa\u672c\u5355\u4f4d":115,"\u662f\u9700\u8981\u4e86\u89e3\u54ea\u4e9b\u6b65\u9aa4\u62d6\u6162\u4e86\u6574\u4f53":134,"\u662fc":72,"\u662fdecoder\u7684\u6570\u636e\u8f93\u5165":139,"\u662fnvidia\u6027\u80fd\u5206\u6790\u5de5\u5177":134,"\u662fpaddlepaddle\u4e2d\u5355\u5c42\u5e8f\u5217\u548c\u53cc\u5c42\u5e8f\u5217\u5b58\u50a8\u793a\u610f\u56fe":115,"\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":139,"\u662fpython\u5c01\u88c5\u7684\u7c7b\u540d":100,"\u662frnn\u72b6\u6001":140,"\u663e\u5f97\u76f8\u5bf9\u6765\u8bf4\u8f83\u4e3a\u8017\u65f6":67,"\u663e\u7136":133,"\u6682\u4e0d\u8003\u8651\u5728\u5185":107,"\u6682\u65e0":30,"\u6682\u65f6\u4e0d\u652f\u6301python3":30,"\u6682\u65f6\u4e0d\u8003\u8651\u591a\u4e2aparamet":37,"\u66b4\u9732\u8fd9\u4e2a\u6982\u5ff5\u5fc5\u8981\u51fd\u6570":72,"\u66f4\u522b\u63d0\u7b80\u5316\u95ee\u9898\u590d\u73b0\u5e26\u6765\u7684\u597d\u5904\u4e86":26,"\u66f4\u591a\u5173\u4e8edocker\u7684\u5b89\u88c5\u4e0e\u4f7f\u7528":104,"\u66f4\u591a\u7684\u8f6c\u6362\u65b9\u6cd5\u8bf7\u53c2\u8003eigen":102,"\u66f4\u597d\u5730\u5b8c\u6210\u4e00\u4e9b\u590d\u6742\u7684\u8bed\u8a00\u7406\u89e3\u4efb\u52a1":139,"\u66f4\u5feb":140,"\u66f4\u65b0":104,"\u66f4\u65b0\u53ef\u80fd\u5bfc\u81f4\u9700\u8981\u65b0\u7684\u5f00\u53d1\u5de5\u5177":26,"\u66f4\u65b0\u6a21\u5f0f":107,"\u66f4\u65b0\u7684\u6587\u6863\u4ee5pr\u7684\u5f62\u5f0f\u63d0\u4ea4\u5230github\u4e2d":103,"\u66f4\u65b0\u7f51\u7edc\u53c2\u6570\u65f6\u5e94\u7528":107,"\u66f4\u65b9\u4fbf\u7684\u8bbe\u7f6e\u65b9\u5f0f":109,"\u66f4\u8be6\u7ec6\u7684\u5b89\u88c5\u548c\u7f16\u8bd1\u65b9\u6cd5\u53c2\u8003":112,"\u66f4\u8fdb\u4e00\u6b65":139,"\u66f4\u9ad8":140,"\u66ff\u6211\u4eec\u5b8c\u6210\u4e86\u539f\u59cb\u8f93\u5165\u6570\u636e\u7684\u62c6\u5206":139,"\u6700\u4e3b\u8981\u7684\u5de5\u4f5c\u5c31\u662f\u89e3\u6790\u51fa":123,"\u6700\u540e":[28,98,100,117],"\u6700\u540e\u4e00\u4e2a":136,"\u6700\u540e\u4e00\u5c42cost\u4e2d\u8bb0\u5f55\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u6240\u6709\u62d3\u6251\u7ed3\u6784":110,"\u6700\u540e\u518d\u8c03\u7528mutabl":102,"\u6700\u540e\u5220\u9664":88,"\u6700\u540e\u6211\u4eec\u4f7f\u7528\u94fe\u5f0f\u6cd5\u5219\u8ba1\u7b97":100,"\u6700\u540e\u8ba1\u7b97softmax":108,"\u6700\u5c0f\u7684ios\u90e8\u7f72\u7248\u672c":143,"\u6700\u5c11\u663e\u793a\u591a\u5c11\u4e2a\u8282\u70b9":129,"\u6700\u5e38\u89c1\u7684\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662fexcept":71,"\u6700\u65b0\u7684\u4ee3\u7801":98,"\u6700\u65b0\u7684paddlepaddl":[28,104],"\u6700\u7ec8":100,"\u6700\u7ec8\u5b9e\u73b0\u4e00\u4e2a\u5c42\u6b21\u5316\u7684\u590d\u6742rnn":139,"\u6700\u7ec8\u6211\u4eec\u53ef\u4ee5\u8c03\u7528trainer\u7684train\u65b9\u6cd5\u542f\u52a8\u8bad\u7ec3":110,"\u6700\u7ec8\u7684\u8f93\u51fa\u7ed3\u679c":139,"\u6709\u4e00\u4e9b\u5fc5\u987b\u914d\u7f6e\u7684\u53c2\u6570":[142,143,144],"\u6709\u4e24\u79cd\u65b9\u6cd5":27,"\u6709\u4e9b\u5c42\u53ef\u80fd\u9700\u8981\u9ad8\u7cbe\u5ea6\u6765\u4fdd\u8bc1\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u6b63\u786e\u6267\u884c":100,"\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":100,"\u6709\u4e9b\u7279\u5f81\u7684\u53d6\u503c\u8fbe\u5230\u6570\u767e\u4e07":107,"\u6709\u4eba\u7528\u865a\u62df\u673a\u6765\u7c7b\u6bd4":26,"\u6709\u4ee5\u4e0b\u5efa\u8bae":[142,143],"\u6709\u5173":137,"\u6709\u5173\u53c2\u6570\u914d\u7f6e\u7684\u8be6\u7ec6\u8bf4\u660e\u89c1":142,"\u6709\u5173\u7ebf\u6027\u56de\u5f52\u7684\u5b9e\u9645\u5e94\u7528":110,"\u6709\u52a9\u4e8e\u5728\u8bad\u7ec3\u65f6\u89c2\u5bdf\u5177\u4f53\u6570\u503c":107,"\u6709\u52a9\u4e8e\u8bca\u65ad\u5206\u5e03\u5f0f\u9519\u8bef":119,"\u6709\u591a\u96be":26,"\u6709\u6548\u63d0\u5347paddlepaddle\u5728\u82f1\u7279\u5c14\u67b6\u6784\u4e0a\u7684\u6027\u80fd":[67,68],"\u6709\u65f6\u5019\u6211\u4eec\u4f1a\u5e0c\u671b\u6e05\u7406\u6389\u5df2\u7ecf\u4e0b\u8f7d\u7684\u7b2c\u4e09\u65b9\u4f9d\u8d56\u4ee5\u53ca\u5df2\u7ecf\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6":26,"\u6709\u65f6\u5019\u6211\u4eec\u53ea\u60f3\u8fd0\u884c\u4e00\u4e2a\u7279\u5b9a\u7684\u5355\u5143\u6d4b\u8bd5":26,"\u6709\u6807\u51c6\u7684":71,"\u6709\u7684\u65f6\u5019":71,"\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":134,"\u6709\u7684\u8bdd\u9700\u8981\u5148\u5378\u8f7d":104,"\u6709\u975e\u5e38\u5927\u7684\u5dee\u522b":133,"\u670d\u52a1\u5668\u4e4b\u95f4\u53ef\u4ee5\u901a\u8fc7\u5c40\u57df\u7f51":127,"\u672a\u6307\u5b9a\u6309\u7167double\u7cbe\u5ea6\u7f16\u8bd1":109,"\u672a\u8bbe\u7f6e":143,"\u672c\u4f8b\u4e2d\u7684\u539f\u59cb\u6570\u636e\u4e00\u5171\u670910\u4e2a\u6837\u672c":137,"\u672c\u5217\u8868\u8bf4\u660epaddlepaddle\u53d1\u7248\u4e4b\u524d\u9700\u8981\u6d4b\u8bd5\u7684\u529f\u80fd\u70b9":88,"\u672c\u5730":[30,104],"\u672c\u5730\u6d4b\u8bd5":128,"\u672c\u5730\u8bad\u7ec3":[115,128],"\u672c\u5730\u8bad\u7ec3\u4e0e\u9884\u6d4b":106,"\u672c\u5730\u8bad\u7ec3\u7684\u5b9e\u9a8c":131,"\u672c\u6559\u7a0b\u4e3b\u8981\u4ecb\u7ecd\u5e26kernel\u7684op\u5982\u4f55\u5199":101,"\u672c\u6559\u7a0b\u5c06\u6307\u5bfc\u4f60\u5982\u4f55\u5728":140,"\u672c\u6587\u4e2d\u6240\u6709\u7684\u4f8b\u5b50":137,"\u672c\u6587\u4e2d\u7684\u4f8b\u5b50\u91cc":26,"\u672c\u6587\u4e2d\u793a\u4f8b\u6240\u4f7f\u7528\u7684\u5355\u5143\u6d4b\u8bd5\u6587\u4ef6\u662f":137,"\u672c\u6587\u4ee5paddlepaddle\u7684\u53cc\u5c42rnn\u5355\u5143\u6d4b\u8bd5\u4e3a\u793a\u4f8b":137,"\u672c\u6587\u5c06\u4ecb\u7ecd\u5728kubernetes\u5bb9\u5668\u7ba1\u7406\u5e73\u53f0\u4e0a\u5feb\u901f\u6784\u5efapaddlepaddle\u5bb9\u5668\u96c6\u7fa4":123,"\u672c\u6587\u6863\u5c06\u4ee5linux":142,"\u672c\u6587\u6863\u63cf\u8ff0paddl":72,"\u672c\u6587\u7684\u5c06\u4ecb\u7ecd\u5728macos\u4e0a":143,"\u672c\u6b21\u8bad\u7ec3\u6587\u4ef6\u6240\u5728\u76ee\u5f55":123,"\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6\u53ef\u4ee5\u5199\u6210":123,"\u672c\u6b21\u8bad\u7ec3\u8981\u6c42\u67093\u4e2apaddlepaddle\u8282\u70b9":123,"\u672c\u793a\u4f8b\u4e2d\u4f7f\u7528\u7684\u539f\u59cb\u6570\u636e\u5982\u4e0b":137,"\u672c\u793a\u4f8b\u610f\u56fe\u4f7f\u7528\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":137,"\u672c\u8282\u5c06\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528paddlepaddle\u5728\u4e0d\u540c\u7684\u96c6\u7fa4\u6846\u67b6\u4e0b\u5b8c\u6210\u5206\u5e03\u5f0f\u8bad\u7ec3":118,"\u673a\u5668\u4e0a\u4ee5\u53ca":144,"\u673a\u5668\u7684\u8bbe\u5907":131,"\u673a\u5668\u7ffb\u8bd1":88,"\u6743\u91cd\u66f4\u65b0\u7684\u68af\u5ea6":129,"\u6765\u4e3a\u4e00\u4e2a":115,"\u6765\u4ee3\u66ff":98,"\u6765\u4f20\u8f93\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6\u4e2d\u5b9a\u4e49\u7684\u7f51\u7edc\u7ed3\u6784\u548c\u76f8\u5173\u53c2\u6570":116,"\u6765\u4f7f\u7528dropout":108,"\u6765\u4f7f\u7528dropout\u7684":108,"\u6765\u4fdd\u8bc1\u8bad\u7ec3\u8fc7\u7a0b\u53ef\u4ee5\u4ece\u4e2d\u95f4\u72b6\u6001\u91cd\u65b0\u542f\u52a8":37,"\u6765\u505a\u68af\u5ea6\u68c0\u67e5":100,"\u6765\u51b3\u5b9a\u662f\u5426\u5f00\u542fmkl":67,"\u6765\u5206\u6790\u6267\u884c\u6587\u4ef6":134,"\u6765\u521d\u59cb\u5316\u53c2\u6570":109,"\u6765\u542f\u52a8\u548c":26,"\u6765\u5b58\u50a8":115,"\u6765\u5b58\u50a8\u6570\u636e":[115,116],"\u6765\u5b8c\u6210\u524d\u5411\u548c\u53cd\u5411\u8ba1\u7b97":116,"\u6765\u5b8c\u6210\u7f51\u7edc\u7684\u8bad\u7ec3":110,"\u6765\u5b9a\u4e49\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":140,"\u6765\u5b9e\u73b0":68,"\u6765\u5b9e\u9645\u5b58\u50a8\u6570\u636e":[115,116],"\u6765\u5bf9\u6bd4\u5206\u6790\u4e24\u8005\u8bed\u4e49\u76f8\u540c\u7684\u539f\u56e0":137,"\u6765\u5f71\u54cdpaddlepaddle\u7684\u7f16\u8bd1\u8fc7\u7a0b":[142,143],"\u6765\u5f97\u5230\u67d0\u4e2a\u7279\u5b9a\u53c2\u6570\u7684\u68af\u5ea6\u77e9\u9635":100,"\u6765\u6267\u884c":26,"\u6765\u63cf\u8ff0\u7684":102,"\u6765\u63cf\u8ff0\u8be5op\u7684\u8f93\u5165":101,"\u6765\u63cf\u8ff0\u8f93\u5165":115,"\u6765\u642d\u5efa\u795e\u7ecf\u7f51\u7edc":110,"\u6765\u663e\u793a\u6027\u80fd\u5206\u6790\u7ed3\u679c":133,"\u6765\u67e5\u770b\u6027\u80fd\u5206\u6790\u7ed3\u679c":133,"\u6765\u6ce8\u518c\u8be5\u5c42":100,"\u6765\u6df7\u5408\u4f7f\u7528gpu\u548ccpu\u8ba1\u7b97\u7f51\u7edc\u5c42\u7684\u53c2\u6570":131,"\u6765\u6e05\u7406\u8fd9\u4e9b\u5185\u5bb9":26,"\u6765\u7279\u6307":115,"\u6765\u7279\u6307\u8c03\u7528paddlepaddl":116,"\u6765\u7279\u6307paddlepaddl":116,"\u6765\u7279\u6307paddlepaddle\u4e2d\u7684\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":115,"\u6765\u7279\u6307paddlepaddle\u4e2d\u7684\u4e8c\u7ef4\u6d6e\u70b9\u578b\u77e9\u9635":115,"\u6765\u7279\u6307paddlepaddle\u4e2d\u795e\u7ecf\u7f51\u7edc\u8ba1\u7b97\u5c42\u4e00\u4e2a\u8f93\u5165":115,"\u6765\u786e\u4fdd\u628a":71,"\u6765\u786e\u5b9a\u7a00\u758f\u77e9\u9635\u7684\u5185\u5bb9":115,"\u6765\u7f16\u8bd1":26,"\u6765\u83b7\u5f97\u8f93\u51fa\u7684\u68af\u5ea6":100,"\u6765\u8868\u793a":140,"\u6765\u8868\u793a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":140,"\u6765\u8868\u793apaddle\u5185\u90e8\u7c7b":71,"\u6765\u89e3\u51b3\u4e0a\u9762\u7684\u95ee\u9898":107,"\u6765\u8ba1\u7b97\u68af\u5ea6":100,"\u6765\u8bb0\u5f55\u8f93\u5165":115,"\u6765\u8bb2\u89e3\u5982\u4f55\u4f7f\u7528\u53cc\u5c42rnn":137,"\u6765\u8bbe\u7f6e":109,"\u6765\u8bbf\u95ee\u7528\u6237\u81ea\u5df1\u7684\u6570\u636e":38,"\u6765\u8bfb\u53d6\u4e00\u4e2a":115,"\u6765\u8c03\u6574":98,"\u6765\u8c03\u7528":26,"\u6765\u8fd0\u884c\u5305\u62ec":26,"\u6765\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5\u4e86":26,"\u6765\u8fd0\u884c\u6027\u80fd\u5206\u6790\u548c\u8c03\u4f18":134,"\u6765\u8fd0\u884c\u955c\u50cf":28,"\u6765\u8fdb\u884c\u8ba8\u8bba":72,"\u6765\u9884\u6d4b\u8fd9\u4e2a\u4e2d\u95f4\u7684\u8bcd":107,"\u6784\u5efa":142,"\u6784\u5efa\u597d\u5f00\u53d1\u955c\u50cf\u540e":142,"\u6784\u5efa\u76ee\u6807\u4e3a":143,"\u6784\u6210\u4e00\u4e2a\u5e8f\u5217":115,"\u6784\u6210\u4e86\u8f93\u51fa\u53cc\u5c42\u5e8f\u5217\u7684\u7b2ci\u4e2a":136,"\u6784\u9020":123,"\u6784\u9020\u51fd\u6570\u542b\u67092\u4e2a\u53c2\u6570":101,"\u6784\u9020\u51fd\u6570\u91cc\u901a\u8fc7":101,"\u67b6\u6784\u7684\u6a21\u62df\u5668\u5e73\u53f0":143,"\u67b6\u6784\u7684iphone\u6216\u8005ipad\u7b49\u7269\u7406\u8bbe\u5907":143,"\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":137,"\u67d0\u4e9b\u53c2\u6570\u53ea\u53ef\u7528\u4e8e\u7279\u5b9a\u7684\u5c42\u4e2d":128,"\u67e5\u770b":98,"\u67e5\u770b\u5305\u7684\u5927\u5c0f":104,"\u67e5\u770b\u5f53\u524d\u72b6\u6001":98,"\u67e5\u770b\u5f53\u524d\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":98,"\u67e5\u770b\u6587\u4ef6\u5177\u4f53\u88ab\u4fee\u6539\u7684\u5185\u5bb9":98,"\u67e5\u770b\u662f\u5426\u662f\u5176\u4ed6\u9519\u8bef\u5f15\u53d1\u7684\u62a5\u9519":105,"\u67e5\u770bjob\u7684\u8be6\u7ec6\u60c5\u51b5":122,"\u67e5\u770blatest":88,"\u6807\u51c6":30,"\u6807\u51c6\u5dee\u4e3a":109,"\u6807\u51c6\u8868\u793apaddlepaddle\u7248\u672c\u53f7":88,"\u6807\u8bc6\u4e86\u4e00\u4e2a\u8f93\u51fa\u7684\u6587\u4ef6\u540d":133,"\u6807\u8bc6\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c\u6587\u4ef6":133,"\u6807\u8bc6\u662f\u5426\u4e3a\u8fde\u7eed\u7684batch\u8ba1\u7b97":129,"\u6807\u8bc6\u88ab\u6027\u80fd\u5206\u6790\u7684\u6e90\u6587\u4ef6":133,"\u6807\u8bc6http\u670d\u52a1\u7684\u7aef\u53e3":133,"\u6807\u8bc6http\u670d\u52a1\u7ed1\u5b9a\u7684ip":133,"\u6838\u4e00\u6837\u591a\u7684\u8fdb\u7a0b\u6765\u5e76\u884c\u7f16\u8bd1":26,"\u6838\u5fc3\u4ee3\u7801\u7f16\u8bd1\u6210\u94fe\u63a5\u5e93":113,"\u6839\u636e\u4e2a\u4eba\u7684\u9700\u6c42\u4fee\u6539\u5b9a\u5236docker\u5bb9\u5668\u6240\u6267\u884c\u7684\u811a\u672c":142,"\u6839\u636e\u4f60\u7684\u4efb\u52a1":131,"\u6839\u636e\u524d\u6587\u7684\u63cf\u8ff0":123,"\u6839\u636e\u7f51\u7edc\u914d\u7f6e\u4e2d\u7684":129,"\u6839\u636e\u8f93\u5165tensor\u7684\u5927\u5c0f\u6765\u8bbe\u7f6e\u8f93\u51fatensor\u7684\u5927\u5c0f":102,"\u6839\u636e\u8fd9\u4e9b\u53c2\u6570\u7684\u4f7f\u7528\u573a\u5408":128,"\u6839\u636e\u9ed8\u8ba4\u503c\u9012\u589e":129,"\u6839\u636e\u9ed8\u8ba4\u7aef\u53e3\u53f7\u9012\u589e":129,"\u6839\u636ejob\u5bf9\u5e94\u7684pod\u4fe1\u606f":122,"\u6839\u636eport":117,"\u683c\u5f0f":129,"\u683c\u5f0f\u4e0d\u5339\u914d\u65f6":68,"\u683c\u5f0f\u5b58\u50a8":115,"\u683c\u5f0f\u7684\u6587\u4ef6\u6765\u5b58\u653e":101,"\u6846\u67b6\u63d0\u4f9b\u7684blas\u51fd\u6570\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":143,"\u6846\u67b6\u8fdb\u884cblas\u77e9\u9635\u8ba1\u7b97":143,"\u68af\u5ea6\u4f1a\u5c31\u5730":100,"\u68af\u5ea6\u4f1a\u6709\u566a\u58f0":118,"\u68af\u5ea6\u53c2\u6570\u7684\u5206\u5757\u6570\u76ee":129,"\u68af\u5ea6\u5c31\u53ef\u4ee5\u901a\u8fc7\u8fd9\u4e2a\u65b9\u7a0b\u8ba1\u7b97\u5f97\u5230":100,"\u68af\u5ea6\u670d\u52a1\u5668\u7684\u6570\u91cf":129,"\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5\u901a\u8fc7\u6709\u9650\u5dee\u5206\u6cd5\u6765\u9a8c\u8bc1\u4e00\u4e2a\u5c42\u7684\u68af\u5ea6":100,"\u68af\u5ea6\u68c0\u67e5\u7684\u8f93\u5165\u6570\u636e\u7684\u6279\u6b21\u5927\u5c0f":100,"\u68c0\u67e5\u70b9\u4fdd\u5b58\u7a0b\u5e8f\u6d41\u7a0b":37,"\u68c0\u67e5\u8f93\u5165\u6570\u636e\u7ef4\u5ea6":101,"\u6982\u5ff5\u4e0a":116,"\u6982\u5ff5\u4e0a\u53ef\u4ee5\u5c06":115,"\u6a21\u5757\u4e0b\u7684\u76f8\u5173":102,"\u6a21\u578b\u4e00\u76f4\u4e0d\u6536\u655b":107,"\u6a21\u578b\u4e2d\u6240\u6709\u53ef\u5b66\u4e60\u53c2\u6570\u4f1a\u88ab\u5b58\u4e3a\u4e00\u4e2a\u538b\u7f29\u6587\u4ef6":116,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\u901a\u8fc7\u5b9a\u671f\u5411\u78c1\u76d8\u4e0a\u4fdd\u5b58\u4e00\u4efd\u5b58\u50a8\u5728paramet":37,"\u6a21\u578b\u6570\u636e\u68c0\u67e5\u70b9\u7684\u5b9e\u73b0":37,"\u6a21\u578b\u6587\u4ef6\u5c06\u88ab\u5199\u5165\u8282\u70b9":119,"\u6a21\u578b\u6765\u6307\u5bfc\u4f60\u5b8c\u6210\u8fd9\u4e9b\u6b65\u9aa4":140,"\u6a21\u578b\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":140,"\u6a21\u578b\u7684\u4ee3\u7801\u53ef\u4ee5\u5728":140,"\u6a21\u578b\u7684\u7f16\u7801\u5668\u90e8\u5206\u5982\u4e0b\u6240\u793a":140,"\u6a21\u578b\u8bad\u7ec3\u7b49\u4efb\u52a1":110,"\u6a21\u578b\u914d\u7f6e":[0,106],"\u6a21\u578b\u914d\u7f6e\u89e3\u6790":71,"\u6a21\u5f0f\u4e0b\u7684\u6027\u80fd\u6d4b\u8bd5\u662f\u6ca1\u6709\u610f\u4e49\u7684":133,"\u6a2a\u5411\u62fc\u63a5":107,"\u6b21\u8fed\u4ee3\u6267\u884c\u7684\u8f6c\u6362\u6b21\u6570\u4e3a":67,"\u6b22\u8fce\u901a\u8fc7":98,"\u6b63\u5728\u7b49\u5f85\u672a\u5b8c\u6210\u7684\u4efb\u52a1":104,"\u6b63\u5e38\u60c5\u51b5\u4e0b\u662f75m":104,"\u6b63\u786e\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":104,"\u6b63\u8d1f\u5bf9\u9a8c\u8bc1":128,"\u6b64\u547d\u4ee4\u5c06\u5728":142,"\u6b64\u5904\u90fd\u4e3a2":137,"\u6b64\u5916":[26,98,108,142],"\u6b64\u6559\u7a0b\u4f1a\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528python\u7684cprofile\u5305":133,"\u6b64\u6559\u7a0b\u5c06\u5411\u60a8\u5206\u6b65\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528\u5185\u7f6e\u7684\u5b9a\u65f6\u5de5\u5177":134,"\u6b64\u65b9\u6cd5\u4e0d\u80fd\u83b7\u53d6":107,"\u6b64\u65f6\u53ea\u9700\u8981":26,"\u6b64\u65f6\u53ef\u4ee5\u5728\u8c03\u7528infer\u63a5\u53e3\u65f6\u901a\u8fc7\u8bbe\u7f6e":107,"\u6b64\u65f6\u53ef\u4ee5\u8df3\u8fc7paddlepaddle\u6a21\u578b\u53c2\u6570\u6587\u4ef6\u7684\u5934\u4fe1\u606f":109,"\u6b64\u65f6\u6bcf\u4e2a\u5c0f\u5206\u652f\u7684":68,"\u6b64\u65f6master\u5c06\u8d1f\u8d23\u542f\u52a8\u4e00\u4e2a\u65b0\u7684train":37,"\u6b64\u76ee\u5f55":116,"\u6b64\u793a\u4f8b":116,"\u6b64\u7c7b\u62a5\u9519\u901a\u5e38\u662f\u7531\u4e8e\u67d0\u4e00\u4e2a\u8282\u70b9\u7684\u9519\u8bef\u5bfc\u81f4\u8fd9\u4e2a\u8282\u70b9\u7684\u8bad\u7ec3\u8fdb\u7a0b\u9000\u51fa":105,"\u6b65\u9aa4":107,"\u6bb5\u843d\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u53cc\u5c42\u7684\u5e8f\u5217":139,"\u6bb5\u843d\u662f\u7531\u53e5\u5b50\u6784\u6210\u7684\u5e8f\u5217":115,"\u6bcf\u4e00\u4e2a":[88,116],"\u6bcf\u4e00\u4e2a\u5916\u5c42\u5e8f\u5217\u53c8\u542b\u6709\u82e5\u5e72\u4e2a\u5185\u5c42\u5e8f\u5217":115,"\u6bcf\u4e00\u4e2a\u5e8f\u5217\u5728\u6574\u4e2a":115,"\u6bcf\u4e00\u4e2a\u6587\u4ef6\u662f\u6570\u636e\u96c6\u7684\u4e00\u4e2ashard":38,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65":137,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u95f4\u7684\u795e\u7ecf\u7f51\u7edc\u5177\u6709\u4e00\u5b9a\u7684\u76f8\u5173\u6027":137,"\u6bcf\u4e00\u4e2a\u8282\u70b9\u90fd\u6709\u76f8\u540c\u7684\u65e5\u5fd7\u7ed3\u6784":119,"\u6bcf\u4e00\u4e2a\u8f93\u5165":[115,116],"\u6bcf\u4e00\u4e2alayer\u8f93\u51fa\u77e9\u9635\u7684\u9ad8\u5ea6":107,"\u6bcf\u4e00\u5217\u7684\u542b\u4e49\u662f":133,"\u6bcf\u4e00\u7ec4\u5185\u7684\u6240\u6709\u53e5\u5b50\u548clabel":137,"\u6bcf\u4e00\u884c\u5143\u7d20\u5728":115,"\u6bcf\u4e2a":68,"\u6bcf\u4e2a\u503c\u7684\u7c7b\u578b\u53ef\u4ee5\u662f\u6574\u5f62":38,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u6bcf\u4e2a\u5355\u5c42rnn":139,"\u6bcf\u4e2a\u53c2\u6570\u670d\u52a1\u5668\u53ea\u4fdd\u5b58\u6574\u4e2a\u795e\u7ecf\u7f51\u7edc\u6240\u6709\u53c2\u6570\u7684\u4e00\u90e8\u5206":118,"\u6bcf\u4e2a\u53e5\u5b50\u53c8\u662f\u5355\u8bcd\u7684\u6570\u7ec4":137,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u5f00\u59cb\u6807\u8bb0\u5f00\u5934":140,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u7ed3\u675f\u6807\u8bb0\u7ed3\u5c3e":140,"\u6bcf\u4e2a\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u4e00\u81f4":137,"\u6bcf\u4e2a\u5c42\u5728\u5176":100,"\u6bcf\u4e2a\u6279\u6b21\u6570\u636e":129,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":139,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u90fd\u7528\u4e86\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u7ed3\u679c":137,"\u6bcf\u4e2a\u6743\u91cd\u5bf9\u5e94\u4e00\u4e2a\u8f93\u5165":100,"\u6bcf\u4e2a\u6837\u672c\u7531\u4e24\u90e8\u5206\u7ec4\u6210":137,"\u6bcf\u4e2a\u6837\u672c\u95f4\u7528\u7a7a\u884c\u5206\u5f00":137,"\u6bcf\u4e2a\u6d4b\u8bd5\u4f1a\u5bf9\u6bd4paddlepaddle\u4e2dcpu\u7b97\u51fa\u7684\u7ed3\u679c\u4e0emkl":68,"\u6bcf\u4e2a\u72b6\u6001":139,"\u6bcf\u4e2a\u7ebf\u7a0b":129,"\u6bcf\u4e2a\u7ebf\u7a0b\u5206\u914d\u5230128\u4e2a\u6837\u672c\u7528\u4e8e\u8bad\u7ec3":129,"\u6bcf\u4e2a\u8bad\u7ec3\u8282\u70b9\u5fc5\u987b\u6307\u5b9a\u4e00\u4e2a\u552f\u4e00\u7684id\u53f7":129,"\u6bcf\u4e2a\u8f93\u5165\u90fd\u662f\u4e00\u4e2a":100,"\u6bcf\u4e2a\u8f93\u51fa\u8282\u70b9\u90fd\u8fde\u63a5\u5230\u6240\u6709\u7684\u8f93\u5165\u8282\u70b9\u4e0a":100,"\u6bcf\u4e2a\u90e8\u5206\u5206\u522b\u7ed9\u6bcf\u4e2atrainer\u4f7f\u7528":118,"\u6bcf\u4e2acommit\u53ea\u505a\u4e86\u5c11\u91cf\u7684\u4fee\u6539":98,"\u6bcf\u4e2adata":38,"\u6bcf\u4e2amkldnnlayer\u90fd\u5305\u542b\u7528\u4e8e\u5185\u90e8\u5b58\u50a8\u548c\u5916\u90e8\u5b58\u50a8\u7684\u4e00\u7cfb\u5217mkldnnmatrix":68,"\u6bcf\u4e2aparamet":37,"\u6bcf\u4e2apod\u5305\u542b\u4e00\u4e2apaddlepaddle\u5bb9\u5668":123,"\u6bcf\u4e2ashard\u5206\u522b\u5b58\u50a8\u5728\u5176\u4e2d\u4e00\u53f0paramet":37,"\u6bcf\u4e2atrainer\u542f\u52a8\u540e\u8bfb\u53d6\u5207\u5206\u597d\u7684\u4e00\u90e8\u5206\u6570\u636e":118,"\u6bcf\u4e2atrainer\u7684\u552f\u4e00id":117,"\u6bcf\u4e2atrainer\u8fdb\u7a0b\u9700\u8981\u80fd\u591f\u8bfb\u53d6\u5c5e\u4e8e\u81ea\u5df1\u7684\u4e00\u4efd\u6570\u636e":117,"\u6bcf\u53f0\u670d\u52a1\u5668\u5177\u6709\u96c6\u7fa4\u4e2d\u552f\u4e00\u7684ip\u5730\u5740":127,"\u6bcf\u5c42\u4e0a\u53ea\u80fd\u4fdd\u5b58\u56fa\u5b9a\u6570\u76ee\u4e2a\u6700\u597d\u7684\u72b6\u6001":129,"\u6bcf\u5c42\u4f7f\u7528\u7684gpu\u53f7\u4f9d\u8d56\u4e8e\u53c2\u6570train":131,"\u6bcf\u6279\u6b21":129,"\u6bcf\u6b21\u63d0\u4ea4\u4ee3\u7801":98,"\u6bcf\u6b21\u63d0\u4ea4\u65f6":98,"\u6bcf\u6b21\u8c03\u7528\u65f6\u5bf9\u539f\u6570\u636e\u7684\u91cd\u590dpacking\u4fbf\u6210\u4e3a\u4e86\u5197\u4f59":67,"\u6bcf\u6b21\u8c03\u7528\u7684\u8017\u65f6\u4e5f\u5f88\u957f":133,"\u6bcf\u6b21\u8f93\u51fa\u4e00\u4e2adata":38,"\u6bcf\u884c\u8868\u793a\u4e00\u4e2a\u6279\u6b21\u4e2d\u7684\u5355\u4e2a\u8f93\u5165":100,"\u6bcf\u8f6e\u4f1a\u5c06\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u8bad\u7ec3\u6837\u672c\u4f7f\u7528\u4e00\u6b21":129,"\u6bcf\u8f6e\u7ed3\u675f\u65f6\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":129,"\u6bcf\u8f6e\u90fd\u4f1a\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c":129,"\u6bcf\u8fd0\u884c\u591a\u5c11\u4e2a\u6279\u6b21\u6267\u884c\u4e00\u6b21\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u68c0\u67e5":129,"\u6bcf\u969410\u5206\u949f":37,"\u6bcfdot":129,"\u6bcflog":129,"\u6bcfsave":129,"\u6bcftest":129,"\u6bd4\u5982":[26,28,38,68,98,105,107],"\u6bd4\u5982\u4e00\u53e5\u8bdd\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5355\u8bcd":137,"\u6bd4\u5982\u53ef\u80fd\u4f1a\u7528openmp\u6539\u8fdbsgd\u7684\u66f4\u65b0\u6027\u80fd":68,"\u6bd4\u5982\u5728":28,"\u6bd4\u5982\u5982\u679c\u8981build\u4e00\u4e2a\u4e0d\u4f9d\u8d56gpu":98,"\u6bd4\u5982\u5c06":88,"\u6bd4\u5982\u5e0c\u671b\u6700\u5c0f\u5316\u751f\u6210\u5e93\u7684\u5927\u5c0f":143,"\u6bd4\u5982\u5e0c\u671b\u6700\u5c0f\u5316\u751f\u6210\u7684\u5e93\u7684\u5927\u5c0f":[142,144],"\u6bd4\u5982\u6bcf\u969410\u5206\u949f\u6700\u65b0\u7684\u5feb\u7167":37,"\u6bd4\u5982\u6d41\u5f0f\u6570\u636e\u5904\u7406":38,"\u6bd4\u5982\u8282\u70b9\u7684id":120,"\u6bd4\u5982\u8bbe\u7f6e\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u53c2\u6570\u521d\u59cb\u5316\u65b9\u5f0f\u548cbias\u521d\u59cb\u5316\u65b9\u5f0f":109,"\u6bd4\u5982cento":30,"\u6bd4\u5982fpe":105,"\u6bd4\u5982ide\u914d\u7f6e\u91cc":98,"\u6bd4\u5982imagenet\u8fd9\u4e2a\u6570\u636e\u96c6\u53ef\u80fd\u88ab\u5206\u62101000\u4e2ashard":38,"\u6bd4\u5982pil\u5e93\u7b49":117,"\u6bd5\u7adf\u5355\u7ebf\u7a0b\u8c03\u8bd5\u66f4\u5bb9\u6613":133,"\u6ca1\u6709\u5fc5\u8981\u5728\u6bcf\u6b21\u524d\u5411\u4e2d\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8ba1\u7b97\u65f6\u5bf9\u6743\u91cd\u8fdb\u884c\u91cd\u590d\u7684packing\u64cd\u4f5c":67,"\u6ca1\u6709\u627e\u5230\u548c\u5f53\u524d\u7cfb\u7edf\u5339\u914d\u7684paddlepaddle\u5b89\u88c5\u5305":[30,104],"\u6ca1\u6709\u8bbe\u7f6e":[142,144],"\u6ce8":[28,37,88,115],"\u6ce8\u518c":101,"\u6ce8\u518ccpu":101,"\u6ce8\u518clayer\u7684\u65f6\u5019\u4fdd\u8bc1":[67,68],"\u6ce8\u518cop":101,"\u6ce8\u518cop\u65f6\u7684\u7c7b\u578b\u540d":101,"\u6ce8\u610f":[27,68,100,103,107,110,116,117,123,133,140,142,143,144],"\u6ce8\u610f\u4e0a\u8ff0\u547d\u4ee4\u4e2d":123,"\u6ce8\u610f\u4e8b\u9879":115,"\u6ce8\u610f\u5230\u6211\u4eec\u5df2\u7ecf\u5047\u8bbe\u673a\u5668\u4e0a\u67094\u4e2agpu":131,"\u6ce8\u610fnode":123,"\u6ce8\u91ca":101,"\u6d41\u7a0b\u6765\u63d0\u4ea4\u4ee3\u7801":98,"\u6d4b\u8bd5":98,"\u6d4b\u8bd5\u5206\u4e3a\u6bcf\u4e2alayer":68,"\u6d4b\u8bd5\u65f6\u6307\u5b9a\u7684\u5b58\u50a8\u6a21\u578b\u5217\u8868\u7684\u6587\u4ef6":129,"\u6d4b\u8bd5\u662f":98,"\u6d4b\u8bd5\u672c\u6b21release\u7684\u6b63\u786e\u6027":88,"\u6d4b\u8bd5\u7684\u6027\u80fd\u5bf9\u6bd4\u7ed3\u679c\u4f1a\u5728":68,"\u6d4b\u8bd5\u7684\u6a21\u578b\u5305\u62ec\u4ece\u7b2cm\u8f6e\u5230\u7b2cn":131,"\u6d4b\u8bd5model_list":128,"\u6d4b\u8bd5oper":101,"\u6d4b\u8bd5save_dir":128,"\u6d6e\u70b9\u578b\u6570\u636e":38,"\u6d6e\u70b9\u578b\u7a00\u758f\u77e9\u9635":115,"\u6d6e\u70b9\u578b\u7a20\u5bc6\u77e9\u9635":115,"\u6d6e\u70b9\u5f02\u5e38\u901a\u5e38\u7684\u539f\u56e0\u662f\u6d6e\u70b9\u6570\u6ea2\u51fa":107,"\u6d6e\u70b9\u6570":115,"\u6d6e\u70b9\u6570\u5411\u91cf\u7b49":115,"\u6d6e\u70b9\u7a00\u758f\u6570\u636e":100,"\u6df1\u5165paddlepaddl":68,"\u6df7\u5408\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790\u6765\u8fdb\u884c\u8c03\u4f18":133,"\u6df7\u5408\u4ee3\u7801\u7684\u6027\u80fd\u74f6\u9888\u4e5f\u662f\u8981\u770b":133,"\u6df7\u5408\u5f53\u524d\u8bcd\u5411\u91cf\u548cattention\u52a0\u6743\u7f16\u7801\u5411\u91cf":140,"\u6dfb\u52a0":67,"\u6dfb\u52a0\u4e86\u4e00\u4e2a\u8f93\u51fa":101,"\u6dfb\u52a0\u542f\u52a8\u811a\u672c":123,"\u6dfb\u52a0\u5e8f\u5217\u4fe1\u606f":115,"\u6dfb\u52a0\u7684\u76f8\u5173\u6587\u4ef6\u548c\u76ee\u5f55\u7ed3\u6784\u5982\u4e0b":[67,68],"\u6dfb\u52a0\u8f93\u5165\u53c2\u6570":101,"\u6dfb\u52a0\u8f93\u51fa\u53c2\u6570":101,"\u6dfb\u52a0op\u7684\u6ce8\u91ca":101,"\u6e05\u7406":26,"\u6e05\u7406\u548c\u7ed3\u675f":116,"\u6e05\u7406\u6389\u8001\u65e7\u7684paddlepaddle\u5b89\u88c5\u5305":104,"\u6e90\u4ee3\u7801\u683c\u5f0f":98,"\u6e90\u5e8f\u5217":140,"\u6e90\u7801\u6811\u6839\u76ee\u5f55":26,"\u6f5c\u5728\u4f1a\u5f15\u8d77\u672a\u5b9a\u4e49\u884c\u4e3a":115,"\u6fc0\u6d3b":100,"\u6fc0\u6d3b\u51fd\u6570\u662f\u72ec\u7acb\u4e8e":68,"\u6fc0\u6d3b\u65b9\u7a0b":100,"\u6fc0\u6d3b\u7684\u7c7b\u578b":100,"\u70b9\u51fb":[30,88],"\u70b9\u51fb\u8fd9\u91cc":103,"\u7136\u540e":[119,134],"\u7136\u540e\u4e0b\u8f7d\u4f18\u5316\u66f4\u65b0\u540e\u7684\u795e\u7ecf\u7f51\u7edc\u53c2\u6570":118,"\u7136\u540e\u4ea4\u7ed9step\u51fd\u6570":139,"\u7136\u540e\u4f7f\u7528":143,"\u7136\u540e\u4f7f\u7528resize\u63a5\u53e3\u8bbe\u7f6etensor\u7684\u5927\u5c0f":102,"\u7136\u540e\u5355\u51fb":98,"\u7136\u540e\u53ef\u4ee5\u4ecehead\u8282\u70b9ssh\u65e0\u5bc6\u7801\u767b\u5f55\u5230openmpi\u7684\u6bcf\u4e2a\u8282\u70b9\u4e0a":124,"\u7136\u540e\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":123,"\u7136\u540e\u5728\u4e0b\u4e00\u4e2a\u65f6\u95f4\u6b65\u8f93\u5165\u7ed9\u53e6\u4e00\u4e2a\u795e\u7ecf\u5143":137,"\u7136\u540e\u5728\u524d\u5411":67,"\u7136\u540e\u5728\u6d4f\u89c8\u5668\u4e2d\u8f93\u5165\u4ee5\u4e0b\u7f51\u5740":28,"\u7136\u540e\u5728dataprovider\u91cc\u9762\u6839\u636e\u8be5\u5730\u5740\u52a0\u8f7d\u5b57\u5178":109,"\u7136\u540e\u5728etcd\u7684":37,"\u7136\u540e\u5b89\u88c5paddle\u7684python\u73af\u5883":104,"\u7136\u540e\u5b9a\u4e49":140,"\u7136\u540e\u5c06\u6784\u5efa\u6210\u529f\u7684\u955c\u50cf\u4e0a\u4f20\u5230\u955c\u50cf\u4ed3\u5e93":123,"\u7136\u540e\u5c06\u8fd9\u4e9blayer\u7684\u53c2\u6570":108,"\u7136\u540e\u5c31\u53ef\u4ee5\u5e76\u53d1\u5199\u5165\u591a\u4e2achunk":54,"\u7136\u540e\u6240\u6709\u7528":98,"\u7136\u540e\u624d\u80fd\u4f7f\u7528pfsclient":54,"\u7136\u540e\u6253\u5370\u8f93\u51fa":110,"\u7136\u540e\u6309\u7167\u4e0a\u8ff0\u7684\u65b9\u6cd5":88,"\u7136\u540e\u63d0\u4ea4\u65b0\u6dfb\u52a0\u7684":98,"\u7136\u540e\u70b9\u51fb":[88,98],"\u7136\u540e\u7533\u660e\u4e00\u4e2a\u5b58\u50a8\u5377":123,"\u7136\u540e\u89c2\u5bdf\u5230\u8f93\u51fa\u7684\u53d8\u5316\u4e3a":100,"\u7136\u540e\u901a\u8fc7\u51fd\u6570":123,"\u7136\u540e\u901a\u8fc7\u81ea\u8eab\u7684ip\u5730\u5740\u5728":123,"\u7136\u540e\u91cd\u65b0cmake\u5373\u53ef":104,"\u7136\u800c":[129,140],"\u7248\u672c":[26,27,30,143],"\u7248\u672c\u4e3acpu_avx_mkl":28,"\u7248\u672c\u4e3acpu_avx_openbla":[30,112],"\u7248\u672c\u5206\u652f":88,"\u7248\u672c\u53f7":88,"\u7248\u672c\u53f7\u5bf9\u5e94\u7684tag\u5373\u53ef":88,"\u7248\u672c\u53f7rc":88,"\u7248\u672c\u5728":98,"\u7248\u672c\u8bf4\u660e":30,"\u7248\u672cfork\u51fa\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f":88,"\u7279\u522b\u662f\u5728lstm\u7b49rnn\u4e2d":107,"\u7279\u6307":116,"\u7279\u6709\u7684\u8bbe\u5907id":68,"\u72ec\u7acb\u5b9a\u5236\u7684\u4e8c\u8fdb\u5236\u65f6\u624d\u9700\u8981\u7f16\u8bd1":29,"\u72ec\u7acb\u5de5\u5177\u94fe":142,"\u72ec\u7acb\u5de5\u5177\u94fe\u6240\u5728\u7684\u7edd\u5bf9\u8def\u5f84":142,"\u73af\u5883\u51c6\u5907":118,"\u73af\u5883\u53d8\u91cf":117,"\u73af\u5883\u53d8\u91cf\u4e2d":113,"\u73af\u5883\u53d8\u91cf\u6765\u6307\u5b9a\u7279\u5b9a\u7684gpu":107,"\u73b0\u9636\u6bb5\u7684\u4f18\u5316\u4e3b\u8981\u9488\u5bf9":67,"\u73b0\u9636\u6bb5paddle\u6709\u4e00\u4e2a\u95ee\u9898\u662f":71,"\u7406\u89e3":26,"\u7406\u89e3\u4e3a\u4e00\u4e2a\u4e00\u7ef4\u7684\u6574\u578b\u6570\u7ec4":115,"\u751a\u81f3\u80fd\u89e3\u91ca\u4e3a\u4ec0\u4e48\u67d0\u4e2a\u64cd\u4f5c\u82b1\u4e86\u5f88\u957f\u65f6\u95f4":134,"\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u8bad\u7ec3\u6570\u636e\u96c6\u901a\u5e38\u4f53\u79ef\u5f88\u5927":38,"\u751f\u4ea7\u73af\u5883\u7684\u65e5\u5fd7\u6570\u636e\u4f1a\u901a\u8fc7\u5b9e\u65f6\u6d41\u7684\u65b9\u5f0f":38,"\u751f\u4ea7\u955c\u50cf":98,"\u751f\u6210":123,"\u751f\u6210\u5404\u79cd\u8bed\u8a00\u7684\u7ed1\u5b9a\u4ee3\u7801":71,"\u751f\u6210\u540e\u7684\u6587\u6863\u5206\u522b\u5b58\u50a8\u5728\u7f16\u8bd1\u76ee\u5f55\u7684":103,"\u751f\u6210\u5e8f\u5217\u7684\u6700\u5927\u957f\u5ea6":140,"\u751f\u6210\u6587\u6863":71,"\u751f\u6210\u7684":38,"\u751f\u6210\u7684\u6027\u80fd\u5206\u6790\u6587\u4ef6\u4e3a":133,"\u751f\u6210\u7684\u6570\u636e\u5c06\u4f1a\u5b58\u50a8\u5728\u8fd9\u4e2avolume\u4e0b":123,"\u751f\u6210\u7684\u6570\u636e\u7f13\u5b58\u5728\u5185\u5b58\u91cc":107,"\u751f\u6210\u7ed9\u5b9a":38,"\u751f\u6210\u7f51\u7edc\u5c42\u914d\u7f6e":100,"\u751f\u6210\u81ea\u5df1\u76ee\u5f55\u4e0b\u7684\u4ed3\u5e93":98,"\u751f\u6210\u8c03\u8bd5\u4fe1\u606f":133,"\u751f\u6210\u968f\u673a\u7684\u8f93\u5165\u6570\u636e":101,"\u751f\u6210api\u6587\u6863":71,"\u751f\u6210pfsclient\u548cpfsserver\u7684\u6846\u67b6\u90e8\u5206":54,"\u751f\u6210python\u6027\u80fd\u5206\u6790\u7684\u547d\u4ee4\u5982\u4e0b":133,"\u7528":54,"\u7528\u4e8e\u521d\u59cb\u5316\u53c2\u6570\u548c\u8bbe\u7f6e":100,"\u7528\u4e8e\u5c06\u53c2\u6570\u4f20\u9012\u7ed9\u7f51\u7edc\u914d\u7f6e":131,"\u7528\u4e8e\u6307\u5b9a\u5176\u8981\u5173\u8054\u7684layer":108,"\u7528\u4e8e\u6307\u5b9a\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":129,"\u7528\u4e8e\u6ce8\u518c\u6ca1\u6709\u53cd\u5411\u7684op":101,"\u7528\u4e8e\u6d4b\u8bd5\u548c\u5bf9\u6bd4\u5728\u4f7f\u7528mkl":68,"\u7528\u4e8e\u7a00\u758f\u7c7b\u578b\u53c2\u6570\u901a\u4fe1\u7684\u7aef\u53e3\u4e2a\u6570":117,"\u7528\u4e8e\u7a00\u758f\u8bad\u7ec3\u4e2d":129,"\u7528\u4e8e\u7ba1\u7406mkl":68,"\u7528\u4e8e\u83b7\u53d6\u7279\u5b9alayer\u4e0a\u4e00\u65f6\u95f4\u6b65\u7684\u8f93\u51fa":108,"\u7528\u4e8e\u8ba1\u7b97\u7f16\u7801\u5411\u91cf\u7684\u52a0\u6743\u548c":140,"\u7528\u4e8e\u8bad\u7ec3\u795e\u7ecf\u7f51\u7edc\u7684\u6570\u636e":118,"\u7528\u4e8e\u9009\u62e9\u662f\u5426\u4f7f\u7528\u76f8\u5173\u529f\u80fd":67,"\u7528\u4e8e\u9009\u62e9\u662f\u5426\u4f7f\u7528mkl":68,"\u7528\u4e8emkl":[67,68],"\u7528\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7f16\u7801":140,"\u7528\u591a\u5bf9\u6548\u679c\u5b8c\u5168\u76f8\u540c\u7684":137,"\u7528\u6237\u4e00\u822c\u901a\u8fc7\u8c03\u7528":133,"\u7528\u6237\u4e0a\u4f20\u6570\u636e\u540e":38,"\u7528\u6237\u4e5f\u53ef\u4ee5\u4e0a\u4f20label":38,"\u7528\u6237\u4e5f\u53ef\u4ee5\u4f7f\u7528paddlepaddle\u63d0\u4f9b\u7684\u5b98\u65b9\u5f00\u53d1\u955c\u50cf":142,"\u7528\u6237\u4eceapp":143,"\u7528\u6237\u53ea\u9700\u5b9a\u4e49rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":139,"\u7528\u6237\u53ef\u4ee5\u5206\u522b\u67e5\u770b\u6700\u65b0\u7684":103,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003\u4e0b\u6587":142,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003sphinx\u6559\u7a0b\u8fdb\u884c\u4e66\u5199":103,"\u7528\u6237\u53ef\u4ee5\u5b89\u5168\u7684\u91ca\u653e\u67d0\u4e2ac":72,"\u7528\u6237\u53ef\u4ee5\u628a\u81ea\u5df1\u7684\u6570\u636e\u5206\u4eab\u7ed9\u522b\u4eba":38,"\u7528\u6237\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u8fd9\u4e2a\u52a8\u6001\u5e93\u6765\u5f15\u5165paddl":72,"\u7528\u6237\u53ef\u4ee5\u81ea\u5b9a\u4e49beam":129,"\u7528\u6237\u53ef\u4ee5\u8bbe\u7f6e":131,"\u7528\u6237\u53ef\u5728\u81ea\u5df1\u719f\u6089\u7684\u5f00\u53d1\u5e73\u53f0\u4e0a\u7f16\u8bd1android\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":142,"\u7528\u6237\u53ef\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e\u5b83\u4eec":27,"\u7528\u6237\u53ef\u5c06":142,"\u7528\u6237\u53ef\u5c06\u5408\u6210\u7684fat\u5e93\u7528\u4e8e\u6df1\u5ea6\u5b66\u4e60\u76f8\u5173\u7684io":143,"\u7528\u6237\u53ef\u6839\u636e\u81ea\u5df1\u7684\u7f16\u8bd1\u76ee\u6807\u67b6\u6784":142,"\u7528\u6237\u53ef\u81ea\u884c\u524d\u5f80\u4e0b\u8f7d\u9884\u7f16\u8bd1\u597d\u7684\u7248\u672c":142,"\u7528\u6237\u53ef\u901a\u8fc7\u5982\u4e0b\u4e24\u79cd\u65b9\u5f0f":142,"\u7528\u6237\u5728\u4f7f\u7528\u8fd9\u4e00\u7c7brecurr":108,"\u7528\u6237\u5728\u4f7f\u7528paddlepaddl":104,"\u7528\u6237\u5728\u672c\u5730\u8f6c\u6362\u597d\u518d\u4e0a\u4f20":38,"\u7528\u6237\u5c06\u53c2\u6570\u8f7d\u5165":109,"\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":123,"\u7528\u6237\u5f3a\u5236\u6307\u5b9a\u7279\u5b9a\u7684python\u7248\u672c":104,"\u7528\u6237\u6587\u4ef6\u53ef\u80fd\u662f\u6bd4\u8f83\u5927\u7684":54,"\u7528\u6237\u8fd8\u53ef\u6839\u636e\u81ea\u5df1\u7684\u9700\u6c42\u8bbe\u7f6e\u5176\u4ed6\u7f16\u8bd1\u53c2\u6570":[142,143,144],"\u7528\u6237\u901a\u8fc7\u53c2\u6570":[108,109],"\u7528\u6237\u901a\u8fc7c":72,"\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":131,"\u7528\u6237\u9700\u8981\u5728cmake\u65f6\u624b\u52a8\u8bbe\u7f6e\u8fd9\u4e9b\u503c":[142,144],"\u7528\u6237\u9700\u8981\u6307\u5b9a\u672c\u673a\u4e0apython\u7684\u8def\u5f84":104,"\u7528\u6237\u9700\u8981\u63d0\u524d\u51c6\u5907\u597d\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883":142,"\u7528\u672c\u673a\u7684\u7b2c\u4e00\u4e2a":26,"\u7528\u6765\u4ece\u53c2\u6570\u670d\u52a1\u5668\u9884\u53d6\u53c2\u6570\u77e9\u9635\u76f8\u5e94\u7684\u884c":100,"\u7528\u6765\u5b58\u50a8\u672c\u6b21\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c":133,"\u7528\u8fd9\u4e2a\u955c\u50cf\u521b\u5efa\u7684\u5bb9\u5668\u9700\u8981\u6709\u4ee5\u4e0b\u4e24\u4e2a\u529f\u80fd":123,"\u7528docker\u7f16\u8bd1\u548c\u6d4b\u8bd5paddlepaddl":29,"\u7528web\u6d4f\u89c8\u5668\u8bbf\u95ee\u5bf9\u5e94\u7f51\u5740":133,"\u7531":[108,115,139],"\u7531\u4e8e":113,"\u7531\u4e8e\u5728\u73b0\u6709\u7684\u67d0\u4e9b\u60c5\u51b5\u4e0b":67,"\u7531\u4e8e\u5b83\u5185\u90e8\u5305\u542b\u4e86\u6bcf\u7ec4\u6570\u636e\u4e2d\u7684\u6240\u6709\u53e5\u5b50":137,"\u7531\u4e8e\u5bf9parameters\u7684\u66f4\u65b0\u9700\u8981\u83b7\u53d6parameters\u5185\u5b58\u7684":37,"\u7531\u4e8e\u6211\u4eec\u60f3\u8981\u7684\u53d8\u6362\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":137,"\u7531\u4e8e\u6211\u4eec\u652f\u6301\u8bad\u7ec3\u6570\u636e\u6709\u4e0d\u540c\u7684\u6279\u6b21\u5927\u5c0f":100,"\u7531\u4e8e\u96c6\u7fa4\u4e2d\u540c\u65f6\u5b58\u5728\u4e24\u53f0\u673a\u5668\u6545\u969c\u7684\u6982\u7387\u6781\u4f4e":37,"\u7531\u4e8earm64\u67b6\u6784\u8981\u6c42android":142,"\u7531\u4e8ec":71,"\u7531\u4e8echunk\u6bd4\u8f83\u5c0f":54,"\u7531\u4e8eeigen":102,"\u7531\u4e8emkl":68,"\u7531\u4e8epypi":88,"\u7531\u4e8estep":139,"\u7531\u4e8etensor\u7684rank\u662f\u6a21\u677f\u53c2\u6570":102,"\u7531\u5206\u652f\u5904\u7684layer\u8d1f\u8d23\u6c42\u548c":68,"\u7531\u8bcd\u8bed\u6784\u6210\u7684\u53e5\u5b50":136,"\u7531\u94fe\u63a5\u65b9\u5f0f\u51b3\u5b9a":113,"\u7533\u8bf7\u7528\u6237\u7a7a\u95f4":54,"\u767b\u5f55\u5230head\u8282\u70b9":124,"\u7684":[26,30,98,102,115,116,119,123,143],"\u7684\u4e00\u4e2a\u5b50\u96c6":68,"\u7684\u4e00\u4e2a\u7b80\u5355\u8c03\u7528\u5982\u4e0b":139,"\u7684\u4e3a0":129,"\u7684\u4efb\u4e00\u4e00\u79cd":107,"\u7684\u4f5c\u7528\u662f\u5ef6\u8fdf\u5206\u914d\u5185\u5b58":102,"\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":136,"\u7684\u4fe1\u606f":68,"\u7684\u503c":[142,143,144],"\u7684\u503c\u81ea\u52a8\u63a8\u5bfc\u5f97\u5230":142,"\u7684\u504f\u7f6e\u5411\u91cf":100,"\u7684\u5171\u4eab\u5df2\u7ecf\u52a0\u8f7d\u7684\u9884\u6d4b\u6a21\u578b":116,"\u7684\u5177\u4f53\u8ba1\u7b97\u903b\u8f91":101,"\u7684\u5185\u5b58":107,"\u7684\u5185\u5bb9\u6765\u5b9a\u5236imag":123,"\u7684\u5185\u6838block\u4f7f\u7528\u60c5\u51b5":134,"\u7684\u522b\u540d":[1,2,4,16,20,23],"\u7684\u5355\u5143\u6d4b\u8bd5":101,"\u7684\u5355\u5143\u6d4b\u8bd5\u548c\u7b80\u5355\u7f51\u7edc\u7684\u6574\u4f53\u6d4b\u8bd5":68,"\u7684\u53cd\u5411\u4f20\u64ad\u5c06\u4f1a\u6253\u5370\u65e5\u5fd7\u4fe1\u606f":129,"\u7684\u53d8\u6362\u77e9\u9635":100,"\u7684\u540d\u79f0\u76f8\u540c":140,"\u7684\u5411\u91cf":100,"\u7684\u542f\u52a8\u53c2\u6570":123,"\u7684\u542f\u52a8\u53c2\u6570\u5e76\u6267\u884c\u8fdb\u7a0b":123,"\u7684\u547d\u4ee4\u548c\u4e00\u822c\u7684":133,"\u7684\u547d\u540d\u98ce\u683c\u5e76\u4e0d\u80fd\u9002\u5e94\u5176\u4ed6\u7b2c\u4e09\u65b9\u8bed\u8a00":71,"\u7684\u5730\u65b9":98,"\u7684\u5747\u5300\u5206\u5e03":109,"\u7684\u57fa\u672c\u903b\u8f91":68,"\u7684\u591a\u79cd\u5b89\u88c5\u65b9\u5f0f":127,"\u7684\u5934\u6587\u4ef6":71,"\u7684\u5b50\u7c7b\u53ea\u9700\u8981\u4f7f\u7528\u5185\u90e8\u5b58\u50a8\u5c31\u53ef\u4ee5\u4e86":68,"\u7684\u5b9e\u73b0":101,"\u7684\u5de5\u4f5c\u6d41\u7a0b\u5982\u56fe1\u6240\u793a":116,"\u7684\u5e73\u5747\u503c":136,"\u7684\u5e8f\u5217":115,"\u7684\u5e8f\u5217\u5f62\u72b6\u4e00\u81f4":137,"\u7684\u5f00\u53d1\u5de5\u4f5c\u90fd\u5e94\u8be5\u5728\u4e00\u4e2a\u65b0\u7684\u5206\u652f\u4e0a\u5b8c\u6210":98,"\u7684\u5f00\u53d1\u6d41\u7a0b":26,"\u7684\u5f00\u59cb\u8bf7\u52a0\u4e0a\u5b8f\u5b9a\u4e49":101,"\u7684\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18\u5206\u4e3a\u4e24\u4e2a\u90e8\u5206":133,"\u7684\u6027\u80fd\u5206\u6790\u5de5\u5177\u975e\u5e38\u591a":133,"\u7684\u6027\u80fd\u6709\u95ee\u9898":133,"\u7684\u60c5\u51b5\u4e0b":67,"\u7684\u63a5\u53e3\u6837\u5f0f":71,"\u7684\u63a5\u53e3\u8bf7\u67e5\u770b":115,"\u7684\u63cf\u8ff0\u8bf4\u660e\u4e2d":98,"\u7684\u64cd\u4f5c":102,"\u7684\u6570\u636e\u6d41\u56fe":38,"\u7684\u6570\u76ee\u4e00\u81f4":136,"\u7684\u6587\u4ef6\u4e5f\u5e26\u5230\u65b0\u5206\u652f\u4e0a":98,"\u7684\u65b9\u7a0b":100,"\u7684\u65f6\u5019":68,"\u7684\u65f6\u95f4\u6b65\u4fe1\u606f\u6210\u6b63\u6bd4":107,"\u7684\u66f4\u8be6\u7ec6\u51c6\u786e\u7684\u5b9a\u4e49":137,"\u7684\u6700\u5c0f\u503c":129,"\u7684\u6700\u65b0\u4ee3\u7801\u5e76\u66f4\u65b0\u5f53\u524d\u5206\u652f":98,"\u7684\u6784\u9020\u51fd\u6570":101,"\u7684\u67b6\u6784\u7684\u793a\u4f8b":140,"\u7684\u6837\u5f0f":98,"\u7684\u6838\u5fc3\u662f\u8bbe\u8ba1step\u51fd\u6570\u7684\u8ba1\u7b97\u903b\u8f91":139,"\u7684\u6839\u76ee\u5f55":143,"\u7684\u683c\u5f0f\u59cb\u7ec8\u662f":68,"\u7684\u683c\u5f0f\u5b58\u50a8":68,"\u7684\u6982\u5ff5":68,"\u7684\u6bb5\u843d\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":139,"\u7684\u6bcf\u4e2a\u8fdb\u7a0b\u90fd\u53ef\u4ee5\u4ececeph\u8bfb\u53d6\u6570\u636e":122,"\u7684\u6e90\u7801\u4ee5\u53ca\u751f\u6210\u6587\u6863\u9700\u8981\u591a\u79cd\u5f00\u53d1\u5de5\u5177":98,"\u7684\u6e90\u7801\u91cc\u4f7f\u7528\u4e86":71,"\u7684\u7248\u672c":[26,88,144],"\u7684\u72b6\u6001":139,"\u7684\u72ec\u7acb\u5de5\u5177\u94fe":142,"\u7684\u77e9\u9635":[100,107],"\u7684\u7a20\u5bc6\u5411\u91cf\u4f5c\u4e3a\u8f93\u5165":100,"\u7684\u7a20\u5bc6\u77e9\u9635":115,"\u7684\u7a20\u5bc6\u77e9\u9635\u662f\u4e00\u4e2a\u7531":115,"\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570":116,"\u7684\u7b2ci\u4e2a\u503c":100,"\u7684\u7b2cj\u4e2a\u503c":100,"\u7684\u7cfb\u7edf":26,"\u7684\u7ed3\u679c":67,"\u7684\u7f16\u5199":117,"\u7684\u7f16\u8bd1\u5de5\u5177\u94fe":142,"\u7684\u7f29\u5199":54,"\u7684\u7f51\u7edc\u6a21\u578b":67,"\u7684\u89c4\u8303":71,"\u7684\u89d2\u5ea6":38,"\u7684\u8ba1\u7b97\u4ee3\u7801":102,"\u7684\u8ba1\u7b97\u8fc7\u7a0b\u4e66\u5199\u66f4\u52a0\u7b80\u5355":101,"\u7684\u8bdd":107,"\u7684\u8be6\u7ec6\u4fe1\u606f":133,"\u7684\u8f93\u5165":139,"\u7684\u8f93\u51fa":[107,134],"\u7684\u8f93\u51fa\u4fe1\u606f\u5165\u624b\u662f\u4e2a\u4e0d\u9519\u7684\u9009\u62e9":134,"\u7684\u8f93\u51fa\u51fd\u6570\u8fd4\u56de\u7684\u662f\u4e0b\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa\u8bcd\u7684":140,"\u7684\u8f93\u51fa\u683c\u5f0f":137,"\u7684\u8f93\u51fa\u88ab\u7528\u4f5c":140,"\u7684\u8f93\u51fab\u662f\u4e00\u4e2a":107,"\u7684\u8fd0\u884c\u73af\u5883":26,"\u7684\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":98,"\u7684\u914d\u7f6e\u5199\u5230\u914d\u7f6e\u6587\u4ef6\u4e2d":38,"\u7684\u96c6\u88c5\u7bb1\u6280\u672f":26,"\u7684\u9875\u9762\u5220\u9664\u8fdc\u7a0b\u4ed3\u5e93\u7684\u5206\u652f":98,"\u7684cpu":101,"\u7684docker\u955c\u50cf":28,"\u7684flag":[67,68],"\u7684linux\u670d\u52a1\u5668\u7ec4\u6210":127,"\u7684paddlepaddle\u5e93":142,"\u7684vanilla":67,"\u76d1\u542c\u7684\u7aef\u53e3\u4e2a\u6570":117,"\u76ee\u524d":139,"\u76ee\u524d\u4f7f\u7528":98,"\u76ee\u524d\u53ea\u8003\u8651":68,"\u76ee\u524d\u53ea\u8003\u8651\u52a8\u6001\u6269\u5bb9trainer\u6570\u91cf":37,"\u76ee\u524d\u5728paddlepaddle\u4e2d":68,"\u76ee\u524d\u5728paddlepaddle\u4e2d\u6570\u636e\u90fd\u662f\u4ee5":68,"\u76ee\u524d\u5d4c\u5165python\u89e3\u91ca\u5668":71,"\u76ee\u524d\u5fc5\u987b\u8bbe\u7f6e\u6210":144,"\u76ee\u524d\u6211\u4eec\u7528cephfs\u6765\u642d\u5efa":54,"\u76ee\u524d\u63d0\u4f9b\u4e09\u79cd\u94fe\u63a5\u65b9\u5f0f":113,"\u76ee\u524d\u652f\u6301":142,"\u76ee\u524d\u652f\u6301\u4e24\u79cd":136,"\u76ee\u524d\u652f\u6301cento":112,"\u76ee\u524d\u652f\u6301fail":129,"\u76ee\u524d\u7684\u4f18\u5316":68,"\u76ee\u524d\u8be5\u53c2\u6570\u4ec5\u7528\u4e8eaucvalidationlayer\u548cpnpairvalidationlayer\u5c42":129,"\u76ee\u524d\u8fd8\u672a\u652f\u6301":139,"\u76ee\u524dpaddle\u7684\u8fdb\u7a0b\u6a21\u578b\u662fc":71,"\u76ee\u524dpaddlepaddle\u7684develop\u5206\u652f\u7684\u6587\u6863\u662f\u81ea\u52a8\u89e6\u53d1\u66f4\u65b0\u7684":103,"\u76ee\u524dpaddlepaddle\u91c7\u7528\u4e86":67,"\u76ee\u5f55":[26,28,119,122,123,142,143,144],"\u76ee\u5f55\u4e0b":[72,100,119],"\u76ee\u5f55\u4e0b\u5bf9\u5e94\u7684\u5730\u65b9":68,"\u76ee\u5f55\u4e0b\u65b0\u589e\u7684":101,"\u76ee\u5f55\u4e0b\u6700\u65b0\u7684":143,"\u76ee\u5f55\u4e0b\u7684\u4ee3\u7801\u793a\u4f8b":116,"\u76ee\u5f55\u4e0b\u7684\u751f\u6210\u6587\u4ef6\u7528\u4e8e\u6df1\u5ea6\u5b66\u4e60\u76f8\u5173android":142,"\u76ee\u5f55\u4e0b\u7684python\u5305":104,"\u76ee\u5f55\u4e2d":[113,116,119],"\u76ee\u5f55\u4e2d\u4f1a\u5305\u542b":[142,144],"\u76ee\u5f55\u4e2d\u4f1a\u5305\u542b\u4ee5\u4e0b\u5185\u5bb9":143,"\u76ee\u5f55\u4e2d\u7684":134,"\u76ee\u5f55\u4e2dpaddl":123,"\u76ee\u5f55\u548c":[142,143,144],"\u76ee\u5f55\u5c31\u6210\u4e3a\u4e86\u5171\u4eab\u5b58\u50a8":123,"\u76ee\u5f55\u751f\u6210\u4e00\u5957\u72ec\u7acb\u7f16\u8bd1\u5de5\u5177\u94fe":142,"\u76ee\u5f55\u91cc\u627e\u5230\u4ea4\u53c9\u7f16\u8bd1\u5668":144,"\u76ee\u6807\u5411\u91cf":140,"\u76ee\u6807\u5de5\u5177\u94fe":142,"\u76ee\u6807\u673a\u7248protobuf\u5e93":144,"\u76ee\u6807\u67b6\u6784":143,"\u76ee\u6807\u67b6\u6784abi":142,"\u76f4\u5230\u8bad\u7ec3\u6536\u655b\u4e3a\u6b62":109,"\u76f4\u63a5\u4f7f\u7528\u4e0a\u8ff0\u5b89\u88c5\u6d41\u7a0b":29,"\u76f4\u63a5\u4f7f\u7528c\u8bed\u8a00\u7684":71,"\u76f4\u63a5\u5220\u9664\u8fd9\u4e2a\u53c2\u6570\u5373\u53ef":72,"\u76f4\u63a5\u5347\u7ea7\u5230\u66f4\u65b0\u7684\u7248\u672c":27,"\u76f4\u63a5\u5bfc\u51fa\u5230c\u7684\u63a5\u53e3\u6bd4\u8f83\u56f0\u96be":71,"\u76f4\u63a5\u8c03\u7528\u76f8\u5e94\u63a5\u53e3\u5373\u53ef":101,"\u76f4\u63a5\u8fd0\u884c":28,"\u76f8\u5173\u5c42":67,"\u76f8\u540c\u540d\u5b57\u7684\u53c2\u6570":109,"\u76f8\u6bd4":101,"\u770b\u5f53\u524dmpi\u96c6\u7fa4\u662f\u5426\u652f\u6301resourc":105,"\u77a7":112,"\u77e9\u9635":128,"\u77e9\u9635\u4e2d\u6bcf\u4e2a\u5143\u7d20\u7684\u503c\u968f\u673a\u751f\u6210":115,"\u77e9\u9635\u4e58\u6cd5\u7684\u516c\u5f0f":101,"\u77e9\u9635\u5927\u5c0f\u662f":67,"\u77e9\u9635\u662f\u5426\u662f\u4e00\u4e2a\u5e8f\u5217":115,"\u77e9\u9635\u7684\u9ad8\u5ea6":115,"\u77e9\u9635\u91cc\u7684\u5143\u7d20\u662f\u6d6e\u70b9\u6570":115,"\u786e\u4fdd\u7f16\u8bd1\u5668\u9009\u9879":98,"\u78c1\u76d8\u4e0d\u591f":26,"\u78c1\u76d8\u7a7a\u95f4\u4e0d\u8db3\u7b49":105,"\u793a\u4f8b":[107,109,116],"\u793a\u4f8b3\u5bf9\u4e8e\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u6570\u636e\u5b8c\u5168\u76f8\u540c":137,"\u793a\u4f8b3\u7684\u914d\u7f6e\u4f7f\u7528\u4e86\u5355\u5c42rnn\u548c\u53cc\u5c42rnn":137,"\u793a\u4f8b3\u7684\u914d\u7f6e\u5206\u522b\u4e3a":137,"\u793a\u4f8b\u4ee3\u7801\u5982\u4e0b":[107,116],"\u793a\u4f8b\u5982\u4e0b":109,"\u793a\u4f8b\u7a0b\u5e8f":117,"\u793e\u533a\u53c2\u4e0e\u56f0\u96be":71,"\u793e\u533a\u8d21\u732e\u4ee3\u7801\u5b66\u4e60\u6210\u672c\u9ad8":71,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u4e00\u4e2a\u8ba1\u7b97\u5c42\u7684\u8f93\u5165":115,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u4e00\u4e2a\u8ba1\u7b97\u5c42\u7684\u8f93\u5165\u8f93\u51fa\u88ab\u7ec4\u7ec7\u4e3a\u4e00\u4e2a":116,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u7684\u53c2\u6570":37,"\u795e\u7ecf\u7f51\u7edc\u4e5f\u9700\u8981\u4e00\u4e9b\u7279\u5b9a\u7684layer\u4f5c\u4e3a\u8f93\u5165\u63a5\u53e3":110,"\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u4ee5\u53ca\u8fed\u4ee3\u65b9\u7a0b":110,"\u795e\u7ecf\u7f51\u7edc\u5728\u8bad\u7ec3\u7684\u65f6\u5019":107,"\u795e\u7ecf\u7f51\u7edc\u672c\u8d28\u4e0a\u662f\u4e00\u4e2a\u8ba1\u7b97\u56fe":102,"\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u5c06\u88ab\u5e8f\u5217\u5316\u5408\u5e76\u5165\u4e00\u4e2a\u6587\u4ef6":116,"\u795e\u7ecf\u7f51\u7edc\u7684\u7f51\u7edc\u7ed3\u6784\u4e2d\u5177\u6709\u6709\u5411\u73af\u7ed3\u6784":137,"\u795e\u7ecf\u7f51\u7edc\u7684\u8bad\u7ec3\u672c\u8eab\u662f\u4e00\u4e2a\u975e\u5e38\u6d88\u8017\u5185\u5b58\u548c\u663e\u5b58\u7684\u5de5\u4f5c":107,"\u79bb\u7ebf\u6279\u5904\u7406":38,"\u79f0\u4e3a":[98,140],"\u79f0\u4e3a\u5f00\u53d1\u955c\u50cf":142,"\u79f0\u4e4b\u4e3a":115,"\u79f0\u4e4b\u4e3a\u53cc\u5c42\u5e8f\u5217\u7684\u4e00\u4e2a\u5b50\u5e8f\u5217":136,"\u79f0\u4e4b\u4e3a\u96c6\u675f\u5927\u5c0f":129,"\u79f0\u4f5c\u6709kernel":101,"\u79f0\u4f5ckernel":101,"\u79fb\u52a8\u7aef\u9884\u6d4b":115,"\u7a00\u758f\u6570\u636e\u7684\u683c\u5f0f":100,"\u7a00\u758f\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":123,"\u7a00\u758f\u768401\u5411\u91cf":110,"\u7a00\u758f\u7684\u5411\u91cf":110,"\u7a00\u758f\u77e9\u9635":115,"\u7a00\u758f\u77e9\u9635\u4f7f\u7528":115,"\u7a00\u758f\u77e9\u9635\u53ca\u76f8\u5173\u7684\u63a5\u53e3":115,"\u7a00\u758f\u77e9\u9635\u5b58\u50a8\u793a\u610f\u56fe":115,"\u7a00\u758f\u77e9\u9635\u7684\u4e58\u79ef\u5e94\u7528\u4e8e\u524d\u5411\u4f20\u64ad\u8fc7\u7a0b":131,"\u7a0b\u5e8f\u4ece\u6b64\u76ee\u5f55\u62f7\u8d1d\u6587\u4ef6\u5230\u5bb9\u5668\u5185\u8fdb\u884c\u8bad\u7ec3":123,"\u7a0b\u5e8f\u4f9d\u8d56":117,"\u7a0b\u5e8f\u505c\u6b62":129,"\u7a0b\u5e8f\u662f\u4e00\u6837\u7684":133,"\u7a0b\u5e8f\u76f4\u63a5\u9000\u51fa":129,"\u7a20\u5bc6\u5411\u91cf":100,"\u7a20\u5bc6\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":123,"\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":110,"\u7a20\u5bc6\u77e9\u9635":115,"\u7acb\u523b\u9000\u51fa":107,"\u7acb\u5373\u6267\u884c\u5355\u5143\u6d4b\u8bd5":27,"\u7ae0\u8282":142,"\u7aef\u53e3":105,"\u7aef\u6570\u636e\u7c7b\u578b":115,"\u7aef\u7684":133,"\u7aef\u8bfb\u53d6\u6570\u636e":107,"\u7b2c\u4e00\u4e2a":98,"\u7b2c\u4e00\u4e2a\u53c2\u6570":101,"\u7b2c\u4e00\u4e2a\u6837\u672c\u540c\u65f6encode\u4e24\u6761\u6570\u636e\u6210\u4e24\u4e2a\u5411\u91cf":137,"\u7b2c\u4e00\u4e2atag\u4e3a":88,"\u7b2c\u4e00\u6b65\u9700\u8c03\u7528":116,"\u7b2c\u4e00\u7ae0\u8282":110,"\u7b2c\u4e09\u4e2a\u53c2\u6570":101,"\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93\u9700\u8981\u6309\u7167\u4e0e\u65b9\u5f0f2\u540c\u6837\u65b9\u6cd5\u663e\u793a\u5730\u8fdb\u884c\u94fe\u63a5":113,"\u7b2c\u4e09\u65b9\u94fe\u63a5\u5e93\u548c\u5934\u6587\u4ef6":113,"\u7b2c\u4e09\u6b65\u5b8c\u6210\u540e":88,"\u7b2c\u4e8c\u4e2a":107,"\u7b2c\u4e8c\u4e2a\u4e3a":88,"\u7b2c\u4e8c\u4e2a\u53c2\u6570":101,"\u7b2c\u4e8c\u7c7b":108,"\u7b2ci\u884c\u7b2cj\u5217\u7684\u6570\u503c":100,"\u7b49":[68,72,101,105,116],"\u7b49\u4e8e\u6837\u672c\u6570":107,"\u7b49\u5168\u90e8\u9759\u6001\u5e93\u4e2d\u7684\u76ee\u6807\u6587\u4ef6\u5168\u90e8\u6253\u5305\u540e\u4ea7\u751f\u7684\u6587\u4ef6":72,"\u7b49\u53c2\u6570":123,"\u7b49\u5f85\u7f16\u8bd1\u5b8c\u6210\u540e":88,"\u7b49\u6587\u4ef6":72,"\u7b49\u7b2c\u4e09\u65b9\u5e93":113,"\u7b80\u5199":101,"\u7b80\u5355\u4ecb\u7ecd\u9700\u8981\u7528\u5230\u57fa\u7c7b":101,"\u7b80\u5355\u603b\u7ed3op\u9700\u8981\u5305\u542b\u7684\u5185\u5bb9\u5982\u4e0b":101,"\u7b80\u5355\u6765\u8bf4":134,"\u7b80\u5355\u7684\u5168\u8fde\u63a5\u7f51\u7edc":109,"\u7b80\u5355\u7684\u6027\u80fd\u5206\u6790":134,"\u7b80\u5355\u7684yaml\u6587\u4ef6\u5982\u4e0b":122,"\u7b97\u6cd5":[107,140],"\u7b97\u6cd5\u4e2d\u7684beam\u5927\u5c0f":140,"\u7ba1\u7406\u4e86\u6bcf\u4e2a\u8ba1\u7b97\u5c42\u8f93\u51fa\u7684\u5b58\u50a8\u7a7a\u95f4":116,"\u7c7b\u4f3c":[72,136],"\u7c7b\u4f5c\u4e3a\u53c2\u6570\u7684\u62bd\u8c61":100,"\u7c7b\u522b\u4e2d\u7684\u53c2\u6570\u53ef\u7528\u4e8e\u6240\u6709\u573a\u5408":128,"\u7c7b\u522b\u6807\u7b7e\u4e4b\u4e00":116,"\u7c7b\u522b\u6807\u7b7e\u5c42":116,"\u7c7b\u540d\u548cc":71,"\u7c7b\u578b":[71,101,115,129],"\u7c7b\u578b\u4e3a":101,"\u7c7b\u578b\u4ecd\u7136\u4e3aeigenvector":102,"\u7c7b\u578b\u53ef\u4ee5\u662fpaddlepaddle\u652f\u6301\u7684\u4efb\u610f\u8f93\u5165\u6570\u636e\u7c7b\u578b":136,"\u7c7b\u578b\u540d\u4e3a":101,"\u7c7b\u578b\u662fnumpy\u7684ndarrai":107,"\u7c7b\u578b\u662fsparse_binary_vector":110,"\u7c7b\u578b\u662fsparse_float_vector":110,"\u7c7b\u578b\u7684":137,"\u7c7b\u578b\u7b49\u662f\u5426\u5408\u6cd5":101,"\u7c7b\u578b\u8fd8\u662f":115,"\u7c7b\u7684\u5b9a\u4e49\u5199\u5728":101,"\u7c7b\u7684\u5bf9\u8c61":116,"\u7c7b\u7684\u6784\u9020\u51fd\u6570\u548c\u6790\u6784\u51fd\u6570":100,"\u7c7b\u91cd\u5199":101,"\u7c7b\u9700\u8981\u5b9e\u73b0\u521d\u59cb\u5316":100,"\u7cfb\u6570":101,"\u7cfb\u7edf\u4e2d\u7684\u74f6\u9888\u53ef\u80fd\u548c\u7a0b\u5e8f\u5458\u5f00\u53d1\u8fc7\u7a0b\u4e2d\u60f3\u8c61\u7684\u74f6\u9888\u76f8\u53bb\u751a\u8fdc":133,"\u7cfb\u7edf\u4f1a\u5bf9\u65b0\u589e\u7684op\u81ea\u52a8\u7ed1\u5b9apython":101,"\u7cfb\u7edf\u4f1a\u63d0\u4f9b\u4e00\u4e2a\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":117,"\u7cfb\u7edf\u4f1a\u6839\u636e\u6587\u4ef6\u540d\u81ea\u52a8\u6784\u5efaop\u548c\u5176\u5bf9\u5e94\u7684python\u6269\u5c55":101,"\u7ebf\u7a0bid\u53f7":131,"\u7ec6\u8282\u63cf\u8ff0":130,"\u7ecf\u5e38\u4f1a\u6d88\u8017\u657010gb\u7684\u5185\u5b58\u548c\u6570gb\u7684\u663e\u5b58":107,"\u7ed3\u5c3e":101,"\u7ed3\u675f\u6807\u8bb0":140,"\u7ed3\u675f\u9884\u6d4b\u4e4b\u540e":116,"\u7ed3\u6784\u4f53":[115,116],"\u7ed3\u679c\u4f1a\u5199\u5165\u5f53\u524d\u8fd0\u884c\u76ee\u5f55\u4e0b\u7684":116,"\u7ed3\u679c\u5982\u4e0b\u56fe\u6240\u793a":133,"\u7ed3\u679c\u8f93\u51fa\u5230":26,"\u7ed3\u8bba":71,"\u7ed9\u4e2a\u7b80\u5355\u7684":98,"\u7ed9\u5b9aencoder\u8f93\u51fa\u548c\u5f53\u524d\u8bcd":139,"\u7edf\u4e00\u7528":38,"\u7ee7\u627f\u81ea":101,"\u7ee7\u627f\u81eaoperatorbas":101,"\u7ef4\u57fa\u767e\u79d1\u4e2d\u6587\u9875\u9762":137,"\u7ef4\u57fa\u767e\u79d1\u9875\u9762":137,"\u7ef4\u7a7a\u95f4":140,"\u7ef4\u7a7a\u95f4\u5b8c\u6210":140,"\u7f13\u5b58\u6c60\u7684\u51cf\u5c0f":107,"\u7f16\u5199":28,"\u7f16\u5199\u4e86\u4e00\u4e2apaddlepaddle\u7684\u7a0b\u5e8f":28,"\u7f16\u5199\u5b8cyaml\u6587\u4ef6\u540e":123,"\u7f16\u5199\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6":123,"\u7f16\u5199\u6df1\u5ea6\u5b66\u4e60\u7a0b\u5e8f":133,"\u7f16\u5199\u7684\u90e8\u5206":30,"\u7f16\u5199\u96c6\u7fa4\u4efb\u52a1\u63d0\u4ea4\u548c\u7ba1\u7406\u811a\u672c":120,"\u7f16\u53f7\u4ece0\u5f00\u59cb":107,"\u7f16\u7801\u5411\u91cf":140,"\u7f16\u7801\u5668\u8f93\u51fa":140,"\u7f16\u7801\u6e90\u5e8f\u5217":140,"\u7f16\u8bd1":[26,28,98,142],"\u7f16\u8bd1\u51fa\u7684paddlepaddle\u9884\u6d4b\u5e93\u548c\u5934\u6587\u4ef6":113,"\u7f16\u8bd1\u540e\u7684\u6587\u4ef6\u5c06\u88ab\u5b58\u50a8\u5728\u5de5\u4f5c\u76ee\u5f55":103,"\u7f16\u8bd1\u548c\u5b89\u88c5paddlepaddl":144,"\u7f16\u8bd1\u548c\u5b89\u88c5paddlepaddle\u9884\u6d4b\u5e93":[142,143],"\u7f16\u8bd1\u5668":[142,143,144],"\u7f16\u8bd1\u5668\u6ca1\u6709":71,"\u7f16\u8bd1\u5668\u8981\u6c42\u7cfb\u7edf\u652f\u6301":142,"\u7f16\u8bd1\u578b\u8bed\u8a00":71,"\u7f16\u8bd1\u5b89\u88c5\u4e0e\u5355\u5143\u6d4b\u8bd5":106,"\u7f16\u8bd1\u5b89\u88c5\u7ed3\u675f\u4e4b\u540e":142,"\u7f16\u8bd1\u5b8c\u6210\u4e4b\u540e":103,"\u7f16\u8bd1\u5b8c\u6210\u540e\u4f1a\u5728build":27,"\u7f16\u8bd1\u5de5\u5177\u94fe":142,"\u7f16\u8bd1\u5de5\u5177\u94fe\u6240\u5728\u7684\u7edd\u5bf9\u8def\u5f84":144,"\u7f16\u8bd1\u6027\u80fd\u4f1a\u548c":133,"\u7f16\u8bd1\u6210\u529f\u540e":101,"\u7f16\u8bd1\u6210\u529f\u540e\u5728":113,"\u7f16\u8bd1\u6210\u52a8\u6001\u5e93":129,"\u7f16\u8bd1\u65f6\u4e00\u5b9a\u8981\u5f00\u542f\u4f18\u5316":133,"\u7f16\u8bd1\u65f6\u4f1a\u628a\u5bf9\u5e94\u7684\u5934\u6587\u4ef6\u548c\u5e93\u653e\u5728":68,"\u7f16\u8bd1\u65f6\u53ef\u80fd\u4f1a\u53bb\u6389\u8c03\u8bd5\u4fe1\u606f":133,"\u7f16\u8bd1\u65f6\u6307\u5b9a":133,"\u7f16\u8bd1\u751f\u6210":103,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684docker\u53d1\u884c\u955c\u50cf":88,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684python":88,"\u7f16\u8bd1c":72,"\u7f16\u8bd1paddlepaddl":27,"\u7f51\u7edc\u5c42\u53ef\u4ee5\u6709\u591a\u4e2a\u8f93\u5165":100,"\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u4e3a":100,"\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":100,"\u7f51\u7edc\u5c42\u7684\u7ec6\u8282\u53ef\u4ee5\u901a\u8fc7\u4e0b\u9762\u8fd9\u4e9b\u4ee3\u7801\u7247\u6bb5\u6765\u6307\u5b9a":100,"\u7f51\u7edc\u5c42\u7684\u8f93\u51fa\u662f\u7ecf\u8fc7\u6fc0\u6d3b\u51fd\u6570\u4e4b\u540e\u7684\u503c":129,"\u7f51\u7edc\u5c42\u914d\u7f6e\u5305\u542b\u4ee5\u4e0b\u51e0\u9879":100,"\u7f51\u7edc\u63a5\u53d7\u4e00\u5e45\u56fe\u7247\u4f5c\u4e3a\u8f93\u5165":116,"\u7f51\u7edc\u7ed3\u6784\u7684\u5e8f\u5217\u5316\u7ed3\u679c\u548c\u6a21\u578b\u53c2\u6570\u5b58\u50a8\u76ee\u5f55":116,"\u7f51\u7edc\u901a\u4fe1":100,"\u7f51\u901f\u6216ssl\u94fe\u63a5\u539f\u56e0":104,"\u800c":[108,133,140],"\u800c\u4e0d\u4f1a\u6539\u53d8\u539f\u6709tensor\u7684shape\u4fe1\u606f":102,"\u800c\u4e0d\u5fc5\u5728\u610fpaddl":72,"\u800c\u4e0d\u652f\u6301pypy\u89e3\u91ca\u5668":71,"\u800c\u4e0d\u662f\u5728layer\u91cc\u5b9e\u73b0":108,"\u800c\u4e0d\u662f\u6e90\u7801\u76ee\u5f55\u91cc":104,"\u800c\u4e0d\u662f\u7279\u5f81\u7684\u96c6\u5408":137,"\u800c\u4e0d\u662f\u76f8\u5bf9":115,"\u800c\u4e0d\u662fc":115,"\u800c\u4e0d\u66b4\u9732\u6982\u5ff5\u7684\u5b9e\u73b0":72,"\u800c\u4e14\u4e2a\u6570\u5e76\u4e0d\u786e\u5b9a":117,"\u800c\u4e14\u5305\u542b\u4e86c":30,"\u800c\u4e14\u5728\u4f20\u8f93\u7684\u8fc7\u7a0b\u4e2d\u4e5f\u53ef\u80fd\u51fa\u73b0\u7f51\u7edc\u4e0d\u7a33\u5b9a\u7684\u60c5\u51b5":54,"\u800c\u4e14cento":30,"\u800c\u4e4b\u524d\u7684\u53c2\u6570\u5c06\u4f1a\u88ab\u5220\u9664":129,"\u800c\u4ece\u5e94\u7528\u7684\u89d2\u5ea6":134,"\u800c\u4f18\u5316\u6027\u80fd\u7684\u9996\u8981\u4efb\u52a1":134,"\u800c\u5176\u4ed6\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":131,"\u800c\u51fa\u73b0\u9636\u6bb5\u6027\u7684\u8fd0\u884c\u505c\u6ede":37,"\u800c\u53cc\u5c42rnn\u662f\u53ef\u4ee5\u5904\u7406\u8fd9\u79cd\u8f93\u5165\u6570\u636e\u7684\u7f51\u7edc\u7ed3\u6784":137,"\u800c\u53cd\u5411\u6d4b\u8bd5\u4e2d":101,"\u800c\u53ea\u9700\u8981\u83b7\u5f97recurr":108,"\u800c\u5728\u8ba1\u7b97\u7ed3\u675f\u4e4b\u540e":102,"\u800c\u5728cpp\u91cc\u9762\u5b9e\u73b0\u8fd9\u4e2ac\u7684\u63a5\u53e3":71,"\u800c\u591a\u8bed\u8a00\u63a5\u53e3\u9700\u8981\u76f4\u63a5\u8bfb\u53d6\u751f\u6210\u7684\u4e8c\u8fdb\u5236":71,"\u800c\u5b89\u88c5\u5305":[30,104],"\u800c\u5b89\u88c5\u5305\u662f":[30,104],"\u800c\u5bf9\u4e8e\u53cc\u5c42\u5e8f\u5217":137,"\u800c\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5185\u5c42\u7279\u5f81\u6570\u636e\u800c\u8a00":137,"\u800c\u5bf9\u4e8egolang":71,"\u800c\u5bf9\u4e8egolang\u9519\u8bef\u5904\u7406\u5e94\u8be5\u4f7f\u7528\u8fd4\u56de\u503c":71,"\u800c\u5c06\u8fd9\u4e2a\u6bb5\u843d\u7684\u6bcf\u4e00\u53e5\u8bdd\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u5411\u91cf":137,"\u800c\u5f53\u524d\u5df2\u7ecf\u67095":134,"\u800c\u662f\u5c06\u8f93\u5165":[115,116],"\u800c\u662f\u76f4\u63a5\u4ece\u5185\u5b58\u7684\u7f13\u5b58\u91cc\u8bfb\u53d6\u6570\u636e":107,"\u800c\u662f\u76f4\u63a5\u4fee\u6539paddl":72,"\u800c\u662f\u76f4\u63a5\u7528api\u7684\u63a5\u53e3\u8fdc\u7a0b\u8bbf\u95ee":38,"\u800c\u66f4\u6df1\u5165\u7684\u5206\u6790":134,"\u800c\u6709\u4e9b\u53c2\u6570\u9700\u8981\u5728\u96c6\u7fa4\u591a\u673a\u8bad\u7ec3\u4e2d\u4f7f\u7528\u7b49":128,"\u800c\u6e90\u5e8f\u5217\u7684\u7f16\u7801\u5411\u91cf\u53ef\u4ee5\u88ab\u65e0\u8fb9\u754c\u7684memory\u8bbf\u95ee":140,"\u800c\u795e\u7ecf\u7f51\u7edc\u662f\u6211\u4eec\u8981\u642d\u5efa\u7684\u5b9d\u5854":110,"\u800c\u7a00\u758f\u66f4\u65b0\u5728\u53cd\u5411\u4f20\u64ad\u4e4b\u540e\u7684\u6743\u91cd\u66f4\u65b0\u65f6\u8fdb\u884c":131,"\u800c\u8ba1\u7b97\u8fc7\u7a0b\u662f\u7531":102,"\u800c\u8fd9\u4e00\u53e5\u8bdd\u5c31\u53ef\u4ee5\u8868\u793a\u6210\u8fd9\u4e9b\u4f4d\u7f6e\u7684\u6570\u7ec4":137,"\u800c\u8fd9\u6bcf\u4e00\u4e2a\u6570\u7ec4\u5143\u7d20":137,"\u800c\u975e\u76f4\u63a5\u56de\u590d\u7684\u65b9\u5f0f":98,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":27,"\u800ceigenvector":102,"\u800crnn\u662f\u6700\u6d41\u884c\u7684\u9009\u62e9":139,"\u800cswig\u53ea\u80fd\u7b80\u5355\u7684\u66b4\u9732c":71,"\u800ctrainer\u9700\u8981\u8bfb\u53d6\u8bad\u7ec3\u6570\u636e\u8fdb\u884c\u8bad\u7ec3":110,"\u800cy_predict\u662f\u63a5\u6536x\u4f5c\u4e3a\u8f93\u5165":110,"\u8054\u901a":127,"\u80fd\u591f\u5904\u7406\u53cc\u5c42\u5e8f\u5217":139,"\u80fd\u591f\u5bf9\u53cc\u5411\u5e8f\u5217\u8fdb\u884c\u5904\u7406\u7684\u6709":139,"\u80fd\u591f\u8bb0\u5f55\u4e0a\u4e00\u4e2asubseq":139,"\u80fd\u591f\u9488\u5bf9cpu\u548cgpu\u7684\u8ba1\u7b97\u505a\u66f4\u591a\u4f18\u5316":108,"\u80fd\u83b7\u53d6":119,"\u811a\u672c":[26,116,142],"\u811a\u672c\u5f00\u59cb\u65f6":123,"\u811a\u672c\u96c6\u6210\u4e86\u5e8f\u5217\u5316\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u7684\u8fc7\u7a0b":116,"\u81ea\u52a8\u5173\u95ed\u5bf9\u5e94\u7684":98,"\u81ea\u52a8\u5730\u5c06\u8fd9\u4e9b\u9009\u9879\u5e94\u7528\u5230":119,"\u81ea\u52a8\u5b8c\u6210\u8fd9\u4e00\u8fc7\u7a0b":139,"\u81ea\u52a8\u6302\u8f7d\u5206\u5e03\u5f0f\u5b58\u50a8\u76ee\u5f55":37,"\u81ea\u52a8\u6784\u5efa\u72ec\u7acb\u5de5\u5177\u94fe":142,"\u81ea\u52a8\u751f\u6210":103,"\u81ea\u52a8\u83b7\u53d6\u4e0a\u4e00\u4e2a\u751f\u6210\u7684\u8bcd":140,"\u81ea\u52a8\u9009\u62e9":143,"\u81ea\u6b64":[142,143],"\u81ea\u7136\u4e5f\u5c31\u6709\u7ba1\u7406\u5458\u6743\u9650":26,"\u81ea\u7136\u8bed\u8a00\u4e2d\u7684\u53e5\u5b50\u662f\u4e00\u4e2a\u5e8f\u5217":115,"\u81ea\u7136\u8bed\u8a00\u4e2d\u7684\u6bb5\u843d\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":115,"\u81ea\u7136\u8bed\u8a00\u5904\u7406\u7b49":131,"\u81f3\u4e8e\u4e3a\u4ec0\u4e48\u9700\u8981c":72,"\u81f3\u5c11\u5305\u542bgcc_3":30,"\u81f3\u5c11\u5305\u542bglibcxx_3":30,"\u81f3\u6b64":[98,137],"\u826f\u597d\u7684\u6587\u6863":71,"\u8282\u70b9":[124,127],"\u8282\u7701\u4e86\u4e0d\u5fc5\u8981\u7684\u64cd\u4f5c":68,"\u82e5":100,"\u82e5\u5728paddlepaddle\u7f16\u8bd1\u65f6":109,"\u82e5\u5e0c\u671b\u5f97\u5230\u6700\u5feb\u7684\u6267\u884c\u901f\u5ea6":143,"\u82e5\u5e0c\u671b\u6700\u5feb\u7684\u6267\u884c\u901f\u5ea6":[142,144],"\u82e5\u5e72\u4e2a\u53e5\u5b50\u6784\u6210\u4e00\u4e2a\u6bb5\u843d":136,"\u82e5\u6709\u4e0d\u4e00\u81f4\u4e4b\u5904":134,"\u82e5\u6709\u5fc5\u8981":100,"\u82e5\u672a\u663e\u5f0f\u6307\u5b9a":143,"\u82e5\u6ca1\u6709\u663e\u5f0f\u8bbe\u7f6e":142,"\u82e5\u73af\u5883\u53d8\u91cf":[142,143,144],"\u82e5\u8981\u5bf9\u8fd9\u51e0\u4e2alayer\u4f7f\u7528dropout":108,"\u82e5\u8f93\u51fa\u662f\u5355\u5c42\u5e8f\u5217":136,"\u82e5\u8f93\u51fa\u662f\u53cc\u5c42\u5e8f\u5217":136,"\u82f1\u6587\u6587\u6863":103,"\u82f1\u6587\u6587\u6863\u76ee\u5f55":103,"\u8303\u56f4":131,"\u83b7\u53d6":98,"\u83b7\u53d6\u53ef\u9009\u7684tag":28,"\u83b7\u53d6\u5f53\u524d\u7cfb\u7edf\u652f\u6301\u7684\u5b89\u88c5\u5305\u683c\u5f0f":30,"\u83b7\u53d6\u5f53\u524d\u7cfb\u7edf\u652f\u6301\u7684python\u5305\u7684\u540e\u7f00":104,"\u83b7\u53d6\u6700\u65b0\u7684\u68c0\u67e5\u70b9\u7684\u6587\u4ef6uuid":37,"\u83b7\u53d6\u6e90\u7801":26,"\u83b7\u53d6\u8f93\u51fa\u65f6":116,"\u83b7\u53d6trainer":123,"\u83b7\u5f97\u53c2\u6570\u5c3a\u5bf8":100,"\u83b7\u5f97\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u67d0\u4e00\u5c42\u7684name":107,"\u83b7\u5f97\u57fa\u672c\u7684docker\u5b89\u88c5\u548c\u4f7f\u7528\u65b9\u6cd5":28,"\u83b7\u5f97\u5f53\u524dmini":107,"\u83b7\u5f97\u7684\u503c\u7c7b\u578b\u5747\u4e3a":107,"\u83b7\u5f97\u8ba1\u7b97\u7ed3\u679c":116,"\u83b7\u5f97\u8fd9\u4e9b\u8282\u70b9\u7684ip\u5730\u5740":119,"\u83b7\u5f97head\u548cnode\u8282\u70b9\u7684ip\u5730\u5740":124,"\u865a\u62df\u673a\u4e0a":26,"\u867d\u7136\u4e0d\u9f13\u52b1\u8fd9\u6837":72,"\u867d\u7136\u5f02\u6b65sgd\u65b9\u5f0f\u4f1a\u63d0\u9ad8\u53c2\u6570\u66f4\u65b0\u5e76\u884c\u5ea6":118,"\u867d\u7136paddle\u770b\u8d77\u6765\u5305\u542b\u4e86\u4f17\u591a\u53c2\u6570":128,"\u884c":115,"\u884c\u504f\u79fb":115,"\u884c\u53f7":133,"\u8865\u5145\u4e0a\u6b21\u7684commit":98,"\u8868\u660e\u4e86\u8fd9\u4e9b\u884c\u7684\u6807\u53f7":100,"\u8868\u660e\u8fd9\u4e2a\u5c42\u7684\u4e00\u4e2a\u5b9e\u4f8b\u662f\u5426\u9700\u8981\u504f\u7f6e":100,"\u8868\u793a":101,"\u8868\u793a\u4e3adeviceid":131,"\u8868\u793a\u5bf9\u8f93\u5165\u6570\u636e":68,"\u8868\u793a\u5c06\u5916\u5c42\u7684outer_mem\u4f5c\u4e3a\u5185\u5c42memory\u7684\u521d\u59cb\u72b6\u6001":137,"\u8868\u793a\u5f53\u524d\u96c6\u7fa4\u4f5c\u4e1a\u7684\u8282\u70b9":119,"\u8868\u793a\u6570\u636e\u7c7b\u578b":101,"\u8868\u793a\u7684\u504f\u79fb\u662f\u4ee5":115,"\u8868\u793a\u8bbe\u5907\u7c7b\u578b":101,"\u8868\u793a\u8bcd\u8bed\u5728\u8bcd\u5178\u4e2d\u7684\u5e8f\u53f7":115,"\u8868\u793a\u8bfb\u8005\u6240\u4f7f\u7528\u7684docker\u955c\u50cf\u4ed3\u5e93\u5730\u5740":123,"\u8868\u793a\u8fd9\u4e2ajob\u7684\u540d\u5b57":123,"\u8868\u793a\u9700\u8981\u6784\u5efa\u63a8\u7406\u5e93":144,"\u88ab":98,"\u88ab\u5207\u5206\u6210\u591a\u4e2a\u90e8\u5206":118,"\u88ab\u6269\u5c55\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"\u88ab\u653e\u5728":100,"\u88ab\u79f0\u4e3a":140,"\u8981\u4f7f\u7528\u547d\u4ee4\u884c\u5206\u6790\u5de5\u5177":134,"\u8981\u5728\u5df2\u6709\u7684kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884cpaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":123,"\u8981\u6c42\u5355\u5c42\u5e8f\u5217\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee":136,"\u8981\u751f\u6210\u7684\u76ee\u6807\u5e8f\u5217":139,"\u8981\u8c03\u7528":100,"\u89c6\u9891\u7b49":115,"\u89e3\u51b3\u529e\u6cd5\u662f":104,"\u89e3\u51b3\u65b9\u6848\u662f":109,"\u89e3\u6790\u73af\u5883\u53d8\u91cf\u5f97\u5230":123,"\u89e3\u7801\u5668\u4f7f\u7528":140,"\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":140,"\u89e3\u7801\u5668\u662f\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":140,"\u89e3\u91ca\u578b\u8bed\u8a00\u53ea\u80fd\u8c03\u7528\u52a8\u6001\u5e93":71,"\u89e3\u91ca\u6027\u8bed\u8a00\u5b9e\u9645\u8fd0\u884c\u7684\u4e8c\u8fdb\u5236\u662f\u89e3\u91ca\u5668\u672c\u8eab":71,"\u8ba1\u5212\u5728":[67,68],"\u8ba1\u7b97":[118,140],"\u8ba1\u7b97\u504f\u7f6e\u7684\u68af\u5ea6":100,"\u8ba1\u7b97\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":140,"\u8ba1\u7b97\u53d8\u6362\u77e9\u9635\u7684\u5927\u5c0f\u548c\u683c\u5f0f":100,"\u8ba1\u7b97\u5f53\u524d\u5c42\u6743\u91cd\u7684\u68af\u5ea6":100,"\u8ba1\u7b97\u6548\u7387\u66f4\u9ad8":108,"\u8ba1\u7b97\u6bcf\u4e2a\u8bcd\u7684\u8bcd\u5411\u91cf":140,"\u8ba1\u7b97\u6fc0\u6d3b\u51fd\u6570\u7684\u68af\u5ea6":100,"\u8ba1\u7b97\u7684\u7ec6\u8282\u5c06\u5728\u4e0b\u9762\u7684\u5c0f\u8282\u7ed9\u51fa":100,"\u8ba1\u7b97\u8282\u70b9":118,"\u8ba1\u7b97\u8282\u70b9\u4e4b\u95f4\u4e5f\u4e0d\u4f1a\u76f8\u4e92\u4f9d\u8d56":118,"\u8ba1\u7b97\u8f6c\u6362\u77e9\u9635\u548c\u8f93\u5165\u7684\u68af\u5ea6":100,"\u8ba1\u7b97\u8f93\u5165\u548c\u53c2\u6570\u7684\u68af\u5ea6":100,"\u8ba1\u7b97\u8f93\u5165\u5c42\u7684\u504f\u5dee":100,"\u8ba1\u7b97\u8f93\u51fa":100,"\u8ba1\u7b97\u8fd9\u4e2a\u6587\u4ef6\u7684md5":37,"\u8ba1\u7b97\u96c6\u7fa4\u901a\u5e38\u7531\u4e00\u7ec4":127,"\u8ba1\u7b97\u9700\u8981\u7684\u6570\u636e\u5b58\u653e\u5728":102,"\u8ba9paddle\u6838\u5fc3\u4e2d":72,"\u8bad\u7ec3":128,"\u8bad\u7ec3\u4e0e\u5e94\u7528":0,"\u8bad\u7ec3\u4efb\u52a1\u7684\u8fd0\u884c\u53ef\u80fd\u4f1a\u5360\u6ee1trainer\u548cparamet":37,"\u8bad\u7ec3\u548c\u7eaf\u4f7f\u7528":88,"\u8bad\u7ec3\u5931\u8d25\u65f6\u53ef\u4ee5\u68c0\u67e5\u9519\u8bef\u65e5\u5fd7":119,"\u8bad\u7ec3\u597d\u4e00\u4e2a\u6df1\u5c42\u795e\u7ecf\u7f51\u7edc\u901a\u5e38\u8981\u8017\u8d39\u975e\u5e38\u957f\u7684\u65f6\u95f4":134,"\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u9ed8\u8ba4\u4fdd\u5b58\u5728\u5f53\u524d\u8fd0\u884c\u76ee\u5f55\u4e0b\u7684":116,"\u8bad\u7ec3\u6570\u636e\u6709\u95ee\u9898":107,"\u8bad\u7ec3\u6570\u636e\u683c\u5f0f\u548c\u8bad\u7ec3\u7a0b\u5e8f\u7684":117,"\u8bad\u7ec3\u65f6":123,"\u8bad\u7ec3\u6a21\u578b\u540e":140,"\u8bad\u7ec3\u6a21\u578b\u6b63\u786e\u6027":88,"\u8bad\u7ec3\u7a0b\u5e8f":117,"\u8bad\u7ec3\u7ed3\u675f\u540e\u67e5\u770b\u8f93\u51fa\u7ed3\u679c":123,"\u8bad\u7ec3\u8282\u70b9\u6570\u91cf":123,"\u8bad\u7ec3\u8bed\u8a00\u6a21\u578b\u8ddd\u79bb":107,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u53c2\u6570\u6216\u8005\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u7684\u68af\u5ea6\u5c3a\u5ea6\u8fc7\u5927":107,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6d4b\u8bd5test_period":128,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4e3a\u672c\u5730\u6a21\u5f0f":129,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4f7f\u7528gpu":129,"\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u8bbe\u5907\u5c5e\u6027\u5c06\u4f1a\u65e0\u6548":129,"\u8bad\u7ec3dot_period":128,"\u8bb0\u5f55\u4e0b\u6240\u6709\u5931\u8d25\u7684\u4f8b\u5b50":88,"\u8bb0\u5fc6\u6a21\u5757":140,"\u8bbe\u4e3a\u5df2\u90e8\u7f72\u7684\u5de5\u4f5c\u7a7a\u95f4\u76ee\u5f55":119,"\u8bbe\u4e3a\u672c\u5730":119,"\u8bbe\u5b9a":108,"\u8bbe\u7f6e":[27,72,107,108,142,143],"\u8bbe\u7f6e\u4e3a":100,"\u8bbe\u7f6e\u4e3a\u4e0d\u540c\u7684\u503c":108,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528\u672c\u5730\u8bad\u7ec3\u6216\u8005\u4f7f\u7528\u96c6\u7fa4\u4e0a\u7684\u4e00\u4e2a\u8282\u70b9":129,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528gpu\u6a21\u5f0f":129,"\u8bbe\u7f6e\u4e86\u76f8\u540c\u7684\u53d6\u503c":108,"\u8bbe\u7f6e\u5176\u53c2\u6570\u5c5e\u6027":109,"\u8bbe\u7f6e\u53c2\u6570\u7684\u540d\u5b57":109,"\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570":107,"\u8bbe\u7f6e\u5b66\u4e60\u7387\u8870\u51cf\u56e0\u5b50\u5206\u6bb5\u51fd\u6570":109,"\u8bbe\u7f6e\u5e8f\u5217\u4fe1\u606f\u7684\u63a5\u53e3":115,"\u8bbe\u7f6e\u6210":109,"\u8bbe\u7f6e\u6210\u4e00\u4e2a\u5c0f\u4e00\u4e9b\u7684\u503c":107,"\u8bbe\u7f6e\u8f93\u51fa\u7684\u5c3a\u5bf8":100,"\u8bbe\u7f6e\u8f93\u51fatensor\u7684\u5f62\u72b6":101,"\u8bbe\u7f6e\u9ed8\u8ba4\u8bbe\u5907\u53f7\u4e3a0":131,"\u8bbe\u7f6egpu":129,"\u8bbf\u95ee\u5bf9\u5e94\u7684\u7f51\u5740":133,"\u8bbf\u95eekubernetes\u7684\u63a5\u53e3\u6765\u67e5\u8be2\u6b64job\u5bf9\u5e94\u7684\u6240\u6709pod\u4fe1\u606f":123,"\u8bc4\u5ba1\u4eba\u4e00\u822c\u4e0d\u505a\u8bc4\u5ba1":98,"\u8bc4\u5ba1\u4eba\u7684\u6bcf\u4e2a\u610f\u89c1\u90fd\u5fc5\u987b\u56de\u590d":98,"\u8bc4\u5ba1\u4eba\u9700\u8981\u9010\u4e00\u67e5\u770b\u6bcf\u4e2acommit\u624d\u80fd\u77e5\u9053\u505a\u4e86\u54ea\u4e9b\u4fee\u6539":98,"\u8bc4\u8bba\u6846\u4e2d\u52a0\u4e0a":98,"\u8bc6\u522b\u6570\u5b57":88,"\u8bcd\u5411\u91cf":88,"\u8bd5\u7740\u8ba9\u8f93\u51fa\u7684\u5206\u6790\u6570\u636e\u548c\u7406\u8bba\u503c\u5bf9\u5e94":134,"\u8be5\u53c2\u6570\u5728\u7f51\u7edc\u914d\u7f6e\u7684output":129,"\u8be5\u53c2\u6570\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":129,"\u8be5\u53c2\u6570\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":129,"\u8be5\u53c2\u6570\u5fc5\u987b\u80fd\u88abflag":129,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u65e5\u5fd7\u622a\u65ad\u4fe1\u606f":129,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u9519\u8bef\u622a\u65ad\u65e5\u5fd7":129,"\u8be5\u53c2\u6570\u7528\u4e8e\u6307\u5b9a\u52a8\u6001\u5e93\u8def\u5f84":129,"\u8be5\u53c2\u6570\u7684\u610f\u601d\u662f\u8bad\u7ec3num":129,"\u8be5\u53c2\u6570\u9ed8\u8ba4\u4e3anull":129,"\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":100,"\u8be5\u622a\u65ad\u4f1a\u5f71\u54cd":129,"\u8be5\u6279\u6b21\u7684\u8f93\u5165\u4e2d\u4ec5\u6709\u4e00\u4e2a\u5b50\u96c6\u662f\u975e\u96f6\u7684":100,"\u8be5\u63a5\u53e3\u53ef\u7528\u4e8e\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":27,"\u8be5\u63a5\u53e3\u63a5\u53d7\u4e24\u4e2a\u53c2\u6570":116,"\u8be5\u6570\u76ee\u662f\u63d0\u524d\u5b9a\u4e49\u597d\u7684":129,"\u8be5\u6587\u4ef6\u5bf9\u76f8\u5173gemm":67,"\u8be5\u65f6\u95f4\u53bb\u9664\u6389\u672c\u51fd\u6570\u8c03\u7528\u5176\u4ed6\u51fd\u6570\u7684\u65f6\u95f4":133,"\u8be5\u6a21\u578b\u7684\u8bf4\u660e\u5982\u4e0b\u56fe\u6240\u793a":140,"\u8be5\u7c7b\u7684":101,"\u8be5\u7c7b\u7684\u5b9e\u73b0\u7ec6\u8282\u5728":100,"\u8be5\u7c7b\u7ee7\u627f\u4e8epaddlepaddle\u7684\u57fa\u7c7b":68,"\u8be5\u811a\u672c\u4e2d\u8bb0\u5f55\u4e86\u4ea4\u53c9\u7f16\u8bd1android\u7248paddlepaddle\u5e93\u5e38\u7528\u7684cmake\u914d\u7f6e":142,"\u8be5\u8bed\u53e5\u4f1a\u4e3a\u6bcf\u4e2a\u5c42\u521d\u59cb\u5316\u5176\u6240\u9700\u8981\u7684\u53d8\u91cf\u548c\u8fde\u63a5":100,"\u8be5layer\u662f\u901a\u8fc7\u53c2\u6570":108,"\u8be6\u7ec6\u4ecb\u7ecd\u53ef\u4ee5\u53c2\u8003":137,"\u8be6\u7ec6\u4ecb\u7ecd\u8bf7\u53c2\u8003\u8bbe\u8ba1\u6587\u6863":101,"\u8be6\u7ec6\u4fe1\u606f\u8bf7\u68c0\u67e5":119,"\u8be6\u7ec6\u53c2\u8003":27,"\u8be6\u7ec6\u53ef\u53c2\u8003":98,"\u8be6\u7ec6\u6587\u6863\u53c2\u8003":107,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":27,"\u8be6\u7ec6\u89c1":136,"\u8be6\u7ec6\u8bbe\u8ba1":54,"\u8bed\u610f\u89d2\u8272\u6807\u6ce8":88,"\u8bed\u8a00\u91cd\u6784\u540e\u7684":133,"\u8bf4\u660e":[27,30,37,115],"\u8bf4\u660e\u63d0\u4ea4\u7684\u4ee3\u7801\u5b58\u5728\u95ee\u9898":98,"\u8bf4\u660e\u8fd9\u4e2a\u5c42\u7684\u8f93\u5165":100,"\u8bf7\u4e0d\u8981\u521b\u5efa\u7a7a\u7684":101,"\u8bf7\u4e0d\u8981\u5fd8\u8bb0\u63d0\u524d\u5728\u7269\u7406\u673a\u4e0a\u5b89\u88c5gpu\u6700\u65b0\u9a71\u52a8":28,"\u8bf7\u4fdd\u8bc1travi":98,"\u8bf7\u5148\u4f7f\u7528":[142,143,144],"\u8bf7\u53c2\u7167\u7f51\u7edc\u914d\u7f6e\u7684\u6587\u6863\u4e86\u89e3\u66f4\u8be6\u7ec6\u7684\u4fe1\u606f":131,"\u8bf7\u53c2\u8003":[72,100,101,104,107,110,116,137],"\u8bf7\u53c2\u8003\u6b64":116,"\u8bf7\u53c2\u89c1":98,"\u8bf7\u53c2\u9605":140,"\u8bf7\u5728\u8be5pull":98,"\u8bf7\u5728\u8f93\u5165\u65f6\u8fdb\u884c\u5408\u6cd5\u6027\u68c0\u67e5":115,"\u8bf7\u60a8\u6bcf\u6b21\u63d0\u4ea4\u4ee3\u7801\u65f6":98,"\u8bf7\u60a8\u9075\u5b88\u4ee5\u4e0b\u7ea6\u5b9a":98,"\u8bf7\u6307\u5b9a\u7684paddlepaddle\u5de5\u4f5c\u76ee\u5f55\u7ed9\u73af\u5883\u53d8\u91cf":103,"\u8bf7\u6307\u5b9a\u8be5\u76ee\u5f55":129,"\u8bf7\u663e\u793a\u5730\u8c03\u7528":101,"\u8bf7\u68c0\u67e5python\u7248\u672c\u662f\u5426\u4e3a2":30,"\u8bf7\u6ce8\u610f":[101,122,140],"\u8bf7\u6ce8\u610f\u662f\u5426\u9700\u8981\u4fee\u6539\u7f51\u7edc\u7ed3\u6784":116,"\u8bf7\u6ce8\u610f\u6bcf\u4e2acommit\u7684\u540d\u79f0":98,"\u8bf7\u6ce8\u610f\u8fd9\u4e2a\u547d\u4ee4\u7ed3\u5c3e\u5904\u7684":26,"\u8bf7\u6ce8\u610fcommit\u7684\u6570\u91cf":98,"\u8bf7\u76f4\u63a5\u586b\u51450":109,"\u8bf7\u770b\u4e0b\u9762\u7684\u4f8b\u5b50":131,"\u8bf7\u786e\u4fdd":98,"\u8bf7\u7ed9\u51fa\u603b\u4f53\u7684\u4fee\u6539\u60c5\u51b5":98,"\u8bf7\u7ed9\u51fa\u60a8\u81ea\u5df1\u7684\u53cd\u9a73\u7406\u7531":98,"\u8bf7\u9009\u62e9\u5408\u9002\u7684\u8bcd\u6c47":98,"\u8bf7\u9009\u62e9\u6b63\u786e\u7684\u7248\u672c":104,"\u8bf7\u9075\u5b88":98,"\u8bf7\u91c7\u7528":98,"\u8bf8\u5982\u56fe\u50cf\u5206\u7c7b":131,"\u8bfb\u53d6\u9700\u8981\u7684\u7ed3\u679c\u5373\u53ef":115,"\u8bfb\u53d6volume\u4e2d\u7684\u6570\u636e\u8fdb\u884c\u8fd9\u6b21\u5206\u5e03\u5f0f\u8bad\u7ec3":123,"\u8bfb\u8005\u53ef\u4ee5\u67e5\u770b":123,"\u8bfb\u8005\u9700\u8981\u66ff\u6362\u6210\u81ea\u5df1\u4f7f\u7528\u7684\u4ed3\u5e93\u5730\u5740":123,"\u8c03\u7528":[100,116,143],"\u8c03\u7528\u5bf9\u5e94":102,"\u8c03\u7528\u65b9\u6cd5\u89c1c":[142,143],"\u8c03\u7528\u7528":133,"\u8c03\u7528\u7684\u4e00\u4e9b\u7528\u6237\u5b9a\u4e49\u7684\u5e93\u51fd\u6570":117,"\u8c03\u7528\u7684\u51fd\u6570\u662f\u5426\u652f\u6301\u4e0d\u540c\u8bbe\u5907":101,"\u8c03\u7528\u8be5\u51fd\u6570\u540e":100,"\u8c03\u7528c":[115,116],"\u8d21\u732e\u6587\u6863":103,"\u8d77\u59cb\u5b58\u50a8\u5730\u5740\u4ee5\u6570\u636e\u7684\u5b58\u50a8\u5927\u5c0f\u4e3a\u5355\u4f4d\u7684\u504f\u79fb":115,"\u8df3\u8f6c\u5230":98,"\u8df3\u8fc7":107,"\u8f6c\u5316\u65b9\u6cd5\u5728\u76f8\u5e94\u7684\u9886\u57df\u90fd\u6709\u901a\u7528\u89e3\u51b3\u65b9\u6848":115,"\u8f6c\u6362\u5185\u5b58\u7684\u5de5\u4f5c":68,"\u8f6c\u6362\u5197\u4f59":67,"\u8f6c\u6362\u51fd\u6570":68,"\u8f6c\u6362\u751f\u6210\u7684\u6587\u4ef6\u540d\u4f1a\u662f\u4ee5\u4e0b\u683c\u5f0f":38,"\u8f6c\u6362\u8017\u65f6":67,"\u8f93\u5165":[114,116,136,140],"\u8f93\u5165\u4e86\u6027\u80fd\u5206\u6790\u7ed3\u679c":133,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u5355\u5c42\u5e8f\u5217":139,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u53cc\u5c42\u5e8f\u5217":139,"\u8f93\u5165\u5e8f\u5217\u4e2d\u5143\u7d20\u7684\u603b\u6570":107,"\u8f93\u5165\u6570\u636e\u4e3a\u4e00\u4e2a\u5b8c\u6574\u7684\u65f6\u95f4\u5e8f\u5217":137,"\u8f93\u5165\u6570\u636e\u4e3a\u5728\u5355\u5c42rnn\u6570\u636e\u91cc\u9762":137,"\u8f93\u5165\u6570\u636e\u53ef\u5206\u4e3a":115,"\u8f93\u5165\u6570\u636e\u6574\u4f53\u4e0a\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":137,"\u8f93\u5165\u6570\u636e\u7684\u5b57\u5178\u7ef4\u6570\u662f1\u767e\u4e07":131,"\u8f93\u5165\u6570\u636e\u7c7b\u578b":115,"\u8f93\u5165\u662f\u5426\u662f\u8f6c\u7f6e\u7684":100,"\u8f93\u5165\u662f\u7531\u4e00\u4e2alist\u4e2d\u7684\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u7ec4\u6210\u7684":100,"\u8f93\u5165\u68af\u5ea6":68,"\u8f93\u5165\u7684\u540d\u5b57":100,"\u8f93\u5165\u7684\u5927\u5c0f":100,"\u8f93\u5165\u7684\u7c7b\u578b":100,"\u8f93\u5165\u9700\u8981\u9884\u6d4b\u7684\u5411\u91cf\u7ec4":110,"\u8f93\u51fa":[101,116,136,140],"\u8f93\u51fa\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":139,"\u8f93\u51fa\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":139,"\u8f93\u51fa\u4fe1\u606f\u6709\u673a\u5730\u7ec4\u7ec7\u5728\u4e00\u8d77":115,"\u8f93\u51fa\u51fd\u6570":140,"\u8f93\u51fa\u521b\u5efa":[115,116],"\u8f93\u51fa\u5e8f\u5217\u7684\u7c7b\u578b":136,"\u8f93\u51fa\u5e8f\u5217\u7684\u8bcd\u8bed\u6570\u548c\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":139,"\u8f93\u51fa\u6240\u643a\u5e26\u7684\u5e8f\u5217\u4fe1\u606f":115,"\u8f93\u51fa\u6570\u636e\u548c\u8f93\u51fa\u68af\u5ea6":68,"\u8f93\u51fa\u6570\u636e\u548c\u8f93\u51fa\u68af\u5ea6\u7684\u8f6c\u6362":68,"\u8f93\u51fa\u6570\u636e\u662f\u5728\u4e0a\u6587\u4ecb\u7ecd\u7684":115,"\u8f93\u51fa\u6570\u636e\u6709\u673a\u5730\u7ec4\u7ec7\u5728\u4e00\u8d77":116,"\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7":[114,116],"\u8f93\u51fa\u7531":115,"\u8f93\u51fa\u7684\u5e8f\u5217\u4fe1\u606f":115,"\u8f93\u51fa\u7684\u68af\u5ea6":129,"\u8f93\u51fa\u7ed3\u679c\u53ef\u80fd\u4f1a\u968f\u7740\u5bb9\u5668\u7684\u6d88\u8017\u800c\u88ab\u5220\u9664":122,"\u8f93\u51fa\u88ab\u7ec4\u7ec7\u4e3a":115,"\u8f93\u51fa\u88ab\u7ec4\u7ec7\u4e3a\u4e00\u4e2a":115,"\u8f93\u51fa\u90fd\u4f1a\u5bf9\u5e94\u6709\u81ea\u5df1\u7684":[115,116],"\u8fbe\u5230\u5bb9\u707e\u7684\u76ee\u7684":37,"\u8fc7\u4e86\u4e00\u4e2a\u5f88\u7b80\u5355\u7684recurrent_group":137,"\u8fc7\u5b8c\u6240\u6709\u8bad\u7ec3\u6570\u636e\u5373\u4e3a\u4e00\u4e2apass":107,"\u8fc7\u7a0b\u4e2d\u6240\u6709\u65f6\u95f4\u6b65":67,"\u8fd0\u884c":116,"\u8fd0\u884c\u4e00\u4e2a":26,"\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u53ef\u4ee5\u8fdb\u884c\u7f16\u8bd1":101,"\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5":26,"\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5\u6d4b\u65f6\u9700\u8981\u7f16\u8bd1\u6574\u4e2a\u5de5\u7a0b":101,"\u8fd0\u884c\u5931\u8d25":131,"\u8fd0\u884c\u5b8c\u6210\u540e":119,"\u8fd0\u884c\u5b8c\u6bd5\u540e\u8f93\u51fa":133,"\u8fd0\u884c\u6027\u80fd\u5206\u6790\u7684\u65f6\u5019":133,"\u8fd0\u884c\u65e5\u5fd7":119,"\u8fd0\u884c\u65f6\u4e5f\u53ef\u80fd\u56e0\u4e3a\u591a\u7ebf\u7a0b\u4ea7\u751f\u6df7\u4e71\u4e0d\u53ef\u8bfb\u7684\u6027\u80fd\u5206\u6790\u7ed3\u679c":133,"\u8fd0\u884c\u65f6\u4f1a\u81ea\u52a8\u627e\u5230\u7cfb\u7edf\u4e2d\u5b89\u88c5\u7684cuda\u548ccudnn\u5e93\u8fdb\u884c\u7f16\u8bd1\u548c\u6267\u884c":27,"\u8fd0\u884c\u65f6c":116,"\u8fd0\u884c\u7684\u4e00\u4e9b\u53c2\u6570\u901a\u8fc7\u8fd9\u79cd\u65b9\u5f0f\u4f20\u9012\u5230\u5bb9\u5668\u5185":123,"\u8fd0\u884c\u8be5\u7f16\u8bd1\u5de5\u5177\u94fe\u9700\u8981\u4e00\u53f0":144,"\u8fd1\u671f\u76ee\u6807":68,"\u8fd4\u56de":[3,4,5,9,10,11,17,18,19,23,25],"\u8fd4\u56de\u7684\u662f":110,"\u8fd4\u56de\u7b2c\u4e8c\u6b65":88,"\u8fd4\u56de\u7b2ci\u4e2a\u8f93\u5165\u77e9\u9635":100,"\u8fd4\u56de\u7c7b\u578b":[3,4,5,9,10,11,18,19,25],"\u8fd4\u56depython\u7aef\u7684\u8ba1\u7b97\u7ed3\u679c":101,"\u8fd8\u4f1a\u4e0b\u8f7dmkl":27,"\u8fd8\u4f1a\u5f3a\u5236\u8bbe\u7f6e\u4e00\u4e9bpaddlepaddle\u53c2\u6570\u7684\u503c":142,"\u8fd8\u4f1a\u8f93\u51fa\u4e00\u4e2a":98,"\u8fd8\u53ef\u4ee5\u901a\u8fc7\u51cf\u5c0f\u5b66\u4e60\u7387\u6216\u8005\u5bf9\u6570\u636e\u8fdb\u884c\u5f52\u4e00\u5316\u5904\u7406\u6765\u89e3\u51b3\u8fd9\u7c7b\u95ee\u9898":107,"\u8fd8\u662f\u4ece":38,"\u8fd8\u662f\u865a\u62df\u673a":26,"\u8fd8\u9700\u8981\u5728\u8282\u70b9\u4e0a\u5b89\u88c5\u5bf9\u5e94\u7684gpu\u9a71\u52a8\u4ee5\u53cacuda":127,"\u8fd8\u9700\u8981\u91cd\u5199":101,"\u8fd9":107,"\u8fd98\u79cdlearning_rate_schedule\u53ca\u5176\u5bf9\u5e94\u5b66\u4e60\u7387\u8ba1\u7b97\u65b9\u5f0f\u5982\u4e0b":109,"\u8fd9\u4e00\u4e2a\u5e93":113,"\u8fd9\u4e00\u5757\u7684\u8017\u65f6\u6bd4\u4f8b\u771f\u7684\u592a\u9ad8":134,"\u8fd9\u4e00\u5c42\u8fdb\u884c\u5c01\u88c5":72,"\u8fd9\u4e00\u6570\u636e\u683c\u5f0f\u7684\u8f6c\u6362\u64cd\u4f5c":67,"\u8fd9\u4e00\u6982\u5ff5\u4e0d\u518d\u7410\u788e":72,"\u8fd9\u4e00\u8282\u5bf9\u56fe1\u4e2d\u9884\u6d4b\u4ee3\u7801\u7f16\u5199\u76845\u4e2a\u6b65\u9aa4\u8fdb\u884c\u4ecb\u7ecd\u548c\u8bf4\u660e":116,"\u8fd9\u4e00\u8ba1\u7b97\u5355\u5143":108,"\u8fd9\u4e00\u8fc7\u7a0b\u5bf9\u7528\u6237\u662f\u5b8c\u5168\u900f\u660e\u7684":139,"\u8fd9\u4e09\u4e2a\u5206\u652f":88,"\u8fd9\u4e24\u4e2a\u6307\u6807\u4ee3\u8868\u4e86\u67d0\u4e00\u4e2a\u51fd\u6570\u771f\u5b9e\u7684\u8fd0\u884c\u65f6\u95f4":133,"\u8fd9\u4e2a":[26,30],"\u8fd9\u4e2a\u4efb\u52a1\u7684\u914d\u7f6e\u4e3a":107,"\u8fd9\u4e2a\u4efb\u52a1\u7684dataprovider\u4e3a":107,"\u8fd9\u4e2a\u4f8b\u5b50\u6709\u4e24\u5904\u4e0d\u540c":101,"\u8fd9\u4e2a\u51fd\u6570\u672c\u8eab\u4f1a\u5728\u8ba1\u7b97\u524d\u5c06\u539f\u6570\u636e\u8f6c\u6362\u4e3a\u66f4\u9002\u5408\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u5185\u90e8\u683c\u5f0f":67,"\u8fd9\u4e2a\u51fd\u6570\u7684":140,"\u8fd9\u4e2a\u51fd\u6570\u8fdb\u884c\u53d8\u6362":137,"\u8fd9\u4e2a\u51fd\u6570\u9700\u8981\u8bbe\u7f6e":140,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u7684\u8fde\u63a5\u53c2\u6570\u4e0epaddle\u7684\u5176\u4ed6\u4e8c\u8fdb\u5236":72,"\u8fd9\u4e2a\u53c2\u6570\u4e5f\u4e0d\u4f1a\u4e00\u5e76\u5220\u9664":72,"\u8fd9\u4e2a\u5730\u5740\u6765\u8868\u793a\u6b64\u6b65\u9aa4\u6240\u6784\u5efa\u51fa\u7684\u955c\u50cf":123,"\u8fd9\u4e2a\u57fa\u7c7b":100,"\u8fd9\u4e2a\u5934\u6587\u4ef6\u4e0d\u5047\u8bbe\u5176\u4ed6\u6587\u4ef6\u7684\u5f15\u7528\u987a\u5e8f":72,"\u8fd9\u4e2a\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":139,"\u8fd9\u4e2a\u60c5\u51b5\u4e0b\u6240\u6709\u7684\u6587\u4ef6\u4f1a\u5b58\u5728\u6574\u7406\u8fc7\u7684\u7684\u6587\u4ef6\u76ee\u5f55":103,"\u8fd9\u4e2a\u63a5\u53e3\u9700\u8981\u505a\u5230":71,"\u8fd9\u4e2a\u6570\u636e\u4e5f\u88ab\u5355\u5c42rnn\u7f51\u7edc\u76f4\u63a5\u4f7f\u7528":137,"\u8fd9\u4e2a\u6587\u4ef6\u5177\u6709\u72ec\u7279\u7684\u8bed\u6cd5":71,"\u8fd9\u4e2a\u662f\u76ee\u524d\u63a8\u8350\u7684\u4f7f\u7528\u65b9\u6cd5":103,"\u8fd9\u4e2a\u73af\u5883\u53d8\u91cf\u5173\u95edopenmp\u4f18\u5316":133,"\u8fd9\u4e2a\u76ee\u5f55\u4e2d\u9664\u4e86":72,"\u8fd9\u4e2a\u793a\u4f8b":116,"\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u5355\u5143\u5c31\u53ebmemori":137,"\u8fd9\u4e2a\u7c7b\u7684\u53c2\u6570\u5305\u62ec":100,"\u8fd9\u4e2a\u7c7b\u9700\u8981\u7ee7\u627f":100,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u4e2d\u7684\u53e6\u4e00\u4e2a\u9879\u76ee\u662f":72,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u5305\u542b\u4e24\u4e2a\u9879\u76ee":72,"\u8fd9\u4e2a\u811a\u672c\u8c03\u7528":26,"\u8fd9\u4e2a\u8f93\u5165\u4e0d\u53c2\u4e0e":101,"\u8fd9\u4e2a\u8fc7\u7a0b\u5bf9\u7528\u6237\u4e5f\u662f\u900f\u660e\u7684":139,"\u8fd9\u4e2a\u8fc7\u7a0b\u9664\u4e86\u7f16\u8bd1paddlepaddle\u4e3a":98,"\u8fd9\u4e2a\u9009\u62e9":[67,68],"\u8fd9\u4e2a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u751f\u6210\u4e00\u7cfb\u5217\u6743\u91cd":140,"\u8fd9\u4e2a\u9759\u6001\u5e93\u5305\u542b\u4e86paddle\u7684\u5168\u90e8\u7b26\u53f7":72,"\u8fd9\u4e2ainstance\u53ef\u4ee5\u662f\u5355\u4e2a\u503c":38,"\u8fd9\u4e2aissu":26,"\u8fd9\u4e2ajob\u624d\u7b97\u6210\u529f\u7ed3\u675f":123,"\u8fd9\u4e2alayer\u7684\u8f93\u51fa\u4f1a\u4f5c\u4e3a\u6574\u4e2a":139,"\u8fd9\u4e5f\u4f1a\u6781\u5927\u51cf\u5c11\u6570\u636e\u8bfb\u5165\u7684\u8017\u65f6":107,"\u8fd9\u4e9b\u4f1a\u5728":[67,68],"\u8fd9\u4e9b\u4f8b\u5b50\u90fd\u53ef\u4ee5\u5728":120,"\u8fd9\u4e9b\u51fd\u6570\u4f1a\u5c06\u5bf9\u5e94\u5185\u5bb9\u6dfb\u52a0\u5230":101,"\u8fd9\u4e9b\u51fd\u6570\u4f1a\u6839\u636e\u8f93\u5165\u53c2\u6570\u91cd\u65b0\u8bbe\u7f6e\u5185\u90e8\u548c\u5916\u90e8\u5b58\u50a8":68,"\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":38,"\u8fd9\u4e9b\u53c2\u6570\u53ef\u4ee5\u901a\u8fc7":117,"\u8fd9\u4e9b\u53c2\u6570\u7684\u5177\u4f53\u63cf\u8ff0":123,"\u8fd9\u4e9b\u540d\u5b57\u5fc5\u987b\u8981\u5199\u5bf9":100,"\u8fd9\u4e9b\u6570\u636e\u4f1a\u88ab\u7528\u6765\u66f4\u65b0\u53c2\u6570":107,"\u8fd9\u4e9b\u6570\u636e\u4f7f\u7528\u7684\u5185\u5b58\u4e3b\u8981\u548c\u4e24\u4e2a\u53c2\u6570\u6709\u5173\u7cfb":107,"\u8fd9\u4e9b\u7279\u5f81\u6570\u636e\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u6709\u610f\u4e49\u7684":137,"\u8fd9\u4e9b\u955c\u50cf\u4e5f\u53ef\u4ee5\u4ece":88,"\u8fd9\u4efd\u6559\u7a0b\u5c55\u793a\u4e86\u5982\u4f55\u5728paddlepaddle\u4e2d\u5b9e\u73b0\u4e00\u4e2a\u81ea\u5b9a\u4e49\u7684\u7f51\u7edc\u5c42":100,"\u8fd9\u4f1a\u63d0\u793a\u5f53\u524d\u76ee\u5f55\u7684\u4e00\u4e9b\u53d8\u5316":98,"\u8fd9\u4f1a\u7ed9\u8bc4\u5ba1\u4eba\u5e26\u6765\u5f88\u5927\u56f0\u6270":98,"\u8fd9\u4f1a\u81ea\u52a8\u8fdb\u884c\u7f51\u7edc\u914d\u7f6e\u4e2d\u58f0\u660e\u7684\u6fc0\u6d3b\u64cd\u4f5c":100,"\u8fd9\u4fbf\u662f\u4e00\u79cd\u53cc\u5c42rnn\u7684\u8f93\u5165\u6570\u636e":137,"\u8fd9\u51e0\u4e2a\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":27,"\u8fd9\u53e5\u8868\u793a\u4f7f\u7528\u57fa\u7c7b":101,"\u8fd9\u53ef\u4ee5\u5e2e\u60a8\u7701\u6389\u82b1\u4e00\u5c0f\u65f6\u5b89\u88c5\u548c\u914d\u7f6e\u5404\u79cd\u5f00\u53d1\u5de5\u5177":26,"\u8fd9\u53ef\u4ee5\u8ba9\u5176\u4ed6\u4eba\u77e5\u9053\u8fd9\u6b21\u63d0\u4ea4\u505a\u4e86\u54ea\u4e9b\u6539\u53d8":98,"\u8fd9\u53ef\u4ee5\u901a\u8fc7":98,"\u8fd9\u548c\u5355\u5c42rnn\u7684\u914d\u7f6e\u662f\u7b49\u4ef7\u7684":137,"\u8fd9\u56db\u4e2a\u5e8f\u5217\u53c8\u5206\u522b\u542b\u67093":115,"\u8fd9\u56db\u6761\u6570\u636e\u540c\u65f6\u5904\u7406\u7684\u53e5\u5b50\u6570\u91cf\u4e3a":137,"\u8fd9\u5728\u6784\u9020\u975e\u5e38\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u662f\u6709\u7528\u7684":140,"\u8fd9\u5bf9\u4e8e\u901a\u5e38\u7684java\u7684\u5f00\u53d1\u8005\u6765\u8bf4":71,"\u8fd9\u5c06\u4f1a\u5bfc\u81f4\u5355\u5143\u6d4b\u8bd5\u51fa\u9519":101,"\u8fd9\u5c06\u4f1a\u5bfc\u81f4\u7f16\u8bd1\u51fa\u9519":101,"\u8fd9\u610f\u5473\u7740":140,"\u8fd9\u610f\u5473\u7740\u9664\u4e86\u6307\u5b9adevic":131,"\u8fd9\u65f6":[107,116],"\u8fd9\u65f6\u5728\u4f7f\u7528":109,"\u8fd9\u65f6\u7684":115,"\u8fd9\u65f6\u7684\u9700\u8981\u540c\u65f6\u63d0\u4f9b":115,"\u8fd9\u65f6\u884c\u504f\u79fb\u548c\u5217\u53f7\u6307\u5b9a\u7684\u5143\u7d20\u9ed8\u8ba4\u5176\u503c\u4e3a1":115,"\u8fd9\u65f6\u8fdb\u884c\u77e9\u9635\u4e58\u6cd5\u8fd0\u7b97\u5c31\u53ef\u80fd\u5bfc\u81f4\u6d6e\u70b9\u6570\u6ea2\u51fa":107,"\u8fd9\u65f6\u9700\u8981\u8c03\u7528\u521b\u5efa\u5e8f\u5217\u4fe1\u606f\u548c\u4e3a":115,"\u8fd9\u662f\u4e00\u79cd\u6309\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5206\u6bb5\u53d6\u503c\u7684\u5b66\u4e60\u7387\u9000\u706b\u65b9\u6cd5":109,"\u8fd9\u662f\u4e00\u79cd\u6309\u5df2\u8bad\u7ec3pass\u6570\u5206\u6bb5\u53d6\u503c\u7684\u5b66\u4e60\u7387\u9000\u706b\u65b9\u6cd5":109,"\u8fd9\u662f\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":136,"\u8fd9\u662f\u56e0\u4e3a":71,"\u8fd9\u662f\u5f00\u6e90\u793e\u533a\u7684\u57fa\u672c\u793c\u8c8c":98,"\u8fd9\u662f\u666e\u901a\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u4ee3\u7801":137,"\u8fd9\u662f\u76ee\u524dcmake\u5bfb\u627epython\u7684\u903b\u8f91\u5b58\u5728\u7f3a\u9677":104,"\u8fd9\u6837":[72,118],"\u8fd9\u6837\u4e0b\u4e00\u4e2acpu":68,"\u8fd9\u6837\u4fdd\u5b58\u5728\u5206\u5e03\u5f0f\u5b58\u50a8\u4e2d\u7684\u6570\u636e\u53ef\u4ee5\u88ab\u96c6\u7fa4\u4e2d\u7684\u6bcf\u4e2a\u8282\u70b9\u8bfb\u53d6\u5230":117,"\u8fd9\u6837\u4fdd\u8bc1":88,"\u8fd9\u6837\u4fdd\u8bc1\u8fd0\u884c\u7ed3\u675f\u4e4b\u540e\u7684":26,"\u8fd9\u6837\u505a\u53ef\u4ee5\u6781\u5927\u7684\u51cf\u5c11\u5185\u5b58\u5360\u7528":107,"\u8fd9\u6837\u53ef\u4ee5\u514d\u53bb\u5355\u72ec\u5b89\u88c5\u7f16\u8bd1\u4f9d\u8d56\u7684\u6b65\u9aa4":27,"\u8fd9\u6837\u53ef\u4ee5\u51cf\u5c0fgpu\u5185\u5b58":131,"\u8fd9\u6837\u5982\u679c\u9047\u5230\u95ee\u9898":26,"\u8fd9\u6837\u5bb9\u5668\u7684":123,"\u8fd9\u6837\u5c31\u53ef\u4ee5\u5728\u4e91\u7aef\u6267\u884c\u591a\u79cd\u6570\u636e\u7c7b\u8ba1\u7b97\u4efb\u52a1":38,"\u8fd9\u6837\u5df2\u7ecf\u4f20\u8f93\u6210\u529f\u7684\u90e8\u5206\u5c31\u4e0d\u7528\u91cd\u65b0\u4f20\u8f93\u4e86":54,"\u8fd9\u6837\u5e26\u6765\u7684\u597d\u5904\u5c31\u662f\u4e0d\u9700\u8981\u4e00\u76f4\u6e05\u7a7amemori":68,"\u8fd9\u6837\u5f53\u8be5pull":98,"\u8fd9\u6837\u65e2\u4f7f\u5f97\u6700\u7ec8\u4fdd\u5b58\u7684\u53c2\u6570\u683c\u5f0f\u4e0epaddlepaddle\u4e00\u81f4":68,"\u8fd9\u6837\u6781\u5927\u5730\u63d0\u9ad8\u4e86\u8ba1\u7b97\u7684\u5e76\u884c\u6027":118,"\u8fd9\u6837\u7684\u88c5\u9970\u5668":100,"\u8fd9\u6837\u7684\u8bdd":122,"\u8fd9\u6837\u8bad\u7ec3\u6587\u4ef6\u7684\u4e2a\u6570\u4f1a\u6bd4\u8f83\u591a":117,"\u8fd9\u6b63\u662f\u5b83\u4eec\u901f\u5ea6\u5feb\u7684\u539f\u56e0":134,"\u8fd9\u7528\u4e8e\u5728\u591a\u7ebf\u7a0b\u548c\u591a\u673a\u4e0a\u66f4\u65b0\u53c2\u6570":100,"\u8fd9\u79cd\u521d\u59cb\u5316\u65b9\u5f0f\u5728\u4e00\u822c\u60c5\u51b5\u4e0b\u4e0d\u4f1a\u4ea7\u751f\u5f88\u5dee\u7684\u7ed3\u679c":109,"\u8fd9\u79cd\u60c5\u51b5\u4e0b\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":100,"\u8fd9\u79cd\u60c5\u51b5\u591a\u51fa\u73b0\u5728\u4f7f\u7528\u591a\u7ebf\u7a0b\u9884\u6d4b\u65f6":116,"\u8fd9\u79cd\u60c5\u51b5\u5e38\u5e38\u53d1\u751f\u5728":107,"\u8fd9\u79cd\u65b9\u5f0f\u5bf9\u5185\u5b58\u6d88\u8017\u8f83\u5927":108,"\u8fd9\u79cd\u65b9\u5f0f\u5fc5\u987b\u4f7f\u7528paddle\u5b58\u50a8\u7684\u6a21\u578b\u8def\u5f84\u683c\u5f0f":131,"\u8fd9\u79cd\u65b9\u5f0f\u6700\u4e3a\u7b80\u4fbf":113,"\u8fd9\u79cd\u751f\u6210\u6280\u672f\u53ea\u7528\u4e8e\u7c7b\u4f3c\u89e3\u7801\u5668\u7684\u751f\u6210\u8fc7\u7a0b":140,"\u8fd9\u79cd\u7c7b\u578b\u7684\u8f93\u5165\u5fc5\u987b\u901a\u8fc7":139,"\u8fd9\u79cd\u94fe\u63a5\u65b9\u5f0f\u4e3b\u8981\u7528\u4e8e\u79fb\u52a8\u7aef\u9884\u6d4b":113,"\u8fd9\u79cd\u96c6\u7fa4\u8282\u70b9\u7ba1\u7406\u65b9\u5f0f\u4f1a\u5728\u5c06\u6765\u4f7f\u7528":123,"\u8fd9\u7bc7":28,"\u8fd9\u7bc7\u6587\u6863":98,"\u8fd9\u7bc7\u6587\u6863\u4e4b\u540e\u90e8\u5206\u4f1a\u4f7f\u7528":116,"\u8fd9\u7bc7\u6587\u6863\u4e4b\u540e\u90e8\u5206\u4f1a\u7edf\u4e00\u4f7f\u7528":115,"\u8fd9\u7bc7\u6587\u6863\u4e4b\u540e\u90e8\u5206\u5c06\u4f1a\u7edf\u4e00\u4f7f\u7528":115,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd":116,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd\u5728":144,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd\u5728\u4f7f\u7528":115,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd\u57fa\u4e8e":26,"\u8fd9\u7bc7\u6587\u6863\u7684\u4e4b\u540e\u90e8\u5206\u4f1a\u4f7f\u7528":116,"\u8fd9\u7bc7\u6587\u7ae0":26,"\u8fd9\u7ec4\u8bed\u4e49\u76f8\u540c\u7684\u793a\u4f8b\u914d\u7f6e\u5982\u4e0b":137,"\u8fd9\u884c\u547d\u4ee4\u4e2d":133,"\u8fd9\u901a\u8fc7\u83b7\u5f97\u53cd\u5411\u5faa\u73af\u7f51\u7edc\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":140,"\u8fd9\u90fd\u9700\u8981\u8fd9\u4e2a\u63a5\u53e3\u6309\u7167\u7ea6\u5b9a\u4fd7\u6210\u7684\u89c4\u5219\u6765\u6ce8\u91ca\u5b8c\u5907":71,"\u8fd9\u91cc":[27,68,98,109,123,140],"\u8fd9\u91cc\u4e0d\u518d\u8d58\u8ff0":101,"\u8fd9\u91cc\u4ecb\u7ecdc":116,"\u8fd9\u91cc\u4f7f\u7528\u4e86\u7528":133,"\u8fd9\u91cc\u4f7f\u7528\u4e86paddlepaddle\u9884\u5b9a\u4e49\u597d\u7684rnn\u5904\u7406\u51fd\u6570":137,"\u8fd9\u91cc\u4f7f\u7528\u7b80\u5355\u7684":107,"\u8fd9\u91cc\u5c06\u4ecb\u7ecdpaddlepaddle\u7684\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":110,"\u8fd9\u91cc\u6211\u4eec\u5c55\u793a\u4e00\u4efd\u7b80\u5316\u8fc7\u7684\u4ee3\u7801":100,"\u8fd9\u91cc\u6211\u4eec\u901a\u8fc7\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2ajob\u6765\u4e0b\u8f7d\u5e76\u5207\u5272\u6570\u636e":123,"\u8fd9\u91cc\u6709\u4e24\u79cd\u6709\u6548\u7684\u89e3\u51b3\u65b9\u6cd5":107,"\u8fd9\u91cc\u68c0\u9a8c\u8fd0\u884c\u65f6\u95f4\u6a21\u578b\u7684\u6536\u655b":119,"\u8fd9\u91cc\u7684dockerimage\u4f5c\u4e3a\u7f16\u8bd1\u73af\u5883\u4ee5\u652f\u6301\u66f4\u591a\u7684linux":88,"\u8fd9\u91cc\u7684eigentensor\u4e4b\u95f4\u7684\u8fd0\u7b97\u53ea\u662f\u6539\u53d8\u4e86\u539f\u6709tensor\u4e2d\u7684\u6570\u636e":102,"\u8fd9\u91cc\u9009\u62e90":88,"\u8fd9\u91cc\u9700\u8981\u7528\u6237\u989d\u5916\u6ce8\u610f":37,"\u8fd9\u9700\u8981\u8054\u5408\u6211\u4eec\u7b2c\u4e8c\u8282":133,"\u8fdb\u4e00\u6b65\u4f18\u5316":68,"\u8fdb\u4e3b\u4ed3\u5e93\u540e":98,"\u8fdb\u5165":88,"\u8fdb\u5165\u5bb9\u5668":122,"\u8fdb\u5165\u5bf9\u5e94\u7684\u76ee\u5f55":104,"\u8fdb\u7a0b\u542f\u52a8\u7684\u5fc5\u8981\u53c2\u6570":123,"\u8fdb\u7a0b\u7684":119,"\u8fdb\u7a0b\u7684\u542f\u52a8\u53c2\u6570":123,"\u8fdb\u7a0b\u7684\u8fd0\u884c\u73af\u5883":123,"\u8fdb\u7a0b\u9700\u8981\u7684":123,"\u8fdb\u800c\u591a\u673a":133,"\u8fdb\u800c\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u547d\u4ee4\u5f00\u542f\u4e00\u4e2ahttp\u670d\u52a1":133,"\u8fdb\u800c\u6307\u5b9a\u4e86python\u53ef\u6267\u884c\u6587\u4ef6\u7684\u8def\u5f84":133,"\u8fdb\u800c\u8fdb\u884c\u4ee3\u7801\u8bc4\u5ba1":88,"\u8fdb\u884c\u4e86":137,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6848":123,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6cd5":123,"\u8fdb\u884c\u524d\u5411\u8ba1\u7b97":116,"\u8fdb\u884c\u56de\u590d":98,"\u8fdb\u884c\u5e8f\u5217\u5316":116,"\u8fdb\u884c\u5f00\u53d1":98,"\u8fdb\u884c\u62c6\u89e3":137,"\u8fdb\u884c\u6fc0\u6d3b\u64cd\u4f5c":100,"\u8fdb\u884c\u7f16\u8bd1\u548c\u5b89\u88c5":142,"\u8fdb\u884c\u8bad\u7ec3":116,"\u8fdb\u884c\u8bbe\u7f6e":101,"\u8fdb\u884c\u94fe\u63a5":113,"\u8fdb\u884c\u9884\u6d4b\u4f9d\u8d56\u4e8e\u5c06":113,"\u8fdb\u884c\u9884\u6d4b\u65f6":116,"\u8fdb\u884cpython\u4e0ec":133,"\u8fdb\u9636\u4f7f\u7528":141,"\u8fdb\u9636\u6307\u5357":110,"\u8fde\u63a5":139,"\u8fde\u63a5\u5230pserver\u7684\u7aef\u53e3":117,"\u8fde\u63a5\u5230pserver\u7684\u7aef\u53e3\u4e2a\u6570":117,"\u9000\u51fa\u5bb9\u5668":122,"\u9009\u62e9\u4e0b\u8f7d\u4f7f\u7528\u4e0d\u540c\u7684blas\u5e93\u7684docker\u955c\u50cf":28,"\u9009\u62e9\u662f\u5426\u7f16\u8bd1mkl":68,"\u9009\u62e9\u76ee\u6807\u5206\u652f":98,"\u9009\u62e9\u8def\u5f84\u6765\u52a8\u6001\u52a0\u8f7dnvidia":129,"\u9009\u62e9\u9700\u8981\u53d1\u5e03\u7684\u7248\u672c":88,"\u9009\u9879":[26,27],"\u900f\u4f20\u7528\u6237\u8eab\u4efd\u7684\u529e\u6cd5":54,"\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":128,"\u901a\u5e38":[72,101,133],"\u901a\u5e38\u4f1a\u4f7f\u7528\u73af\u5883\u53d8\u91cf\u914d\u7f6ejob\u7684\u914d\u7f6e\u4fe1\u606f":123,"\u901a\u5e38\u4f1a\u4f7f\u7528mapreduce\u4efb\u52a1\u7684\u8f93\u51fa\u7ed3\u679c\u4f5c\u4e3a\u8bad\u7ec3\u7ed3\u679c":117,"\u901a\u5e38\u4f7f\u7528\u7a00\u758f\u8bad\u7ec3\u6765\u52a0\u901f\u8ba1\u7b97\u8fc7\u7a0b":131,"\u901a\u5e38\u4f7f\u7528cento":30,"\u901a\u5e38\u505a\u6cd5\u662f\u4ece\u4e00\u4e2a\u6bd4\u8f83\u5927\u7684learning_rate\u5f00\u59cb\u8bd5":109,"\u901a\u5e38\u5305\u542b\u4e00\u4e2acpu\u7248\u672c\u548c\u4e00\u4e2agpu\u7248\u672c":88,"\u901a\u5e38\u540d\u5b57\u662f":98,"\u901a\u5e38\u60c5\u51b5\u4e0b":134,"\u901a\u5e38\u6211\u4eec\u4f1a\u5b89\u88c5ceph\u7b49\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u6765\u5b58\u50a8\u8bad\u7ec3\u6570\u636e":122,"\u901a\u5e38\u6307\u5c06\u4e00\u4e2a\u6574\u4f53\u62c6\u5206\u6210\u591a\u4efd\u7684\u5176\u4e2d\u7684\u4e00\u4efd":37,"\u901a\u5e38\u6709\u4e24\u4e2a\u65b9\u6cd5\u6765\u6784\u5efa\u57fa\u4e8e":144,"\u901a\u5e38\u7528\u4e8e\u8868\u793a\u79bb\u6563\u7684\u7c7b\u522b\u6807\u7b7e":115,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u4f7f\u7528":140,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u5c06\u914d\u7f6e\u5b58\u4e8e":100,"\u901a\u5e38\u8981\u6c42\u65f6\u95f4\u6b65\u4e4b\u95f4\u5177\u6709\u4e00\u4e9b\u4f9d\u8d56\u6027":137,"\u901a\u5e38\u89c2\u5bdf\u70ed\u70b9\u51fd\u6570\u95f4\u7684\u8c03\u7528\u5173\u7cfb":133,"\u901a\u5e38\u90fd\u4f1a\u4f7f\u7528\u4e0b\u9762\u8fd9\u4e9b\u547d\u4ee4\u884c\u53c2\u6570":131,"\u901a\u5e38\u9700\u8981\u53bb\u6389\u7f51\u7edc\u4e2d\u7684":116,"\u901a\u7528":128,"\u901a\u8fc7":[98,100,101,107,115,137],"\u901a\u8fc7\u4e24\u4e2a\u5d4c\u5957\u7684":139,"\u901a\u8fc7\u4f7f\u7528":27,"\u901a\u8fc7\u4f7f\u7528\u8fd9\u4e9bapi":67,"\u901a\u8fc7\u51fd\u6570":123,"\u901a\u8fc7\u547d\u4ee4\u884c\u53c2\u6570":107,"\u901a\u8fc7\u591a\u4e2a\u7ebf\u7a0b\u5171\u4eab\u540c\u4e00\u4e2a\u6a21\u578b\u6765\u51cf\u5c11\u5185\u5b58\u5f00\u9500":116,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u7684\u8f93\u51fa":139,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa":139,"\u901a\u8fc7\u6240\u6709\u5355\u5143\u6d4b\u8bd5":98,"\u901a\u8fc7\u6a21\u578b\u63a8\u65adapi\u7684\u5b9e\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u6837\u4f8b":72,"\u901a\u8fc7\u7075\u6d3b\u4f7f\u7528\u4ee5\u4e0a\u4e24\u4e2a\u63a5\u53e3":116,"\u901a\u8fc7\u7ec4\u5408\u4e0d\u540c\u7684layer":110,"\u901a\u8fc7\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u6765\u6307\u5b9a":100,"\u901a\u8fc7\u8ba1\u7b97\u8282\u70b9\u548c\u53c2\u6570\u670d\u52a1\u5668\u7684\u5206\u5e03\u5f0f\u534f\u4f5c":118,"\u901a\u8fc7\u8be5\u53c2\u6570\u53ef\u83b7\u53d6\u5230\u8f93\u5165\u8f93\u51fa\u4ee5\u53ca\u5c5e\u6027":101,"\u901a\u8fc7\u8c03\u7528":[115,116],"\u901a\u8fc7\u8c03\u7528\u4ee5\u4e0b\u63a5\u53e3\u521b\u5efa\u7a00\u758f\u77e9\u9635":115,"\u901a\u8fc7data":139,"\u901a\u8fc7ssh\u7b49\u65b9\u5f0f\u767b\u5f55\u5230raspberri":144,"\u903b\u8f91\u4e0a\u9ad8\u4e8e\u4e8c\u7ef4\u7684\u6570\u636e":115,"\u903b\u8f91\u5212\u4e0a\u6587\u4ef6\u5206\u5757\u7684\u5355\u4f4d":54,"\u9047\u5230\u8be5\u9519\u8bef\u65f6":108,"\u9075\u5b88\u4ee5\u4e0b\u7ea6\u5b9a":98,"\u9075\u5faa\u4ee5\u4e0b\u6d41\u7a0b":88,"\u90a3\u4e48":[72,100,139],"\u90a3\u4e480\u5c42\u5e8f\u5217\u5373\u4e3a\u4e00\u4e2a\u8bcd\u8bed":139,"\u90a3\u4e48\u4f1a\u643a\u5e26\u6709":115,"\u90a3\u4e48\u53ef\u4ee5\u8ba4\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":109,"\u90a3\u4e48\u5728":101,"\u90a3\u4e48\u5982\u4f55\u5224\u65ad\u8bad\u7ec3\u4e0d\u6536\u655b\u5462":109,"\u90a3\u4e48\u5bf9\u5e94\u7684\u5185\u90e8\u5b58\u50a8\u4e5f\u4f1a\u4e0e\u5b83\u4eec\u5171\u4eab\u5185\u5b58":68,"\u90a3\u4e48\u5c31\u4f1a\u4f7f":68,"\u90a3\u4e48\u5e38\u6570\u8f93\u51fa\u6240\u80fd\u8fbe\u5230\u7684\u6700\u5c0fcost\u662f":109,"\u90a3\u4e48\u6211\u4eec\u4e5f\u5c31\u4e0d\u9700\u8981":26,"\u90a3\u4e48\u6211\u4eec\u53ef\u4ee5\u5224\u65ad\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":109,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528":140,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u65b9\u6cd5":140,"\u90a3\u4e48\u6536\u655b\u53ef\u80fd\u5f88\u6162":109,"\u90a3\u4e48\u6700\u597d\u5c06\u6570\u636e\u6587\u4ef6\u5728\u6bcf\u6b21\u8bfb\u53d6\u4e4b\u524d\u505a\u4e00\u6b21shuffl":107,"\u90a3\u4e48\u7528\u6237\u9700\u8981\u62c9\u53d6\u6240\u6709\u7684\u8fdc\u7a0b\u5206\u652f\u5230\u672c\u673a":104,"\u90a3\u4e48\u7f16\u8bd1\u8fc7\u7a0b\u53ea\u4f1a\u4ea7\u751f":26,"\u90a3\u4e48\u8bad\u7ec3\u6709\u53ef\u80fd\u4e0d\u6536\u655b":109,"\u90a3\u4e48\u8be5\u4f18\u5316\u7b97\u6cd5\u81f3\u5c11\u9700\u8981":107,"\u90a3\u4e48fc1\u548cfc2\u5c42\u5c06\u4f1a\u4f7f\u7528\u7b2c1\u4e2agpu\u6765\u8ba1\u7b97":131,"\u90a3\u4e5f\u5c31\u4e0d\u9700\u8981\u6025\u7740\u4f18\u5316\u6027\u80fd\u5566":134,"\u90a3\u4f30\u8ba1\u8fd9\u91cc\u7684\u6f5c\u529b\u5c31\u6ca1\u5565\u597d\u6316\u7684\u4e86":134,"\u90a3\u51cf\u5c11\u5b66\u4e60\u738710\u500d\u7ee7\u7eed\u8bd5\u9a8c":109,"\u90a3\u6211\u4f1a\u671f\u671b\u5206\u6790\u5de5\u5177\u7edf\u8ba1\u5230\u901f\u5ea6\u662f100gb":134,"\u90a3\u7a0b\u5e8f\u5206\u6790\u5de5\u5177\u662f\u5fc5\u4e0d\u53ef\u5c11\u7684\u5229\u5668":134,"\u90fd\u4e0d\u4f1a\u60f3\u8981\u77e5\u9053next":68,"\u90fd\u4e0d\u9700\u8981":26,"\u90fd\u4f1a\u4ea7\u751f\u5f53\u524d\u5c42\u72b6\u6001\u7684\u6240\u6709\u7ee7\u627f\u7ed3\u679c":129,"\u90fd\u4f1a\u7ba1\u7406\u7ef4\u62a4\u4e00\u4efd\u8bad\u7ec3\u597d\u7684\u6a21\u578b":116,"\u90fd\u4f1a\u9020\u6210\u8bad\u7ec3\u4e2d\u7684\u6570\u636e\u4ecec":107,"\u90fd\u4f7f\u7528":115,"\u90fd\u53ea\u662f\u4ecb\u7ecd\u53cc\u5c42rnn\u7684api\u63a5\u53e3":137,"\u90fd\u53ef\u4ee5\u8fd0\u884c":26,"\u90fd\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528\u4e0b\u9762\u7684\u63a5\u53e3\u4e3a\u539f\u6709\u7684\u6570\u636e\u8f93\u5165\u9644\u52a0\u4e0a\u5e8f\u5217\u4fe1\u606f":115,"\u90fd\u5e94\u4f7f\u7528c":115,"\u90fd\u662f\u4e94\u4f4d\u7684\u6570\u5b57":38,"\u90fd\u662f\u4ee5ext\u5f00\u5934":68,"\u90fd\u662f\u5bf9layer1\u5143\u7d20\u7684\u62f7\u8d1d":136,"\u90fd\u662f\u5c06\u6bcf\u4e00\u53e5\u5206\u597d\u8bcd\u540e\u7684\u53e5\u5b50":137,"\u90fd\u662fabi\u8c03\u7528\u6807\u51c6\u7684":71,"\u90fd\u7528":98,"\u90fd\u7ee7\u627f\u4e8epaddlepaddle\u7684\u57fa\u7c7b":67,"\u90fd\u9700\u8981\u5199\u63d0\u4ea4\u8bf4\u660e":98,"\u90fd\u9700\u8981\u8c03\u7528\u4e00\u6b21":100,"\u914d\u5236\u7f16\u8bd1\u9009\u9879":113,"\u914d\u7f6e\u6253\u5f00":134,"\u914d\u7f6e\u6587\u4ef6\u63a5\u53e3\u662ffc_layer":100,"\u914d\u7f6e\u6587\u4ef6\u91cc\u52a0\u4e24\u884c":26,"\u914d\u7f6e\u7684\u65b9\u6cd5\u53c2\u8003":54,"\u914d\u7f6e\u7b80\u5355\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u4f8b\u5b50":140,"\u914d\u7f6e\u7f51\u7edc\u5c42\u7684\u8f93\u5165":100,"\u914d\u7f6eapi":136,"\u91c7\u7528\u5747\u5300\u5206\u5e03\u6216\u8005\u9ad8\u65af\u5206\u5e03\u521d\u59cb\u5316":129,"\u91c7\u7528multi":109,"\u91ca\u653e\u5bf9paramters\u5185\u5b58\u7684\u9501\u5b9a":37,"\u91cc":26,"\u91cc\u53ef\u4ee5\u6807\u51c6\u5316\u7f16\u8bd1\u73af\u5883":26,"\u91cc\u5b8c\u6210":101,"\u91cc\u6240\u6709\u7684\u7b26\u53f7\u90fd\u5199\u5165\u81ea\u5df1\u7684\u7a0b\u5e8f\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cc":71,"\u91cc\u7684":26,"\u91cc\u7684\u65e5\u5fd7":119,"\u91cc\u7684\u6e90\u7801":26,"\u91cc\u8fd0\u884c\u7684\u7f16\u8bd1\u5de5\u5177\u5b9e\u9645\u4e0a\u90fd\u662f\u5728\u672c\u673a\u7684":26,"\u91cc\u9009\u62e9\u9700\u8981\u53d1\u5e03\u7684\u5206\u652f":88,"\u91cc\u9762":101,"\u91cc\u9762\u6db5\u76d6\u4e86\u4ea4\u53c9\u7f16\u8bd1android\u7248paddlepaddle\u5e93\u9700\u8981\u7684\u6240\u6709\u7f16\u8bd1\u5de5\u5177":142,"\u91cc\u9762\u6dfb\u52a0":68,"\u91ccstep\u7684\u5185\u5bb9":107,"\u91cd\u5199\u7236\u7c7blayer\u7684":68,"\u91cd\u547d\u540d\u6210":71,"\u91cd\u65b0\u7f16\u8bd1paddlepaddl":134,"\u9488\u5bf9\u4e0d\u540c\u7684":143,"\u9488\u5bf9\u4efb\u52a1\u8fd0\u884c\u5b8c\u6210\u540e\u5bb9\u5668\u81ea\u52a8\u9000\u51fa\u7684\u573a\u666f":122,"\u9488\u5bf9\u5185\u5b58\u548c\u663e\u5b58":107,"\u94fe\u63a5":113,"\u94fe\u63a5\u4e2d\u627e\u5230":30,"\u94fe\u63a5\u4f55\u79cdblas\u5e93\u7b49":27,"\u94fe\u63a5\u5230\u81ea\u5df1\u7684\u7a0b\u5e8f\u91cc":71,"\u94fe\u63a5\u76f8\u5bf9\u5bb9\u6613":113,"\u94fe\u63a5\u9009\u9879":113,"\u94fe\u63a5\u9759\u6001\u5e93":113,"\u9519\u8bef":104,"\u9519\u8bef\u5904\u7406":71,"\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662f\u8fd4\u56de\u503c":71,"\u9519\u8bef\u5904\u7406\u7684\u65b9\u5f0f\u4e5f\u4e0d\u5c3d\u76f8\u540c":71,"\u9519\u8bef\u7684define_py_data_sources2\u7c7b\u4f3c":109,"\u952e\u6765\u542f\u52a8\u7f16\u8bd1\u4e86":26,"\u955c\u50cf\u91cc\u6709paddlepaddle\u7684\u6e90\u7801\u4e0edemo":122,"\u957f\u5ea6":107,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570":140,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u8f93\u51famemori":140,"\u9644\u52a0\u4e0a\u5e8f\u5217\u4fe1\u606f":115,"\u9650\u5236\u5957\u63a5\u5b57\u53d1\u9001\u7f13\u51b2\u533a\u7684\u5927\u5c0f":129,"\u9650\u5236\u5957\u63a5\u5b57\u63a5\u6536\u7f13\u51b2\u533a\u7684\u5927\u5c0f":129,"\u9664\u4e86\u53ef\u4ee5\u81ea\u52a8\u7f16\u8bd1\u6587\u6863":103,"\u9664\u4e86boot_lay":137,"\u9664\u6784\u9020\u67d0\u79cd\u7c7b\u578b\u7684\u51fd\u6570":72,"\u9664\u6b64\u4e4b\u5916":107,"\u9664\u96f6\u7b49\u95ee\u9898":107,"\u968f\u540e\u53ef\u4ee5\u7528\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf\u5f00\u59cbbuild":98,"\u968f\u673a\u6570\u7684\u79cd\u5b50":129,"\u968f\u673a\u6570seed":128,"\u9694\u5f00":117,"\u96c6\u6210\u5230":67,"\u96c6\u6210\u5230paddlepaddl":68,"\u96c6\u675f\u641c\u7d22\u4f7f\u7528\u5e7f\u5ea6\u4f18\u5148\u641c\u7d22\u7684\u65b9\u5f0f\u6784\u5efa\u67e5\u627e\u6811":129,"\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673a\u4f7f\u7528cpu\u7684paddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":122,"\u96c6\u7fa4\u4e2d\u7684\u6bcf\u53f0\u8ba1\u7b97\u673a\u901a\u5e38\u88ab\u6210\u4e3a\u4e00\u4e2a":127,"\u96c6\u7fa4\u4efb\u52a1":119,"\u96c6\u7fa4\u4f5c\u4e1a\u5c06\u4f1a\u5728\u51e0\u79d2\u540e\u542f\u52a8":119,"\u96c6\u7fa4\u6d4b\u8bd5":128,"\u96c6\u7fa4\u7ba1\u7406\u5de5\u5177":120,"\u96c6\u7fa4\u8bad\u7ec3":128,"\u96c6\u7fa4\u8bad\u7ec3\u4e0e\u9884\u6d4b":106,"\u96c6\u7fa4\u8fdb\u7a0b":119,"\u9700\u52a0\u8be5\u6a21\u677f\u53c2\u6570":101,"\u9700\u5728nvvp\u754c\u9762\u4e2d\u9009\u4e0a\u624d\u80fd\u5f00\u542f":134,"\u9700\u6307\u5b9a":113,"\u9700\u63d0\u4f9b\u975e\u96f6\u5143\u7684\u503c":115,"\u9700\u6ce8\u610f":113,"\u9700\u8981":[26,38,101,116],"\u9700\u8981\u4e3a":101,"\u9700\u8981\u4f7f\u7528":107,"\u9700\u8981\u4f7f\u7528\u5176\u5236\u5b9a\u7684\u65b9\u5f0f\u6302\u8f7d\u540e\u5e76\u5bfc\u5165\u6570\u636e":123,"\u9700\u8981\u4f7f\u7528\u6700\u65b0\u7684pip":30,"\u9700\u8981\u4f7f\u7528\u8005\u81ea\u5df1\u4e86\u89e3\u5e76\u5b8c\u6210\u8f6c\u5316":115,"\u9700\u8981\u4fdd\u6301\u5f53\u524d\u5206\u652f\u76ee\u5f55":98,"\u9700\u8981\u4fee\u6539build":88,"\u9700\u8981\u521b\u5efa\u5e76\u586b\u5199":115,"\u9700\u8981\u5347\u7ea7pip\u7248\u672c\u5230\u6700\u65b0":[30,104],"\u9700\u8981\u5355\u72ec":28,"\u9700\u8981\u53ef\u4ee5\u8de8\u5e73\u53f0\u6267\u884c":54,"\u9700\u8981\u540c\u65f6\u63d0\u4f9b\u6bcf\u4e00\u4e2a\u5185\u5c42\u5e8f\u5217\u5728\u6574\u4e2a":115,"\u9700\u8981\u540c\u6b65\u539f\u4ed3\u5e93":98,"\u9700\u8981\u542f\u52a8\u7684\u8282\u70b9\u4e2a\u6570\u4ee5\u53ca":123,"\u9700\u8981\u548c\u8be5op\u7684\u540d\u5b57\u4e00\u6837":101,"\u9700\u8981\u54ea\u4e9b\u5c42\u7684\u8ba1\u7b97\u7ed3\u679c\u4f5c\u4e3a\u8f93\u51fa":116,"\u9700\u8981\u5728":101,"\u9700\u8981\u5728\u521b\u5efa\u5bb9\u5668\u524d\u6302\u8f7d\u5377\u4ee5\u4fbf\u6211\u4eec\u4fdd\u5b58\u8bad\u7ec3\u7ed3\u679c":122,"\u9700\u8981\u5728\u7cfb\u7edf\u91cc\u5148\u5b89\u88c5\u597ddocker\u5de5\u5177\u5305":103,"\u9700\u8981\u5728cmake\u7684\u65f6\u5019":72,"\u9700\u8981\u5728macos\u7cfb\u7edf\u4e0a\u8fdb\u884c":143,"\u9700\u8981\u5c06\u5176parameter\u8bbe\u7f6e\u6210":107,"\u9700\u8981\u5c06\u7f51\u7edc\u7ed3\u6784\u4f7f\u7528":116,"\u9700\u8981\u5c06bugfix\u7684\u5206\u652f\u540c\u65f6merge\u5230":88,"\u9700\u8981\u5c06cuda\u76f8\u5173\u7684\u5e93\u8bbe\u7f6e\u5230":113,"\u9700\u8981\u5c06paddl":113,"\u9700\u8981\u5f15\u7528":72,"\u9700\u8981\u5f3a\u8c03\u7684\u662f":26,"\u9700\u8981\u601d\u8003\u5b8c\u6210\u4ee5\u4e0b\u5de5\u4f5c":[115,116],"\u9700\u8981\u624b\u52a8\u8fdb\u884c\u89e3\u538b":116,"\u9700\u8981\u6267\u884c":[27,30,112],"\u9700\u8981\u6307\u5b9a":113,"\u9700\u8981\u6307\u5b9a\u4e0e\u67d0\u4e00\u4e2a\u8f93\u5165\u7684\u5e8f\u5217\u4fe1\u606f\u662f\u4e00\u81f4\u7684":137,"\u9700\u8981\u6307\u5b9alayer\u7684\u8f93\u5165\u6765\u6e90":110,"\u9700\u8981\u660e\u786e\u6307\u5b9a":129,"\u9700\u8981\u663e\u5f0f\u5730\u94fe\u63a5":113,"\u9700\u8981\u663e\u793a\u5730\u94fe\u63a5":113,"\u9700\u8981\u663e\u793a\u5730\u94fe\u63a5mkl\u7684\u52a8\u6001\u5e93":113,"\u9700\u8981\u6709\u7a33\u5b9a\u7684\u5bfc\u51fa\u7b26\u53f7":71,"\u9700\u8981\u6839\u636e\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u6765\u7ed1\u5b9a\u4e00\u4e2a":123,"\u9700\u8981\u6ce8\u610f":101,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[68,88,107,129],"\u9700\u8981\u6ce8\u610f\u7684\u662f\u68af\u5ea6\u68c0\u67e5\u4ec5\u4ec5\u9a8c\u8bc1\u4e86\u68af\u5ea6\u7684\u8ba1\u7b97":100,"\u9700\u8981\u6ce8\u610f\u7684\u662fpaddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5b50\u5e8f\u5217\u6570\u76ee\u4e00\u6837\u7684\u591a\u8f93\u5165\u53cc\u5c42rnn":137,"\u9700\u8981\u7528\u5230\u7684\u7f16\u8bd1\u5de5\u5177\u548c\u7cfb\u7edf\u5e93":142,"\u9700\u8981\u7528\u6237\u663e\u5f0f\u8bbe\u5b9a":108,"\u9700\u8981\u7d2f\u52a0\u4e0d\u540clayer\u4f20\u8fc7\u6765\u7684\u68af\u5ea6":68,"\u9700\u8981\u81ea\u5df1\u94fe\u63a5mkl\u94fe\u63a5\u5e93":113,"\u9700\u8981\u88ab\u66b4\u9732\u5230\u5176\u4ed6\u8bed\u8a00":72,"\u9700\u8981\u8bf7\u7ba1\u7406\u5458\u5b89\u88c5\u548c\u914d\u7f6e\u597d":26,"\u9700\u8981\u9075\u5faa\u4ee5\u4e0b\u7ea6\u5b9a":139,"\u9700\u8981\u91cd\u547d\u540dwheel\u5305\u4e2dplatform\u76f8\u5173\u7684\u540e\u7f00":88,"\u9700\u8981\u989d\u5916\u6ce8\u610f\u7684\u662f":102,"\u9700\u9644\u52a0\u53cc\u5c42\u5e8f\u5217\u4fe1\u606f":115,"\u9700\u9644\u52a0\u5e8f\u5217\u4fe1\u606f":115,"\u975e\u5e38\u6570":100,"\u975e\u5e8f\u5217\u8f93\u5165\u4e0d\u643a\u5e26":115,"\u975e\u5e8f\u5217\u8f93\u5165\u65e0\u9700\u6784\u9020":115,"\u975e\u96f6\u5143\u4e2a\u6570":115,"\u975e\u96f6\u5143\u7d20\u7684\u503c":115,"\u975e\u96f6\u5143\u7d20\u7684\u5217\u53f7":115,"\u975e\u96f6\u6570\u5b57\u7684\u4e2a\u6570":100,"\u9762\u5411\u67b6\u6784\u4e3a32\u4f4darm\u67b6\u6784":142,"\u9762\u5411\u67b6\u6784\u4e3a64\u4f4darm64\u67b6\u6784":142,"\u9879\u76ee\u5728\u52aa\u529b\u5f00\u59cb\u652f\u6301\u5176\u4ed6\u4e0d\u9700\u8981":26,"\u987a\u5e8f":137,"\u9884\u63d0\u4ea4\u94a9\u5b50":98,"\u9884\u6d4b\u4ee3\u7801\u66f4\u591a\u8be6\u7ec6\u793a\u4f8b\u4ee3\u7801\u8bf7\u53c2\u8003":116,"\u9884\u6d4b\u4f7f\u7528\u7684\u7f51\u7edc\u7ed3\u6784\u5f80\u5f80\u4e0d\u540c\u4e8e\u8bad\u7ec3":116,"\u9884\u6d4b\u65f6":116,"\u9884\u6d4b\u65f6\u53ea\u9700\u52a0\u8f7d\u4e00\u4e2a\u6587\u4ef6\u4fbf\u4e8e\u53d1\u5e03":116,"\u9884\u6d4b\u7a0b\u5e8f\u5f00\u53d1\u4e24\u5927\u90e8\u5206":116,"\u9996\u5148":[100,137,140],"\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":27,"\u9996\u5148\u5b89\u88c5\u5e76\u5728\u5f53\u524d\u76ee\u5f55\u8fd0\u884c\u5b83":98,"\u9996\u5148\u5b9a\u4e49":101,"\u9996\u5148\u5bf9\u8f93\u5165\u505a\u4e00\u4e2a\u5c0f\u7684\u6270\u52a8":100,"\u9996\u5148\u6211\u4eec\u9700\u8981\u63a8\u5bfc\u8be5\u7f51\u7edc\u5c42\u7684":100,"\u9996\u5148\u6784\u9020\u5934\u4fe1\u606f":109,"\u9996\u5148\u901a\u8fc7":98,"\u9996\u5148\u9700\u8981\u52a0\u8f7d\u76f8\u5e94\u7684python\u5e93":110,"\u9a71\u52a8":103,"\u9ad8\u4eae\u90e8\u5206":137,"\u9ad8\u5ea6":115,"\u9ad8\u5ea6\u652f\u6301\u7075\u6d3b\u548c\u9ad8\u6548\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":140,"\u9ad8\u65af\u5206\u5e03":109,"\u9ed8\u8ba4":129,"\u9ed8\u8ba40":117,"\u9ed8\u8ba41":117,"\u9ed8\u8ba4127":117,"\u9ed8\u8ba4256k":54,"\u9ed8\u8ba47164":117,"\u9ed8\u8ba4\u4e0d\u663e\u793a":129,"\u9ed8\u8ba4\u4e0d\u8bbe\u7f6e":139,"\u9ed8\u8ba4\u4e3a0":[129,131],"\u9ed8\u8ba4\u4e3a1":[115,131],"\u9ed8\u8ba4\u4e3a100":131,"\u9ed8\u8ba4\u4e3a4096mb":129,"\u9ed8\u8ba4\u4e3a\u7b2c\u4e00\u4e2a\u8f93\u5165":139,"\u9ed8\u8ba4\u4e3anull":129,"\u9ed8\u8ba4\u4f1a\u5c06a\u548cb":107,"\u9ed8\u8ba4\u4f7f\u7528concurrentremoteparameterupdat":129,"\u9ed8\u8ba4\u4f7f\u7528mkl":27,"\u9ed8\u8ba4\u503c":[27,131,136,143],"\u9ed8\u8ba4\u503c\u4e3a":[142,143,144],"\u9ed8\u8ba4\u503c\u4e3a\u73af\u5883\u53d8\u91cf":143,"\u9ed8\u8ba4\u521d\u59cb\u72b6\u4e3a0":139,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b":[109,119],"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u6309\u7167float\u7cbe\u5ea6\u8ba1\u7b97":109,"\u9ed8\u8ba4\u6307\u5b9a\u7b2c\u4e00\u4e2a\u8f93\u5165":137,"\u9ed8\u8ba4\u662f\u4f7f\u7528mkl\u7684\u955c\u50cf":28,"\u9ed8\u8ba4\u6ca1\u6709\u5b89\u88c5vim":28,"\u9ed8\u8ba4\u7684paddlepaddle\u751f\u4ea7\u73af\u5883\u955c\u50cf":122,"\u9ed8\u8ba4\u7f16\u8bd1\u6240\u6709\u67b6\u6784":143,"\u9ed8\u8ba4\u8bbe\u7f6e\u4e3a":67,"\u9ed8\u8ba4\u8bbe\u7f6e\u4e3a\u771f":131,"\u9ed8\u8ba4\u8bbe\u7f6e\u6210\u73af\u5883\u53d8\u91cf":[142,144],"\u9ed8\u8ba4\u8c03\u7528":26,"\u9ed8\u8ba4fals":117,"abi\u7684paddlepaddle\u5e93":142,"abstract":[45,46,53,57,78,87,89],"android\u5b98\u65b9\u63d0\u4f9b\u7684":142,"android\u5e73\u53f0\u4e0a\u4f7f\u7528\u7684c":142,"android\u5e73\u53f0\u53ef\u9009\u914d\u7f6e\u53c2\u6570":142,"android\u7684docker\u5f00\u53d1\u955c\u50cf\u5411\u7528\u6237\u63d0\u4f9b\u4e24\u4e2a\u53ef\u914d\u7f6e\u7684\u53c2\u6570":142,"api\u4e0d\u4f1a\u76f4\u63a5\u52a0\u8f7d":116,"api\u4e0d\u5c0f\u4e8e21":142,"api\u4e2d":115,"api\u4e2d\u4f7f\u7528":71,"api\u4e2d\u7684\u4e00\u7ef4\u6570\u7ec4":115,"api\u4e2d\u7684\u77e9\u9635\u6765\u8868\u793a":115,"api\u4e2d\u795e\u7ecf\u7f51\u7edc\u7684\u4e00\u4e2a\u8f93\u5165":116,"api\u4f7f\u7528\u4e2d\u7684\u4e00\u4e2a\u91cd\u8981\u6982\u5ff5":116,"api\u4f7f\u7528\u6d41\u7a0b":114,"api\u4f7f\u7528\u6d41\u7a0b\u793a\u610f\u56fe":116,"api\u4f7f\u7528\u793a\u4f8b":116,"api\u521b\u5efa\u7684gradientmachine\u7c7b\u7684\u5bf9\u8c61":116,"api\u53ef\u4ee5\u901a\u8fc7\u5206\u522b\u6307\u5b9a\u5e8f\u5217\u5316\u540e\u7684\u7f51\u7edc\u7ed3\u6784\u6587\u4ef6\u548c\u53c2\u6570\u76ee\u5f55\u6765\u52a0\u8f7d\u8bad\u7ec3\u597d\u7684\u6a21\u578b":116,"api\u53ef\u4ee5\u901a\u8fc7\u6307\u5b9a":116,"api\u5b8c\u6210\u5206\u5e03\u5f0f\u8bad\u7ec3":117,"api\u5bf9\u6bd4\u4ecb\u7ecd":138,"api\u5bfc\u51fa\u7684\u52a8\u6001\u5e93":72,"api\u5bfc\u51fa\u7684\u9759\u6001\u5e93":72,"api\u5e93\u5c06\u88ab\u5b89\u88c5\u5230":142,"api\u5f00\u53d1\u5305\u5e76\u5b89\u88c5":30,"api\u5f00\u53d1\u9884\u6d4b\u7a0b\u5e8f\u65f6":113,"api\u5f00\u53d1\u9884\u6d4b\u7a0b\u5e8f\u9700\u8981\u4e00\u4e2a\u8bad\u7ec3\u597d\u7684\u6a21\u578b":116,"api\u6240\u9700\u7684\u4f9d\u8d56":113,"api\u63a5\u53d7\u7684\u7c7b\u578b\u5168\u662f":72,"api\u63a5\u53e3":54,"api\u63a5\u53e3\u751f\u6210":101,"api\u63a5\u53e3\u7684\u53c2\u6570\u8f6c\u53d1\u7ed9":72,"api\u63a5\u53e3\u7684\u751f\u6210":101,"api\u63d0\u4f9b\u7684":116,"api\u652f\u6301\u7684\u6240\u6709\u8f93\u5165\u6570\u636e\u7c7b\u578b\u548c\u4ed6\u4eec\u7684\u7ec4\u7ec7\u65b9\u5f0f":116,"api\u6587\u6863":[142,143],"api\u65f6":72,"api\u65f6\u4e3a\u8f93\u51fa":116,"api\u65f6\u6240\u552f\u4e00\u9700\u8981\u5f15\u5165\u7684\u5934\u6587\u4ef6":72,"api\u662f\u591a\u8bed\u8a00api\u7684\u57fa\u7840\u90e8\u5206":72,"api\u66b4\u9732\u7684\u7c7b\u578b":72,"api\u6765\u9884\u6d4b":[142,143],"api\u751f\u6210\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u4f1a\u88ab\u5b89\u88c5\u5230":72,"api\u7684\u5934\u6587\u4ef6":[142,143,144],"api\u7684\u5b9e\u4f8b":72,"api\u7684\u5b9e\u73b0\u7ec6\u8282":72,"api\u7684\u63a5\u53e3":72,"api\u7684\u65f6\u5019\u63a8\u8350paddle\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":72,"api\u7684\u7f16\u8bd1\u9009\u9879\u9ed8\u8ba4\u5173\u95ed":72,"api\u76ee\u5f55\u7ed3\u6784\u5982\u4e0a\u56fe\u8868\u6240\u793a":72,"api\u76f8\u5173\u63a5\u53e3":115,"api\u7ea7\u522b":142,"api\u7ea7\u522b\u4e3a21":142,"api\u83b7\u5f97\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u5b9e\u4f8b":72,"api\u8bad\u7ec3":116,"api\u9700\u8981\u521b\u5efa\u7684\u6570\u636e\u7c7b\u578b":115,"api\u9759\u6001\u5e93":143,"api\u9884\u6d4b\u5e93":[132,143],"api\u9884\u6d4b\u65f6":116,"apis\u505a\u4e86\u5c01\u88c5":67,"app\u4e2d":[142,143],"apple\u5b98\u65b9\u4e3aios\u5f00\u53d1\u63d0\u4f9b\u4e86\u5b8c\u6574\u7684\u4ea4\u53c9\u7f16\u8bd1\u5de5\u5177\u548c\u96c6\u6210\u5f00\u53d1\u73af\u5883":143,"async_sgd\u8fdb\u884c\u8bad\u7ec3\u65f6":109,"avx\u662f\u4e00\u79cdcpu\u6307\u4ee4\u96c6":28,"avx\u7248\u672c":28,"avx\u7684\u955c\u50cf":28,"batch\u4e2d\u5305\u542b":107,"batch\u7684\u6743\u91cd":107,"batches\u4e2a\u6279\u6b21\u4fdd\u5b58\u4e00\u6b21\u53c2\u6570":129,"batches\u6b21":129,"block\u6784\u6210\u4e00\u4e2amodel":37,"book\u4e00\u5b9a\u662f\u60a8\u6700\u597d\u7684\u9009\u62e9":28,"book\u4e2d\u6240\u6709\u7ae0\u8282\u529f\u80fd\u7684\u6b63\u786e\u6027":88,"book\u662f\u4e3a\u7528\u6237\u548c\u5f00\u53d1\u8005\u5236\u4f5c\u7684\u4e00\u4e2a\u4ea4\u4e92\u5f0f\u7684jupyt":28,"book\u7684":110,"book\u7684docker\u955c\u50cf":28,"boolean":[18,53,55,63,71,94],"break":[9,35,92,94,96,97],"bugfix\u5206\u652f\u4e5f\u662f\u5728\u5f00\u53d1\u8005\u81ea\u5df1\u7684fork\u7248\u672c\u5e93\u7ef4\u62a4":88,"bugfix\u5206\u652f\u9700\u8981\u5206\u522b\u7ed9\u4e3b\u7248\u672c\u5e93\u7684":88,"byte":[9,11,54,70,109],"c99\u662f\u76ee\u524dc\u6700\u5e7f\u6cdb\u7684\u4f7f\u7528\u6807\u51c6":71,"c\u6709\u6807\u51c6\u7684abi":71,"c\u8bed\u8a00\u662f\u6709\u5bfc\u51fa\u7b26\u53f7\u7684\u6807\u51c6\u7684":71,"case":[4,18,39,45,47,48,53,57,66,72,78,82,84,85,94,121,134,145],"cc\u4e2d\u7684":102,"cells\u7b49":108,"char":41,"ci\u73af\u5883\u4f7f\u7528":88,"ci\u7f16\u8bd1wheel\u5b8c\u6210\u540e\u4f1a\u81ea\u52a8\u5c06docker\u955c\u50cfpush\u5230dockerhub":88,"class":[1,2,3,4,5,6,7,9,10,13,14,15,18,21,25,31,34,45,46,47,48,51,52,56,57,58,59,61,62,64,66,71,75,76,80,81,85,86,87,89,90,91,93,96,100,101,102,109,135],"cmake\u4e2d\u5c06":134,"cmake\u5b98\u65b9\u5bf9android\u5e73\u53f0\u7684\u4ea4\u53c9\u7f16\u8bd1\u63d0\u4f9b\u4e86\u901a\u7528\u7684\u652f\u6301":142,"cmake\u627e\u5230\u7684python\u5e93\u548cpython\u89e3\u91ca\u5668\u7248\u672c\u53ef\u80fd\u6709\u4e0d\u4e00\u81f4\u73b0\u8c61":104,"cmake\u7cfb\u7edf\u5bf9\u4ea4\u53c9\u7f16\u8bd1\u63d0\u4f9b\u4e86\u652f\u6301":142,"cmake\u7f16\u8bd1\u65f6":27,"cmake\u7f16\u8bd1\u7684\u76ee\u6807\u5e73\u53f0":[142,143,144],"cmake\u914d\u7f6e\u4e2d\u5c06":134,"cmake\u914d\u7f6e\u5b8c\u6210\u540e":[142,143,144],"compute\u51fd\u6570":67,"const":[34,39,41,46,56,58,64,65,79,80,82,86,89,91,93,95,96,100,101,102],"container\u4e2d":122,"core\u4e2d\u7684\u6a21\u578b\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u53c2\u6570":72,"core\u4e2d\u8fd9\u4e00\u7c7b\u578b\u63a5\u53e3\u7684\u667a\u80fd\u6307\u9488":72,"core\u662f\u5426\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u5b9e\u4f8b":72,"core\u6982\u5ff5":72,"cost\u63a5\u6536y_predict\u4e0ey\u4f5c\u4e3a\u8f93\u5165":110,"cost\u8fd8\u5927\u4e8e\u8fd9\u4e2a\u6570":109,"count\u4e2agpu\u4e0a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97\u67d0\u4e00\u5c42":131,"count\u548cgpu":131,"csr\u5b58\u50a8\u683c\u5f0f\u901a\u8fc7":115,"cuda\u5171\u4eabkernel\u5b9e\u73b0\u5728":101,"cuda\u5b9e\u73b0\u5171\u4eab\u540c\u4e00\u4e2a":101,"cuda\u5b9e\u73b0\u5728":101,"cuda\u5e93":129,"cuda\u7684\u4ee3\u7801\u53ef\u4ee5\u590d\u7528":101,"cuda\u76f8\u5173\u5e93\u4f1a\u5728\u9884\u6d4b\u7a0b\u5e8f\u8fd0\u884c\u65f6\u52a8\u6001\u88c5\u8f7d":113,"cudnn\u5e93":[27,129],"cumtime\u7684\u6bcf\u6b21\u8c03\u7528\u5e73\u5747\u65f6\u95f4":133,"data\u5230\u5206\u5e03\u5f0f\u5b58\u50a8\u8865\u5145\u8bad\u7ec3\u6570\u636e":38,"data\u63a5\u53e3\u5206\u914d\u5b9e\u9645\u7684\u5185\u5b58":102,"data\u76ee\u5f55\u4e2d\u5b58\u653e\u5207\u5206\u597d\u7684\u6570\u636e":123,"dataprovider\u5171\u8fd4\u56de\u4e24\u4e2a\u6570\u636e":137,"dataprovider\u5171\u8fd4\u56de\u4e24\u7ec4\u6570\u636e":137,"dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":107,"decoder\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4f1a\u5f15\u7528\u5168\u90e8\u7ed3\u679c":139,"decoder\u63a5\u53d7\u4e24\u4e2a\u8f93\u5165":139,"decoder\u6bcf\u6b21\u9884\u6d4b\u4ea7\u751f\u4e0b\u4e00\u4e2a\u6700\u53ef\u80fd\u7684\u8bcd\u8bed":139,"decoer\u67b6\u6784":139,"default":[2,3,4,5,6,7,9,10,11,17,18,19,22,25,31,34,35,45,47,51,60,64,70,73,74,82,83,89,90,91,94,97,121,122,123,131,142,145],"device\u5c31\u80fd\u62ff\u5230\u6b63\u786e\u7684\u6570\u636e":68,"dist\u76ee\u5f55\u4e0b\u751f\u6210\u8f93\u51fa\u7684whl\u5305":27,"dnn\u4e09\u8005\u5173\u7cfb\u5982\u4e0b\u8868":68,"dnn\u4e2d\u7684":68,"dnn\u4e2d\u7684\u6392\u5217\u65b9\u5f0f\u4e0d\u6b62\u8fd9\u4e00\u79cd":68,"dnn\u4f1a\u4f5c\u4e3a\u7b2c\u4e09\u65b9\u5e93\u96c6\u6210\u8fdbpaddlepaddl":68,"dnn\u4f1a\u7528\u5230":68,"dnn\u5171\u540c\u4f7f\u7528":68,"dnn\u524d\u540e\u7684cnn\u7f51\u7edc\u6027\u80fd":68,"dnn\u5728\u53d1\u5e03":68,"dnn\u5b9e\u73b0":68,"dnn\u5e0c\u671b\u7684\u683c\u5f0f":68,"dnn\u6570\u5b66\u5e93":27,"dnn\u6570\u636e\u7684\u4e0d\u540c\u683c\u5f0f\u4ee5\u53ca\u76f8\u4e92\u4e4b\u95f4\u7684\u8f6c\u6362":68,"dnn\u7684":68,"dnn\u7684\u5e93\u76ee\u524d\u53ea\u6709\u52a8\u6001\u5e93":68,"dnn\u7684\u6027\u80fd":68,"dnn\u7684\u60c5\u51b5\u4e0b":68,"dnn\u7684\u64cd\u4f5c\u90fd\u662f\u76f4\u63a5\u8986\u76d6\u7684\u5f62\u5f0f":68,"dnn\u7684\u6d4b\u8bd5":68,"dnn\u7684\u73af\u5883\u4e0b":68,"dnn\u7684\u76f8\u5173\u529f\u80fd":68,"dnn\u7684\u7ed3\u679c":68,"dnn\u7684\u9ad8\u6027\u80fd\u683c\u5f0f\u4e0epaddlepaddle\u539f\u6709\u7684":68,"dnn\u7684layer":68,"dnn\u7684layers\u90fd\u4f1a\u7ee7\u627f\u4e8e":68,"docker\u5b89\u88c5\u65b9\u5f0f\u53ef\u4ee5\u8fdb\u5165docker\u5bb9\u5668\u6267\u884c":127,"docker\u5b89\u88c5\u8bf7\u53c2\u8003":103,"docker\u5b89\u88c5\u8bf7\u53c2\u8003docker\u7684\u5b98\u7f51":103,"docker\u5b98\u7f51":28,"docker\u5bb9\u5668\u4e2d\u5c06\u9ed8\u8ba4\u4f7f\u7528":142,"docker\u7684\u5b98\u7f51":103,"docker\u7f16\u8bd1\u73af\u5883\u955c\u50cf\u5b8c\u6210\u7f16\u8bd1":27,"docker\u80fd\u5728\u6240\u6709\u4e3b\u8981\u64cd\u4f5c\u7cfb\u7edf":142,"docker\u955c\u50cf":28,"docker\u955c\u50cf\u4e3a\u4e86\u51cf\u5c0f\u4f53\u79ef":28,"docker\u955c\u50cf\u9ed8\u8ba4":28,"dockerhub\u7f51\u7ad9":28,"double\u7c7b\u578b\u65f6\u4e3a8":109,"dropout\u7684\u6bd4\u4f8b":100,"eigenscalar\u7684\u8f6c\u6362":102,"encode\u6210\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":137,"encoder\u548cdecoder\u53ef\u4ee5\u662f\u80fd\u591f\u5904\u7406\u5e8f\u5217\u7684\u4efb\u610f\u795e\u7ecf\u7f51\u7edc\u5355\u5143":139,"encoder\u8f93\u51fa":139,"entropy\u4f5c\u4e3acost":109,"enum":[39,41,47,73,80,81,90,91,97],"export":[28,57,62,103,104,117],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":110,"fc1\u548cfc2\u5c42\u5728gpu\u4e0a\u8ba1\u7b97":131,"fc3\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":131,"final":[4,5,18,32,33,48,62,74,75,92,96],"flatten\u65b9\u6cd5\u662f\u628apaddle\u4e2d\u7684\u4e00\u4e2atensor\u8fdb\u884creshape\u64cd\u4f5c":102,"float":[2,3,4,6,9,18,19,51,56,64,91,93,95,101,102,115,134],"float\u7b49":131,"forward\u7684output\u7684\u503c":107,"from\u65b9\u6cd5\u662f\u628apaddle\u4e2d\u7684\u4e00\u7ef4tensor\u8f6c\u4e3aeigen\u7684\u4e00\u7ef4tensor":102,"from\u662feigentensor\u6a21\u677f\u63d0\u4f9b\u7684\u4e00\u4e2a\u63a5\u53e3":102,"full\u53c2\u6570\u63d0\u4ea4":105,"function":[4,5,9,10,18,19,23,31,33,34,36,40,41,42,44,45,46,47,48,51,52,56,58,61,64,69,74,75,78,79,80,81,82,84,85,86,87,89,91,96,133,140,145],"function\u4f7f\u7528":108,"git\u6d41\u5206\u652f\u6a21\u578b":98,"github\u9996\u9875":98,"glibc\u81f3\u5c11\u5305\u542bglibc_2":30,"golang\u53ef\u4ee5\u4f7f\u7528":71,"golang\u7684":71,"google\u5f00\u6e90\u7684\u5bb9\u5668\u96c6\u7fa4\u7684\u8c03\u5ea6\u6846\u67b6":120,"gpu\u4e8c\u8fdb\u5236\u6587\u4ef6":27,"gpu\u5219\u8fd8\u9700\u8981\u9ad8\u5e76\u884c\u6027":134,"gpu\u6027\u80fd\u8c03\u4f18":132,"gpu\u6267\u884c":102,"gpu\u6838\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u6307\u5b9a":129,"gpu\u7684docker\u955c\u50cf\u7684\u65f6\u5019":104,"gpu\u7b49":88,"group\u6559\u7a0b":138,"group\u7684\u5b9e\u73b0\u65b9\u5f0f":108,"gru\u6216lstm":140,"h\u5e76\u4e0d\u56f0\u96be":71,"html\u5373\u53ef\u8bbf\u95ee\u672c\u5730\u6587\u6863":103,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":110,"id\u6307\u5b9a\u4f7f\u7528\u54ea\u4e2agpu\u6838":129,"id\u6307\u5b9a\u7684gpu":131,"id\u65e0\u6548":129,"image\u91cc":122,"images\u6570\u636e\u96c6\u4e0a\u4f20\u5230\u4e91\u7aef\u7684":38,"imikolov\u6570\u636e\u96c6":117,"import":[3,4,30,31,34,35,45,47,50,58,60,62,63,69,74,75,81,89,92,101,110,112,116,117,121],"infer\u63a5\u53e3\u7684\u8fd4\u56de\u503c\u662f\u4e00\u4e2apython":107,"ingress\u9700\u8981\u628apfsclient\u7684\u8eab\u4efd\u4fe1\u606f\u4f20\u7ed9pfsserv":54,"instance\u4e0e\u751f\u6210\u6570\u636e\u96c6\u65f6":38,"instance\u5305\u6db5\u4e24\u4e2a\u503c":38,"instance\u662f\u4e00\u6a21\u4e00\u6837\u7684":38,"int":[2,3,4,5,9,10,11,18,19,33,34,39,40,41,44,45,47,49,63,64,67,68,69,71,72,80,81,83,84,90,91,93,96,97,100,102,115,117,131],"interface\u6587\u4ef6\u7684\u5199\u6cd5\u975e\u5e38":71,"ios\u5e73\u53f0\u53ef\u9009\u914d\u7f6e\u53c2\u6570":143,"ip\u548c\u4efb\u52a1\u8282\u70b9\u4e2a\u6570\u7b49":120,"issue\u7f16\u53f7":98,"job\u662f\u672c\u6b21\u8bad\u7ec3\u5bf9\u5e94\u7684job":123,"job\u7684\u540d\u5b57":123,"kernel\u5b9e\u73b0":101,"kernel\u6ce8\u518ccpu\u5b9e\u73b0\u5728":101,"kernel\u7684\u5b9e\u73b0\u57fa\u4e8eeigen":101,"kubernetes\u4e3a\u8fd9\u6b21\u8bad\u7ec3\u521b\u5efa\u4e863\u4e2apod\u5e76\u4e14\u8c03\u5ea6\u5230\u4e863\u4e2anode\u4e0a\u8fd0\u884c":123,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":120,"kubernetes\u5355\u673a\u8bad\u7ec3":120,"kubernetes\u53ef\u4ee5\u901a\u8fc7yaml\u6587\u4ef6\u6765\u521b\u5efa\u76f8\u5173\u5bf9\u8c61":123,"kubernetes\u5c31\u4f1a\u521b\u5efa3\u4e2apod\u4f5c\u4e3apaddlepaddle\u8282\u70b9\u7136\u540e\u62c9\u53d6\u955c\u50cf":123,"kubernetes\u6709job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u652f\u6301":122,"label\u662f\u539f\u59cb\u6570\u636e\u4e2d\u5bf9\u4e8e\u6bcf\u4e00\u53e5\u8bdd\u7684\u5206\u7c7b\u6807\u7b7e":137,"labels\u662f\u6bcf\u7ec4\u5185\u6bcf\u4e2a\u53e5\u5b50\u7684\u6807\u7b7e":137,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":136,"layer\u4f5c\u4e3a\u4e00\u4e2a\u6574\u4f53\u6765\u5b9e\u73b0":108,"layer\u62ff\u5230\u7684\u7528\u6237\u8f93\u5165":139,"layer\u65f6":[68,108],"layer\u662f\u6211\u4eec\u7684\u79ef\u6728":110,"layer\u7684\u540e\u9762\u63a5\u6709cpu":68,"layer\u7c7b\u53ef\u4ee5\u81ea\u52a8\u8ba1\u7b97\u4e0a\u9762\u7684\u5bfc\u6570":100,"layer\u8ba1\u7b97\u7684\u8f93\u51fa":108,"linux\u4e2d":28,"list\u4e2d":116,"list\u4f5c\u4e3a\u68c0\u67e5\u5217\u8868":88,"list\u5982\u4e0b\u6240\u793a":131,"list\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b\u5217\u8868":131,"long":[4,5,9,18,47],"memory\u4e0d\u80fd\u72ec\u7acb\u5b58\u5728":139,"memory\u4e5f\u53ef\u4ee5\u5177\u6709":140,"memory\u4e5f\u53ef\u4ee5\u662f\u5e8f\u5217":140,"memory\u53ea\u80fd\u5728":139,"memory\u53ef\u4ee5\u7f13\u5b58\u4e0a\u4e00\u4e2a\u65f6\u523b\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u8f93\u51fa":137,"memory\u6307\u5411\u4e00\u4e2alay":139,"memory\u662f\u5728\u5355\u6b65\u51fd\u6570\u4e2d\u5faa\u73af\u4f7f\u7528\u7684\u72b6\u6001":140,"memory\u662fpaddlepaddle\u5b9e\u73b0rnn\u65f6\u5019\u4f7f\u7528\u7684\u4e00\u4e2a\u6982\u5ff5":137,"memory\u7684":140,"memory\u7684\u521d\u59cb\u72b6\u6001":139,"memory\u7684\u65f6\u95f4\u5e8f\u5217\u957f\u5ea6\u4e00\u81f4\u7684\u60c5\u51b5":137,"memory\u7684\u66f4\u591a\u8ba8\u8bba\u8bf7\u53c2\u8003\u8bba\u6587":139,"memory\u7684\u8f93\u51fa\u5b9a\u4e49\u5728":140,"memory\u7684i":139,"memory\u9ed8\u8ba4\u521d\u59cb\u5316\u4e3a0":139,"mkl\u5e93\u7684":67,"mklml\u4ee5\u53camkl":68,"mklml\u53ef\u4ee5\u4e0emkl":68,"mklml\u7684\u5e93\u76ee\u524d\u90fd\u662f\u52a8\u6001\u5e93":68,"mnist\u624b\u5199\u6570\u5b57\u8bc6\u522b\u76ee\u5f55":116,"mode\u4e0b\u7684\u7ed3\u679c":67,"model\u505a\u5206\u652f\u7ba1\u7406":88,"name\u7ec4\u5408\u53ef\u4ee5\u627e\u5230\u672c\u6b21\u8bad\u7ec3\u9700\u8981\u7684\u6587\u4ef6\u8def\u5f84":123,"ndarray\u7c7b\u578b\u7684\u503c\u548c\u6574\u578b\u7684\u503c":38,"ndk\u4e2d\u5305\u542b\u4e86\u6240\u6709android":142,"new":[4,9,18,32,33,34,35,36,39,40,41,42,43,46,47,48,51,56,57,66,67,69,73,75,78,83,84,85,87,91,92,96,98,100,121,145],"note\u7684\u4e66\u5199":88,"null":[62,100,115,129],"num\u51b3\u5b9a\u603b\u7aef\u53e3\u4e2a\u6570":117,"num_gradient_servers\u53c2\u6570":123,"num_samples_processed\u4e3a\u5df2\u8bad\u7ec3\u6837\u672c\u6570":109,"only\u7684\u4e8c\u8fdb\u5236":27,"op\u4e0d\u9700\u8981\u5b9a\u4e49opprotomak":101,"op\u5355\u5143\u6d4b\u8bd5\u7ee7\u627f\u81ea":101,"op\u5b9a\u4e49":101,"op\u6709\u8ba1\u7b97\u51fd\u6570":101,"op\u6ce8\u518c\u5b9e\u73b0\u5728":101,"op\u7684\u4fe1\u606f":68,"op\u8ba1\u7b97\u51fd\u6570\u7684\u57fa\u7c7b":101,"openmp\u7528\u4e8e\u63d0\u9ad8mklml\u7684\u6027\u80fd":68,"opprotomake\u5b9a\u4e49":101,"org\u5de5\u5177\u7684\u8be6\u7ec6\u4fe1\u606f":103,"org\u76ee\u524d\u9075\u5faa":88,"outer_mem\u662f\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":137,"output\u53ef\u4ee5\u662f\u4efb\u610f\u7ef4\u5ea6\u7684tensor":102,"output\u6587\u4ef6\u5939\u5b58\u653e\u8bad\u7ec3\u7ed3\u679c\u4e0e\u65e5\u5fd7":123,"output\u7684\u539f\u6709shape\u4fe1\u606f\u4e0d\u53d8":102,"packages\u91cc\u9762":104,"packages\u91cc\u9762\u7684python\u5305":104,"packed\u4f18\u5316\u540elayer\u7684\u6d4b\u8bd5":67,"packed\u76f8\u5173\u529f\u80fd":67,"paddepaddle\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":27,"paddle\u4e00\u4e2a\u52a8\u6001\u5e93\u53ef\u4ee5\u5728\u4efb\u4f55linux\u7cfb\u7edf\u4e0a\u8fd0\u884c":71,"paddle\u4e2d\u7ecf\u5e38\u4f1a\u5c06\u65f6\u95f4\u5e8f\u5217\u6210\u4e3a":137,"paddle\u4e8c\u8fdb\u5236\u5728\u8fd0\u884c\u65f6\u6355\u83b7\u4e86\u6d6e\u70b9\u6570\u5f02\u5e38":107,"paddle\u5185\u5d4c\u7684python\u89e3\u91ca\u5668\u548c\u5916\u90e8\u4f7f\u7528\u7684python\u5982\u679c\u7248\u672c\u4e0d\u540c":71,"paddle\u5185\u90e8\u7684\u7c7b\u4e3ac":71,"paddle\u7684\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0\u5305\u62ec\u4e00\u4e0b\u51e0\u4e2a\u65b9\u9762":71,"paddle\u7684\u7c7b\u578b\u5168\u90e8\u9000\u5316\u6210":72,"paddle\u7684\u94fe\u63a5\u65b9\u5f0f\u6bd4\u8f83\u590d\u6742":71,"paddle\u7684c":72,"paddle\u8bad\u7ec3\u4efb\u52a1":38,"paddle\u8def\u5f84\u4e0b":72,"paddle\u9700\u8981\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3":71,"paddle\u9700\u8981\u66b4\u9732\u7684api\u5f88\u591a":72,"paddle\u9759\u6001\u5e93\u94fe\u63a5\u590d\u6742":71,"paddle_\u7c7b\u578b\u540d":72,"paddle_\u7c7b\u578b\u540d_\u51fd\u6570\u540d":72,"paddlepaddle\u4e2d":[136,139],"paddlepaddle\u4e2d\u4e00\u4e2a\u8ba1\u7b97\u5c42\u7684\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f\u548c\u8f93\u5165\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f\u5b8c\u5168\u76f8\u540c":115,"paddlepaddle\u4e2d\u7684\u8bb8\u591alayer\u5e76\u4e0d\u5728\u610f\u8f93\u5165\u662f\u5426\u662f\u65f6\u95f4\u5e8f\u5217":137,"paddlepaddle\u4e2d\u7684cudnn\u90e8\u5206\u4f7f\u7528\u7684\u4e5f\u662f":68,"paddlepaddle\u4e2d\u795e\u7ecf\u7f51\u7edc\u8ba1\u7b97\u5c42\u8f93\u5165":115,"paddlepaddle\u4e2d\u8fd8\u5305\u542b":108,"paddlepaddle\u4e2d\u901a\u8fc7reader\u6765\u52a0\u8f7d\u6570\u636e":110,"paddlepaddle\u4e3a\u4ea4\u53c9\u7f16\u8bd1\u63d0\u4f9b\u4e86\u5de5\u5177\u94fe\u914d\u7f6e\u6587\u6863":[142,143],"paddlepaddle\u4e3a\u6df1\u5ea6\u5b66\u4e60\u7814\u7a76\u4eba\u5458\u63d0\u4f9b\u4e86\u4e30\u5bcc\u7684api":110,"paddlepaddle\u4e3ano":28,"paddlepaddle\u4e3b\u8981\u4f7f\u7528":27,"paddlepaddle\u4f1a\u81ea\u52a8\u8bbe\u5b9a":108,"paddlepaddle\u4f7f\u7528\u540c\u6b65\u5c4f\u969c":118,"paddlepaddle\u4f7f\u7528\u5747\u503c0":109,"paddlepaddle\u4f7f\u7528avx":104,"paddlepaddle\u4f7f\u7528git":88,"paddlepaddle\u4fdd\u5b58\u7684\u6a21\u578b\u53c2\u6570\u6587\u4ef6\u5185\u5bb9\u753116\u5b57\u8282\u5934\u4fe1\u606f\u548c\u7f51\u7edc\u53c2\u6570\u4e24\u90e8\u5206\u7ec4\u6210":109,"paddlepaddle\u4fdd\u5b58\u7684\u6a21\u578b\u53c2\u6570\u6587\u4ef6\u524d16\u5b57\u8282\u4e3a\u5934\u4fe1\u606f":109,"paddlepaddle\u53d1\u5e03\u7684\u5b89\u88c5\u5305\u4f1a\u5c3d\u91cf\u5bf9\u9f50":30,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528\u591a\u79cd\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u6784\u5efa\u5206\u5e03\u5f0f\u8ba1\u7b97\u4efb\u52a1":120,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528\u5e38\u7528\u7684python\u5305\u7ba1\u7406\u5de5\u5177":30,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528cudnn":27,"paddlepaddle\u53ef\u4ee5\u540c\u65f6\u652f\u6301\u540c\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":118,"paddlepaddle\u53ef\u4ee5\u6bd4\u8f83\u7b80\u5355\u7684\u5224\u65ad\u54ea\u4e9b\u8f93\u51fa\u662f\u5e94\u8be5\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684":137,"paddlepaddle\u53ef\u4ee5\u901a\u8fc7\u8be5\u673a\u5236\u5224\u65ad\u662f\u5426\u5df2\u7ecf\u6536\u96c6\u9f50\u6240\u6709\u7684\u68af\u5ea6":100,"paddlepaddle\u5728\u5b9e\u73b0rnn\u7684\u65f6\u5019":137,"paddlepaddle\u5728\u6fc0\u6d3b\u51fd\u6570\u91cc\u5b9e\u73b0dropout":108,"paddlepaddle\u5728\u7f16\u8bd1\u65f6":27,"paddlepaddle\u5b58\u7684\u662f\u6709\u503c\u4f4d\u7f6e\u7684\u7d22\u5f15":110,"paddlepaddle\u5b89\u88c5\u5305\u7531\u4e8e\u4e0d\u4ec5\u4ec5\u5305\u542b":30,"paddlepaddle\u5c06\u4f1a\u6839\u636e":143,"paddlepaddle\u5c06\u4f1a\u81ea\u52a8\u9009\u62e9":143,"paddlepaddle\u5c06\u6839\u636e":142,"paddlepaddle\u5c06\u81ea\u52a8\u4e0b\u8f7d\u548c\u7f16\u8bd1\u6240\u6709\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93":[142,143,144],"paddlepaddle\u5e93\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6210":143,"paddlepaddle\u5f00\u53d1\u8fc7\u7a0b\u4f7f\u7528":88,"paddlepaddle\u63d0\u4f9b":29,"paddlepaddle\u63d0\u4f9b\u4e13\u7528\u7684":38,"paddlepaddle\u63d0\u4f9b\u7684":108,"paddlepaddle\u652f\u6301":27,"paddlepaddle\u652f\u6301\u4e0d\u540c\u7c7b\u578b\u7684\u8f93\u5165\u6570\u636e":110,"paddlepaddle\u652f\u6301\u4f7f\u7528pip\u5feb\u901f\u5b89\u88c5":112,"paddlepaddle\u652f\u6301\u975e\u5e38\u591a\u7684\u4f18\u5316\u7b97\u6cd5":107,"paddlepaddle\u652f\u6301sparse\u7684\u8bad\u7ec3":107,"paddlepaddle\u6587\u6863\u4f7f\u7528":103,"paddlepaddle\u662f\u6e90\u4e8e\u767e\u5ea6\u7684\u4e00\u4e2a\u6df1\u5ea6\u5b66\u4e60\u5e73\u53f0":110,"paddlepaddle\u6bcf\u6b21\u53d1\u65b0\u7684\u7248\u672c":88,"paddlepaddle\u6bcf\u6b21\u53d1\u7248\u672c\u9996\u5148\u8981\u4fdd\u8bc1paddlepaddl":88,"paddlepaddle\u7684":122,"paddlepaddle\u7684\u4e3b\u7248\u672c\u5e93\u9075\u5faa":88,"paddlepaddle\u7684\u5185\u5b58\u5360\u7528\u4e3b\u8981\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u65b9\u9762":107,"paddlepaddle\u7684\u53c2\u6570\u4f7f\u7528\u540d\u5b57":109,"paddlepaddle\u7684\u5404\u7248\u672c\u955c\u50cf\u53ef\u4ee5\u53c2\u8003":122,"paddlepaddle\u7684\u5b89\u88c5\u53ef\u4ee5\u53c2\u8003":127,"paddlepaddle\u7684\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6210":142,"paddlepaddle\u7684\u6240\u6709layer\u90fd\u6709\u552f\u4e00\u7684nam":108,"paddlepaddle\u7684\u6587\u6863\u5305\u62ec\u82f1\u6587\u6587\u6863":103,"paddlepaddle\u7684\u6587\u6863\u6784\u5efa\u6709\u4e09\u79cd\u65b9\u5f0f":103,"paddlepaddle\u7684\u6e90\u7801":98,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":27,"paddlepaddle\u7684activation\u4f1a\u76f4\u63a5\u4f7f\u7528":68,"paddlepaddle\u7684bas":100,"paddlepaddle\u7684c":142,"paddlepaddle\u7684cmake\u7cfb\u7edf\u4f1a\u81ea\u52a8\u7f16\u8bd1\u6240\u6709\u7684\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93":143,"paddlepaddle\u7684cmake\u7cfb\u7edf\u5c06\u6839\u636e\u8be5\u503c\u81ea\u52a8\u63a8\u5bfc\u548c\u8bbe\u7f6e\u9700\u8981\u4f7f\u7528\u7684\u4ea4\u53c9\u7f16\u8bd1\u5668":142,"paddlepaddle\u7684cmake\u7cfb\u7edf\u5c06\u6839\u636e\u8be5\u503c\u81ea\u52a8\u8bbe\u7f6e\u9700\u8981\u4f7f\u7528\u7684\u4ea4\u53c9\u7f16\u8bd1\u5668":144,"paddlepaddle\u7684cmake\u7cfb\u7edf\u624d\u8ba4\u4e3a\u5728\u662f\u5728\u4ea4\u53c9\u7f16\u8bd1raspberri":144,"paddlepaddle\u7684cmake\u7cfb\u7edf\u624d\u8ba4\u4e3a\u662f\u5728\u4ea4\u53c9\u7f16\u8bd1android\u7cfb\u7edf\u7684\u7248\u672c":142,"paddlepaddle\u7684dock":122,"paddlepaddle\u7684softmax\u4e0d\u80fd\u6307\u5b9a\u8ba1\u7b97\u7ef4\u5ea6":108,"paddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":137,"paddlepaddle\u76ee\u524d\u63d0\u4f9b\u4e24\u79cd\u53c2\u6570\u521d\u59cb\u5316\u7684\u65b9\u5f0f":109,"paddlepaddle\u76ee\u524d\u652f\u63018\u79cdlearning_rate_schedul":109,"paddlepaddle\u7f16\u8bd1\u9700\u8981\u4f7f\u7528\u5230\u4e0b\u9762\u7684\u4f9d\u8d56":27,"paddlepaddle\u82e5\u68c0\u6d4b\u5230\u7528\u6237\u4f7f\u7528\u7684cmake\u7248\u672c\u4e0d\u4f4e\u4e8e3":142,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u68af\u5ea6\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":139,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u8bef\u5dee\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":139,"paddlepaddle\u955c\u50cf\u9700\u8981\u63d0\u4f9b":123,"pass\u4e2a\u6a21\u578b\u5230\u7b2c":129,"pass\u5c06\u4e0d\u8d77\u4f5c\u7528":129,"pass\u8f6e\u5f00\u59cb\u8bad\u7ec3":129,"pass\u8f6e\u7684\u6a21\u578b\u7528\u4e8e\u6d4b\u8bd5":129,"passes\u8f6e":129,"patch\u53f7":88,"patch\u53f7\u52a0\u4e00":88,"path\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b":131,"perftools\u6765\u8fdb\u884c\u6027\u80fd\u5206\u6790":133,"period\u4e2a\u6279\u6b21\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":129,"period\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":129,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u53c2\u6570\u7edf\u8ba1":129,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u7b26\u53f7":129,"period\u6574\u9664":129,"period\u8f6e\u4fdd\u5b58\u8bad\u7ec3\u53c2\u6570":129,"pfsclient\u9700\u8981\u548cingress\u4e4b\u95f4\u505a\u53cc\u5411\u9a8c\u8bc1":54,"pfsclient\u9700\u8981\u5728\u4f20\u8f93\u5b8c\u6bd5\u6700\u540e\u4e00\u4e2achunk\u7684\u65f6\u5019\u68c0\u67e5destination\u6587\u4ef6\u7684md5\u503c\u662f\u5426\u548csource\u6587\u4ef6\u4e00\u81f4":54,"pfsserver\u63d0\u4f9brest":54,"pi\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u7684\u65b9\u6cd5\u548c\u6b65\u9aa4":144,"pi\u7248\u672c\u7684\u5e93":144,"pi\u7248\u672cpaddlepaddle\u5e93\u65f6":144,"pi\u7684\u914d\u7f6e\u4fe1\u606f\u5728":144,"pi\u7cfb\u7edf\u4e0a\u6765\u6784\u5efa":144,"pi\u7cfb\u7edf\u7684\u7248\u672c":144,"pip\u548cdocker\u7684\u5b89\u88c5\u65b9\u5f0f":29,"pserver\u5730\u5740\u7b49\u53c2\u6570\u4f7ftrainer\u53ef\u4ee5\u6b63\u786e\u8fde\u63a5\u5230pserv":117,"pserver\u76d1\u542c\u7684\u8d77\u59cb\u7aef\u53e3":117,"public":[10,18,34,46,56,59,64,80,86,89,91,92,93,96,100,101,102,121,122],"pwd\u53d8\u91cf\u4f1a\u5c55\u5f00\u4e3a\u5f53\u524d\u8def\u5f84\u7684\u7edd\u5bf9\u8def\u5f84":28,"py\u4e2d":88,"py\u7a0b\u5e8f":30,"pydataprovider\u4f7f\u7528\u7684\u662f\u5f02\u6b65\u52a0\u8f7d":107,"pypi\u4e0a\u7684package\u540d\u79f0\u4e3apaddlepaddle\u548cpaddlepaddl":88,"pypi\u4e0d\u652f\u6301\u8986\u76d6\u4e0a\u4f20":88,"pypi\u5b89\u88c5\u5305\u53ef\u4ee5\u5728":30,"python\u5b89\u88c5\u5305\u652f\u6301linux":104,"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":100,"python\u5e93yep":133,"python\u6807\u51c6\u5e93\u4e2d\u63d0\u4f9b\u4e86\u6027\u80fd\u5206\u6790\u7684\u5de5\u5177\u5305":133,"reader\u7684\u4f7f\u7528\u65b9\u5f0f\u90fd\u662f\u4e00\u81f4\u7684":38,"reader\u8f93\u51fa\u7684data":38,"recommendation\u6587\u4ef6\u5939\u5185\u5b58\u653e\u8bad\u7ec3\u6587\u4ef6":123,"request\u524d":98,"request\u7684":98,"request\u88ab\u5408\u5e76\u540e":98,"resnet\u7684mkl":68,"return":[2,4,5,7,9,10,11,17,18,23,25,31,32,33,34,38,39,41,44,45,46,47,52,56,58,59,60,62,64,65,66,69,74,75,76,80,81,82,86,89,91,93,96,100,102,107,109,110,121,123,140],"rnn\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":137,"rnn\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u901a\u8fc7\u4e86\u4e00\u4e2alstm\u7f51\u7edc":137,"rnn\u603b\u662f\u5f15\u7528\u4e0a\u4e00\u65f6\u523b\u9884\u6d4b\u51fa\u7684\u8bcd\u7684\u8bcd\u5411\u91cf":139,"rnn\u6a21\u578b":132,"rnn\u90e8\u5206\u4e2d":67,"rnn\u914d\u7f6e":138,"root\u66ff\u6362\u4e3apaddlepaddle\u9884\u6d4b\u5e93\u7684\u5b89\u88c5\u8def\u5f84":113,"s3\u4e4b\u7c7b\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u4e4b\u4e0a":38,"search\u7684\u65b9\u6cd5":129,"sentences\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684\u6570\u636e":137,"seq\u53c2\u6570\u5fc5\u987b\u4e3afals":139,"server\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":129,"server\u4e4b\u4e0a":37,"server\u4e4b\u95f4\u7684\u7f51\u7edc\u5e26\u5bbd":37,"server\u4f1a\u6682\u505c\u53c2\u6570\u66f4\u65b0\u5e76\u7b49\u5f85":37,"server\u4f1a\u83b7\u53d6parameters\u5185\u5b58\u7684":37,"server\u5185\u5b58\u4e2d\u7684\u6a21\u578b\u6570\u636e\u7684\u5b8c\u6574\u955c\u50cf":37,"server\u540c\u6b65\u7684\u4fdd\u5b58\u4e00\u4e2a\u7279\u5b9a\u65f6\u95f4\u70b9\u7684\u5168\u5c40\u68c0\u67e5\u70b9":37,"server\u5728\u96c6\u7fa4\u4e2d\u542f\u52a8\u540e":37,"server\u6545\u969c\u540e\u88abkubernetes\u91cd\u65b0\u542f\u52a8":37,"server\u6b64\u65f6\u8fd8\u9700\u8981\u901a\u8fc7\u7f51\u7edc\u8bbf\u95ee\u5206\u5e03\u5f0f\u5b58\u50a8\u4ee5\u4fdd\u5b58\u5feb\u7167":37,"server\u751f\u6210\u4e00\u4e2auuid":37,"server\u7684\u5355\u70b9\u6216\u591a\u70b9\u540c\u65f6\u6545\u969c":37,"server\u7684\u6570\u636e\u5feb\u7167":37,"server\u7684\u68c0\u67e5\u70b9\u5404\u81ea\u72ec\u7acb\u4fdd\u5b58":37,"server\u7b2c\u4e00\u6b21\u542f\u52a8\u6216\u4efb\u610f\u65f6\u95f4paramet":37,"short":[4,5,56,60,83,89,92,96],"simd\u6307\u4ee4\u63d0\u9ad8cpu\u6267\u884c\u6548\u7387":104,"size\u4e3a512":129,"size\u53ef\u80fd\u4f1a\u5bf9\u8bad\u7ec3\u7ed3\u679c\u4ea7\u751f\u5f71\u54cd":107,"size\u672c\u8eab\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u8d85\u53c2\u6570":107,"softmax\u6fc0\u6d3b\u7684\u8f93\u51fa\u7684\u548c\u603b\u662f1":100,"sparse\u8bad\u7ec3\u9700\u8981\u8bad\u7ec3\u7279\u5f81\u662f":107,"static":[25,41,72,89,91,121,145],"step\u51fd\u6570\u4e2d\u7684memori":139,"step\u51fd\u6570\u5185\u90e8\u53ef\u4ee5\u81ea\u7531\u7ec4\u5408paddlepaddle\u652f\u6301\u7684\u5404\u79cdlay":139,"store\u4e0b\u8f7d\u5b89\u88c5xcode\u5373\u53ef":143,"subseq\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":136,"super":[83,100],"swig\u652f\u6301\u7684\u8bed\u8a00\u6216\u8005\u89e3\u91ca\u5668\u6709\u5c40\u9650":71,"swig\u66b4\u9732\u7684\u63a5\u53e3\u4fdd\u7559\u4e86c":71,"swig\u751f\u6210\u7684\u4ee3\u7801\u4e0d\u80fd\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":71,"swig\u76f4\u63a5\u8bfb\u53d6c":71,"swig\u9700\u8981\u5199\u4e00\u4e2ainterface\u6587\u4ef6":71,"switch":[34,72,121],"tag\u4e3a":88,"tag\u53ef\u4ee5\u662flatest\u6216latest":88,"tag\u7684\u66f4\u65b0\u65f6\u95f4\u662f\u5426\u5728\u4e0a\u8ff0\u7f16\u8bd1wheel\u5305\u5b8c\u6210\u540e\u662f\u5426\u6700\u65b0":88,"tensor\u5230\u5bf9eigentensor\u7684\u8f6c\u6362":102,"tensor\u5230eigentensor":102,"tensor\u5b9a\u4e49\u5728framework\u76ee\u5f55\u4e0b":102,"tensor\u662f\u4e00\u4e2a\u6b63\u5728\u5f00\u53d1\u4e2d\u7684\u6a21\u5757":102,"tensor\u6a21\u5757\u5bf9el":102,"tensor\u6a21\u5757\u6765\u5b9e\u73b0":101,"tensor\u6a21\u5757\u7684\u6587\u6863\u8f83\u5c11":102,"tensor\u6a21\u5757\u7684\u8be6\u7ec6\u4ecb\u7ecd\u8bf7\u53c2\u8003":102,"tests\u7684paddlepaddl":98,"tflops\u4e86":134,"throw":121,"tottime\u7684\u6bcf\u6b21\u8c03\u7528\u5e73\u5747\u65f6\u95f4":133,"trainer\u542f\u52a8\u9700\u8981\u4f20\u5165\u7aef\u53e3":117,"trainer\u63a5\u6536\u4e09\u4e2a\u53c2\u6570":110,"trainer\u8282\u70b9\u4e2a\u6570":117,"trainer\u9700\u8981\u548cpserver\u4fdd\u6301\u7f51\u7edc\u8054\u901a\u4ee5\u5b8c\u6210\u8bad\u7ec3":117,"true":[2,3,4,5,6,7,9,10,11,17,18,19,21,25,31,33,34,39,57,63,67,76,81,82,83,84,88,91,96,100,107,109,115,116,121,123,131,140],"true\u8868\u793a\u53cd\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":140,"try":[35,36,39,40,41,57,62,66,69,84,89,92,104],"type\u5b57\u6bb5\u5747\u4e0d\u5c3d\u76f8\u540c":72,"type\u6307\u5b9a\u4e3a":133,"ubuntu\u4e0b\u5b89\u88c5\u547d\u4ee4\u4e3a":133,"unit\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u8ba1\u7b97\u5f97\u5230\u7684\u4e2d\u95f4\u503c":108,"unsupported\u6a21\u5757":101,"update\u53c2\u6570\u65f6\u624d\u6709\u6548":129,"v1\u7248\u672c":104,"var":[17,18,33,34,45,46,47,51,58,59,61,63,76,81,82,83,85,89,94,95,96,103],"vector\u662frank\u4e3a1\u7684tensor":102,"void":[34,39,41,46,53,56,58,59,64,69,70,71,72,81,82,90,91,93,95,100,101,102],"wheel\u5305":88,"while":[2,4,9,34,43,46,47,57,62,65,66,75,78,79,84,87,89,93,123],"wise\u8ba1\u7b97\u63d0\u4f9b\u4e86\u5f3a\u5927\u7684\u652f\u6301":102,"wmt14\u6570\u636e\u7684\u63d0\u4f9b\u6587\u4ef6\u5728":140,"words\u5373\u4e3a\u8fd9\u4e2a\u6570\u636e\u4e2d\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":137,"words\u662f\u539f\u59cb\u6570\u636e\u4e2d\u7684\u6bcf\u4e00\u53e5\u8bdd":137,"x86_64\u548cmaco":104,"x\u4e0ey\u4e3a\u4e4b\u524d\u63cf\u8ff0\u7684\u8f93\u5165\u5c42":110,"x\u548cwindow":142,"y\u8868\u793a\u8f93\u5165\u6570\u636e\u662f\u4e00\u4e2a\u7ef4\u5ea6\u4e3a1\u7684\u7a20\u5bc6\u5411\u91cf":110,"yaml\u6587\u4ef6\u4e2d\u5404\u4e2a\u5b57\u6bb5\u7684\u5177\u4f53\u542b\u4e49":123,"yaml\u6587\u4ef6\u63cf\u8ff0\u4e86\u8fd9\u6b21\u8bad\u7ec3\u4f7f\u7528\u7684docker\u955c\u50cf":123,"zero\u4e09\u79cd\u64cd\u4f5c":129,AGE:[121,122],AWS:[9,38,120,125,126],Abs:18,Added:92,And:[3,4,6,9,10,11,18,25,39,43,44,53,60,62,73,77,80,84,89,93,121],But:[4,5,9,18,32,59,65,73,80,89,104,145],EOS:4,For:[3,4,5,6,9,18,25,31,33,34,40,41,42,44,45,46,48,51,52,57,58,59,61,64,66,70,73,74,75,78,79,80,81,82,83,84,85,86,87,90,91,92,93,95,97,134,145],IDE:26,IDs:[10,18,43,47,75],IRs:48,Into:121,Its:[3,4,58,90,121],K8s:145,NMS:4,NOT:[18,83],Not:[31,36,66,92,145],OPs:[48,50],One:[3,5,18,25,32,43,70,73,89,92,95],Ops:[85,87,91],PFS:54,QoS:122,Such:[64,83,92,96],TLS:[31,54,121],That:[4,9,46,77],The:[1,2,3,4,5,6,9,10,11,17,18,19,22,25,31,32,33,35,36,40,42,43,44,46,47,48,50,51,52,55,56,58,62,65,66,69,70,72,74,75,77,78,80,81,82,83,84,87,89,90,91,92,93,95,96,97,100,101,102,115,121,123,135],Their:[4,36],Then:[4,5,18,45,48,59,64,66,77,80,82,121],There:[3,4,10,18,25,31,34,35,36,41,43,44,47,48,55,56,57,62,66,73,74,75,78,79,80,83,87,89,90,93,121],These:[3,11,18,33,34,46,51,56,61,76,87,90,91,92],Use:[3,9,18,31,49,55,84,85,92,121],Used:[5,21,85,93],Uses:66,Using:[18,36,57,78,84,85,87,89],VPS:121,Will:[9,25,110],With:[4,5,45,46,51,57,77,81,92,96],YES:44,Yes:[28,68],___embedding_0__:123,___embedding_1__:123,___fc_layer_0__:121,__align__:56,__cuda_align__:56,__device__:56,__doc__:91,__file__:44,__forceinline__:56,__fp16:56,__global__:56,__gradient_machines__:25,__hadd:56,__half:56,__half_raw:56,__impl__:91,__init__:[51,52,60,66,76,83,96,100,133],__main__:60,__name__:60,__param_conf__:25,__rnn_step__:140,__square_error_cost_0__:123,__tmp_params__:25,__va_args__:86,__x:56,_addup_repetitive_outputs_:33,_append_backward_ops_:[33,51],_append_backward_vars_:33,_binari:35,_create_global_var:83,_def:66,_dtype:62,_filer:69,_filter:69,_fwd:69,_fwd_pd:69,_input:69,_librari:35,_link:5,_live_in:66,_live_out:66,_loss:60,_op:[62,101],_output:69,_presucessor:66,_program:66,_proj:4,_recurrent_group:140,_remove_no_grad_branch_:33,_reorder_input:69,_source_language_embed:140,_src_input:69,_sucessor:66,_target_language_embed:140,_test:35,_update_op:52,_use:66,_value_index:62,a75:56,a_op:101,a_prev:92,aaaaa:38,aaaaaaaaaaaaa:121,abbrevi:11,abc:4,abil:60,abl:[4,31,33,46,47,48,64,76,80,83,145],about:[5,11,18,32,34,35,44,50,55,58,66,74,84,89,91,92,93,121],abov:[3,4,18,31,32,33,34,35,36,40,45,47,48,56,57,58,59,61,69,74,75,76,77,78,80,81,83,91,92,94,95,96,121,134,145],abs:[5,32,60,107],abs_numerical_grad:32,absolut:[32,107],acc:48,acceler:[4,37,68,77,78],accept:[2,4,9,18,31,85],access:[4,5,18,31,35,40,43,44,45,47,48,83],accessmod:121,accessor:83,accord:[3,4,11,18,32,33,41,48,50,61,75,85,96],accordingli:[3,4],account:[85,145],accoust:92,accrodingli:39,accumul:[14,36,41,52,77,78,92],accur:[32,43,80],accuraci:[3,52,92],achiev:[46,50,77,78,92,93],acl2016:10,acquir:57,across:[4,9,33,48,74,92],act1:62,act2:62,act:[4,5,18,19,34,46,48,62,75,83,96,107,110,112,135,140],act_output:91,act_typ:62,actgat:18,action:[47,121],activ:[5,18,19,24,35,62,66,75,80,83,87,91,107,110,112,140],activi:5,actnod:18,actual:[4,18,39,51,57,60,62,69,73,78,91,93,95],actual_layout:69,adadelta:107,adagrad:[78,90],adagradoptim:[20,76],adam:[31,41,48,60,109],adamaxoptim:20,adamoptim:20,adapt:[3,6,18,25,80,95],add:[4,5,9,18,22,23,25,32,33,34,35,39,43,46,47,48,50,52,56,59,63,65,76,78,82,83,85,87,89,93,94,98,102,104,135],add_activ:83,add_bia:83,add_depend:35,add_execut:35,add_input:[74,100],add_memori:74,add_output:74,add_scalar:[34,75,81],add_sum:83,add_test:[35,100],add_to:108,add_two:[34,74],add_unittest_without_exec:100,addattr:[91,101],addbia:100,addcom:[91,101],added:[3,4,10,18,25,34,50,51,56,73,77,78,87,98],adding:87,addinput:[91,101],addit:[4,5,33,47,77,80,85,87,96],addition:74,addmemori:69,addop:59,addoutput:101,addprimit:69,addprimitivedesc:69,addr:36,address:[36,41,45,48,115,145],addrow:100,addtolay:4,addtyp:91,adjust:[33,51],admin:145,administr:[43,145],adopt:[56,60],advanc:32,advantag:[32,56,57,78,84],adversari:[60,84],affect:[4,18,34],affili:75,afford:40,aforement:35,after:[4,5,10,11,18,33,34,35,40,41,43,48,49,50,51,53,55,56,66,69,77,79,80,83,92,98,121],aftern:80,again:[31,32,36,78],against:121,age:[10,123],agg_level:[4,136,137],aggreg:[52,77,121],aggregatelevel:[136,137],ago:35,agre:[107,110],ahead:92,alex:18,alexnet_pass1:131,alexnet_pass2:131,algo:69,algo_hrnn_demo:137,algorithm:[4,6,18,33,40,51,66,69,75,78,87,92,95],alia:18,alias:18,align:[4,5,9],all:[2,3,4,17,18,19,25,31,33,34,35,36,39,41,43,44,45,46,47,48,49,51,53,55,57,60,61,62,65,66,69,70,72,73,74,75,76,77,78,80,81,83,85,91,92,93,104,107,110,121,123,135,139,145],all_output_nam:33,alloc:[2,41,44,66,69,93,102,135],allow:[31,41,45,46,48,51,57,78,87,121],allow_only_one_model_on_one_gpu:[128,129,131],allreduc:77,almost:18,along:[4,11,18,19],alpha:[18,35,87],alreadi:[18,22,35,36,57,69,83,89,104,121],also:[4,5,10,18,21,31,33,34,35,39,42,46,47,48,56,57,59,60,61,62,65,66,73,74,75,78,79,80,81,82,83,84,87,89,91,92,93,96,97,134,145],although:[22,33,77],altogeth:145,alwai:[4,5,18,25,35,47,70,90,121,123],amazon:[121,122],amazonaw:121,amazonec2fullaccess:121,amazonelasticfilesystemfullaccess:121,amazonroute53domainsfullaccess:121,amazonroute53fullaccess:121,amazons3fullaccess:121,amazonvpcfullaccess:121,ambigu:[84,92],amd64:121,amd:73,amend:98,amodei:92,among:[18,121],amort:77,amount:18,analys:80,analysi:80,analyz:66,ancestor:[81,83],andd:121,andrew:66,android:142,android_abi:142,android_api:142,android_arm_neon:142,android_native_api_level:142,android_standalone_toolchain:142,android_toolchain:142,androideabi:142,ani:[4,5,9,18,21,31,35,36,41,43,44,45,47,48,53,56,57,64,66,70,75,77,78,83,84,86,87,92,95,107,110,121],announc:56,anoth:[4,9,18,25,31,33,34,44,46,47,57,58,69,75,83,89,91,93,95,121],anroid_arm_mod:142,ans:121,answer:[45,57,121],anthor:18,anymor:77,anyth:[9,75,84,121],anytim:60,apach:[68,107,110],apart:18,api:[10,25,30,31,33,35,41,42,44,45,46,52,54,59,60,62,74,79,80,85,88,96,97,113,115,116,117,121,123,133,134,142,145],api_pydataprovider2:107,api_shar:35,api_test:35,api_trainer_config_helpers_lay:140,apiserv:121,apivers:[121,122,123],appar:33,appear:[45,57,61,93],appel:66,append:[18,25,33,51,52,75,83,84,92,117,123,140],append_backward:[33,76,133],append_batch_s:18,append_clip_op:51,append_gradient_machin:25,append_op:[51,65,83],append_oper:83,appleyard:134,appli:[4,18,23,60,61,77,80,89],applic:[22,45,47,56,57,58,61,83,85,107,110,121,122,134,145],applyl1:39,appreci:92,approach:[4,48,49,50,77,78,79,87,92,145],approxim:[18,78],apt:[28,133],arbitrari:[4,48,70],arch:142,archetectur:92,architectur:[56,92],archiv:[10,18,71,72,113],area:60,arg:[3,5,15,22,33,62,76,91,101,109,123],arg_nam:4,argmax:18,argpars:123,args_ext:123,argu:82,argument:[4,9,11,18,22,33,34,39,40,48,76,79,82,83,113,115,116,123,142],argumentpars:123,arithmet:56,arm64:[142,143],arm64_standalone_toolchain:142,arm:[56,142,143,144],arm_standalone_toolchain:142,armeabi:142,armv7:[56,143],armv8:56,arn:121,around:[4,43,66,83,121,145],arrai:[2,4,9,11,18,25,41,45,47,61,75,81,83,84,85,96,101,107,109,110,115],arrang:[18,96],arrari:[18,115],array_to_lod_tensor:66,arrow:60,articl:[10,18,58,61,98],artifact:[88,121],artifici:66,arxiv:[5,18,60,92],as_row_vector:4,as_step_input:34,asgd:78,ask:[33,36,43],asr:92,assgin:66,assign:[3,4,19,33,40,45,49,51,56,58,77,92,115,121,145],assigne:92,assignmemt:66,associ:[79,86],assum:[3,4,18,34,48,69],assumpt:48,ast:45,astyp:[84,101],asyc:36,async:[36,50,128],async_count:[128,129],async_lagged_grad_discard_ratio:129,async_lagged_ratio_default:[128,129],async_lagged_ratio_min:[128,129],asynchron:[36,47,77,80],atom:49,att_seq:5,attach:5,attend:5,attended_sequ:5,attenion:5,attent:[4,5,19],attr1:4,attr2:4,attr:[2,4,5,18,34,45,62,65,69,81,82,83,91,101,107,108,109,140],attr_map:91,attrdesc:81,attribu:69,attribut:[4,5,18,24,33,34,50,51,65,81,83,85,89,91,96],attributemap:101,attrproto:91,attrtyp:[81,91,101],attrvalu:91,auc:[52,128],audio:92,augment:92,authent:121,author:[10,54,92,107,110,121],auto:[27,34,39,49,58,69,71,82,85,89,95,96,100,101,102,134],autom:121,automat:[4,18,31,33,41,48,50,51,59,76,85,91,92,121],avail:[10,36,41,50,56,57,66,121,145],ave:22,averag:[3,4,7,14,18,22,25,40,107],average_test_period:[128,129],avg:[134,136],avg_cost:[48,135],avg_x:18,avgpool:4,avoid:[18,32,34,36,47,48,65,69,77,78,79,134],avx:28,awai:57,await:122,awar:[31,45,47,52,58,74,83,121],awk:124,awni:92,aws:54,aws_account_id:121,awsaccountid:121,awskeymanagementservicepowerus:121,axi:[4,18,107],axiom:47,axis:4,b363:122,b8561f5c79193550d64fa47418a9e67ebdd71546186e840f88de5026b8097465:122,ba5f:121,back:[4,18,25,33,36,48,56,60,78],background:[3,4,87,92],background_id:[3,4],backpropag:[32,33],backward:[1,4,5,18,23,32,34,39,41,51,60,67,68,76,78,79,82,86,87,94,100,101,135],backward_first:140,backward_op:32,backwardactiv:100,bacth:46,baidu:[18,57,92,122],bake:48,balanc:[50,77,121],bandwidth:[56,77],bare:[122,145],barrier:118,barrierstatset:134,basci:62,base:[3,4,7,9,10,18,26,31,40,46,51,52,56,57,64,69,73,76,77,78,80,85,86,87,93,96,121,135],baseactiv:5,baseerrorclipattr:51,baseev:25,baselin:92,basematrix:100,basenam:3,basepoolingtyp:[4,5],basestr:[2,3,4,5,7,25],bash:[26,27,28,98,121,122,123,127],basi:[107,110],basic:[4,25,48,62,69,80,81,85,86,92,96],batch:[4,5,9,11,14,18,19,25,31,34,36,38,39,46,47,48,52,53,57,60,63,73,74,75,77,78,92,96,107,110,115,121,122,123],batch_id:[25,60,107,110],batch_im:60,batch_images_from_tar:11,batch_label:60,batch_norm:[60,92],batch_norm_lay:5,batch_norm_typ:4,batch_read:[38,46,84],batch_ref:18,batch_siz:[9,18,48,60,67,75,107,110],batch_szi:60,batch_z:60,batchnorm:[18,60,92],batchread:46,batchsiz:[4,100],bazel:35,bbbbb:38,bbox:3,bcd:4,bcebo:10,bcm2708:144,bdist_wheel:88,beacus:62,beam:[4,18,140],beam_gen:[4,140],beam_search:[25,75,139,140],beam_siz:[4,18,75,128,129,131,140],becaus:[3,4,10,31,32,34,35,36,41,56,75,79,83,84,87,89,90,94,95,96,97,121],becom:[18,49,50,89,93],been:[4,5,18,33,35,40,47,57],befor:[4,5,18,33,36,43,47,51,55,58,61,73,78,79,80,84,87,101,104,107,121,145],begin:[3,4,18,22,33,39,41,46,52,55,61,75,77],begin_norm_axi:18,beginiter:[25,31],beginn:140,beginpass:[25,31],begintrain:31,behavior:[18,47],behind:[57,96],being:[18,33,43,51,57,82,84],belong:[3,4,46,48,89],below:[18,34,36,41,48,50,56,57,70,79,84,87,96,97,121],benchmark:[70,92],benefit:[5,43,44,75],besid:[4,10,21,48,66,73,77],best:[35,69],besteffort:122,beta1:6,beta2:6,beta:[18,60],better:[5,35,57,66,69,75,95,121,145],between:[3,4,11,18,25,33,35,36,41,48,50,56,57,69,72,77,79,86,89,121],bgr:11,bi_gru:5,bi_lstm:5,bia:[4,5,18,75,83,100],bias:[4,18],bias_attr:[4,5,18,83,107,109,140],bias_initi:18,bias_param_attr:5,biases_:100,biasparameter_:100,biassiz:100,bidi:122,bidirect:[4,5,92],bidirectional_lstm:108,big:[45,50,66,145],bigger:[18,36],bilinear:4,bilinear_interpol:4,bilinearfwdbwd:134,bin:[28,116,117,121,122,123,127],binari:[3,4,9,18,35,44,48,56,58,60,70,116,121],bind:[45,47,56,59,89,93],bioinf:18,bit:56,bitcod:143,black:60,blank:[4,18,121],blob:10,block0:66,block1:66,block2:66,block:[4,18,33,37,39,41,45,47,48,49,50,51,52,53,57,64,66,73,74,76,93,97,102],block_expand:92,block_i:[4,18],block_id:[45,53],block_x:[4,18],blockdesc:[23,34,61,83,85],blockdescbind:64,blockingcount:49,blueprint:75,bn_bias_attr:5,bn_layer_attr:5,bn_param_attr:5,book:[10,28,85,92,103,135,140],booktitl:10,bool:[2,3,4,5,6,7,9,10,11,17,18,25,34,46,56,63,65,67,68,69,82,83,90,91,96,97,100,115,129,131],boost:[73,92,93],boot:[4,18,139,140,145],boot_bia:4,boot_bias_active_typ:4,boot_lay:140,boot_stat:96,boot_with_const_id:4,bootstrapp:145,borrow:[60,96],bos_id:[4,140],both:[1,2,4,5,11,18,22,31,34,35,36,43,46,47,48,50,56,57,60,64,66,73,75,77,80,82,90,92,93,121],bottl:77,bottleneck:80,bottom:[25,92],bound:[4,18,66],boundari:48,boundri:3,box:[4,60],brace:[34,61],brain:43,branch:[4,18,31,34,35,48,57,63,81,88,94,98],break_if:96,brief:[35,41,56,93,102],bring:[57,66],broadcast:[18,36,77,85,145],broken:98,browser:121,bsd:[26,47,77],bsp:47,bucket_nam:121,buf:39,buf_siz:[9,48],buffer:[9,39,47,69,70,78,84,89,135],buffer_s:[9,47],buffered_read:84,bufsiz:9,bug:[98,121],build:[4,10,17,26,27,35,44,48,61,62,66,68,78,87,88,91,92,98,103,104,113,121,123,125,126,127,133,142,143,144],build_dict:10,build_doc:103,build_model:60,buildtool:88,built:[35,45,48,56,58,66,73,77,91,92,96,133,134,145],bulk:47,bunch:70,button:121,c11:71,c703c041:98,c99:72,c99e:121,cach:[56,107],cache_pass_in_mem:107,cachetyp:107,cacul:[5,52],caff:[34,57],caffe2:[34,45,47,57],caffe_poli:109,calcul:[3,4,5,18,32,33,36,41,49,52,56,66],calcut:66,calendar:80,call:[3,4,5,9,18,22,25,31,32,33,34,39,40,41,42,44,45,47,48,51,58,60,61,66,74,75,76,80,83,85,86,89,91,93,96,121,123,133,134],callabl:[2,4,9,10,17],callback:[51,100],caller:[32,121],can:[2,3,4,5,9,10,11,17,18,19,21,22,25,31,32,33,34,35,36,39,40,43,44,45,46,47,48,50,51,53,56,57,58,59,60,61,62,64,65,66,69,73,74,75,76,77,78,80,81,82,83,84,85,86,87,91,93,95,96,97,121,134,145],cancel:43,candid:[4,18,75,92],candidate_activ:18,cannot:[46,85,89,95,96,104],cantain:62,capabl:[56,79,85],capac:[87,121],capi:[71,113],capi_priv:113,capi_prvi:72,caption:75,captur:4,card:77,care:[5,44,66,84,92,93,145],carefulli:92,caret:25,carpedm20:60,carri:18,cast:[56,95],cast_to_op_attr:91,cat:[9,11,28,123,124],categori:[4,10,36],categorig:10,categoryfil:122,caus:[36,55],caution:121,cbla:[67,113],cc_:35,cc_binari:35,cc_test:35,cclient:42,cde:4,cdn:10,cduadevicecontext:[73,93],ceil:4,ceil_mod:4,cell:[4,5,18],cell_activ:18,cell_t_prev:18,cell_valu:18,center:11,center_crop:11,cento:[30,145],central:87,ceph:[9,38],cephf:[38,44,54],cer:92,certain:[18,46,65,73,76,80,89,93],certif:[31,54,104,121],cffi:71,cfg:[66,122],cgo:71,ch1:47,ch2:47,chain:[9,33,61],challeng:[4,32,36,57,63,93],chan:47,chanc:[31,56],chang:[4,10,35,40,44,48,57,69,79,81,84,86,88,89,92,98,121],changes:69,channel:[4,5,11,18,45,134],channel_shar:4,chapter:[74,75,92],chapter_data:74,chapter_out:74,charact:[47,92],check:[9,33,34,35,51,69,82,85,94,98,100,104,109,115,121,129],check_align:9,check_attr:91,check_eq:100,check_grad:[32,101],check_l:100,check_output:101,check_sparse_distribution_batch:[128,129],check_sparse_distribution_in_pserv:[128,129],check_sparse_distribution_ratio:[128,129],check_sparse_distribution_unbalance_degre:[128,129],checker:[32,85],checkgrad:129,checkgrad_ep:129,checkmark:145,checkout:98,checkpoint:[50,82],checksum:54,child:34,chip:57,chmod:121,choic:[18,35,57],choos:[18,47,65],chosen:[18,60,73],chunk:[18,40,54],chunk_ev:14,chunk_schem:[3,14,18],chunktyp:3,chw:11,circl:61,circular:47,circumst:93,cite:10,claim:121,claimnam:[121,123],clang:[56,71,98,142],clarifi:[3,18],clariti:75,classdim:18,classic:[4,66,92],classif:[4,18,61],classifi:[4,60],classification_cost:107,classification_error_evalu:3,classification_evalu:3,claster:121,clean:[34,35,53,79,85,98,104],clear:[3,22,35,75,79,89,95],clearer:[79,83],clearli:89,cli:121,click:121,client:[18,39,42,85],clip:[2,5,129],clip_op:51,clip_op_desc:51,clipbynorm:18,clock:4,clone:[4,26,27,103,113,133,142,144],close:[47,84,98],close_channel:47,cloud:[35,36,44,54,55,85,145],cloud_read:9,cludform:121,cluster:[9,25,31,34,36,41,48,92,117,119,123],cluster_test_fil:117,cluster_train:[107,119],cluster_train_fil:117,cluster_train_v2:[119,120,124],cm469:121,cmake:[26,27,72,98,101,103,104,113,134,142,143,144],cmake_build_typ:[142,143,144],cmake_c:[142,143],cmake_system_nam:[142,143,144],cmakefil:104,cmakelist:[35,67,68,100],cmatrix:[71,72],cmu:18,cname:121,cnn:[4,18,122],coars:59,code:[4,9,31,33,35,43,46,47,48,50,53,56,59,60,61,65,70,73,76,78,79,80,82,84,85,86,87,91,96,100,121,122],codebas:85,coded_stream:109,codedinputstream:109,coeff:4,colindic:115,collabor:36,collect:[4,10,25,80],collectbia:100,color:11,colour:10,colum:115,column:[3,4,18,61,84,115],column_evalu:3,com:[4,5,10,18,26,27,28,35,60,88,98,103,104,113,121,122,133,135,142,144,145],combin:[3,4,5,9,25,66,76,85,89],come:[48,52,66,81,92,96],comma:[18,22,25,41],command:[9,22,26,35,39,44,55,100,121,122,123,125,126,131],commandlin:[123,134],comment:[18,35,62,91,92,123],commit:[35,122],common:[11,38,87,93],commonli:[55,87],commun:[36,41,42,47,48,50,77,121],compani:57,compar:[32,35,45,85],comparison:[35,57],compat:[19,56,59,77],compil:[4,26,35,48,57,62,64,66,73,77,86,90,91,97,127,142,143,144],complaint:35,complet:[4,5,10,18,25,33,34,36,40,41,51,54,61,70,73,85,121,122,123,145],complex:[5,18,43,47,66,75,85],complianc:[107,110],complic:[4,48,59,84,95,96],compon:[47,48,62,92,96,97],compos:[9,19,31,47,59,62,74,83,85],composenotalign:9,composit:59,compress:[40,115],compris:33,comput:[4,5,14,18,19,22,31,32,36,47,48,50,53,56,57,58,62,66,70,73,76,77,78,80,86,89,92,93,95,97,101,102,121,135],computation:4,computationgraph:62,con:77,concat:[60,140],concaten:[4,5,18,60,74,96,107],concentr:85,concept:[3,31,45,46,47,57,59,60,62,69,74,75,78,79,81,89,96,97],conceptu:[47,53,57,60,62],concern:[31,47,52],concis:[60,96],conckerneltrac:22,concret:[85,93],concurr:[36,43,50,80],cond:[18,34,57,63,81],condit:[4,18,40,48,57,63,69,92,94,107,110,122],condtion:60,conf:[4,109,119,123,137],conf_paddle_gradient_num:[121,123],conf_paddle_n:[121,123],conf_paddle_port:[121,123],conf_paddle_ports_num:[121,123],conf_paddle_ports_num_spars:[121,123],confid:4,confidence_threshold:4,config:[2,4,22,38,55,75,100,110,113,121,122,123,128,129,131,145],config_:[39,129],config_arg:[128,129,131],config_bas:[3,4,25],config_lay:100,config_len:41,config_pars:[67,68,100],config_proto:41,configmap:48,configprotostr:109,configur:[4,18,25,33,39,41,43,44,48,50,57,62,65,83,92,93,95,100,112,145],confirm:55,conflict:[89,98],confus:[11,60,65],conll:10,connect:[5,18,44,45,48,50,92,105,121,122,145],connectionist:[4,18],consequ:[4,5],consid:[3,4,33,82,93,145],consider:[4,5,73,92],consist:[3,4,10,11,18,40,46,47,58,70,81,84,85,86,91,92,97],consol:121,consolid:34,constant:[4,18,62,64,65,73,109],constantiniti:[16,18],constraint:89,construct:[3,18,31,53,62,66,74,83,85,89,91],constructbackwardgraph:61,constructoptimizationgraph:61,constructor:[18,51,56,80,83,85,89,91],consum:36,consumpt:66,contact:43,contain:[3,4,5,7,9,10,11,18,25,26,31,33,34,40,53,60,62,69,70,73,79,80,83,85,86,89,90,91,92,96,97,121,122,123],containerport:121,content:[18,41,55,70,75,103],content_dir:103,content_len:41,context:[4,5,10,18,19,51,69,89,90,93,101,102,107,135,140],context_attr:5,context_len:[4,5],context_proj_layer_nam:5,context_proj_param_attr:5,context_project:5,context_start:[4,5],contin:121,continu:[3,33,36,70,92],contrast:[4,18],contrib:87,contribut:[87,92],contributor:85,control:[2,34,45,47,94,121,122,145],controlflowgraph:66,conv2d:60,conv:[5,18,60,69,95],conv_act:5,conv_batchnorm_drop_r:5,conv_bias_attr:5,conv_filter_s:5,conv_fwd:69,conv_layer_attr:5,conv_num_filt:5,conv_op:4,conv_pad:5,conv_param_attr:5,conv_pool_2:48,conv_strid:5,conv_with_batchnorm:5,conveni:[31,33,62,76,91,92],convent:[18,33,41],convers:[56,57],convert:[10,18,38,48,49,50,56,57,58,69,84,86,92],convlay:4,convlut:92,convlution2d:18,convolut:[4,5,9,18,19,60,73,83,93],convolution2d:18,convolution_algorithm_opt:69,convoper:4,convproject:4,convtranslay:4,convtransproject:4,cool:98,cooper:92,coordin:[36,41],copi:[18,25,31,40,43,55,61,74,75,77,78,96,107,110,121],copy_from:51,copyright:[107,110],copyvariablewithtensor:95,core:[2,18,33,62,65,72,78,79,96,135],coreo:[121,145],corespond:18,corner:85,corpu:[10,92],correct:[4,18,32,33,56,77,121],correctli:[3,9,33,56,60],corresond:56,correspend:18,correspoind:31,correspond:[4,17,18,19,31,33,34,35,51,56,62,63,69,73,74,75,83,85,86,87,91,93,94,109],corss_entropi:31,cortex:56,cos:[4,91],cosin:[4,18,91],cosineop:91,cosineopproto:91,cosineopprotomak:91,cost:[18,25,31,33,48,61,76,77,81,82,95,107,110,135],cost_id:4,cost_np:82,could:[4,9,25,31,32,40,45,47,48,49,50,56,57,58,74,76,78,79,81,83,84,86,94,121],count:[3,18,36,44,52,82,84,92,117,122,129,131,134],counter:[14,22,36,40,49,61],cours:[3,44,73],covari:4,cover:[57,92],cp27:30,cp27m:[30,88],cp27mu:[30,88],cpp:[32,39,59,67,68,71,72,79,85,97,100,109,123,134,137],cprofil:133,cprofilev:133,cpu:[2,4,18,22,26,32,44,56,65,73,78,79,80,85,87,88,93,95,101,102,122,131,134,135],cpu_avx_mkl:30,cpu_avx_openbla:30,cpu_kernel:65,cpu_noavx_openbla:30,cpu_ns_:80,cpu_per_pserv:48,cpu_per_train:48,cpudevicecontext:[73,93,101],cpuelapsedu:80,cpuengin:68,cpuinfo:28,cpuplac:[22,48,65,69,73,93,95,101,102,135],cpusparsematrix:72,crash:[36,134],creat:[2,9,10,18,23,25,31,32,34,36,41,45,46,49,51,52,53,54,55,56,57,59,60,61,69,73,74,76,77,78,79,83,86,87,92,98,100,103,107,109,110,115,122,123,124,145],create_backward_pass:76,create_bias_paramet:100,create_block:83,create_doc_str:91,create_input_paramet:100,create_local_scop:53,create_oper:59,create_optimization_pass:76,create_paramet:83,create_python_ops_creatation_funct:91,create_rnn:34,create_rnn_op:74,create_tmp_var:83,create_tmp_vari:83,create_var:83,create_whileloop:96,creategradientoper:86,creatememori:69,createop:91,createoper:34,createprimitivedesc:69,createstack:121,createvari:34,creation:[59,121],creationd:121,creator:[9,10,38,85,86],creator_:86,credenti:55,crf:[93,95],critic:60,crlf:98,crop:[11,93],crop_grad:93,crop_siz:11,crope:11,cropgradkernel:93,cropkernel:93,cross:[4,18,83,109,142,143,144],cross_entropi:[4,31,48,60,66,95],cross_entropy_with_selfnorm:4,crt:54,csc:100,csr:[100,115],csv:[22,109],ctc:[3,18],ctc_error_evalu:92,ctc_evalu:3,ctest:[26,27,98,101],ctor:83,ctrl:[26,119],ctx:[69,95,101,102],cubla:73,cublas_handle_:93,cublashandle_t:93,cuda7:[30,112],cuda8:[27,28,30],cuda:[22,26,35,58,73,80,85,93,101,129,134],cuda_context:58,cuda_dir:[128,129],cuda_fp16:56,cuda_so:[28,104],cuda_visible_devic:107,cudaconfigurecal:134,cudadevicecontext:[58,73,93,101],cudadevicegetattribut:134,cudaelapsedu:80,cudaevent_t:80,cudaeventcr:134,cudaeventcreatewithflag:134,cudafre:134,cudagetdevic:134,cudagetdevicecount:134,cudagetdeviceproperti:134,cudagetlasterror:134,cudahostalloc:134,cudalaunch:134,cudamalloc:134,cudamemcpi:134,cudaplac:[22,73,93,95],cudaprofilerstart:134,cudaprofilerstop:134,cudaprofilestop:134,cudaruntimegetvers:134,cudasetdevic:134,cudasetupargu:134,cudastream_t:93,cudastreamcr:134,cudastreamcreatewithflag:134,cudastreamsynchron:134,cudeviceget:134,cudevicegetattribut:134,cudevicegetcount:134,cudevicegetnam:134,cudevicetotalmem:134,cudnn:[4,7,18,35,65,69,73,93,95],cudnn_batch_norm:4,cudnn_conv:4,cudnn_conv_workspace_limit_in_mb:[128,129],cudnn_convt:4,cudnn_dir:[128,129],cudnn_kernel:65,cudnnavginclpadpool:4,cudnnavgpool:4,cudnnv5:27,cudrivergetvers:134,cuinit:134,cumtim:133,cumul:4,cur_mem:75,curl:121,curli:[34,61],current:[4,18,33,34,35,36,39,41,45,50,52,57,65,73,74,75,78,79,80,83,89,95,96,103,121],current_block:[81,83],current_oper:81,current_word:[107,140],curv:31,custom:[31,44,56,60,75,78,85,92,121],custom_batch_read:84,custom_neg_class:18,cut:[9,96],cut_lin:9,cutoff:10,cv2:11,cxx:[142,143],cxx_compil:[142,143,144],cxx_flag:[142,143],cxxabi_1:30,cycl:36,cyclic:4,cython:71,d3e0:121,d_b0:60,d_b1:60,d_b2:60,d_block:60,d_f:60,d_g:60,d_h0:60,d_h0_bn:60,d_h0_relu:60,d_h1:60,d_h1_bn:60,d_h1_relu:60,d_h2:60,d_loss:60,d_loss_fak:60,d_loss_real:60,d_optim:60,d_step:60,d_t:60,d_w0:60,d_w1:60,d_w2:60,dandroid_abi:142,dandroid_arm_mod:142,dandroid_arm_neon:142,dandroid_standalone_toolchain:142,dario:92,darwin:121,dash:60,dat:38,data:[3,10,11,17,25,31,32,34,38,39,40,47,50,52,54,56,57,60,61,62,64,65,66,69,70,73,74,75,76,77,78,79,81,83,85,87,89,90,91,92,93,96,97,102,107,110,112,115,118,122,123,125,128,135,140],data_batch:107,data_feed:12,data_fil:11,data_i:60,data_lay:[39,107],data_layout:18,data_layout_:95,data_read:[9,84],data_reader_creator_random_imag:84,data_shar:96,data_typ:[9,10,70,90,92,95,97,108,110,112,115,140],data_type_:[65,73,95],data_x:60,databas:10,datacent:[38,55],datacenter1:38,datacenter2:38,datacenter_1:38,datacenter_2:38,datacenter_nam:38,datadim:4,datafeed:135,dataflow:62,dataflow_analysi:66,datalayout:95,dataparallel:48,dataprovid:[107,109,123],dataprovider_convert:92,datasci:4,dataset:[18,38,44,48,78,84,92,110,112,117,133,140],dataset_nam:11,datatransform:95,datatyp:[10,18,65,69,90,92,95,97],dcgan:60,dcmake_build_typ:[103,113],dcmake_install_prefix:[113,142,143,144],dcmake_system_nam:[142,143,144],dcs:10,dcuda_arch_nam:27,dcudnn_root:27,ddim:[46,73,93,102],dead:36,deal:[33,145],deb:98,debug:[32,33,48,55,57,83,103,133],debug_str:62,decai:[6,23],decar:9,decayedadagradoptim:20,decayr:39,decent:40,decid:[31,43,60,70,78,86,87,90],declar:[18,34,60,74],decod:[4,5,18,92,139,140],decoder_boot:140,decoder_dim:75,decoder_group_nam:140,decoder_input:[75,107,140],decoder_mem:[75,140],decoder_prev:5,decoder_s:[107,140],decoder_st:[5,140],deconv:[4,60],deconvolut:4,decor:[9,46],decrement:49,decrementcount:49,decrypt:121,deduc:85,deep:[4,18,33,43,47,53,60,61,66,68,80,85,87,92,93,134],deeper:58,deepspeech2:67,def:[4,9,31,32,33,38,44,51,52,59,60,62,65,66,74,75,76,83,84,91,96,100,101,107,109,110,123,140],def_block:60,defalut:[18,22],default_block:60,default_decor:123,default_devic:131,default_initi:18,default_main_program:[17,135],default_param_attr:83,default_st:96,default_startup_program:135,default_valu:131,defaultdict:66,defaultinfervartyp:64,defect:79,defer:43,defin:[4,5,9,25,31,33,34,35,36,43,45,46,49,50,51,56,57,58,59,60,62,65,66,73,74,77,81,83,84,85,89,91,93,96,101,107,110,135],define_py_data_sources2:109,definit:[33,34,36,40,48,53,58,65,81,86,91,96,135],definiton:59,degener:18,degre:4,delai:[78,93],delet:[18,44,54,98],deletestack:121,delimit:[3,109],deliv:145,delta:[4,32],delv:4,demand:[36,93],demo:[4,10,85,122,125],denot:18,dens:[4,9,41,42,90,92,121],dense_arrai:[9,108],dense_vector:[9,110,112,115],dense_vector_sequ:[9,115],dense_vector_sub_sequ:115,densescann:92,dep:35,depart:92,depend:[18,34,35,36,44,46,47,48,50,62,77,82,90,145],dependent_var:82,deploi:[4,145],deploy:[62,70,85,121,145],deprec:[4,92],depth:[34,57,92],dequeu:50,deriv:[1,31,46,48,51,63,76],desc:[34,51,69,70,83,91,96],desc_:34,descend:[18,96],descent:[4,36,78],descproto:70,describ:[17,31,33,34,35,40,45,48,53,58,65,69,70,74,75,79,81,83,85,90,91,95,97,121,122],describestack:121,describestackev:121,describestackresourc:121,descripotor:69,descript:[3,10,34,35,64,68,70,73,86,90,92,95,97,121,123],descriptor:[47,69,95],deseri:[25,70,79],deserializ:85,desgin:61,design:[4,9,18,33,39,46,65,66,71,78,80,87,145],desir:[9,36,48,78,121,122],destin:[18,41,55],destroi:[34,53],destruct:89,destructor:80,det_output:3,detail:[2,3,4,5,6,10,18,32,33,40,44,48,50,55,57,60,62,66,69,70,73,74,80,83,87,89,93,95,96,97,121,145],detect:[18,64,98],detection_evalu:3,detection_output:3,determin:[4,9,18,34,48,66,73,85],dev:[26,28,85,98,104,133,142,145],dev_ctx:[34,69,80],devel:88,develop:[33,35,57,64,79,80,83,86,88,92,98,135,143],deviat:2,devic:[2,28,32,45,48,52,56,62,68,69,73,77,79,80,85,95,102,104,131,135],device_:80,device_context:[69,101],device_count:18,device_typ:18,devicecontext:[34,73,80,101],deviceid:[68,131],deviceid_:68,deviceplac:93,devid:4,devot:92,dhcp:145,diagon:18,diagram:74,diamond:60,dic:11,dict:[3,10,25,33,83,109,123],dict_dim:107,dict_fil:3,dict_siz:[10,18,39,75],dictionari:[3,4,10,18,25,31,32,83,107],did:79,diff:[98,107],diff_mat:32,differ:[3,4,18,22,25,32,33,34,35,36,41,43,46,48,49,50,51,52,53,56,57,60,62,63,66,69,73,75,77,78,80,82,86,89,92,94,95,96,97,121],differenti:59,difficult:[3,32,57],dig:121,digit:4,digraph:62,dilat:[4,18,69],dilation_h:18,dilation_i:4,dilation_w:18,dim0:101,dim1:101,dim:[4,9,18,19,21,39,69,70,74,85,90,93,97,100,101,102],dim_:[93,102],dimens:[1,4,5,7,9,18,19,21,60,85,90,92,93,96,102,107],dimension:[4,18,115],dimes:4,dios_arch:143,dios_enable_bitcod:143,dios_platform:143,dios_use_veclib_for_bla:143,dir:[104,123,142],direcit:92,direct:[4,5,11,18,57,66,78,92],directli:[5,35,42,44,46,48,56,65,79,91,95,96],directori:[4,17,35,38,43,54,55,93,98,103,122,134],dirnam:17,disabl:[18,80,109],disadvantag:[78,83],discard:[9,36,40,75,98,129],discexp:109,discount:4,discov:36,discoveri:121,discret:4,discrim:60,discuss:[31,34,40,41,42,48,69,92],disk:70,dispatch:[48,79],dispens:18,displac:18,displai:[44,55],dissimilar:18,dist:[27,88,104],dist_train:[31,44],distanc:[3,4,18],distinguish:35,distribut:[4,18,34,40,41,42,43,45,47,52,58,77,85,92,97,107,110,120,125,126,129,145],distribute_test:[128,129],distributedli:48,disucss:31,div:18,divid:[4,6,18,33,52,91,97],divis:18,divisor:18,diy_beam_search_prob_so:[128,129],dnn:[69,92,104],dns:121,do_forward_backward:84,doc:[9,62,74,96,101,103,117,119,123],doc_cn:103,docker:[27,28,88,98,103,104,121,122,123,125,126,127,142,145],docker_build:31,docker_clust:[119,124],docker_push:31,dockerfil:[26,98,123,142,144],document:[4,5,18,32,46,48,54,61,74,75,80,85,92],doe:[5,18,36,40,41,43,44,45,46,48,50,53,56,62,66,74,79,83,85,86,87,135],doesn:[2,4,9,31,34,45,47,84],doing:[39,43,48,61],domain:121,don:[5,31,35,59,61,66,84,92,121],done:[3,4,5,33,35,36,40,41,48,49,64,66,70,78,86,87,92,98,121,123,134],dot:[4,5,19,101],dot_period:[123,129,131],dotmuloper:4,dotmulproject:4,doubl:[18,48,56,61,80,95,101,129],down:[92,134],download:[10,35,36,39,43,54,104,122,145],doxygen:98,dozen:35,dpython_execut:104,dpython_include_dir:104,dpython_librari:104,draw:75,drive:89,drop:[4,5,18,19,46,75],drop_fc:108,drop_rat:[2,108],drope:18,dropout:[2,5,19,108],dropout_prob:18,dropout_r:[4,18,19,108],drpi_arm_neon:144,drpi_toolchain:144,drwxr:122,ds2:92,dst:[41,69],dst_primitive_desc:69,dtoh:134,dtype:[18,47,48,62,83,109,135],due:[40,43,60,66,75,83,133],dummi:[25,40],dump:[70,116],dump_config:116,dump_v2_config:116,duplic:[18,50],durat:40,dure:[4,5,17,18,33,34,36,40,43,44,52,57,66,77,78,80,83,85,92,97,121,145],duse_eigen_for_bla:142,dwith_c_api:[72,113,142,143,144],dwith_doc:103,dwith_golang:113,dwith_gpu:[27,103,113,144],dwith_mkl:[103,113],dwith_profil:134,dwith_python:[72,113,144],dwith_swig_pi:[72,113,142,143,144],dwith_test:[27,101,143],dwith_tim:134,dynam:[18,41,72,74,83,84],dynamic_cast:100,dynamic_recurrent_op:96,dyogatam:18,e2e:145,each:[3,4,5,7,9,10,18,19,25,32,33,35,36,39,40,41,43,44,45,46,47,48,51,52,53,58,61,64,66,69,73,74,75,77,79,80,82,83,84,85,86,89,90,91,92,93,94,95,96,97,121,145],each_feature_vector:1,each_time_step_output:1,eager:57,earli:[56,58],eas:[9,64],easi:[32,33,75,78,84,85,87],easier:[31,50,56,57,84,94,96],easili:[31,60,77,80,84,86,89,93],echo:[28,104],edg:[11,66],edit:[3,18,47,121],editdist:18,editor:83,edu:[10,18,121,122],eeoi3ezpr86c:121,effect:[4,18,25,121],effici:[4,48,70,84,92,93],effort:[48,92],efg:4,efs:121,efs_dns_nam:121,efsvol:121,egd:66,eigen:[56,73,78,85,87,93,101],eigen_device_:93,eigen_use_gpu:101,eigenmatrix:102,eigentensor:102,eigenvector:102,either:[4,5,9,18,25,31,48,60,63,64,74,78,87,107,110],elabor:92,elb:121,elbapis:121,electr:66,electron:122,elem_dim:4,elememt:4,element:[3,4,5,9,11,18,25,32,40,47,50,62,75,85,101],element_typ:41,elementari:85,elementwis:[18,19],elif:[31,91,94],elliott:10,els:[28,31,39,44,47,48,50,51,57,60,63,64,65,66,89,91,94,100],elsewher:80,emac:26,emailweixu:35,emb1:39,emb2:[39,137],emb:[107,109,122],emb_para:109,emb_param_fil:109,emb_sum:107,embed:[31,34,39,50,64,75,90,96,109,140],embedding_lay:[39,107,137],embedding_nam:[4,140],embedding_s:[4,140],empir:[4,18],emplace_back:100,emploi:[33,51,91],empti:[3,9,18,33,36,46,47,75],emul:56,enabl:[2,4,18,26,34,35,40,45,50,51,62,80,121,134],enable_grad_shar:[128,129],enable_parallel_vector:129,enableonstart:22,enc_proj:[5,140],enc_seq:5,enc_vec:140,encapsul:41,encod:[5,40,75],encoded_proj:[5,140],encoded_sequ:[5,140],encoded_vector:140,encoder_ctx:75,encoder_ctx_expand:75,encoder_ctx_proj:75,encoder_dim:75,encoder_last:4,encoder_out_seq:75,encoder_s:140,encount:[18,39],encourag:[48,53],encrypt:121,encrypt_decrypt:121,end2end:145,end:[3,4,10,18,22,25,33,34,48,51,58,62,66,75,79,80,84,89,92,94,98,140],end_id:18,end_pass:31,end_po:4,endforwardbackward:[25,107],endian:70,endif:[73,80],enditer:[25,31,107,110],endpass:[25,31,110],endpoint:[9,18,38,121],endtrain:31,engin:[44,68,69,92],english:[4,10,92],enough:[33,34,65,66,73],enqueu:50,ensembl:5,ensur:[36,69,77,89],enter:[34,53],enterpris:85,entir:[4,5,10,41,43],entiti:[3,34,89],entranc:53,entri:[9,18,40,44,64,121],entropi:[4,18,83],entry_point:44,enumer:[1,73,109],env:[103,107,121,123,133],environ:[31,46,48,104,121,122,134],environmenterror:117,eol:98,eos_id:[4,140],epoch:60,epol:47,epsilon:[4,6,18],equal:[4,5,18,19,36,96,101],equat:[3,4,5,6,18,66,101],equival:[31,34,45,47,51,57,63,91,145],eras:46,erlang:47,error:[2,3,4,5,18,31,32,40,55,56,57,69,89,92,104,109,113,121,129],error_clip:51,error_clip_callback:51,error_clipping_threshold:[2,107],errorclipbyvalu:51,espeaci:18,especi:[4,5,17,68],essenc:[31,33],essenti:[4,31,53,56],establish:45,estim:[4,18,31,50,78],eta:122,etal:[10,18],etc:[3,9,18,34,47,48,52,69,77,78,84,89,92,121,145],etcd:[9,25,36,40,41,43],etcd_addr:41,etcd_endpoint:9,eth0:[121,123],etyp:47,euclidean:4,eval:[3,34,52,60,85],eval_program:52,eval_result:52,evalu:[4,12,18,24,25,43,62,82,92,116,134,135],evaluate_difficult:3,even:[31,56,77,83,84],evenli:[41,121],event:[22,107,110,122],event_:80,event_block:80,event_handl:[25,31,107,110],eventkind:80,eventlist:80,eventu:[48,96],everi:[3,4,5,9,18,31,36,40,41,43,51,52,61,62,64,66,69,73,74,77,83,89,91,95,110,135],everyth:[48,50,60],evid:79,evolv:57,exactli:[4,5,46,107,121],exampl:[3,4,5,9,10,11,18,19,25,34,44,46,48,50,52,55,57,58,59,60,61,62,64,66,69,73,74,75,79,80,81,83,84,85,86,87,90,93,94,95,96,121,135],example_read:9,exc_path:104,exceed:4,except:[4,10,21,23,43,45,57,61,80,92,96,110],excess:66,exchang:79,exclud:4,exclude_mod:4,exclude_param:25,excluded_chunk_typ:[3,14,18],exclus:18,exconv:4,exconvt:4,exdb:10,exe:[48,135],execut:[4,18,22,35,36,40,44,45,46,47,48,52,53,58,60,62,66,69,77,80,86,97,121],executioncontext:[69,95,101,102],executor:[12,17,45,48,52,56,57,58,60,76,81,83,95,97,133,135],exist:[18,31,34,36,46,55,57,75,83,84,86,91,93,96,102,121],exit:[41,55,122],exp:109,expand:[18,75],expand_a:[4,136],expand_level:[4,136],expandconvlay:4,expandlevel:136,expans:4,expect:[4,95],expected_desc:69,expected_kernel_kei:95,experi:[70,92],expert:35,expir:36,explain:[3,18,36,45,57,59,61],explan:[4,18,44,45,48,89,95],explicit:[46,80,96,100],explicitli:[31,48,53],explod:51,explor:[4,75,87],expon:4,exponenti:[1,18],expos:[33,42,47,69,70,93,96,121],express:[31,50,52,62,66,107,110,121],extend:[3,78,96],extens:[43,50,75],extent:72,extern:[35,68,71,72,85,92],extern_mklml:104,external_librari:35,extingrad_:68,extinval_:68,extoutgrad_:68,extoutval_:68,extra:[2,4,5,21,48,87,93,145],extra_lay:25,extraattr:[2,131],extraattribut:4,extraattributenon:4,extract:[3,4,18,57,79,92,121],extralayerattribut:[2,5,107,108],extralayeroutput:5,extrem:[4,45,57],f1205:109,f120da72:122,f7e3:121,fa0wx:122,fabric:[119,120],face:[35,87],fact:[18,45,57,77,81,83],factor:[2,4,6,18],factor_s:4,factori:71,fail:[36,40,75,104,109,122,129],failur:[36,41],fake:60,fake_imag:84,faked_imag:60,fall:[56,82],falloc:54,fals:[2,3,4,5,6,9,10,18,32,33,34,57,63,65,67,74,81,82,84,90,97,100,101,107,110,112,115,117,122,131,140],false_block:[34,63,81],false_label:84,false_neg:52,false_posit:52,false_read:84,familiar:47,fan_in:18,faq:141,far:[51,96],fashion:48,fast:[4,40,57,134],faster:[4,5,18,36,57],fastest:57,father:33,fault:[25,40,85],fbd1f2bb71f4:122,fc1:[62,100,131],fc1_bia:62,fc1_weight:62,fc2:[62,131],fc3:[62,131],fc4:131,fc8a365:121,fc8a:121,fc_act:5,fc_attr:5,fc_bias_attr:5,fc_layer:[83,91,107,109,131],fc_layer_nam:5,fc_mat:25,fc_op:91,fc_out:[18,34],fc_output:91,fc_param_attr:5,fc_without_b:34,fclayer:100,fcop:59,feasibl:78,featur:[1,4,9,10,18,33,48,56,62,77,80,92,98],feed:[5,17,25,31,48,61,74,87,110,135],feed_dict:60,feed_list:[13,135],feed_target_nam:17,feeded_var_nam:17,feeder:[9,48,135],fetch:[10,36,39,46,48,82,98,104,135],fetch_list:[48,83,135],fetch_op:82,fetch_target:17,few:[35,36,47,48,66,78,84,90,92],fewer:[4,18,47,83],fft:92,fg0:4,field1:25,field2:25,field:[4,10,21,25,34,62,64,70,82,83,86,90,91,107,121],fifth:61,figur:[31,35,48,50,60,68,74,80,83,92],file:[3,4,9,10,11,17,22,25,31,33,35,36,38,40,41,43,44,46,47,54,55,57,58,62,70,72,84,85,92,93,97,98,101,107,110,135,145],file_nam:109,file_typ:9,filelist:92,filenam:[11,38,83,107,133],fileoffset:54,filesystem:[43,44,48,54,121],fill:[4,36,40,73,83,115,121],fill_zero_grad:85,fill_zeros_like_op:33,filter:[4,5,17,18,51,69],filter_s:[4,5,18,19],filter_size_h:18,filter_size_i:4,filter_size_w:18,filter_strid:18,find:[4,18,34,36,43,47,56,62,69,75,89,95],find_var:32,findmemori:69,findop:34,findprimit:69,findprimitivedesc:69,findvar:[34,89],fine:[2,40,59],fingerprint:121,finish:[18,36,40,43,44,53,66,77,91,121,122],first:[4,18,22,25,31,33,34,36,40,43,44,45,48,53,55,57,60,61,62,69,74,75,81,82,83,85,90,91,92,93,94,96,101,102,115,121,145],first_n:18,first_seq:140,firstli:[3,4],firstn:9,firstseen:122,fit:[10,17,56,65,66,70,75,85],five:81,fix:[2,4,18,48,66,71,83,92,98,115],flag:[4,10,18,22,67,68,80],flatten0:62,flatten:[18,62,81,83,102],flatten_result:107,flexibl:[4,5,22,31,41,48,57,61,65,74,75,78,84,93,96],flip:11,flist:117,fliud:45,float16:47,float16_t:56,float32:[9,18,48,56,59,60,83,84,101,109,135],float64:18,float_16:18,float_to_half_rn:56,floor:[4,109],flow:[18,34,45,47,74,80,88,94],fluid:[0,13,14,15,16,17,18,19,20,21,22,23,33,46,48,50,53,73,80,83,93,94,95,133],fluid_cuda_create_tensor:58,fluid_cuda_mult:58,fluid_cuda_read:58,fly:33,fmt:[47,109],fname:109,fnt03:121,focu:[47,62],folder:[35,38,44,55,121],follew:18,follow:[3,4,5,6,9,10,11,18,19,25,31,32,33,34,35,36,40,44,45,46,47,48,50,53,56,57,58,59,60,61,62,63,64,66,69,73,74,75,77,78,80,81,82,83,84,85,86,87,89,90,91,92,93,94,95,96,121,125,126,135,145],footprint:58,forbid:31,forc:[18,77,83,95],force_cpu:[18,65],force_cudnn:65,force_load:71,forest:34,forev:47,forget:[6,18,31],forget_bia:18,fork:4,form:[4,5,18,46,47,52],formal:95,format:[3,9,10,11,18,22,25,32,40,46,48,56,57,73,75,92,96,98,100,112,115,121],former:[31,35,57,66,78],formul:18,formula:[4,5,6,18,32,66],formular:4,forth:60,forward:[1,4,5,18,32,33,34,39,41,51,57,60,67,68,69,70,76,79,81,84,85,86,87,90,100],forward_infer:69,forward_list:80,forward_op:32,forward_proj:18,forward_train:69,forwardactiv:100,forwardbackward:25,found:[56,81,87,89],four:[3,18,47,52,57,61,69,73],fourth:18,foward:82,fp16:[56,85,97],fp32:[73,85,95,97],fp64:[73,97],fparam:109,fpga:[73,135],fpgadevicecontext:93,fpgaengin:68,fpgaplac:[73,93],frac:18,frame:[3,53,85,92,96],framework:[31,33,34,47,51,52,56,57,62,73,77,78,80,81,85,87,89,91,93,101,133,135],frank:10,free:[10,58,93,145],freememoryop:58,frequenc:[10,92],frequent:[40,84,85,87,93],fresh:43,friend:89,friendli:60,from:[3,4,5,9,10,11,14,17,18,22,25,32,33,34,35,36,38,39,40,41,45,46,47,48,50,51,52,55,56,57,59,60,61,62,63,65,66,69,74,75,76,77,79,81,83,84,85,86,89,92,93,94,95,96,101,102,104,116,121,122,133,134,139,145],from_no_sequ:[4,136],from_sequ:[4,136],from_tar:[25,110],fromfil:[84,109],fromstr:109,front:[62,66],fuction:22,fulfil:134,full:[4,18,36,43,47,74,77,78,145],full_matrix_project:[5,140],fulli:[18,48,50,92,145],fullmatrixproject:4,fullsiz:39,fully_matrix_project:5,fullyconnect:[62,83],fullyconnectedlay:100,func:[9,40,45,58,86],funciton:[5,18],functor:[59,62],fundament:[46,47,50,56,85],funtion:18,further:[4,46,91,145],furthermor:18,futur:[4,18,43,48,56,66,74,85],future_context_s:18,fvs:91,fwd_desc:69,fwd_op:86,fwd_primit:69,fwd_primitive_desc:69,fwd_var:51,g_b0:60,g_b1:60,g_b2:60,g_block:60,g_command_config_arg:[67,68],g_h0:60,g_h0_bn:60,g_h0_relu:60,g_h1:60,g_h1_bn:60,g_h1_relu:60,g_h2:60,g_im:60,g_loss:60,g_optim:60,g_program:83,g_state:80,g_step:60,g_w0:60,g_w1:60,g_w2:60,gain:[4,18],gan:31,gangliao:35,gate:[4,5,18,19],gate_act:[4,5],gate_activ:18,gate_attr:4,gate_bias_attr:4,gate_param_attr:4,gate_recurr:4,gate_v:18,gatedrecurrentlay:67,gather:[4,18,33,66,77,79,101],gauss:2,gaussian_normal_random:60,gcc:[26,27,56,58,71,85,142,144],gcc_3:30,gcreators_:91,gemm:67,gen:4,gen_proto_pi:103,gen_rand_param:109,gender:[10,123],gendrated_id:75,gener:[3,4,5,9,25,31,32,33,34,35,36,38,40,41,43,45,48,57,59,64,66,69,73,77,78,81,82,83,84,85,86,90,91,92,93,94,96,110,115,121,123,131,134],generated_id:75,generated_scor:75,generated_word_embed:4,generatedinput:[4,139,140],genr:123,gereat:3,german:10,germani:10,get:[3,4,10,17,18,25,28,32,33,34,35,36,40,41,43,44,46,54,57,60,62,65,66,67,68,69,73,74,75,80,83,85,86,89,91,95,96,100,107,115,119,121,122,124,133],get_all_op_proto:91,get_block:83,get_config_arg:131,get_data:122,get_dict:10,get_dim:32,get_embed:10,get_float_el:32,get_grad:[25,107],get_grad_op_desc:33,get_input_lay:100,get_lin:9,get_movie_title_dict:10,get_numeric_gradi:32,get_numerical_gradi:32,get_output:32,get_program:66,get_sample_from_lin:107,get_shap:25,get_support:[30,104],get_symbol:62,get_tensor:32,get_var:18,get_vari:34,get_word_dict:10,get_worker_addr:45,getactualkerneltyp:65,getattr:51,getbatchs:100,geteigendevic:102,getengin:69,getenv:[31,44,117,123],getexpectedkerneltyp:[65,69,95],gethostbynam:123,gethostnam:123,getidmap:123,getinfervartyp:64,getinput:100,getinputgrad:100,getinputvalu:100,getkerneltyp:56,getkerneltypeforvar:95,getlayeroutput:[25,107],getlibrari:69,getmat:39,getoptconfig:39,getoutputgrad:100,getoutputvalu:100,getparam:39,getparameterconfig:39,getparameterptr:100,getparameterspars:39,getparametersremot:39,getplac:[69,93,101,102],getpodlist:123,getsiz:100,gettask:40,gettempl:121,gettensor:95,gettranspos:100,getw:100,getweight:100,getwgrad:100,gflag:113,gflags_complet:113,gflags_declar:113,gist:5,git:[26,27,88,98,103,104,113,142,144],github:[5,10,18,26,27,35,60,73,88,98,103,104,113,133,135,142,144],give:[18,36,74,83,85,95,121],given:[4,9,18,19,25,33,41,43,47,50,51,57,59,60,75,84,87,96],glibc:[142,144],glibc_2:30,glibcxx_3:30,glide:35,global:[2,18,23,26,31,34,35,36,58,62,65,79,80,85,89,91,93,95,121,134],global_block:83,global_learning_r:2,global_pool:18,global_step:18,globalstat:134,globalstatinfo:134,glog:113,gnueabihf:144,go_librari:35,go_test:35,goal:[46,47,50,56,61,77,85,92],gob:40,godep:35,godoc:71,goe:[5,36,57,63,89,135],going:[33,59,78,145],golang:35,good:[47,60,78,83,84,87,145],googl:[18,31,80,85,109,113,133,142],googleapi:121,googlenet:68,goroutin:[45,47],got:[65,89],govern:[107,110],gpg2:121,gpg:121,gprof:133,gprotos_:91,gpu:[2,4,7,22,26,28,30,32,44,47,52,56,66,73,77,78,79,80,85,87,88,93,95,104,112,115,117,131,134,135,145],gpu_id:[107,129,131],gpu_per_train:48,gpudevic:93,gpugpu_id:128,gpukernel:85,gpustarttimestamp:22,grab:36,grad:[32,33,41,51,68,83,90,107,129],grad_info_map:33,grad_n:51,grad_nam:51,grad_op:51,grad_op_class:85,grad_op_desc:51,grad_op_maker_:86,grad_op_typ:[85,86],grad_op_type_:86,grad_s_block:33,grad_share_block_num:[128,129],grad_to_var:[33,51],grad_var_nam:32,gradient:[2,3,4,6,18,23,25,36,40,47,49,51,61,64,76,77,78,79,83,85,90,116,117,118,123,129],gradient_clip:21,gradient_clipping_threshold:[2,107],gradient_evalu:3,gradient_flat:32,gradient_machin:[25,72,113],gradientmachin:[25,72,79,123],gradientmachine_:39,gradopdescmak:[64,86],gradopdescmakerbas:86,gradopmak:86,grai:11,grain:59,gram:92,grandient:25,grant:121,graph:[4,18,25,33,34,35,36,45,47,48,49,50,52,57,60,74,77,78,81],great:[50,92,145],greater:[4,18,51,78,115],greaterthan:91,greedi:[18,92],green:[45,60],grep:[28,124],gridsize3d:22,groudtruth:140,ground:[3,4,18],group:[5,18,40,62,69,93,145],group_input1:140,group_input2:140,group_input:140,grouplen:10,grpc:145,gru:[4,18,75,92],gru_bias_attr:5,gru_decod:140,gru_decoder_with_attent:140,gru_layer_attr:5,gru_memori:5,gru_out:75,gru_param_attr:5,gru_step:[75,140],gru_step_lay:5,grumemori:[5,108,140],gserver:[4,67,68,100],gsizex:134,gtx:66,guarante:[69,83],guard:39,guest:30,guid:[22,54,66,85,121,122],gutmann10a:18,gzip:[40,122],h0_bn:60,h_0:18,h_f:18,h_prev:34,hadoop:31,half:[4,18,19,56,121],half_to_float:56,hand:[66,85,92,93],handi:35,handl:[9,31,33,44,45,48,62,66,69,73,79,84,89,93,96,135],handler:[25,34],hannun:92,happen:[18,40,91],hard:[48,57,75,92,96,121],hardshrink:18,hardsigmoid:18,hardwar:[57,58,93],has:[3,4,5,10,18,19,22,31,32,33,34,35,36,40,41,43,46,47,48,50,51,52,56,57,60,62,66,70,73,75,77,80,81,85,90,91,93,95,115,121,134,135,145],has_kei:[25,33,51],has_selected_colum:4,has_var_recurs:33,hasdependentvar:82,hash:[73,77],hasn:57,hasnext:46,have:[4,5,9,18,19,22,31,32,33,34,35,36,40,41,43,44,46,47,48,50,51,53,56,57,58,59,60,61,65,66,69,70,73,74,75,77,78,79,80,81,83,84,85,86,89,90,92,93,95,97,121,145],haven:57,hdf:[9,38],head:[19,101,124],header:[41,70,72,85,93,109],headip:124,height:[4,9,11,18,34,71,84,100,101,109,115],height_:90,held:36,hello:31,help:[4,18,34,55,57,62,69,75,84,85,96,98],helper:[18,48,69,86,96],henc:[48,78,83,86,87,89],here:[2,3,4,5,9,18,19,21,31,35,36,42,47,50,51,53,55,57,61,62,69,73,74,84,87,91,115,121,145],heterogen:[48,50,80],heurist:[4,50,75],hidden:[4,5,18,76,83,108,109,121],hidden_a:109,hidden_b:109,hidden_dim:18,hidden_out:34,hidden_s:5,hidden_t_prev:18,hidden_v:18,hidden_valu:18,hierach:139,hierarch:[4,81,83,85],hierarchi:85,high:[2,56,77,92,93,145],higher:[59,74,96],highest:[9,34],highli:[10,92,96],him:31,hint:65,histor:59,histori:6,hl_get_sync_flag:100,hold:[31,33,36,40,42,46,47,56,60,62,64,66,89,91,93,95,102,121],holder_:[93,102],home:[28,38,55,121,122,123,124,133],honor:40,hook:2,hookattr:2,hookattribut:2,horizont:[4,11],host:[35,44,80,121,122],host_c:[142,143,144],hostfil:124,hostnam:121,hostnetwork:123,hostpath:[122,123],hostport:121,hot:18,hous:[10,112],how:[2,4,18,31,34,36,40,45,46,47,48,53,55,57,59,62,65,69,74,75,79,80,87,91,95,121],howev:[4,5,18,32,33,43,47,48,53,57,66,73,78,79,83,84,86,87,90,91,92,93,95,121],howto:[117,119,123],hpp:[56,71],html:10,htod:134,http:[4,5,10,18,26,27,28,35,44,60,88,98,103,104,107,110,113,121,122,133,135,142,144,145],hub:88,huber:4,huge:78,human:[4,80,92],hwc:11,hyp:18,hyper:[4,60],hyperparamet:[4,87],hyperplan:9,hypothesi:18,i1116:123,i1117:134,i386:143,iOS:143,iamfullaccess:121,iamusersshkei:121,icc:58,iclrworkshop2016:18,icml:92,ics:10,id_input:3,id_rsa:124,idea:[35,47,57,58,78,84,87],ideal:[48,95],ident:[4,18,86,121],identifi:[4,63,73],identityoffsetproject:4,identityproject:4,idmap:123,ids:[3,4,18,75,107,115],ids_arrai:115,idx:[40,46,60,66,100],ies:55,if_els:94,if_else_op:33,ifdef:[73,80],ifels:[34,81],ifelseop:81,ignor:[4,17,18],ignored_token:18,iii:92,iil:109,illustr:[3,18,36,41,47,48,59,74],im2col:18,im_siz:60,imag:[7,8,9,10,18,26,31,48,57,60,61,75,76,81,84,92,98,121,122,123,125,126,145],image_a:84,image_b:84,image_conv_lay:92,image_fil:84,image_h:18,image_lay:84,image_nam:31,image_path:84,image_reader_cr:84,image_w:18,imagenet:[4,38],imagepullpolici:[121,123],imageri:4,images_reader_cr:84,imagin:61,img2label:11,img:[4,5],img_conv_lay:5,img_pool_lay:5,imgsiz:134,imgsizei:134,imgsizex:134,immedi:[66,69,78,87,121],immutable_paramet:31,imper:45,imperfect:85,implement:[4,5,9,18,19,23,34,40,41,42,43,44,45,47,48,50,57,59,62,63,64,66,69,71,72,73,75,79,82,89,91,92,93,94,95,96],implemet:39,impli:[35,107,110],implicitli:45,imposs:[46,75,145],impractic:95,improv:[4,49,50,66,85,92,121],in_arg:115,in_fals:18,in_plac:18,in_tru:18,inarg:39,inbound:121,inc_path:104,includ:[3,4,5,10,11,18,31,34,35,41,44,46,47,56,57,60,62,66,71,72,74,75,80,81,83,85,91,98,101,113,121,134,142,143,144],inclus:[18,75],incom:[45,65],incorpor:4,incorrect:4,increas:[36,40,56,109],increment:[47,52,61,66],incupd:100,inde:[9,47],independ:[4,18,32,33,41,49,89,93,145],index:[3,4,7,9,10,18,25,32,33,34,36,40,45,81,83,96,121],indexslot:4,indiact:18,indic:[3,4,10,18,21,33,34,41,53,60,74,81,86,90,93,96,115,121],indice_map:96,indices_map:96,individu:[36,77,121],industri:[36,70,145],ineffici:[79,95],infer:[11,17,18,31,33,34,36,52,57,63,64,65,66,67,71,73,82,83,85,90,92,95,110,112],infer_shap:83,infer_var_type_:64,inferer:92,inferfac:64,inferior:43,infershap:[34,83,85,101,102],infershapecontext:[101,102],infervartypefn:64,info:[3,4,10,56,74,100,107,110,119,123,145],infom:4,inform:[4,10,18,25,34,44,55,62,65,66,69,70,73,74,77,83,87,89,90,121],infrastructur:[57,121],ingrad_:68,ingredi:[47,92],inherit:[22,34,46,76,85,93],ininst:31,init:[2,18,25,34,49,60,68,74,75,100,110,112,117,121,123],init_attr:83,init_batch_dim_idx:18,init_from_tar:25,init_model_path:[128,129,131],init_valu:18,initi:[2,4,5,10,12,18,21,25,33,35,40,45,48,49,50,52,61,74,77,78,83,87,91,96,112,129,135],initial_max:[2,109],initial_mean:[2,4,109],initial_min:[2,109],initial_std:[2,4,109],initialize_op_attr:83,initrd:145,inlcud:5,inlin:[93,102,121],inner:[4,107,115],inner_param_attr:5,inner_pos_arrai:115,inner_seq_pos_arrai:115,inproj_attr:4,inproj_bias_attr:4,inproj_param_attr:4,input0:102,input1:[4,5,102],input2:4,input:[1,3,4,5,7,9,11,14,17,18,19,25,32,33,34,39,43,45,46,48,49,50,51,52,56,57,58,59,60,61,62,64,65,66,68,69,73,74,75,78,79,82,83,84,85,86,89,91,92,93,95,96,100,101,102,107,108,109,110,112,115,123,131,135,136,139,140],input_conf:4,input_data:100,input_data_target:100,input_dim_idx:18,input_dtyp:18,input_featur:1,input_hassub_sequence_data:100,input_id:4,input_imag:5,input_index:100,input_label:100,input_lay:100,input_loc:4,input_nam:31,input_proj_bias_attr:5,input_proj_layer_attr:5,input_seg:96,input_seq:[4,18],input_sequence_data:100,input_sequence_label:100,input_sparse_float_value_data:100,input_sparse_non_value_data:100,input_t:100,input_to_check:32,input_typ:107,input_valu:32,input_var:[32,83],inputbuff:39,inputdef:100,inputgradi:86,inputlayers_:100,inputs_to_check:32,inputsizechang:69,inputtyp:9,insert:[18,33,51,58,77,82,85,86,98],insid:[3,5,33,36,48,50,51,52,65,69,79,80,84,85,86,121],inspir:80,instal:[4,18,27,28,30,44,68,88,98,103,104,112,113,122,127,133,142,143,144],install_android:142,instanc:[4,18,32,34,36,38,42,45,46,48,49,51,53,58,63,69,74,75,78,83,85,86],instance_ip:121,instanti:[36,53,135],instead:[4,5,7,32,33,35,39,44,45,47,48,56,57,61,62,92],instrins:56,instruct:[34,61],int16:97,int32:[18,73,81,96,97,129],int64:[18,48,54,73,90,95,97],int64_t:80,int8:73,integ:[3,4,9,18,40,44,45,47,56,71,75,110],integer_sequ:107,integer_valu:[9,107,110,115],integer_value_sequ:[9,75,92,115,140],integer_value_sub_sequ:115,integr:[3,18,145],intel:[57,73,93],intellig:66,inteloptimizedpaddl:68,intens:92,inter:[4,18,48],interact:[4,48,121],intercept:4,interchang:[61,85],interconnect:77,interest:[45,56,77],interfac:[2,4,5,22,25,34,40,44,46,55,62,77,79,85,86,92,93,121,145],intergr:4,intermedi:[18,48,55,58,60,66,76,92],intern:[4,5,25,56,92,121,133],internel:68,internet:[35,36,145],interpret:[3,18,53,57,58,97],interv:18,intrins:[45,53,56],introduc:[4,11,34,36,46,60,67,70,87,89,91],intuit:[43,85],inval_:68,invalid:[84,89],invent:57,invoc:[35,59,85],invok:[4,25,33,46,48,51,79,83,85,86,91,95,121,134],involv:75,iob:3,ioe:3,ios:143,ios_arch:143,ios_deployment_target:143,ios_development_root:143,ios_enable_bitcod:143,ios_platform:143,ios_sdk_root:143,ios_use_veclib_for_bla:143,ip_str:123,ips:[121,123],ipt:[4,83,91,109,140],ipx:145,ipython:31,is_bia:18,is_color:11,is_cpu_plac:69,is_inf:116,is_loc:25,is_mkldnn_librari:69,is_revers:18,is_scalar_condit:18,is_seq:[4,140],is_spars:18,is_stat:[2,109],is_target:82,is_tensor:91,is_test:[18,69],is_traget:82,is_train:11,isbinari:115,isinst:[51,107,110],ismkldnnkernel:69,ispodallrun:123,isspars:100,issu:[18,35,60,92],issue_numb:98,issuecom:18,istag:88,item:[4,9,18,25,43,56,84,112,123,145],iter:[4,5,6,9,25,31,36,48,57,58,66,69,78,80,84,92,96],iter_multiple_input_and_param:83,its:[4,5,18,25,31,33,34,36,40,45,46,47,50,51,52,57,58,60,61,62,64,66,70,74,75,77,78,79,82,83,85,86,89,90,91,93,95,121,134],itself:[33,36,43,46,58,69,78,89],ivector:[115,116],ivs:91,java:[34,71,81,85],jeremi:134,jku:18,jmlr:18,job:[10,33,43,45,48,51,85,123,128,129,131],job_desc:48,job_dispatch_packag:119,job_id:10,job_nam:[44,121,123],job_namespac:[121,123],job_path:[121,123],job_path_output:123,job_workspac:119,jobdesc:48,jobnam:[48,123],jobpath:[121,123],jobport0:121,jobport1:121,jobport2:121,jobport3:121,jobselector:123,jobserv:44,join:36,jointli:5,jpg:[11,46],json:[62,92,121,122],jth:5,judg:4,juditski:78,jupyt:44,just:[1,3,4,5,10,18,35,40,41,45,48,57,58,60,64,69,78,79,83,84,85,86,87,89,90,121],jx4xr:121,jypyt:31,k8s:[45,123,145],k8s_data:[121,123],k8s_job:31,k8s_token:31,k8s_train:[121,123],k8s_user:31,kafka:38,kcpu:80,kcuda:80,kdisabl:80,kebilinearinterpbw:134,kebilinearinterpfw:134,keep:[4,9,11,18,36,47,57,58,61,75,78,83,89,91,145],keep_dim:18,keep_top_k:4,kei:[10,11,19,22,25,26,32,33,34,36,38,40,54,56,65,69,85,86,91,92,96,98,107,123,134],kenlm:92,kept:[4,66,83],kera:87,kernel:[4,18,22,32,47,56,58,65,68,78,80,87,90,92,93,95,101,102],kernel_hint:65,kernel_type_for_var:95,kerneltyp:[65,69],key1:129,key2:129,key_pair_nam:121,keyid:121,keymetadata:121,keypair:121,keyserv:121,keystat:121,keyusag:121,keyword:[18,83,94,123],kforcecpu:65,kill:[36,121],kind:[31,32,36,42,46,48,51,58,61,65,69,76,77,80,93,95,97,107,110,121,122,123],kind_:80,kitten:18,kmark:80,kms:121,knchw8c:73,knchw:73,knhwc:73,know:[18,31,40,45,46,66,70,121],knowledg:92,known:[33,34,47,57,59,74],kpoprang:80,kpushrang:80,kqueue:47,kriz:10,krizhevski:18,kselectedrow:90,ksimonyan:5,kstate:80,kube_cluster_tl:31,kube_ctrl_start_job:31,kube_get_workers_addr:45,kube_list_containers_in_job_and_return_current_containers_rank:31,kubeconfig:121,kubectl:[119,122,123,124],kuberent:[36,121],kubernet:[31,36,45,48,85,120,123,125,126,145],kubernetes_service_host:31,kusecudnn:65,kusemkldnn:65,kvp:22,kwarg:[5,6,9,14,18,21,52,62,83,91],kwd:[15,22],l1_rate:2,l1_regularization_op:87,l1decayregular:23,l2_rate:2,l2_regularization_op:87,l2_sim:4,l2decayregular:23,l2regular:107,l93:39,label:[3,4,9,10,11,14,18,25,48,52,57,60,61,62,76,81,84,92,95,107,110,122,135],label_dim:4,label_fil:84,label_lay:84,label_path:84,labelselector:123,lag:129,lambda:[18,45,51],lambdacost:4,lambdarank:4,lan:127,lang:10,languag:[4,10,19,45,47,57,61,66,80,85,89,92,94,107,110],larg:[7,10,18,48,50,51,66,70,78,92,98],larger:[2,3,4,66,94],larger_than:[34,63,81],largest:18,last:[3,4,5,18,33,51,66,74,80,81,136],last_seq:[75,137],last_time_step_output:4,lastseen:122,latenc:[4,56,92,121],latent:4,later:[35,85,87,92,93,102,121],latest:[4,28,34,35,36,43,88,98,103,104,122,123,142],latter:[78,96],launch:[69,121],launcher:31,law:[107,110],layer1:[4,5,107,136],layer2:[4,107,136],layer3:4,layer:[2,3,5,7,9,12,24,25,33,34,39,45,47,48,50,57,60,61,63,76,78,81,84,85,87,91,92,93,96,100,107,109,110,112,115,116,135,136,139,140],layer_0:100,layer_att:108,layer_attr:[4,107,108,131,140],layer_help:65,layer_nam:107,layer_num:131,layer_s:115,layer_typ:[4,67,68],layerbas:100,layerconfig:100,layergradutil:100,layerhelp:[18,65,83],layermap:100,layeroutout:4,layeroutput:5,layers_test:104,layout:[11,69,95],layout_:[65,73],layouttyp:65,lazi:[78,87],lbl:3,ld_library_path:113,lead:[46,66,73],leaki:60,leakyrelu:18,learing_r:76,learn:[2,3,4,5,6,10,18,19,28,31,33,41,43,47,48,50,53,60,61,66,68,75,77,78,80,84,85,87,93,134],learnabl:[4,18,19,25],learning_r:[2,21,41,48,107,109,135],learning_rate_arg:109,learning_rate_decay_a:109,learning_rate_decay_b:109,learning_rate_schedul:109,leas:36,least:[3,18,36],leav:[34,121],lectur:66,lecun:10,left:[4,18,34],left_right_flip:11,left_scor:107,legal:91,len:[4,18,41,45,54,57,83,100,112,123],length:[4,5,9,10,11,18,41,56,67,70,74,75,85,92,96,122],leran:66,less:[4,18,31,51,145],less_equ:94,less_than:[31,66],lesser:18,let02:122,let:[3,4,18,31,34,43,45,47,58,59,61,65,69,73,74,75,76,86,93,95,121],level:[2,4,18,56,59,62,70,74,75,80,93,96,97,115,139],levenshtein:18,lgtest:35,lgtest_main:35,lib64:[28,104,129],lib:[27,72,113,133,142,143,144],lib_path:104,libapi:35,libari:72,libc:30,libcuda:[28,104],libgcc_:30,libgflag:113,libglog:113,libgoogl:133,libiomp5:68,libmkldnn:68,libmklml_intel:68,libnvidia:[28,104],libopenbla:113,libpaddl:[71,72,85,98,133],libpaddle_capi:72,libpaddle_capi_engin:113,libpaddle_capi_lay:113,libpaddle_capi_shar:113,libpaddle_capi_whol:113,libpaddle_gserv:72,libpaddle_math:72,libprotobuf:[109,113],librari:[4,18,35,42,47,48,68,69,72,77,92,95,129],library_:73,library_type_:95,librarydevicecontext:73,librarytyp:95,libstdc:30,libz:113,licens:[68,77,107,110],life:36,lifecycl:[80,145],lifetim:89,lightweight:59,like:[3,4,9,10,18,33,34,35,36,39,44,45,47,53,57,58,59,60,61,62,64,69,73,77,78,83,84,85,86,87,89,90,92,94,95,96,121,135,145],limit:[4,9,18,57,66,70,75,85,87,107,109,110,134],linaro:144,line:[3,9,22,35,39,44,47,55,61,78,81,83,85,87,98,107,109,121,131],line_break:9,line_count:109,linear:[4,18,19,75,107,109,110,112],linearli:18,lineno:133,link1:56,link2:56,link:[4,5,18,35,54,55,89,121,139,145],linux:[9,26,30,47,54,121,142,144],linux_x86_64:[30,88,104],lipo:143,list:[3,4,5,9,11,18,22,23,25,31,33,34,35,40,44,45,53,55,57,60,73,76,79,80,83,86,89,96,107,121,131,133],listdir:117,listen:[36,45,48],listen_and_do:45,listen_and_serv:18,listenanddo:45,lite:113,littl:[41,65,70],live:135,live_in:66,live_out:66,load:[11,17,31,36,48,60,77,83,110,121,123],load_and_transform:11,load_file_nam:17,load_imag:11,load_image_byt:11,load_missing_parameter_strategi:[128,129,131],load_mnist:60,load_paramet:109,loadsave_parameters_in_pserv:[39,128,129],loc:3,local:[2,18,25,27,32,34,36,42,43,47,61,66,74,81,83,85,123,128,129],local_scop:32,localhost:[28,103],localip:123,localpath:55,locat:[4,18,25,35,57,73,80,93,96],lock:[35,36,40,41],lod:[18,47,70,74,90,96,97],lod_desc:[90,97],lod_expand:75,lod_level:[18,83,90,97],lod_rank_table_obj:18,lod_tensor:[18,74,90,97],lod_tensor_aarri:18,lod_tensor_arrai:18,lodrankt:18,lodtenosr:46,lodtensor:[18,19,46,47,64,70,85,97],lodtensorarrai:18,lodtensordesc:[70,90],log:[30,40,48,55,60,100,105,109,117,119,121,122,123,124,129],log_barrier_abstract:[128,129],log_barrier_lowest_nod:[128,129],log_barrier_show_log:[128,129],log_clip:[128,129],log_error_clip:[128,129],log_period:[122,123,129,131],log_period_serv:[128,129],logarithm:[1,18],logger:107,logic:[43,48,50,51,60,64,76,77,79,89,96],logist:18,logit:[18,60,95],longer:[36,48,66],look:[3,18,34,44,45,47,57,58,61,78,83,86,87,92,121,135],lookahead:[4,18,92],lookup:[18,64,75,135],lookup_t:66,loop:[32,34,46,57,66,80,84,89],loop_var:96,loss:[4,18,33,48,60,62,76,78,87,92],lot:[48,73,75,78,83,87,93,145],low:[4,76,77,92,93,96],low_rnn:74,lower:[4,18,56,74,75],lower_level_rnn:74,lpaddle_capi_engin:113,lpaddle_capi_lay:113,lpaddle_capi_shar:72,lpaddle_capi_whol:72,lrelu:60,lstm:[4,18,122,140],lstm_bias_attr:5,lstm_cell_attr:5,lstm_group:5,lstm_last:137,lstm_layer_attr:5,lstm_step:5,lstm_unit_op:18,lstmemori:[5,108,140],lstmemory_group:[4,108],lstmemory_unit:108,lstmlayer:67,lstmp:18,ltr:4,luckili:66,mac:[72,142],machin:[4,5,10,25,48,50,57,60,66,68,77,78,87,107,116,121,124,139,145],machine_transl:140,maco:[26,30],macro:[59,73,86],made:[10,36,41,57],mai:[4,5,18,22,32,34,48,52,56,58,65,66,69,77,80,84,85,89,92,95,107,110,121],main:[18,45,47,51,57,58,62,77,81,85,113,121,133],main_program:[17,18,33,52],mainli:[42,66,73,93],maintain:[4,34,40,78,83,85,121],majel:35,major:[48,56,95],make:[3,4,18,26,27,31,33,34,35,36,40,41,43,47,48,49,56,57,61,74,75,78,79,83,84,85,87,92,96,98,100,101,103,104,113,121,134,142,143,144,145],make_chan:47,make_channel:47,make_ddim:102,make_function_oper:59,make_vari:91,maker:[85,86],malloc:93,man:54,manag:[25,36,41,42,45,47,48,55,80,89,93,103],mandarin:[4,92],mani:[5,11,33,35,40,45,47,57,60,65,66,75,79,80,83,85,86,89,90,91,94,95,96],manili:62,manipul:[57,83,86],manner:[4,78,87,92,93],mantain:66,manual:[48,76,78,86,109,145],manufactur:57,manylinux1:30,manylinux1_x86_64:[30,88,104],manylinux:88,map:[3,4,9,19,25,31,34,40,51,69,73,83,86,89,91,93,96,110,113,145],map_fn:96,map_read:9,mapper:9,mapreduc:31,margin:18,mark:[10,33,50,60,61,74,75,80,89,145],marker:80,market:56,mask:[2,4,18],master:[10,31,43,85,88,144],mastermind:35,mat:[71,72,115],mat_cache_row:39,mat_norm:39,mat_normal_shar:39,mat_param_attr:5,mat_sparse_row:39,mat_sparse_row_auto_grow:39,mat_sparse_row_id:39,mat_sparse_row_prefetch:39,mat_sparse_row_prefetch_full_s:39,mat_value_shar:39,match:[18,35,56,94,107],matchbox:145,math:[5,18,68,71,85,100,101,134],mathemat:87,matirx:4,matmul:[34,62,74,96,101],matric:18,matrix:[3,4,5,9,18,19,25,39,71,72,100,101,113,115,116],matrixptr:100,matrixtyp:72,mattyp:39,max:[2,4,9,10,18,19,22,32,49,51,66,83,109,131,134,136],max_diff:32,max_id:[4,25],max_job_id:10,max_length:[4,75,140],max_movie_id:10,max_norm:18,max_relative_error:[32,101],max_seq_len:18,max_sort_s:4,max_user_id:10,max_x:18,maxframe_evalu:3,maxid:3,maxid_evalu:3,maxim:[4,51],maximum:[3,4,10,18,22,34,41],maxinum:7,maxoutfunctor:93,maxpool:4,mayb:[34,69],mchw:18,md5:[10,37],mean:[2,3,4,5,6,7,9,11,22,25,33,35,48,49,51,62,75,82,84,89,92,95,107,121,129,135,145],mean_op:18,mean_var_nam:4,meant:96,measur:[18,52],mechan:[4,5,19,33,42,46,52,69,83,86,121],mem:[4,18,34,44,75],mem_per_pserv:48,mem_per_train:48,member:[4,10,31,51,61,62,73,79,83,89],memcpi:[79,134],memor:4,memori:[5,18,33,34,39,40,44,56,58,68,69,70,73,75,78,80,85,95,102,122,134,135,140],memory_boot:5,memory_nam:[4,108],memory_optim:66,memory_test:26,memory_threshold_on_load_data:[128,129],memoryalloc:93,memorydesc:69,mention:[18,33,35,40,48,50,57,74,77,78,80],mere:5,merg:[4,18,41,43,49,52,68,74,77,79,98,116],merge_model:116,merge_v2_model:116,merge_v2_modelss:116,messag:[18,34,45,47,53,57,58,61,70,80,81,82,83,85,86,90,97,98,104,122],metadata:[54,121,122,123],metal:145,metaphor:61,metaplotlib:31,method:[4,6,25,32,34,43,45,46,48,49,51,56,60,61,62,65,76,77,83,84,85,89,90,95,96,133],methodolog:78,metric:[52,80],mfs:123,microarchitectur:56,might:[4,34,35,45,47,57,66,81,92,121],million:10,min:[2,4,18,22,49,51,83,121,131,134],min_block:34,min_count:50,min_desc:34,min_pool_s:107,min_word_freq:10,mini:[4,9,14,18,25,34,36,47,52,53,57,63,74,107],mini_batch:84,minibatch:[4,18,34,52,61,63,81],minim:[34,48,50,51,57,60,76,85,135],minimum:[4,18,22,92],minsizerel:[142,143,144],minu:86,minus_grad:86,minusgradop:86,minusop:86,minusopgradmak:86,minusopprotoandcheckermak:86,minut:[36,43,121],mip:142,mirror:35,mislead:41,mismatch:18,miss:60,mistak:57,mit:121,mix:[5,80,96,140],mixed_lay:5,mixed_layer_attr:5,mixedlayertyp:4,mkdir:[27,55,103,113,121,124],mkl:[27,69,85,93,95,104,113],mkl_packed_:67,mkldnn:[4,68,73,95],mkldnn_:68,mkldnn_batch_norm:4,mkldnnactiv:68,mkldnnbase:68,mkldnnlayer:68,mkldnnmatrix:68,mkldnnstream:68,mkldnntester:68,mklml:[68,104],mklml_lnx_2018:104,mklpack:67,mklpackedgatedrecurrentlay:67,mklpackedgemm:67,mklpackedlstmlay:67,mklpackedrecurrentlay:67,mlp:62,mmt16_task1_test:10,mnist:[38,48,60,61,81,84,85,116,133],mnist_random_image_batch_read:84,mnist_train:84,mnist_train_batch_read:84,mnist_v2:116,mnt:123,mobil:[56,57,66,85,103],mode:[4,22,25,56,67,77,98,123],model:[4,5,10,17,18,19,25,33,34,36,37,45,48,50,51,52,61,66,67,76,77,78,85,87,92,95,96,103,110,112,116,121,131],model_list:[129,131],model_path:131,modelparallel:48,modern:66,modif:92,modifi:[4,23,48,56,62,87,98,121],modul:[5,10,25,48,59,60,75,92,96,101,109,133],modular:75,modulo:4,moment:18,momentum:[2,18,89,107,110],momentumop:133,momentumoptim:20,mon:122,monitor:[47,80],mono:4,month:35,more:[3,4,5,9,10,18,22,31,32,33,35,36,40,43,44,46,47,48,50,55,56,57,58,59,61,65,66,69,73,74,75,76,80,83,84,85,87,92,93,94,96,109,134,135,145],mose:10,mosesdecod:10,most:[4,9,25,31,33,35,43,47,48,58,61,62,73,75,78,80,84,87,92,93,94,135,145],mostli:[56,145],motiv:85,mount:[44,121],mountpath:[121,122,123],move:[4,36,40,46,55,57,78,121,145],movi:10,movidiu:57,movie_categori:10,movie_id:123,movie_info:10,movie_review:10,movieinfo:10,moving_average_fract:4,moving_mean_nam:18,moving_variance_nam:18,mpi:[47,77,124],mpirun:124,mse:[57,61,76,81],much:[4,18,36,57,69,76,84,87,96],mul:[59,66,83,100,101],mul_grad:101,mul_op:[18,101],mul_ratio:4,mul_result:83,mulgradkernel:101,mulkernel:101,mulop:[59,101],mulopgrad:101,mulopmak:101,mult:[45,58],multi30k:10,multi:[4,19,52,77,79,95,145],multi_binary_label_cross_entropi:4,multidimension:18,multigradientmachin:79,multilingu:10,multimod:10,multinomi:4,multip:19,multipl:[3,4,5,9,14,18,19,25,31,32,40,41,43,45,47,48,50,52,57,58,59,65,77,80,85,92,95,97,110,121],multiple_input:83,multiple_param_attr:83,multipli:[3,4,18,45],multiprocess:9,must:[1,3,4,5,9,11,18,33,41,46,51,66,69,70,73,80,82,83,84,85,91,97,100,101,102,107,117,121],mutabl:[93,102],mutable_data:[69,93,101,102],mutex:47,mutuable_data:93,mutual:18,mxnet:[34,45,47,57],my_cluster_nam:121,my_cost:109,my_external_dns_nam:121,my_lib:117,myerrorclip:51,myfil:9,mypaddl:[122,123],naiv:45,name:[2,3,4,5,7,11,17,18,21,22,25,31,32,33,34,36,38,39,41,44,45,46,48,52,56,59,62,65,68,69,70,72,73,75,80,81,83,85,88,90,91,96,97,100,109,110,112,122,123,125,126,131,134,135,140,145],name_:80,name_prefix:38,namespac:[34,63,71,83,100,101,122,123],nativ:[4,18,56],natur:[18,40,43,50,75,96],ncall:133,nccl1:77,nccl2:77,ncclinit:77,nchw8:95,nchw8c:95,nchw:[4,18,68,73],ndarrai:[11,18,25,38,107],ndcg:4,ndcg_num:4,ndk:142,nearest:56,nearli:[32,46],necess:96,necessari:[4,33,34,41,43,51,52,66,70,75,79,83,91,96],necessarili:45,neck:77,need:[3,4,5,9,17,18,19,23,31,32,33,35,39,40,41,43,44,46,47,48,50,51,52,55,57,58,59,60,65,66,69,73,75,76,77,78,79,80,82,83,85,86,87,89,90,91,92,93,96,104,121,123,134,145],need_tran:109,neg:[3,4,18],neg_distribut:4,neg_overlap:4,neg_pos_ratio:4,neglect:4,neighberhood:77,neither:4,neon:56,ner:3,nervana:57,nessesari:92,nest:[4,9,33,34,80,81,97,115],net:[4,5,12,18,34,60,74,89,116],netop:[34,85],network:[2,3,4,9,18,19,24,25,31,32,33,34,36,39,48,50,52,60,62,66,67,68,74,76,78,80,83,84,87,89,91,92,93,95,97,108,110,112,115,116,123,131,137,145],network_config:131,networkadministr:121,neural:[4,5,9,18,25,31,33,34,36,48,62,66,67,68,74,78,87,89,93,95,97,112,137,139],neuralnetwork:79,neuron:18,never:[9,66,84,89,121,122,123],new_block_idx:83,new_dim:18,new_op_desc:51,new_scop:95,new_stat:74,newblock:83,newbuff:69,newest:41,newli:[56,145],newop:34,newopdesc:83,newprogram:83,newscop:95,newvardesc:83,next:[4,10,33,36,42,46,47,51,75,77,96,121],nextlay:68,nfs4:121,nfs:[121,123],nfsdir:123,nfsver:121,ngram:10,nic:[123,128,129],nil:[40,47],nine:10,nlp:4,nltk:10,nms_threshold:4,nms_top_k:4,nmt_without_attent:107,nnz:[100,115],no_grad_dict:33,no_grad_set:[32,33,101],no_gradi:33,no_sequ:[4,110],node0:123,node1ip:124,node2ip:124,node3ip:124,node:[4,35,43,45,48,50,62,66,75,77,85,121,122,123,124,145],node_0:[121,123],node_1:[121,123],node_2:[121,123],node_id:117,nodeattr:62,nodeentri:62,nodefil:119,nodesep:62,nohup:117,nois:[4,18,36,60],noisi:[4,60],non:[4,18,36,47,56,57,90,121],none:[2,3,4,5,6,7,11,13,14,17,18,19,21,22,23,25,31,32,33,34,47,51,52,60,62,63,74,75,76,81,83,91,96,135,140],noneedtran:69,nontranspos:18,nor:45,norm:[5,18,60,73],norm_by_tim:[4,18],normal:[4,5,10,18,21,78,92,122,123],normaliniti:16,normalizaiton:18,normliz:18,notat:[4,66],note:[2,4,5,7,10,11,18,25,31,33,34,39,40,44,66,70,73,77,84,85,93,95,102,121],notebook:[28,44],noteworthi:57,noth:[1,25,65,83,89,98],notic:[4,51,57,77,86],notimplementederror:51,notin:95,notingradi:101,notion:96,notori:32,now:[9,33,35,36,46,47,50,60,70,73,78,85,86,87,89,121,139],np_arrai:9,nproc:26,nullptr:[69,80,86,89,100],num:[4,5,117,123,129],num_channel:[4,5],num_chunk_typ:[3,14,18],num_class:[4,5,18,62],num_col_dim:18,num_filt:[4,5,18,19],num_flatten_dim:18,num_gradient_serv:[117,128,129],num_head:19,num_hidden:62,num_neg_sampl:[4,18],num_or_sect:18,num_parameter_serv:31,num_pass:[25,110,122,123,128,129,131],num_per_batch:11,num_pserv:48,num_repeat:4,num_result:3,num_results_per_sampl:4,num_row:90,num_samples_process:109,num_shard:38,num_step:96,num_total_class:18,num_train:48,num_true_class:18,number:[3,4,5,9,10,11,14,18,19,22,34,36,38,50,52,66,78,80,84,85,91,96,121],numchunktyp:3,numdevices_:131,numer:4,numeric_grad:32,numerical_grad:32,numlogicaldevices_:131,numofallsampl:3,numofwrongpredict:3,numpi:[2,9,11,18,25,27,38,56,60,83,84,101,107,109,110],numreal:39,numsampl:134,numtagtyp:3,numtimeout:40,nv_:35,nv_gpu:26,nv_librari:35,nv_test:35,nvcc:[35,56,58],nvidia:[26,28,56,73,77,93,104],nvlink:77,nvprof:80,obei:3,obj:109,object:[2,4,5,9,18,25,31,39,46,48,51,52,60,62,66,71,76,80,83,85,87,89,134],observ:4,obtain:[4,18,43,46,78,93,107,110],obvious:[35,73],occup:[66,123],occupi:[56,80],occur:[10,25,66],occurr:34,oct:122,odd:4,odoti:18,off:[26,27,72,98,103,113,127,142,143,144,145],offer:[34,85,91],offici:[4,35,121],offlin:[36,38,145],offset:[4,18,39,115],often:[4,39,62,66,73],ograd:100,old:[32,41,43,75,85],older:[18,57],omega:87,omit:[18,107],omp_num_thread:133,ompi_comm_world_rank:117,onc:[4,36,40,45,48,50,52,57,61,78,94,121],one:[1,3,4,5,7,9,18,22,25,31,32,33,34,36,39,40,41,43,44,45,46,47,48,51,52,53,56,57,58,59,60,62,64,65,69,70,73,74,75,76,77,78,79,81,82,83,84,85,86,89,90,92,93,94,95,96,115,121,135,145],onehotcrossentropyopkernel:101,ones:[59,60,85],onli:[3,4,5,7,11,18,25,31,32,33,35,39,40,41,42,43,44,45,46,48,50,51,52,53,55,56,57,60,61,66,69,74,75,76,77,79,80,83,85,90,91,92,93,94,95,96,115,121,139,145],onlin:[4,6,36,38,66,84],only_cpu:32,onnx:57,onto:[18,48,50,121],op1:[66,95],op1_2_op2:95,op1_to_op2:95,op2:[66,95],op3:66,op_:101,op_check:101,op_class:[85,91],op_desc:[51,64,82],op_info:135,op_kei:69,op_maker_class:[85,91],op_proto:91,op_registri:135,op_siz:51,op_test:101,op_typ:[85,101],op_unique_kei:69,opattrcheck:101,opcreat:91,opdesc:[34,51,61,81,82,83,85,86,91,97],opdescbind:[64,86],opdescbuild:34,open:[4,11,18,31,38,57,60,68,84,107,109,110,121],openbla:[27,28,113],opencv:11,openmpi:[120,124],opensourc:77,oper:[4,5,9,11,18,22,23,32,34,45,47,48,49,50,52,53,56,57,58,60,61,62,64,65,74,75,76,77,80,82,87,89,92,93,95,97,101,102,121,135],operand:[18,56],operartor:102,operator_grad:32,operator_list:80,operatorbas:[34,59,85,86,91,101],operatorwithkernel:[95,101],opinfo:[64,85,86],opinfomak:64,opinfomap:86,opkernel:[101,102],opkernelkei:85,opkerneltyp:[73,95],opmak:91,opposit:18,opproto:101,opprotoandcheckermak:[86,101],opprotomak:[91,101],opregist:91,opregistri:91,ops:[32,33,34,35,45,46,58,61,62,78,81,82,83,85,93,101,145],ops_:34,ops_test:35,opt:[27,31,76,82,91,123],opt_op_list:76,optest:101,optim:[2,12,23,24,25,32,33,48,49,50,58,60,77,78,79,81,85,87,90,92,107,109,110,135],optimis:76,optimize_op_attr:83,optimizer_mod:18,optimzi:107,option:[3,4,10,18,22,31,35,48,60,65,70,81,82,83,85,90,91,92,97,142,145],optmization_op_list:76,opts_np:82,optyp:[64,91],opwithkernel:90,order:[4,5,9,11,18,22,25,33,61,70,80,84,87,96,121,123,133,145],ordereddict:25,orderli:18,oregon:121,org:[3,4,5,10,18,38,54,60,107,110],organ:[3,4],orient:91,origin:[4,5,9,10,17,18,32,56,60,89,96,98],other:[3,4,5,9,18,22,34,36,41,45,55,56,57,58,64,66,69,73,74,78,82,87,89,91,92,93,95,121,135,145],otherchunktyp:3,otherwis:[4,9,10,11,18,22,25,31,33,36,41,43,60,64,69,84,92],our:[18,31,33,35,46,47,48,50,60,64,66,73,77,78,89,96,121],out:[4,18,25,31,34,35,40,43,46,48,51,57,62,66,69,74,75,83,95,101,102,107,121,133,139,140],out_dir:[121,123],out_fals:18,out_left:4,out_mem:140,out_memori:5,out_right:4,out_size_i:4,out_size_x:4,out_tru:18,outer:4,outer_mem:137,outgrad_:68,outlier:4,outout_lay:25,outout_layer1:25,outout_layer2:25,output:[1,2,3,4,5,7,9,14,18,19,22,25,31,32,33,34,38,43,45,46,50,51,55,58,59,60,61,62,63,64,66,69,70,74,75,78,81,82,83,84,85,86,89,90,91,93,95,96,101,102,107,110,116,119,122,123,131,140],output_:[4,68],output_all_step:74,output_arg_nam:51,output_dim_idx:18,output_dtyp:18,output_fil:[22,116],output_height:18,output_id:4,output_lay:[25,107,110,112],output_max_index:7,output_mem:[4,140],output_mod:22,output_nam:32,output_num:74,output_path:38,output_s:18,output_seg:96,output_width:18,outputbuff:39,outputgradi:86,outputh:4,outputw:4,outsid:[4,5,48,89],outter:115,outter_pos_arrai:115,outter_seq_pos_arrai:115,outupt:96,outv:100,outval_:68,over:[4,5,18,25,31,57,66,77,78,96],overal:[60,78,80,145],overfit:[18,87],overflow:18,overlap:[3,4],overlap_threshold:[3,4],overload:[56,65],overrid:[34,36,55,69,93,101,102],overview:[40,41,42,93],overwrit:55,own:[4,18,33,41,43,51,53,62,64,76,77,78,87,91,121],pack:[96,109],packag:[9,10,40,44,45,59,68,88,104,121,133],pad:[5,18,69,92],pad_c:4,pad_h:4,pad_w:4,padding_attr:4,padding_down:18,padding_h:18,padding_i:4,padding_idx:18,padding_left:18,padding_right:18,padding_up:18,padding_w:18,padding_x:4,paddl:[1,2,3,4,5,6,7,9,10,11,13,14,15,16,17,18,19,20,21,22,23,25,26,27,28,30,31,34,35,36,38,44,46,48,55,58,59,60,63,67,68,69,70,71,72,74,75,79,81,85,87,88,91,92,93,96,98,100,101,103,107,109,110,112,113,115,116,117,119,121,122,123,124,127,131,133,134,135,140,142,145],paddle_arguments_get_sequence_start_po:115,paddle_arguments_set_id:115,paddle_arguments_set_sequence_start_po:115,paddle_arguments_set_valu:115,paddle_begin_init_param:41,paddle_capi:113,paddle_dir:101,paddle_doc:103,paddle_docs_cn:103,paddle_element_typ:41,paddle_element_type_float32:41,paddle_element_type_float64:41,paddle_element_type_int32:41,paddle_element_type_int64:41,paddle_element_type_uint32:41,paddle_element_type_uint64:41,paddle_enforc:[34,46,69],paddle_enforce_eq:[101,102],paddle_error:[71,72],paddle_exampl:44,paddle_finish_init_param:41,paddle_get_param:41,paddle_gradi:41,paddle_gradient_machine_create_shared_param:116,paddle_gradient_machine_forward:116,paddle_gradient_machine_load_parameter_from_disk:116,paddle_init:116,paddle_init_num_gradient_serv:117,paddle_init_param:41,paddle_init_port:117,paddle_init_ports_num:117,paddle_init_ports_num_for_spars:117,paddle_init_pserv:117,paddle_init_trainer_count:117,paddle_init_trainer_id:117,paddle_init_use_gpu:117,paddle_ivector:115,paddle_ivector_cr:115,paddle_job:44,paddle_manylinux_devel:27,paddle_matrix:[71,72,115,116],paddle_matrix_cr:[72,115],paddle_matrix_create_spars:115,paddle_matrix_get_row:115,paddle_matrix_get_shap:71,paddle_matrix_shap:71,paddle_matrix_sparse_copy_from:115,paddle_n:123,paddle_new_etcd_pserver_cli:41,paddle_new_pserver_cli:41,paddle_on_cloud:44,paddle_output:122,paddle_paramet:41,paddle_port:123,paddle_ports_num:123,paddle_ports_num_spars:123,paddle_process_by_paddl:123,paddle_pserver2:119,paddle_pserver_cli:41,paddle_pserver_client_releas:41,paddle_r:115,paddle_root:113,paddle_save_model:41,paddle_send_grad:41,paddle_server_num:123,paddle_train:[72,88,119,123],paddle_with_cuda:80,paddle_with_mkldnn:73,paddlepaddl:[4,5,9,10,11,18,25,26,27,28,30,35,36,38,41,42,43,44,45,48,54,55,59,60,61,63,65,70,74,75,76,79,80,83,84,85,89,96,97,98,107,110,112,113,115,116,119,120,122,123,125,126,127,133,134,140,142,144,145],paddlepaddle_gpu:30,paddlepaddlebook:28,paddlepaddlehub:28,page:[10,121],pair:[3,18,19,22,23,33,34,48,61,76,80,85],pairwis:4,pakcag:35,panic:47,paper:[4,10,18,60,92],para:39,paradigm:[45,53,85],paragraph:74,paragraph_data:74,paragraph_out:74,parallel:[18,45,47,48,50,77,80,85,95,121,122,123,131,134],parallel_for:45,parallel_nn:[2,128,129],paralleldo:49,parallelfor:45,paralleliz:92,param:[2,4,5,9,17,18,32,34,41,79,83,93,102,107,109],param_attr:[4,5,12,18,19,39,83,107,109,140],param_config_proto:41,param_fil:[109,116],param_initi:18,paramattr:[2,4,18,107,109,140],paramet:[3,4,5,6,7,9,17,18,19,23,24,32,33,34,35,37,39,43,45,48,49,51,53,55,57,58,60,61,62,64,70,74,76,77,81,84,89,91,92,96,107,109,110,112,115,118,119,123,129,135],parameter_block_s:[128,129],parameter_block_size_for_spars:[128,129],parameter_learning_r:2,parameter_list:[33,76],parameter_nam:[25,31],parameter_serv:31,parameter_valu:39,parameterattribut:[2,4,5,39],parameterclient2:123,parameterclient_:39,parametermap:100,parametermutex_:39,parameters_:100,parameters_and_grad:[23,76],parameterserver2:39,parameterset:31,parameterupdat:79,parameterupdater_:39,parametr:4,params_grad:76,params_pass_4:116,params_pass_90:110,params_pass_:110,paramt:121,paraspars:100,parent:[34,45,81,83,85],parent_:[34,89],parent_idx:83,parenthes:85,pars:[9,10,26,35,48,62,121],parse_known_arg:123,parsefromstr:109,parser:[9,123],part:[3,4,18,19,33,34,43,48,57,69,70,81,83,92,93,145],partial:[4,25],partial_sum:4,particular:[61,70,85,92,95],partit:[36,38,48,50,85,121],paserv:123,pass:[4,18,25,33,34,36,47,51,52,57,60,66,70,76,78,79,82,83,84,85,87,89,92,94,96,98,107,110,121,122,123,129,131,134],pass_gener:4,pass_id:[25,48,107,110],pass_idx:84,pass_manu:109,passtyp:100,past:[31,121],patch:[18,54],path:[3,9,10,11,17,25,36,40,41,44,66,75,84,92,113,121,122,123,129,142,143,144],path_to_paddlepaddle_working_directori:103,pattern:[10,36,71,78,87,121],paus:[36,43],pcie:77,pd_api:115,pdf:[5,18],peephol:18,peer:[77,105],pem:[31,38,121],pend:[36,40],pep425tag:[30,104],per:[3,4,10,11,18,36,41,77,78,84,87],percal:133,perf_test:133,perfectli:92,perform:[4,5,18,32,41,47,48,52,56,57,60,66,77,79,80,84,85,87,92,93,95,128,133,134],perftool:[80,133],period:[36,43,129],perl:10,perm:18,permiss:[107,110,121],permut:18,peroid:[4,11],persist:[17,18,53,90,92,97,121],persistentvolum:121,persistentvolumeclaim:[121,123],person:[3,31,65],perspect:85,perturb:32,pex:145,pfs:[38,55],pfsclient:38,pfspath:55,pgp:121,phase:[18,69,75,77,78,84,86,92,145],philosophi:[78,87],photo:60,phrase:18,physic:145,pick:[18,121],pickl:[117,124],pictur:77,piec:[5,45,80],pil:11,pillow:44,pip:[27,30,88,98,103,104,112,133],pipe:9,pipelin:[52,92],piperead:9,pivot:69,pixel:[4,9,10,48],place:[13,15,18,22,33,34,36,43,48,50,53,65,69,77,85,95,102,135],place_:[65,73,93,95],placehold:[60,93,102],placement:50,plain:[3,4,9,44,70,72,73],plan:[36,45,69,85,92],platform:[30,34,58,69,73,80,93,95,101,102,121,135,142,143],pleas:[2,4,5,6,10,11,18,31,36,40,41,42,45,47,58,62,73,74,83,84,85,92,93,97,103,104,121,123],plot:31,plu:4,plug:[77,78],pnpairvalid:128,pod:[38,44,45,121,122,123],pod_nam:121,podip:123,podlist:123,point:[18,34,36,44,47,56,66,69,77,93,102,115,134,145],pointer:[34,41,62,66,73,83,85,89,93,102,115],polar:10,poli:109,polici:[18,121],polit:18,poll:47,pollut:43,polyak:78,ponit:62,pool3:100,pool:[5,18,24,49,66,92],pool_attr:5,pool_bias_attr:5,pool_layer_attr:5,pool_pad:[5,18],pool_siz:[4,5,18,19],pool_size_i:4,pool_strid:[5,18,19],pool_typ:[4,5,18,19],pooled_height:4,pooled_width:4,pooling_lay:[5,107],pooling_typ:[4,107,136],poolingtyp:7,pop:[34,53],popul:41,popular:[35,60,62,80],port:[18,35,45,117,121,122,123,128,129],port_num:128,portabl:62,portal:103,ports_num:[117,123,129],ports_num_for_spars:[39,117,123,128,129,131],pose:36,posit:[3,4,5,18,115],positive_label:3,possibl:[31,34,40,47,50,66,83,87,97],post:[44,54],postpon:87,potenti:[56,134],pow:109,power:[56,66,77,92,145],ppo_workspac:103,pprof:133,pre:[4,5,10,31,41,65,66,121],pre_activ:83,pre_bia:83,pre_id:18,pre_mem:18,pre_stat:[74,96],preambl:83,precis:[3,14,18,52,56,78],precision_evalu:3,precompil:53,pred:[62,66],predecessor:66,predetermin:4,predic:[10,17],predict:[3,4,18,25,48,87,107,112,116],predict_fil:[128,129],predict_lay:25,predict_output_dir:[128,129],prediction1:25,prediction2:25,prefer:[57,65],prefetch:[39,100],prefix:[3,5,18,36,38,75,92,121],pregel:47,pregrad:100,prepand:83,prepar:[32,44,79,92,117,124,125],prepend:[18,83],prepend_oper:83,preprocess:[10,11,92,96],present:[31,33,34,80,96,98],preserv:55,prev_batch_st:[128,129],prev_cel:18,prev_cell_data:18,prev_hidden:18,prev_hidden_data:18,prevent:[6,18,31,36,40,43,51,87],preview:85,previou:[4,5,18,22,25,33,36,50,55,74,75,121],previous:4,previous_memori:34,price:[10,85,112],prim:69,primari:[57,61],primarili:[78,87],primit:[56,68,69,77,96],primitive_desc:69,primitivedesc:69,principl:[18,31,35,73],print:[2,22,25,30,31,47,48,57,62,83,94,104,110,112,124],print_graphviz:62,print_phas:18,print_tensor_lod:18,print_tensor_nam:18,print_tensor_shap:18,print_tensor_typ:18,printallstatu:134,printer:3,println:47,printstatu:134,priorbox:4,prioriti:85,prite:3,privat:[34,72,80,83,89,90,91,93,96,98,102],privileg:121,pro:77,prob:[3,25,112],probabalist:18,probabilist:4,probabl:[3,4,18,25,75,92],problem:[4,31,32,35,43,46,57,60,61,78,85,87],proc:28,proce:[9,36,84,121],procedur:[34,70],proceed:[10,18],process:[2,4,5,9,18,31,33,34,38,39,40,43,45,46,47,48,52,53,57,58,62,66,68,70,77,87,91,95,107,109,121,123],process_num:9,processor:[56,134],prod:98,produc:[4,5,9,18,36,57,62,84],product:[4,5,18,19,44,57,110,121],productgraph:122,prof:133,profil:[12,55,80,92,133,134],profilerst:80,profl:133,proflier:[80,134],prog:123,program:[9,13,17,18,22,31,33,38,41,43,48,50,53,61,63,66,76,77,80,84,85,89,94,97,123,134],programdesc:[45,48,53,57,66,70,82,83,86,97],programm:[48,57,83],progress:[36,40],proj:4,proj_activ:18,proj_dim:18,proj_out:18,proj_siz:18,project:[4,5,18,19,44,72,92],promis:[4,5,75],prompt:[55,57],prone:31,pronunc:92,prop_kind:69,propag:[4,6,33,57,78],proper:65,properli:65,properti:[62,87],propos:[18,34,49,50,75,76,77,78,96],proprietari:68,protect:[46,56,91,100,101],proto:[7,47,65,70,73,81,85,91,97,101],proto_:91,protobuf:[25,34,44,45,48,53,57,58,61,62,66,70,81,83,85,86,91,109,113,116],protocol:[3,135,145],protomak:101,provi:117,provid:[4,10,18,31,34,41,44,45,52,53,56,57,60,62,64,65,73,77,78,80,83,87,91,92,93,94,96,107,112,121,128,145],provis:[121,145],prune:[4,17,34],ps_desir:36,pserver:[25,39,41,42,44,85,117,119,121,123,128,129],pserver_addr:41,pserver_cpu:44,pserver_id:37,pserver_mem:44,pserver_num_thread:[39,128,129],pserver_spec:25,pseudo:[31,33,44,86,96],pseudocod:96,psize:100,ptr:[72,93],pub:[18,124],pull:[28,35,85,88,142],purpos:[4,36,48,50,65,134],push:[34,53,57,80,88,123],push_back:100,put:[35,36,39,50,66,69,83,93],pvc:121,pwd:[26,27,28,98,103,142],pxe:145,py_paddl:104,pybind:[34,47,56],pydataprovid:107,pydataprovider2:123,pyramid:4,pyramid_height:4,python2:133,python:[10,18,25,26,27,30,31,34,42,46,47,52,53,57,59,60,61,62,65,71,75,79,80,85,88,93,96,98,100,101,103,104,107,112,116,117,124,133,135,140],pythonpath:104,pytorch:[57,80],qualcomm:56,queri:[3,4,19,121],query_id:3,quest:10,question:[4,31,45,50,91,121],queue:[47,50],quick:62,quick_start:[44,121,122,123,125],quick_start_data:122,quickli:[75,83,85],quickstart:122,quit:75,r14b:142,r_h_val:18,r_t:4,rais:[9,18,19,23,51,62,117],rajathkmp:60,ran:[50,134],rand:[60,109,115,129,131,134],rand_max:115,random:[2,4,9,18,38,47,60,73,79,83,84,101,109],random_crop:11,random_imag:38,randomli:[11,18,43,115],rang:[4,9,18,38,45,48,56,60,66,80,84,91,123],rank0:77,rank1:77,rank:[4,18,31,96,121],rank_tabl:18,rankdir:62,ranktabl:18,rapid:86,raspberrypi:144,raspbian:144,rasspberri:144,rate:[2,3,4,5,6,10,19,41,92,107,123],rather:[33,44,60,96,121],ratio:[4,129],raw:[4,18,70],rdma_tcp:[128,129],reach:[36,66,77],read:[9,11,18,25,31,33,36,38,45,46,47,48,50,57,58,84,85,92,96,109,121,145],read_from_arrai:66,read_from_realistic_imag:31,read_from_rng:31,read_lock:37,read_minibatch:57,read_mnist_imag:31,read_next_from_fil:107,read_paramet:109,read_ranking_model_data:31,readabl:[80,85],reader:[10,25,38,48,56,60,61,81,92,110,117,133],reader_cr:38,reader_creator_bool:84,reader_creator_random_imag:[9,84],reader_creator_random_image_and_label:[9,84],readi:[36,47,121,122,145],readlockguard:39,readm:[72,98],readnext:46,readwritebuffer_:39,readwritemani:121,real:[4,18,39,60,84],realist:31,realiti:92,realiz:[34,74],realli:[57,87],rearrang:18,reason:[5,31,32,36,47,57,122],recal:[3,14,18],receiv:[18,36,44,47,48,50,74],recent:[66,78],recognit:[4,92],recommend:[5,18,31,123],record:[9,22,40,69,80,91,121],recordev:80,recordio:[9,10,31,38,40,46],recov:[36,96],recover:85,recoveri:40,rectifi:4,recurr:[18,67,74,89,92,137,138],recurrent_group:[5,92,107,108,137,139,140],recurrent_lay:5,recurrent_op:96,recurrentgradientmachin:[72,75,96],recurrentgroup:3,recurrentlay:67,recurs:[33,34,35,55,66,85],recv:[45,48,50,77,121],recvparametertyp:39,red:60,redirect:9,reduc:[4,18,50,56,77,85,133],reduce_by_kei:85,reduce_mean:60,ref:18,ref_batch_dim_idx:18,refactor:[48,50,61,75,78,79,83,87,96],refer:[2,4,5,6,11,18,19,22,34,36,40,41,42,45,56,62,69,73,74,77,81,83,85,87,89,93,96,97],referenc:40,refine_unknown_arg:123,reflect:40,reformul:18,reg:91,regard:[18,46,145],region:[4,89],regist:[47,66,73,86,93,95],register_gpu_profil:134,register_lay:100,register_op:[59,85,86,91,101],register_op_cpu_kernel:[93,101],register_op_cuda_kernel:[93,101],register_op_without_gradi:[85,101],register_oper:[64,86],register_tim:39,register_timer_info:134,registerop:91,registr:135,registri:[44,64,93,122,145],regress:4,regular:[2,12,18,21,33,107,121],reinit:46,reiniti:[46,69],rel:[5,18,32,43,87],relat:[36,43,44,56,73,80,89,95,145],relationship:[86,93],releas:[88,92,104,113,121,142,143,144],reli:[32,45,75,76,78,87],reliabl:[36,87],relu1:62,relu2:62,relu:[4,60,62,66],relwithdebinfo:133,remain:[18,96],remaind:18,rememb:4,remot:[2,35,39,48,85,98,121,129,131],remoteparameterupdat:[39,42],remov:[9,18,33,48,55,57,75,98],renam:[33,55,56],reorder:[18,69],reorder_primit:69,reorderlodtensorbyrankt:18,reorgan:4,repeat:[18,34,61,81,82,90,91,97],repeatedli:[61,66],replac:[18,35,40,64,78,86,92],replic:48,replicaset:44,repo:[35,144],report:[40,56,57,80],reportdataset:40,repositori:[4,103,142],reprenset:18,repres:[4,5,18,33,34,40,45,48,50,51,57,62,70,73,75,78,83,85,87,90,93,96,97,121],represent:[4,18,41,48,58,60,61,66,73,75,90],request:[35,36,39,43,45,85,88,121,122,145],requir:[3,4,31,33,36,41,43,44,46,48,50,51,55,56,62,66,68,74,78,80,81,82,85,87,90,91,92,97,103,107,110,121,145],requisit:66,research:[10,18,48,57],reserv:[18,55,107,110],reserveoutput:100,reset:[4,18,36,52,105],reset_program:52,resetingrad:68,resetinvalu:68,resetoutgrad:68,resetoutvalu:68,resetxxx:68,reshap:[32,84,109],reshape_s:4,resid:18,residu:47,resiz:[11,39,93,101,102],resize_s:11,resize_short:11,resolv:[35,98,122],resourc:[48,53,77,80,93,121],respect:[18,19,32,46,51,56,60,74],respons:[4,39,47,48,52,60,77,78,79,87,121,122],rest:[18,34,44,54,58,95,145],restart:[36,41,121,122,145],restartpolici:[121,122,123],restor:[32,78],restrict:[87,89,133],result:[1,3,4,17,18,22,25,32,33,40,47,52,60,61,62,66,70,75,76,77,79,110,121,134,135],result_fil:3,resum:43,retran:121,retriev:[34,75,89],return_seq:5,reuqest:88,reus:[34,43,75,84,85],rev:26,revamp:48,reveal:31,revers:[4,5,10,18,33,139,140],review:[10,45,98,122],reviews_electronics_5:122,rewrit:[35,47],rgb:[4,11],rho:6,rid:[46,57],right:[4,32,33,34,35,44,52,66,85,87,107,110],right_scor:107,ring:77,risk:33,rkt:[26,44],rmsprop:[78,107],rmspropoptim:78,rnn:[4,5,18,34,57,60,75,83,85,89,92,128,139,140],rnn_bias_attr:140,rnn_layer_attr:140,rnn_out:140,rnn_output:96,rnn_step:4,rnn_use_batch:[67,128,129],rnnalgorithm:75,rnnlm:10,rnnstep:96,roadmap:[92,96],robust:4,rocmplac:73,roi:4,role:[10,31,40,41,48,77,121],rollback:83,root:[6,7,33,77,121,122,123],rot:4,roughli:92,round:[56,77],routin:[56,68,77],row:[3,4,9,18,39,47,115],row_id:4,row_offset:115,rowoffset:115,rows_:90,rpc:40,rpcserver:40,rpi:144,rpi_arm_neon:144,rpi_toolchain:144,rsize:121,rstrip:123,rtk:145,rtype:9,rule:[3,18,33,48,51,57,61,121],run:[18,26,27,28,31,32,33,34,35,36,44,45,46,47,48,49,50,52,56,57,58,59,60,61,62,66,69,73,74,76,77,78,80,81,82,83,85,88,89,90,92,93,94,95,98,102,103,104,119,121,122,123,125,126,127,133,134,142,145],run_test:27,runinitfunct:[123,134],runnabl:50,running_on_cloud:44,runserv:103,runtim:[22,34,45,47,48,64,74,85,97,104],runtime_table_:34,s_block:33,safe:44,sai:[4,46,58,61,63,66,84],said:57,same:[3,4,5,18,19,23,25,31,32,40,41,43,45,46,47,48,59,60,62,65,66,74,75,77,83,85,86,89,92,95,96,102,121],samping_id:4,sampl:[3,9,10,18,52,60,83,91,115],sample_id:3,sample_num:3,sample_weight:18,sampler:60,satifi:[3,66],satisfi:[35,69,90,121],save:[4,9,17,25,36,38,40,41,44,45,48,61,62,66,70,78,90,97,121,122],save_dir:[122,123,129,131],save_file_nam:17,save_only_on:[128,129],save_parameter_to_tar:[25,110],savetxt:109,saving_period:[123,128,129],saving_period_by_batch:[128,129,131],scalabl:85,scalar:[4,18,33,34,63,94,96],scale:[1,19,48,50,78,86,91,92,101],scale_a:18,scale_b:18,scaleop:101,scaleopmak:[85,101],scalingproject:4,scan:[18,33,40,66,85],scatter:[4,33,77],scenario:75,schdule:121,schedul:[40,44,50,121],scheduler_factor:2,scheme:[3,6,39,87],scienc:66,scope:[15,32,45,49,53,58,95,135],score:[3,4,14,18,75,107],score_diff:107,scorer:92,scp:124,script:[10,27,77,103,119,121,124,142],sdk:143,search:[4,18,36,89,140],second:[4,18,31,45,55,57,60,62,74,75,81,82,84,89,91,101],secret:121,section:[33,50,57,83,121],see:[4,5,10,18,31,33,36,45,47,50,56,57,83,92,107,109,110,121],seed:[18,109,129,134],seem:[35,47,56,57,92],seen:[19,87],segment:[3,18,74,96],sel:47,sel_fc:4,selcet:4,select:[4,18,75,121],selected_generation_scor:75,selected_id:[4,75],selected_indic:4,selected_row:[90,97],selected_rows_desc:[90,97],selected_scor:75,selectedrow:[64,97],selectiv:4,selector:122,self:[32,51,52,60,62,66,67,68,70,76,83,96,100,101],self_addr:45,selfnorm:4,semant:[10,31,75,88],semaphor:47,semat:31,send:[36,41,45,48,50,65,77,85,91,121],send_back_parameter_typ:39,send_var:18,sendbackparameterspars:39,sendbackparametertyp:39,sendparameterrequest:39,sendparameterrespons:39,sens:[78,87],sensit:4,sent:[31,41,45,48,85,91,97,122],sentanc:107,sentenc:[4,10,57,74,75,96,140],sentence_input:96,separ:[3,17,18,22,41,48,59,78,86,87],seper:[18,96],seq0:18,seq1:18,seq2:18,seq3:18,seq:[4,10],seq_len:96,seq_po:115,seq_pool:[4,136],seq_pos_arrai:115,seq_silc:4,seq_text_print:3,seq_typ:9,seqtext_evalu:3,seqtoseq:[4,109],seqtoseq_net:4,sequenc:[1,3,4,5,7,9,10,18,33,34,45,53,57,61,67,76,81,92,96,107,110,115,137,139],sequence_group:4,sequence_layer_group:137,sequence_nest_group:4,sequence_nest_layer_group:137,sequence_start_posit:115,sequencegen:137,sequencesoftmaxop:18,sequencestartposit:4,sequencetextprint:3,sequencetyp:[4,110],sequenti:[4,34,45,47],seri:[5,46,137],serial:[25,34,40,70,79,85],serializ:[85,97],serv:[48,56,85,96,121],server:[18,27,31,35,39,42,43,48,58,77,85,105,117,118,119,123,129,145],serverless:36,servic:145,sess:[60,62,76],session:[62,76,82],set:[2,3,4,5,9,10,11,18,19,21,22,23,25,26,31,33,36,44,46,60,64,66,69,73,74,75,80,82,83,85,86,89,92,93,96,101,102,103,107,109,110,115,121,122,134],set_active_typ:100,set_attr:51,set_default_parameter_nam:2,set_drop_r:100,set_float_el:32,set_input:[4,51],set_output:51,set_shap:46,set_siz:100,set_typ:[51,100],setdatatyp:90,setdefault:101,setp:121,setq:26,settotalbyteslimit:109,setup:[48,78,88,100,101,145],seven:92,sever:[3,4,32,39,48,50,60,74,75,77,79,80,83,90,93,96,121],sexstant:145,sgd:[25,31,36,44,50,78,79,90,110,118,128,135],sgd_optim:135,sgdoptim:20,shall:[33,35],shape:[3,4,9,18,19,25,32,33,34,46,48,60,63,73,74,81,83,85,90,92,93,110,135],shapes_:46,shard:[36,37,38,39,40,41,43,48,50,118,121],share:[4,18,35,46,60,72,79,83,85,87,92,93,96],shared_bia:5,shared_bias:4,shared_librari:35,shared_ptr:[69,71,72,89,93,102],shef:10,shell:121,shift:[4,18],shorten:4,shorter:11,shoul:32,should:[2,3,4,9,11,18,19,22,25,31,32,33,34,41,44,47,48,51,52,56,58,59,60,64,65,69,73,74,75,76,78,79,80,81,84,85,86,87,90,91,92,94,95,96,97,101,103,121,139],should_be_fals:31,should_be_tru:31,show:[3,6,26,33,34,36,46,47,55,57,63,66,70,74,77,78,81,94,96,121],show_check_sparse_distribution_log:[128,129],show_layer_stat:[128,129],show_parameter_stats_period:[122,128,129,131],shown:[4,18,31,48,52,77,80,92,121],shrink_rnn_memori:18,shrunk:51,shuf:107,shuffl:[9,46,48,107],shuffleread:46,sid:121,side:[4,18,25,48,52,66,79],sig:121,sigint:119,sigma:18,sigmod:91,sigmod_op:91,sigmod_output:91,sigmoid:[4,19,34,91,96],sigmoidactiv:5,sigmoidcrossentropywithlogit:18,sign:[54,70,121],signatur:121,signific:92,sima:10,similar:[4,18,34,45,47,48,50,53,57,75,78,80,84,85,87,92,93,95,96,121,145],similarli:[4,9,57,66],simpl:[1,3,4,5,9,10,18,25,45,50,56,58,61,62,66,74,78,81,87,89,91,92,96,123],simple_attent:140,simple_gru:140,simple_lstm:[4,108],simple_rnn:[4,140],simple_transform:11,simpler:79,simplest:121,simpli:[4,11,31,41,48],simplifi:[31,75,83,91,92],simul:[57,143],simultan:121,sinc:[4,5,18,36,40,42,43,47,48,49,50,57,64,66,69,73,78,83,84,86,87,96,121,145],singl:[3,5,9,17,18,33,36,46,48,50,52,56,65,77,85,89,92],singleton:[45,49],sinlg:25,sit:[18,48],site:[35,121,133],sitten:18,sittin:18,situat:[33,82,95],size:[3,4,5,9,10,11,18,19,25,36,38,39,41,47,48,56,60,66,70,75,78,83,84,90,91,92,93,96,100,101,102,107,109,110,112,115,135,140],size_a:4,size_b:4,size_in_byt:69,size_t:[39,46,93,96,100],sizeof:[34,115],skip:[33,84,98,109,121],sleep:123,slice0:18,slice1:18,slice2:18,slice3:18,slice:[4,18,45],sliceproject:4,slide:[4,6,10,36],slight:57,slightli:60,slope:[4,18],slopeinterceptlay:4,small:[4,10,18,32,45,58,60,68,75],small_messag:[128,129],smaller:[18,32,36,56,75],smart:89,smooth:4,smt:10,snap:122,snapdragon:56,snapshot:[37,43,121],snippet:[47,59,76,121],sock:44,sock_recv_buf_s:[128,129],sock_send_buf_s:[128,129],socket:123,soft:18,soft_label:18,softmax:[4,5,18,31,34,48,50,57,62,63,75,81,100,107,140],softmax_param:109,softmax_param_attr:5,softmax_selfnorm_alpha:4,softmaxoutput:62,softrelu:18,softwar:[56,80,107,110,145],solid:60,solut:[77,145],solv:[31,33,46,66,85],some:[2,4,9,11,18,25,31,33,34,35,39,40,41,43,44,46,47,48,50,51,56,58,59,60,61,65,66,69,73,74,75,76,81,82,83,84,85,86,89,93,95,96,121,145],some_c_api_funct:72,some_inst:72,some_lay:18,some_op:[64,74,96],some_python_class:71,somecppclass:71,somedata:25,somegotyp:71,someth:[33,39,83],sometim:[4,80,84],somewhat:41,somewher:89,soon:36,sort:[4,10,18,22,96,121,123],sort_by_length:96,sortagrad:92,sorted_kei:22,sourc:[4,10,18,32,35,55,57,60,68,70,72,75,84,85,121],source_dict_dim:[75,140],source_dict_s:75,source_language_word:[75,140],space:[3,4,50,56,83,87,92],space_seperated_tokens_from_dictionary_according_to_seq:3,space_seperated_tokens_from_dictionary_according_to_sub_seq:3,span:80,spars:[2,4,6,9,18,39,47,100,107,115,117,121,123,129,131],sparse_binary_vector:[9,107,110,115],sparse_binary_vector_sequ:[9,115],sparse_binary_vector_sub_sequ:115,sparse_float_vector:[9,110],sparse_float_vector_sequ:9,sparse_non_value_slot:9,sparse_remot:39,sparse_upd:[2,39,107],sparse_value_slot:9,sparse_vector:[107,115],sparse_vector_sequ:115,sparse_vector_sub_sequ:115,sparseparam:100,sparseprefetchrowcpumatrix:100,spatial:4,spatial_scal:4,spec:[121,122,123],specia:10,special:[4,10,18,33,41,48,56,58,64,73,75,76],specialvartypeinfer:64,specif:[10,18,19,25,33,35,36,46,48,51,55,58,75,85,89,93,107,110],specifi:[3,4,10,17,18,22,31,32,39,40,41,44,45,47,48,49,51,52,53,55,60,70,80,83,89,91,96,103,121],spectrogram:92,speech:[4,92],speed:[4,5,56,70,77,78,145],speedup:80,sphinx:[71,103],split:[4,9,19,43,45,46,49,57,63,75,85,96,117,121],split_count:[117,121,123],sport:18,spread:33,sqrt_x:18,squar:[4,6,7,62],square_error_cost:[110,135],squash:98,srand:[115,129],src:[35,69,104,117,119,123],src_backward:140,src_dict:109,src_dict_path:109,src_dict_siz:10,src_embed:[75,140],src_forward:140,src_lang:10,src_primitive_desc:69,src_word_id:[75,140],src_word_vec:75,sreializ:97,srl:10,ssd:4,ssh:[121,124],ssh_server:119,sstabl:31,stabil:[4,18,32,66],stabl:[88,121],stack:[18,53,85,96,121],stackexchang:4,stage:[35,42,49,60,66,69,92,97,98],stale:36,stamp:104,stand:18,standalon:142,standard:[2,9,18,47,57,85,87,92],stanford:[10,32,122],star:35,start:[4,5,10,18,25,33,35,36,39,40,41,43,44,47,48,49,75,77,79,80,98,104,115,122,123,129],start_mpi_train:124,start_op_idx:33,start_paddl:123,start_pass:[128,129],start_po:4,start_pserv:[128,129],startpaddl:123,startup:[18,36,44,57,121],startup_program:18,stat:[129,134],state:[4,5,18,22,34,36,52,53,74,75,80,89,92,96,108,122,139],state_act:[4,5],statem:66,statement:[47,57,61,66,121],statfulset:123,static_cast:[69,102],staticinput:[4,139,140],statist:[4,18,52,80],statmt:10,statset:134,statu:[18,44,75,98,121,122,123,134],status:122,std:[25,35,39,46,62,64,65,69,71,72,80,82,85,86,89,91,93,100,101,102,129],stdbuf:117,stderr:119,stdout:[9,119],step1:18,step:[4,5,7,18,32,34,36,41,47,48,50,52,57,60,61,67,75,78,79,83,85,91,92,96,121,137,139,140,145],step_gradi:33,step_id:96,step_input:96,step_net:34,step_output:96,step_scop:85,stepnet:[34,74,85,89],still:[33,40,43,48,57,66,86],stirng:83,stmt1482205552000:121,stmt1482205746000:121,stochast:[6,36,40,43,78],stop:[4,83],stop_gradi:[18,83],storag:[54,56,121],store:[3,4,10,18,25,32,34,35,39,53,62,64,70,73,75,79,81,83,85,86,87,89,96,121],str:[11,18,25,33,44,96,123,131],straight:[81,84,90],straightforward:69,strategi:[7,36,83,129],stream:[9,48,69,80,93],stream_:93,streamid:22,street:4,strict:84,stride:[4,5,18,69,73,92],stride_h:18,stride_i:4,stride_w:18,stride_x:4,string:[3,4,9,10,11,18,22,25,33,34,40,55,62,65,70,80,81,82,83,85,86,89,90,91,97,100,101,121,129],strip:109,struct:[40,41,54,56,64,65,72,73,80,86,91,95,109],structur:[33,34,40,57,60,70,75,81,83,85,90,121],sts:121,stuff:98,style:[4,85,91],sub:[3,4,9,18,31,33,43,45,50,60,66,74,77,79,83],sub_block:33,sub_nest_seq:4,sub_sequ:[4,110],subclass:83,subcommand:55,subgradi:6,subgraph:[50,60],submiss:48,submit:[69,85,121],subnet0:121,subnet:[31,121],subobjectpath:122,subscript:18,subseq:[136,139],subsequ:[4,77],subsequenceinput:[4,137],subset:18,substitut:18,subtract:32,succ:66,succeed:[40,122],success:[4,41,121,122],successfulcr:122,sucess:66,sucessor:66,sudo:[26,121],suffer:32,suffix:[44,117],suggest:[4,35],suit:145,suitabl:[90,93],sum:[4,6,19,33,34,37,49,64,83],sum_op:33,sum_x:18,sume:18,summar:[18,60,80],summari:80,summat:18,sumopgradmak:86,sumpool:107,supercomput:66,suppli:[18,90],support:[2,3,4,6,7,9,11,18,30,32,34,36,43,44,45,47,48,50,57,59,60,66,69,70,73,75,78,79,80,82,84,85,86,87,90,92,95,121,145],support_inplac:66,suppos:[5,18,19,35,45,59,90,115],suppress:[4,55],sure:[18,121],surpass:4,svs:91,swagger:54,swig:[27,42,71,72],switch_ord:108,switchop:34,sychron:77,symbol:[4,34,62,72],symbols_ready_:34,symbolt:[34,85],symlink:98,sync:[36,78,87],sync_with_cpp:133,syncflag:100,synchron:[36,40,47,69,77,80,121],syntax:[45,53,57,75,84],sysroot:142,system:[34,35,36,41,43,47,48,50,54,59,60,66,68,92,107],t_max:18,t_min:18,tabl:[3,4,18,34,45,57,64,70,90,97],tablelookup:90,tablelookupgrad:90,tablelookupop:90,tableproject:4,tag:[3,10,28,88,98,104,127],tagtyp:3,tail:75,tainer_id:123,take:[3,4,5,9,18,25,31,33,34,35,36,43,45,46,47,48,51,53,56,58,60,61,63,64,66,69,73,78,81,82,83,84,85,86,93,95,96,121],taken:[4,51,62,66,73,96],talk:[41,58],tanh:[4,5,60,75,100,140],tanhactiv:5,tanhshrink:18,tar:[10,11,25,104,110,116,121],tarbal:121,target:[4,10,18,25,33,34,35,51,53,60,62,76,82,85],target_block:[33,51],target_dict_dim:140,target_dict_s:75,target_dictionary_dim:4,target_language_embed:4,target_language_word:140,target_link_librari:35,target_var:17,target_word:75,targetinlink:[4,137],task13:92,task14:92,task1:10,task:[3,4,10,18,48,70,75,80,91,116],task_queu:40,taskentri:40,taskqueu:40,tbd:[42,69,92,137],tcp:[121,129],tear:134,technic:[33,36],techniqu:[18,66],technolog:[18,57],tee:122,tell:[22,36,40,41,75,91],templat:[59,69,91,93,101,102,122,123,145],tempor:[4,18,92],temporari:[33,44,53,66,78,83],tempori:66,tensor:[19,32,35,45,47,49,50,56,57,58,60,62,64,65,69,70,73,74,75,90,96,97,101,135],tensor_arrai:45,tensor_array_read:96,tensor_array_s:96,tensor_array_stack:96,tensor_array_unstack:96,tensor_array_writ:96,tensor_data:70,tensor_in:95,tensor_s:32,tensor_test:35,tensor_to_check:32,tensorarrai:49,tensorarraydesc:96,tensordesc:[70,90],tensorflow:[34,45,47,48,50,57,60,63,87,96],term:[4,5,18,36,86,87,92],termin:122,terminolog:66,tessorarrai:96,test100:10,test10:10,test1:38,test:[4,9,10,11,18,25,31,32,35,62,72,78,84,88,98,100,101,102,112,115,117,124,129,131,134,135],test_:101,test_all_data_in_one_period:122,test_check_grad_ingore_i:101,test_check_grad_ingore_x:101,test_check_grad_norm:101,test_check_output:101,test_compar:104,test_comparespars:104,test_comparetwonet:104,test_comparetwoopt:104,test_config_pars:104,test_data_dir:117,test_fcgrad:100,test_gpuprofil:134,test_layergrad:100,test_list:109,test_mkldnn:68,test_mklpack:67,test_mul_op:[27,101],test_networkcompar:104,test_pass:[128,129,131],test_period:[128,129,131],test_predict:104,test_pydataprovid:104,test_pydataprovider2:104,test_pydataproviderwrapp:104,test_recurrent_machine_gener:104,test_recurrentgradientmachin:[104,137],test_swig_api:104,test_train:104,test_traineronepass:104,test_wait:[128,129],testa:31,testb:31,testbilinearfwdbwd:134,testconfig:100,testfcgrad:100,testfclay:100,testlayergrad:100,testmulop:101,testq:31,testresult:25,testutil:100,text1:55,text:[3,5,9,18,31,70,74,80,92,121],text_fil:9,tflop:134,tftp:145,tgz:[10,30,104],than:[2,3,4,5,18,33,36,44,45,46,51,57,58,59,60,83,85,87,94,96,109,115,121,145],the_step:57,theano:57,thehalf:18,thei:[4,18,31,33,35,36,41,43,45,46,47,50,51,55,57,60,61,65,66,75,76,80,83,85,91,96,97,121,134],them:[3,4,5,11,18,31,32,33,35,36,39,44,46,47,50,51,57,58,59,64,65,66,75,83,84,85,86,89,90,91,95,96,97,121,134],themselv:[33,35],theori:57,therefor:[33,66,78],therein:[4,18,34],theta:60,theta_d:60,theta_g:60,thi:[2,3,4,5,6,9,10,11,17,18,22,23,25,30,31,32,33,34,35,36,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,56,57,58,59,60,61,62,65,66,69,73,74,75,76,77,78,79,80,81,83,84,85,86,87,90,91,92,93,95,96,107,110,115,121,134,135,145],thin:64,thing:[48,60,85,93],think:[31,35],third:[4,18,36,62,115],third_parti:[4,68,104,113,142,143,144],those:[4,34,35,36,59,61,62,63,81],though:[96,145],thought:[18,35],thread:[45,47,49,80,134],thread_count:49,thread_id:80,thread_id_:80,thread_local_rand_use_global_se:[128,129],thread_pool:49,threadblocks:22,threadid:131,threadloc:134,threadpool:45,three:[3,4,10,18,19,32,33,36,47,52,56,57,58,61,69,75,76,79,80,81,84,92,93],threshold:[2,3,4,18,36,40,51,129],thresholdedrelu:18,through:[4,19,33,35,36,40,42,52,66,76,78,103,135],throughout:53,throughput:134,thrust:85,thu:[4,18,43,52,62,66,92,121],tier:122,time:[4,5,7,9,18,22,25,31,32,35,36,40,43,46,47,48,50,51,57,59,64,66,67,73,74,75,77,80,83,84,85,86,90,91,92,96,97,115,122,123,129,133,134,137,145],timelin:[4,80,85],timeo:121,timeout:[36,40],timeout_sec:9,timer:[22,134],timestamp:[4,37],timestep:[4,89],titan:66,titl:[10,123],tls:54,tmp:83,to_chw:11,to_no_sequ:[4,136],to_sequ:[4,136,137],to_tar:25,to_your_paddle_clone_path:103,todo:[3,9,10,34,36,40,43,75,91,92],toend:4,togeth:[4,5,9,25,33,96],token:[3,4,10,18,31,92,140],toler:[25,32],too:[10,32,45,47,51,69,95,96],took:145,tool:[80,103,121,123,142,144],toolchain:142,toolkit:92,top:[3,18,25,74,75,92],top_k:[3,18,75],top_level_rnn:74,topic:69,topk_generated_scor:75,topk_id:75,topk_indic:18,topk_out:18,topk_scor:75,toplevel:26,topolog:[31,36,48,62,66,70,79],topoloi:62,topolopi:25,torch:[34,57],toronto:10,tostr:109,total:[18,22,25,36,50,52,77,80,84,122,134,145],total_pass:84,tottim:133,touch:104,toward:57,trace:[34,58,60],track:[36,40,62,83,98],tractabl:4,tradit:[4,34,56,92],traffic:48,trail:9,train100:10,train10:10,train:[2,3,4,9,10,11,18,28,33,34,38,40,41,43,45,46,51,52,53,57,58,60,61,66,67,70,77,78,79,80,81,82,83,85,87,90,92,93,97,105,107,110,117,119,120,122,123,124,125,126,129,131,140],train_arg:123,train_args_dict:123,train_args_list:123,train_config_dir:[121,123],train_data:117,train_data_dir:117,train_i:110,train_id:121,train_list:[109,117],train_loop:57,train_read:[48,110],train_x:110,trainabl:[4,21,70,83],trainer:[31,37,38,39,40,42,48,50,58,67,68,78,79,85,100,110,117,118,123,129,131],trainer_config:[116,121,122,123],trainer_config_help:100,trainer_count:[107,112,117,121,122,123,128,129,131],trainer_cpu:44,trainer_cr:44,trainer_gpu:44,trainer_id:[117,121,123,129],trainer_intern:39,trainer_mem:44,trainer_packag:44,trainer_prog:48,trainerconfighelp:109,trainerid:[43,123],trainingjob:48,trainonebatch:39,tran:[69,100],trans_var:95,transact:[36,40],transcript:92,transfer:[66,80],transform:[4,5,11,18,85,92],transform_param_attr:5,transformed_st:5,translat:[4,5,10,66,107],translation_id:75,translation_scor:75,transpar:75,transpil:45,transpos:[4,11],transpose_i:18,transpose_x:18,transposedfullmatrixproject:4,travers:[33,61,66],travi:98,treat:[4,18,34,41,66],treatment:[41,56],tree:[4,34,45,53,57,83,123,135,144],trg_dic_siz:75,trg_dict_siz:10,trg_embed:[75,140],trick:75,tricki:71,trigger:[43,46,79],trim:4,trivial:[75,96],true_block:[34,63,81],true_imag:84,true_label:84,true_neg:52,true_posit:52,true_read:84,truth:[3,4,18],tune:[2,92,128,133],tupl:[4,5,9,10,11,18,25,33,83,84],ture:4,turn:[4,18,83,84,139],tutori:[22,121,123,124,125,126],twice:[50,60],twine:88,two:[3,4,5,18,19,31,33,41,42,43,44,45,46,47,48,52,55,56,57,58,60,61,64,66,70,73,75,78,80,81,84,85,86,87,89,90,91,92,95,96,97,101,102,121,134],txt:[35,44,55,67,68,100,103,117,121,124],type:[3,4,5,7,9,10,18,23,31,33,34,36,39,40,43,44,46,48,54,55,56,58,64,65,69,70,71,72,74,75,81,82,83,84,85,86,87,90,91,92,93,95,97,100,101,102,110,112,115,121,122,131,140],type_nam:91,typedef:[41,56,71,72,73,93],typeerror:51,typeid:91,typenam:[59,91,93,101,102],typic:[3,48],ubuntu:[30,88,112],ubyt:84,uci:10,uci_h:112,ufldl:[4,18],uid:122,uint16_t:56,uint32:[54,70],uint32_t:80,uint64:[70,71],uint64_t:[71,115],unawar:41,unbias:32,unblock:47,unbound:66,unbuff:47,unchang:18,unclear:43,uncreat:33,under:[18,35,40,50,77,95,107,110,121],underli:[18,46,75],understand:[57,83,92,145],understand_senti:140,undeterminist:134,uni:92,unidirect:[4,92],unifi:[46,53,62,90],uniform:[2,4,9,18,38,60,83,84],uniform_random:83,uniforminiti:16,uniniti:33,uninstal:[27,104],uniqu:[31,34,36,43,44,69,73,83,89,121],unique_nam:83,unique_name_gener:83,unique_ptr:[86,89,93,100],unit:[4,5,18,19,35,78,80,87,93],unitest:18,unittest:[72,101,104],unix:47,unk:[10,90,97],unknown:[4,10,18,23],unless:[18,107,110],unlik:[4,18,75],unnecessari:[33,92],unnorm:18,unordered_map:89,unpack:96,unrol:74,unscal:18,unseen:87,unseg:4,unsign:[41,56],unstack:96,unstack_from:96,unsupervis:60,unsupport:101,until:[36,41,47,49,50,57,66,89,121,123],untrack:98,unus:18,unzip:142,updat:[2,4,6,18,33,36,40,41,48,54,56,60,74,75,76,77,78,79,89,92,96,98,131,133],update_equ:[25,110],update_hook:2,update_memori:34,update_op:76,updatecallback:100,updatestack:121,upgrad:[30,77,104],upload:[36,44,47,54,88],upon:36,upper:4,upstream:[98,104],uri:121,url:[9,10],usag:[3,4,5,11,18,25,56,63,66,79,83,94,119,123],use:[2,3,4,5,7,9,10,11,18,22,25,31,32,34,35,36,42,46,47,48,49,50,53,56,60,62,64,65,66,69,73,75,76,77,79,80,83,89,90,91,92,95,96,97,98,101,107,110,115,117,121,123,134],use_cpu:65,use_cudnn:[18,19,65],use_eigen_bla:142,use_eigen_for_bla:[142,143],use_etcd:25,use_global_stat:4,use_gpu:[107,110,112,117,122,123,128,129,131],use_mkl_pack:67,use_mkldnn:[4,65,68],use_old_updat:[39,128,129],use_peephol:18,use_sparse_remote_updat:39,used:[3,4,5,6,7,9,10,11,18,22,25,31,32,34,35,36,42,43,46,47,48,51,53,56,57,60,62,66,74,75,78,79,80,83,84,85,87,89,91,93,95,96,121,134],useful:[4,5,32,56,66,83,89,95],usegpu:[100,115],user:[2,4,5,9,10,11,18,22,25,31,32,33,34,35,38,40,43,44,45,48,49,50,51,52,53,55,59,60,61,62,64,65,69,73,75,76,77,78,80,83,84,85,86,87,89,91,93,95,96,121,145],user_id:123,user_info:10,user_nam:38,usercert:38,userinfo:10,userkei:38,usernam:[38,98,142],uses:[4,18,32,36,43,45,47,48,56,66,73,74,75,79,80,93,95,121],using:[2,4,5,9,10,14,18,25,31,32,33,34,35,36,40,41,43,44,47,48,53,55,56,57,59,60,62,64,66,74,76,78,81,83,84,86,87,89,91,92,93,101,112,121],usr:[27,28,104,117,121,123,129],usual:[4,18,25,33,44,66,73,80,81,87,93,121,134],util:[48,67,68,77,116,123,134,145],uuid:[37,43],v7a:142,v8a:142,val:33,valgrind:133,valid:[4,10,11,18,84,85,89,121],valu:[2,3,4,7,9,10,11,18,19,22,25,32,33,34,36,45,47,51,52,62,63,66,68,70,74,75,76,78,79,81,85,89,90,91,94,96,97,100,107,115,121,123,131,142],value1:129,value2:129,value_:90,value_evalu:3,value_rang:9,valueerror:[18,19,62,107],values_:96,vanilla:140,var_nam:[33,95],var_recurs:51,vardesc:[34,61,81,83,85,90],vardescbuild:34,vari:121,variabl:[6,9,10,17,18,19,31,32,34,45,46,47,48,50,51,52,53,58,60,61,62,63,64,73,74,75,76,78,81,82,86,87,90,91,92,95,96,121,122,135],variablenamemap:101,varialbl:60,varianc:[4,18],variant:[4,64,73,93,96],varibal:33,varibl:62,varienc:96,varient:96,variou:[34,47,56,66,87],varproto:91,vars_:[34,89],vartyp:[18,90,97],vartypeinfer:64,vec1:4,vec2:4,vec2seq:92,vec:109,veclib:143,vecter:18,vector:[4,5,9,10,18,31,34,39,41,46,62,63,69,74,75,80,83,85,86,90,92,94,96,113,115],vendor:35,verb:10,verbos:55,veri:[4,7,35,40,45,50,53,57,59,60,66,69,75,79,84,87,89,92,93,96],verifi:34,version:[4,5,10,33,35,44,48,51,55,58,60,62,63,70,75,88,92,98,107,110,121,127,128,129,134,143],versu:31,vertic:4,vgg:5,via:[18,33,36,73,98,121,145],view:[4,70,73],vim:28,virtual:[46,51,64,65,86,93],virtualenv:26,visibl:[43,89],vision:10,visit:[25,33],visual:[4,75],vl16:10,vlog:39,vocabulari:92,volum:[103,122,123],volumemount:[121,122,123],volumn:121,vutbr:10,w_f:18,wai:[3,5,18,31,33,41,43,47,53,57,65,66,75,78,83,84,87,96],wait:[36,41,47,49,123],wang:18,wangkuiyi:35,want:[4,22,31,44,45,47,52,60,65,73,78,80,82,84,87,89,93,94,95,96],warn:[25,55,104,109,123],warp:[4,18],warp_ctc:92,warpctc:4,warranti:[107,110],wast:77,watch:36,wbia:121,websit:10,weight:[3,4,5,6,18,19,21,23,67,70,87,100],weight_act:5,weightlist:100,weights_:100,weights_primitive_desc:69,weights_t:100,welcom:[35,92],well:[18,33,44,47,48,50,57,59,60,87,90,92,121],wer:92,were:[3,18,35,47,57],west:121,wget:[104,142],wgt:69,what:[2,4,35,46,57,60,75,83,91,95,98,145],wheel:30,when:[2,3,4,6,9,18,19,25,33,34,35,36,39,40,41,44,45,48,50,51,52,53,55,56,57,58,62,75,77,78,79,80,81,83,85,93,96,121,134,145],whenev:[18,83,92],where:[4,5,6,18,19,31,33,34,36,43,45,48,57,58,61,73,74,75,78,81,85,87,93,96,135],wherea:[18,34,40,59,63,93,94,97],whether:[3,4,11,18,25,32,33,34,46,53,80,84,90,96,115],which:[2,3,4,5,9,10,11,17,18,19,21,22,25,31,32,33,34,35,36,38,40,41,43,44,45,46,47,48,49,51,53,56,57,58,59,60,62,64,66,69,70,73,74,75,76,77,79,81,82,83,84,85,86,89,90,91,94,95,96,97,110,121,133,145],while_grad:66,while_loop:[75,96],while_op:[18,33],whileloop:96,whileop:34,white:92,whl:[27,30],who:[33,59,61,77,83],whoever:41,whole:[3,9,33,46,60,63,66,71,72,74,77,82,91,92,113,121,145],wholli:46,whose:[4,9,18,32,33,36,43,51,74,85,86,91,96],why:[5,32,72,134],wide:[35,51,60,119,124],width:[3,4,9,11,18,39,71,84,100,101,109,115],wiil:18,wiki:[4,35],wikipedia:[4,10],window:[4,7,10,26,78,92],wirt:62,wise:[4,11,18,50,85,92],with_avx:[27,98,127,142,143],with_bia:91,with_c_api:[27,113,142,143,144],with_doc:27,with_doubl:[27,100,127],with_dso:27,with_golang:[27,113,142],with_gpu:[26,27,98,113,127,142,143],with_mkl:[27,67,68,113,142],with_mkldnn:68,with_mklml:68,with_profil:134,with_python:[27,113,127,142,143],with_rdma:[127,142,143],with_style_check:[27,98],with_swig_pi:[27,113,142,143],with_test:[27,98,101],with_tim:[127,134],within:[4,18,40,48,57,92],without:[3,4,33,36,41,46,47,80,83,84,85,92,107,110],wloop:96,wmt14:140,wmt16_files_mmt:10,wmt_shrinked_data:10,word2vec:[44,107,117,119],word:[3,4,10,33,50,61,64,66,74,75,85,91,92,96,107,139],word_dict:[117,124],word_dim:109,word_id:107,word_idx:10,word_vector_dim:[4,75,140],wordcount:92,words_freq_sort:10,work:[4,9,18,28,31,34,35,36,48,53,56,57,65,76,78,80,83,98,103,121,122,123,145],worker:[50,97,121],workercount:121,workflow:[85,121],workshop:10,workspac:[117,119],worri:32,worth:135,would:[22,25,34,35,36,43,47,48,49,50,57,59,60,61,69,76,78,79,83,84,90,92,96,121,145],wouldn:[57,61],wrap:[18,46,57,59,60,77,145],wrapper:[5,35,46,47,59,77,78,86,96,134],write:[9,18,31,36,43,45,46,47,48,50,56,57,58,59,62,64,69,76,78,83,84,85,86,93,96,107,109,110,121],write_lock:37,write_to_arrai:66,writer:[31,83],written:[18,22,33,34,45,50,53,60,70,78,85,86,90],wrong:84,wrote:62,wsize:121,www:[10,18,107,110],x64:144,x86:[142,143],x86_64:[142,143],x_first_step:18,x_last_step:18,x_neg:32,x_num_col_dim:18,x_po:32,x_reshap:18,x_t:18,x_t_data:18,x_transpos:18,xarg:[3,28,100,104,124],xavier:18,xavieriniti:[16,18],xcode:143,xcodebuild:143,xgbe0:129,xgbe1:129,xmap_read:9,xpu:57,xrang:[32,57,60,80,84,100,110,112],xx_layer:65,xxx:[31,96],xxxx:37,xxxxxxxxx:121,xxxxxxxxxx:121,xxxxxxxxxxxxx:121,xxxxxxxxxxxxxxxxxxx:121,y_dim:60,y_neg:32,y_num_col_dim:18,y_po:32,y_predict:[18,110,112,135],yaml:[35,119,121,122,123,124,145],yancey1989:44,yann:10,yapf:98,year:[10,57],yeild:25,yep:[80,133],yet:[57,92,145],yield:[9,31,38,46,84,107,110],you:[2,4,5,9,10,18,19,22,25,32,44,48,56,89,107,110,121,145],your:[4,9,10,25,31,35,39,44,55,85,104,121,142,143,144,145],your_access_key_id:121,your_param_nam:109,your_repo:123,your_secrete_access_kei:121,your_source_root:72,yuang:57,yuyang18:[9,10],yuyang:133,z_dim:60,z_size:60,zero:[2,4,5,6,9,10,32,33,36,47,60,75,79,83,90,121,129],zip:[10,83,123,142],zlib:113,zone:121,zxf:104,zxvf:121},titles:["API","Activation","Parameter Attribute","Evaluators","Layers","Networks","Optimizer","Pooling","Data Reader Interface and DataSets","Data Reader Interface","Dataset","Image Interface","Fluid","data_feeder","evaluator","executor","initializer","io","layers","nets","optimizer","param_attr","profiler","regularizer","Model Configuration","Training and Inference","\u7528Docker\u7f16\u8bd1\u548c\u6d4b\u8bd5PaddlePaddle","\u4ece\u6e90\u7801\u7f16\u8bd1","\u4f7f\u7528Docker\u5b89\u88c5\u8fd0\u884c","\u5b89\u88c5\u4e0e\u7f16\u8bd1","\u4f7f\u7528pip\u5b89\u88c5","PaddlePaddle Design Doc","Auto Gradient Check Design","Backward Building","Design Doc: Block and Scope","Required CMake Function","Design Doc: Distributed Training","\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\uff08Checkpointing\uff09","\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1","Alalysis of large model distributed training in Paddle","Design Doc: Master Server","Design Doc: The Client Library of Parameter Server","Design Doc: Remote Parameter Updater for Cluster Train","Design Doc: Save Model","Submit a Distributed Training Job","Design Doc: Concurrent Programming with Fluid","C++ Data Feeding","Design Doc: CSP in PaddlePaddle Fluid","Design Doc: Distributed Training Architecture","Design Doc: Execute the Program with Multi CPU","Design Doc: Parameter Server","Error Clip","Evaluator Design","Executor Design Doc","FileManager\u8bbe\u8ba1\u6587\u6863","PFSClient","Design Doc: float16","Design Doc: PaddlePaddle Fluid","PaddlePaddle Fluid: Towards a Compiled Programming Language","Design Doc: Functions, Operators, and Layers","Design for GAN","Design Doc: Computations as a Graph","Survey on Graph","The IfElse Operator","Design Doc: InferVarType","Problem","Memory Optimization","Intel\u00ae MKL Packed on PaddlePaddle: Design Doc","Intel\u00ae MKL-DNN on PaddlePaddle: Design Doc","Design Doc: Add MKLDNN Kernel in Fluid Operator","Design Doc: Model Format","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Design Doc: The Keys of Operator Kernel Type","RNNOp design","Design: Sequence Decoder Generating LoDTensors","Optimizer Design","Design Doc: NCCL support in Paddle Fluid","Averaging Parameter in PaddlePaddle","Design Doc: The C++ Class Parameters","Introduction","Design Doc: PaddlePaddle Programs","Prune","Design Doc: Python API","Python Data Reader Design Doc","Design Doc: Refactorization Overview","Design Doc: Gradient Operators Registration","Regularization in PaddlePaddle","PaddlePaddle\u53d1\u884c\u89c4\u8303","Design of Scope in Paddle","Design Doc: Selected Rows","Interaction between C++ and Python","DeepSpeech2 on PaddlePaddle: Design Doc","Design Doc: Supporting new Device/Library","Design Doc: Switch","Background","Design for TensorArray","Background","\u5982\u4f55\u8d21\u732e\u4ee3\u7801","\u5f00\u53d1\u6807\u51c6","\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42","\u5982\u4f55\u5199\u65b0\u7684Operator","\u5728Paddle\u4e2d\u5982\u4f55\u4f7f\u7528Eigen","\u5982\u4f55\u8d21\u732e\u6587\u6863","\u7f16\u8bd1\u5b89\u88c5\u4e0e\u5355\u5143\u6d4b\u8bd5","\u96c6\u7fa4\u8bad\u7ec3\u4e0e\u9884\u6d4b","FAQ","\u672c\u5730\u8bad\u7ec3\u4e0e\u9884\u6d4b","\u6a21\u578b\u914d\u7f6e","\u53c2\u6570\u8bbe\u7f6e","\u57fa\u672c\u4f7f\u7528\u6982\u5ff5","\u65b0\u624b\u5165\u95e8","\u5feb\u901f\u5f00\u59cb","\u5b89\u88c5\u4e0e\u7f16\u8bd1C-API\u9884\u6d4b\u5e93","C-API\u9884\u6d4b\u5e93","\u8f93\u5165/\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7","C-API\u4f7f\u7528\u6d41\u7a0b","\u542f\u52a8\u53c2\u6570\u8bf4\u660e","\u5206\u5e03\u5f0f\u8bad\u7ec3","\u4f7f\u7528fabric\u542f\u52a8\u96c6\u7fa4\u8bad\u7ec3","\u5728\u4e0d\u540c\u96c6\u7fa4\u4e2d\u8fd0\u884c","Distributed PaddlePaddle Training on AWS with Kubernetes","Kubernetes\u5355\u673a\u8bad\u7ec3","Kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3","\u5728OpenMPI\u96c6\u7fa4\u4e2d\u63d0\u4ea4\u8bad\u7ec3\u4f5c\u4e1a","<no title>","<no title>","\u73af\u5883\u51c6\u5907","\u53c2\u6570\u6982\u8ff0","\u7ec6\u8282\u63cf\u8ff0","\u547d\u4ee4\u884c\u53c2\u6570\u8bbe\u7f6e","\u4f7f\u7528\u6848\u4f8b","\u8fdb\u9636\u4f7f\u7528","Python\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790","GPU\u6027\u80fd\u8c03\u4f18","PaddlePaddle Fluid Source Code Overview","\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684Layer","\u5355\u53cc\u5c42RNN API\u5bf9\u6bd4\u4ecb\u7ecd","RNN\u6a21\u578b","Recurrent Group\u6559\u7a0b","RNN\u914d\u7f6e","PaddlePaddle \u6587\u6863","Android\u5e73\u53f0\u7f16\u8bd1\u6307\u5357","iOS\u5e73\u53f0\u7f16\u8bd1\u6307\u5357","Raspberry Pi\u5e73\u53f0\u7f16\u8bd1\u6307\u5357","Cluster bootstrapping tool survey"],titleterms:{"\u4e00\u4e9b\u7ec6\u8282\u7684\u8865\u5145":123,"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":38,"\u4e0b\u8f7d\u6570\u636e":122,"\u4e0b\u8f7dmklml\u5e93\u5931\u8d25":104,"\u4e0d\u4f7f\u7528":71,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":71,"\u4e0d\u540c\u7684":108,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":71,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":71,"\u4e24\u79cd\u4f7f\u7528":108,"\u4e3a\u4ec0\u4e48\u8981":26,"\u4e3a\u4ec0\u4e48\u9700\u8981\u6027\u80fd\u5206\u6790":134,"\u4ec0\u4e48\u662f\u6027\u80fd\u5206\u6790":134,"\u4ec5\u4ec5\u4f7f\u7528void":71,"\u4ece\u5feb\u7167\u6062\u590d":37,"\u4ece\u6e90\u7801\u7f16\u8bd1":27,"\u4ee3\u7801\u8981\u6c42":98,"\u4f7f\u7528":[98,122],"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":71,"\u4f7f\u7528\u6848\u4f8b":131,"\u4f7f\u7528\u6a21\u578b\u521d\u59cb\u5316\u7f51\u7edc":131,"\u4f7f\u7528\u6d41\u7a0b":116,"\u4f7f\u7528\u73af\u5883\u53d8\u91cf":123,"\u4f7f\u7528\u8f6c\u6362\u5e93":38,"\u4f7f\u7528docker\u542f\u52a8paddlepaddl":28,"\u4f7f\u7528docker\u5b89\u88c5\u8fd0\u884c":28,"\u4f7f\u7528docker\u6267\u884cgpu\u8bad\u7ec3":28,"\u4f7f\u7528docker\u6784\u5efa":103,"\u4f7f\u7528fabric\u542f\u52a8\u96c6\u7fa4\u8bad\u7ec3":119,"\u4f7f\u7528paddlepaddl":103,"\u4f7f\u7528pip\u5b89\u88c5":30,"\u4fdd\u6301\u672c\u5730\u4ed3\u5e93\u6700\u65b0":98,"\u4fee\u6539\u542f\u52a8\u811a\u672c":122,"\u514b\u9686":98,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":72,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":72,"\u5185\u7f6e\u5b9a\u65f6\u5668":134,"\u5199\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":100,"\u51c6\u5907\u4e00\u4e2alinux\u96c6\u7fa4":119,"\u51c6\u5907\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883":[142,143],"\u51c6\u5907\u6570\u636e\u96c6":117,"\u51c6\u5907\u8bad\u7ec3\u6570\u636e":123,"\u51c6\u5907\u8bad\u7ec3\u7a0b\u5e8f":117,"\u51c6\u5907\u9884\u6d4b\u6a21\u578b":116,"\u51c6\u5907openmpi\u96c6\u7fa4":124,"\u51cf\u5c11\u6570\u636e\u8f7d\u5165\u7684\u8017\u65f6":107,"\u51cf\u5c11dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":107,"\u51fa\u73b0":108,"\u5206\u5757\u6587\u4ef6\u4f20\u8f93":54,"\u5206\u5e03\u5f0f\u8bad\u7ec3":118,"\u5206\u652f\u89c4\u8303":88,"\u521b\u5efa\u672c\u5730\u5206\u652f":98,"\u521b\u5efa\u795e\u7ecf\u7f51\u7edc\u8f93\u5165":116,"\u521b\u5efajob":123,"\u521b\u5efapaddlepaddl":122,"\u521d\u59cb\u5316paddlepaddle\u8fd0\u884c\u73af\u5883":116,"\u5220\u9664\u672c\u5730\u5206\u652f":98,"\u5220\u9664\u8fdc\u7a0b\u5206\u652f":98,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":107,"\u5230\u8fdc\u7a0b\u4ed3\u5e93":98,"\u5236\u4f5c\u955c\u50cf":123,"\u5236\u4f5cdocker\u955c\u50cf":122,"\u524d\u5411\u8ba1\u7b97":116,"\u524d\u5411operator\u5355\u6d4b":101,"\u52a0\u8f7d\u6a21\u578b":116,"\u52a0\u8f7dpaddlepaddl":110,"\u52a0\u901f\u6267\u884c":37,"\u52a0\u901f\u8bad\u7ec3\u901f\u5ea6":107,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":71,"\u52a8\u6001\u6269\u5bb9":37,"\u5355\u5143\u6d4b\u8bd5":129,"\u5355\u53cc\u5c42rnn":137,"\u539f\u56e0":71,"\u539f\u56e0\u5217\u8868":71,"\u53c2\u6570\u5185\u5b58":107,"\u53c2\u6570\u670d\u52a1\u5668\u548c\u5206\u5e03\u5f0f\u901a\u4fe1":129,"\u53c2\u6570\u6982\u8ff0":128,"\u53c2\u6570\u8bbe\u7f6e":109,"\u53c2\u8003\u6587\u6863":54,"\u53c2\u8003\u8d44\u6599":134,"\u53cc\u5c42rnn":137,"\u53cc\u5c42rnn\u4ecb\u7ecd":139,"\u53cc\u5c42rnn\u7684\u4f7f\u7528":139,"\u53cd\u5411operator\u5355\u6d4b":101,"\u53d1\u5e03docker\u955c\u50cf":88,"\u53d1\u5e03wheel\u5305\u5230pypi":88,"\u53ef\u80fd\u78b0\u5230\u7684\u95ee\u9898":26,"\u5404\u4e2a\u7248\u672c\u6700\u65b0\u7684whl\u5305":30,"\u540d\u8bcd\u89e3\u91ca":54,"\u5411\u91cf":129,"\u542f\u52a8\u4efb\u52a1":123,"\u542f\u52a8\u53c2\u6570\u670d\u52a1\u5668":117,"\u542f\u52a8\u53c2\u6570\u8bf4\u660e":117,"\u542f\u52a8\u8ba1\u7b97\u8282\u70b9":117,"\u542f\u52a8\u96c6\u7fa4\u4f5c\u4e1a":[119,124],"\u5440":26,"\u547d\u4ee4\u884c\u53c2\u6570\u8bbe\u7f6e":130,"\u548c":136,"\u5728\u4e0d\u540c\u8bbe\u5907\u4e0a\u6307\u5b9a\u5c42":131,"\u5728\u4e0d\u540c\u96c6\u7fa4\u4e2d\u8fd0\u884c":120,"\u5728docker\u4e2d\u6267\u884cpaddlepaddle\u8bad\u7ec3\u7a0b\u5e8f":28,"\u5728openmpi\u96c6\u7fa4\u4e2d\u63d0\u4ea4\u8bad\u7ec3\u4f5c\u4e1a":124,"\u5728paddle\u4e2d\u5982\u4f55\u4f7f\u7528eigen":102,"\u57fa\u4e8edocker\u5bb9\u5668\u7684\u7f16\u8bd1\u65b9\u5f0f":142,"\u57fa\u4e8elinux\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883\u7684\u7f16\u8bd1\u65b9\u5f0f":142,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":[110,115],"\u57fa\u672c\u539f\u7406":139,"\u57fa\u672c\u8981\u6c42":71,"\u5982\u4f55\u4e66\u5199\u6587\u6863":103,"\u5982\u4f55\u4f7f\u7528":108,"\u5982\u4f55\u5171\u4eab\u53c2\u6570":109,"\u5982\u4f55\u5199\u65b0\u7684oper":101,"\u5982\u4f55\u51cf\u5c11\u5185\u5b58\u5360\u7528":107,"\u5982\u4f55\u521d\u59cb\u5316\u53c2\u6570":109,"\u5982\u4f55\u52a0\u8f7d\u9884\u8bad\u7ec3\u53c2\u6570":109,"\u5982\u4f55\u52a0\u901f\u8bad\u7ec3\u901f\u5ea6":107,"\u5982\u4f55\u548c\u660e\u6587\u8fdb\u884c\u76f8\u4e92\u8f6c\u5316":109,"\u5982\u4f55\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u53c2\u6570\u7684\u6743\u91cd\u548c\u68af\u5ea6":107,"\u5982\u4f55\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u67d0\u4e00\u4e2alayer\u7684output":107,"\u5982\u4f55\u6307\u5b9agpu\u8bbe\u5907":107,"\u5982\u4f55\u66f4\u65b0www":103,"\u5982\u4f55\u6784\u5efa\u6587\u6863":103,"\u5982\u4f55\u8bbe\u7f6e\u5b66\u4e60\u7387\u9000\u706b":109,"\u5982\u4f55\u8c03\u7528":107,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":98,"\u5982\u4f55\u8d21\u732e\u6587\u6863":103,"\u5982\u4f55\u8fdb\u884c\u6027\u80fd\u5206\u6790":134,"\u5982\u4f55\u9009\u62e9sgd\u7b97\u6cd5\u7684\u5b66\u4e60\u7387":109,"\u5b50\u5e8f\u5217\u95f4\u65e0memori":137,"\u5b50\u5e8f\u5217\u95f4\u6709memori":137,"\u5b58\u50a8\u7684\u53c2\u6570\u683c\u5f0f\u662f\u4ec0\u4e48":109,"\u5b89\u88c5":30,"\u5b89\u88c5\u4e0e\u7f16\u8bd1":29,"\u5b89\u88c5\u4e0e\u7f16\u8bd1c":113,"\u5b89\u88c5\u4ea4\u53c9\u7f16\u8bd1\u5668":144,"\u5b89\u88c5\u6d41\u7a0b":29,"\u5b9a\u4e49operator\u7c7b":101,"\u5b9a\u4e49opkernel\u7c7b":101,"\u5b9a\u4e49protomaker\u7c7b":101,"\u5b9e\u73b0":71,"\u5b9e\u73b0\u5355\u5143\u6d4b\u8bd5":101,"\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42":100,"\u5b9e\u73b0\u65b9\u5f0f":72,"\u5b9e\u73b0\u8ba1\u7b97":102,"\u5b9e\u73b0c":[100,101],"\u5b9e\u73b0python\u5c01\u88c5":100,"\u5bfb\u627e\u6027\u80fd\u74f6\u9888":133,"\u5bfc\u51fac":71,"\u5c06\u547d\u4ee4\u53c2\u6570\u4f20\u7ed9\u7f51\u7edc\u914d\u7f6e":131,"\u5de5\u5177":134,"\u5e38\u89c1\u95ee\u9898\u548c\u89e3\u51b3\u65b9\u6cd5":30,"\u5e38\u89c1\u95ee\u9898\u89e3\u7b54":29,"\u5e76\u5b8c\u6210":98,"\u5efa\u7acb":98,"\u5f00\u53d1\u6807\u51c6":99,"\u5f00\u59cb\u5f00\u53d1":98,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":129,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":37,"\u5feb\u901f\u4f7f\u7528":112,"\u5feb\u901f\u5b89\u88c5":112,"\u5feb\u901f\u5f00\u59cb":112,"\u6027\u80fd\u5206\u6790\u5c0f\u6280\u5de7":134,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u4ecb\u7ecd":134,"\u6027\u80fd\u8c03\u4f18":129,"\u603b\u4f53\u6d41\u7a0b":26,"\u603b\u7ed3":115,"\u6216\u8005\u662f":104,"\u6267\u884c\u5355\u5143\u6d4b\u8bd5":27,"\u627e\u5230\u7684pythonlibs\u548cpythoninterp\u7248\u672c\u4e0d\u4e00\u81f4":104,"\u62a5importerror":104,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":71,"\u63a5\u53e3\u8f93\u51fa\u591a\u4e2alayer\u7684\u9884\u6d4b\u7ed3\u679c":107,"\u63a8\u5bfc\u65b9\u7a0b":100,"\u63a8\u6d4b\u6267\u884c":37,"\u63d0\u4ea4":98,"\u63d0\u4ea4\u4ee3\u7801\u7684\u4e00\u4e9b\u7ea6\u5b9a":98,"\u63d0\u4ea4\u955c\u50cf":122,"\u642d\u5efa\u795e\u7ecf\u7f51\u7edc":110,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":136,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":38,"\u6570\u636e\u652f\u6301":129,"\u6574\u4f53\u65b9\u6848":123,"\u6587\u4ef6\u4f20\u8f93\u4f18\u5316":54,"\u6587\u4ef6\u8bbf\u95ee\u65b9\u5f0f":38,"\u6587\u4ef6\u8bbf\u95ee\u7684\u6743\u9650":38,"\u6587\u4ef6\u9884\u5904\u7406":38,"\u6587\u6863":141,"\u65b0\u624b\u5165\u95e8":111,"\u65e5\u5fd7\u4e2d\u4fdd\u5b58\u5747\u4e3a\u7f51\u7edc\u901a\u4fe1\u7c7b\u9519\u8bef":105,"\u65f6\u95f4\u5e8f\u5217":137,"\u65f6\u95f4\u6b65":137,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":72,"\u672c\u5730\u6d4b\u8bd5":131,"\u672c\u5730\u8bad\u7ec3":131,"\u672c\u5730\u8bad\u7ec3\u4e0e\u9884\u6d4b":107,"\u672f\u8bed":37,"\u6784\u5efa\u548c\u6d4b\u8bd5":98,"\u6784\u5efapaddlepaddle\u7684android\u5f00\u53d1\u955c\u50cf":142,"\u67b6\u6784\u56fe":54,"\u67e5\u770b\u6027\u80fd\u5206\u6790\u6587\u4ef6":133,"\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":122,"\u67e5\u770b\u8f93\u51fa":123,"\u6846\u67b6\u751f\u6210":54,"\u6848\u4f8b\u4e00":131,"\u6848\u4f8b\u4e8c":131,"\u68c0\u67e5\u6a21\u578b\u8f93\u51fa":119,"\u68c0\u67e5\u96c6\u7fa4\u8bad\u7ec3\u7ed3\u679c":119,"\u6982\u5ff5\u7b80\u4ecb":101,"\u6982\u5ff5\u89e3\u91ca":38,"\u6982\u8ff0":[113,136,139],"\u6a21\u5757":54,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":37,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":72,"\u6a21\u578b\u914d\u7f6e":[108,137],"\u6a21\u578b\u914d\u7f6e\u7684\u6a21\u578b\u914d\u7f6e":137,"\u6ce8\u518coper":101,"\u6ce8\u610f\u4e8b\u9879":[101,116],"\u6d41\u7a0b\u4ecb\u7ecd":38,"\u6d4b\u8bd5":129,"\u6df7\u5408\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790":133,"\u6e05\u7406":116,"\u73af\u5883\u51c6\u5907":127,"\u751f\u6210\u5e8f\u5217":140,"\u751f\u6210\u6027\u80fd\u5206\u6790\u6587\u4ef6":133,"\u751f\u6210\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":139,"\u751f\u6210sparse\u6587\u4ef6":54,"\u7528\u6237\u4f7f\u7528\u6d41\u7a0b":54,"\u7528docker\u7f16\u8bd1\u548c\u6d4b\u8bd5paddlepaddl":26,"\u7684\u533a\u522b":108,"\u7684\u53c2\u6570":108,"\u7684\u65b9\u6cd5\u6709\u4f55\u533a\u522b":108,"\u76ee\u5f55\u7ed3\u6784":72,"\u76ee\u6807":54,"\u76f4\u63a5\u6784\u5efa":103,"\u76f8\u5173\u6982\u5ff5":139,"\u77e9\u9635":129,"\u793a\u4f8b1":137,"\u793a\u4f8b2":137,"\u793a\u4f8b3":137,"\u793a\u4f8b4":137,"\u793a\u4f8b\u7a0b\u5e8f":38,"\u795e\u7ecf\u5143\u6fc0\u6d3b\u5185\u5b58":107,"\u7a00\u758f\u8bad\u7ec3":131,"\u7aef\u6570\u636e\u7c7b\u578b\u8bf4\u660e":115,"\u7b26\u53f7":71,"\u7b80\u5355\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":140,"\u7c7b":[71,100,101],"\u7ebf\u6027\u56de\u5f52\u5b8c\u6574\u793a\u4f8b":110,"\u7ec4\u7ec7\u5e8f\u5217\u4fe1\u606f":115,"\u7ec4\u7ec7\u8f93\u5165\u6570\u636e":[115,116],"\u7ec6\u8282\u63cf\u8ff0":129,"\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":119,"\u7ed1\u5b9apython":101,"\u7f16\u5199\u9884\u6d4b\u4ee3\u7801":116,"\u7f16\u5199yaml\u6587\u4ef6":122,"\u7f16\u8bd1":101,"\u7f16\u8bd1\u4f9d\u8d56":27,"\u7f16\u8bd1\u548c\u5b89\u88c5":[142,143,144],"\u7f16\u8bd1\u548c\u6267\u884c":101,"\u7f16\u8bd1\u5b89\u88c5\u4e0e\u5355\u5143\u6d4b\u8bd5":104,"\u7f16\u8bd1\u5b89\u88c5\u540e\u6267\u884c":104,"\u7f16\u8bd1\u65b9\u6cd5":27,"\u7f16\u8bd1\u6d41\u7a0b":29,"\u7f16\u8bd1\u9009\u9879":[27,72],"\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":27,"\u7f16\u8bd1\u9009\u9879\u8bf4\u660e":27,"\u7f16\u8bd1paddlepaddl":142,"\u7f29\u5bb9":37,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":71,"\u80cc\u666f":71,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":129,"\u83b7\u53d6paddlepaddle\u7684docker\u955c\u50cf":28,"\u8986\u76d6\u4e0d\u4e00\u81f4\u7684\u90e8\u5206":54,"\u8bad\u7ec3":129,"\u8bad\u7ec3\u56e0\u6b64\u9000\u51fa\u600e\u4e48\u529e":107,"\u8bad\u7ec3\u6570\u636e\u5b58\u50a8":38,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":38,"\u8bad\u7ec3\u6a21\u578b":110,"\u8bad\u7ec3\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":139,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0":107,"\u8bcd\u6c47\u8868":137,"\u8be6\u7ec6\u6559\u7a0b":134,"\u8bfb\u53d6\u53cc\u5c42\u5e8f\u5217\u6570\u636e":137,"\u8f6c\u6362\u5e93":38,"\u8f93\u5165":[115,139],"\u8f93\u5165\u4e0d\u7b49\u957f":137,"\u8f93\u5165\u793a\u4f8b":139,"\u8f93\u51fa":139,"\u8f93\u51fa\u6570\u636e":115,"\u8f93\u51fa\u6570\u636e\u7c7b\u578b":115,"\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7":115,"\u8fd0\u884c\u5bb9\u5668":122,"\u8fd0\u884c\u73af\u5883\u4f9d\u8d56":30,"\u8fd0\u884cdocker":104,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":71,"\u8fdb\u884c\u8bad\u7ec3":[38,122],"\u8fdb\u9636\u4f7f\u7528":132,"\u901a\u7528":129,"\u9047\u5230":104,"\u914d\u7f6e\u4ea4\u53c9\u7f16\u8bd1\u53c2\u6570":[142,143,144],"\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":140,"\u914d\u7f6e\u7f51\u7edc":110,"\u94a9\u5b50":98,"\u94fe\u63a5\u8bf4\u660e":113,"\u9519\u8bef\u600e\u4e48\u529e":108,"\u968f\u673a\u6570":129,"\u96c6\u7fa4\u591a\u8282\u70b9\u8bad\u7ec3":105,"\u96c6\u7fa4\u8bad\u7ec3":131,"\u96c6\u7fa4\u8bad\u7ec3\u4e0e\u9884\u6d4b":105,"\u9700\u8981\u7684\u8f6f\u786c\u4ef6":26,"\u975e\u6cd5\u6307\u4ee4":104,"abstract":[48,49,50,77,145],"android\u5e73\u53f0\u7f16\u8bd1\u6307\u5357":142,"api\u4f7f\u7528\u6d41\u7a0b":116,"api\u5bf9\u6bd4\u4ecb\u7ecd":137,"api\u5e93":142,"api\u9884\u6d4b\u5e93":[113,114],"beam_search\u7684\u751f\u6210":137,"book\u4e2d\u6240\u6709\u7ae0\u8282":88,"book\u6559\u7a0b":28,"case":33,"class":[60,79,83],"cmake\u6e90\u7801\u7f16\u8bd1":104,"filemanager\u8bbe\u8ba1\u6587\u6863":54,"final":65,"float":107,"function":[35,59,60,83],"gpu\u548ccpu\u6df7\u5408\u4f7f\u7528":131,"gpu\u6027\u80fd\u8c03\u4f18":134,"gpu\u955c\u50cf\u51fa\u73b0":104,"group\u6559\u7a0b":139,"import":104,"ios\u5e73\u53f0\u7f16\u8bd1\u6307\u5357":143,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":123,"kubernetes\u5355\u673a\u8bad\u7ec3":122,"new":93,"org\u5de5\u5177":103,"paddle\u52a8\u6001\u5e93\u4e2d":71,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":71,"paddle\u7248\u672c\u53f7\u4e3a0":104,"paddlepaddle\u53d1\u884c\u89c4\u8303":88,"paddlepaddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":88,"paddlepaddle\u662f\u5426\u652f\u6301\u7ef4\u6570\u53ef\u53d8\u7684\u6570\u636e\u8f93\u5165":108,"paddlepaddle\u73af\u5883\u4f9d\u8d56":30,"paddlepaddle\u7684softmax\u80fd\u5426\u6307\u5b9a\u8ba1\u7b97\u7684\u7ef4\u5ea6":108,"paddlepaddle\u7f16\u8bd1\u4f9d\u8d56":27,"pi\u5e73\u53f0\u7f16\u8bd1\u6307\u5357":144,"pod\u95f4\u901a\u4fe1":123,"python\u4e0ec":133,"python\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790":133,"python\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86":104,"return":[83,84],"rnn\u6a21\u578b":138,"rnn\u914d\u7f6e":140,"switch":[69,93,94],"tensor\u4f7f\u7528\u6837\u4f8b":102,"tensor\u5230eigentensor\u7684\u8f6c\u6362":102,"tensor\u6a21\u5757":102,"while":18,AWS:121,Abs:1,DNS:121,EFS:121,For:35,KMS:121,The:[34,41,45,53,57,60,61,63,64,73,76,79,85,86,94],Use:[34,81],Using:[35,41],With:44,about:60,abs:18,absolut:75,access:121,account:121,accuraci:[14,18],action:[67,68],activ:[1,4,68],actor:47,adadelta:6,adagrad:[6,20],adam:[6,20],adamax:[6,20],add:[66,69,121],address:121,addto:4,advanc:93,aggreg:4,aggregatelevel:4,alalysi:39,algorithm:[32,36,48,74,82],all:[89,96],analog:45,analysi:[48,66],anneal:109,api:[0,48,67,68,72,76,78,83,87,91],append_regularization_op:23,appendix:145,arbitrari:57,architectur:[48,80],argument:[55,84],arrai:32,array_length:18,array_read:18,array_to_lod_tensor:18,array_writ:18,asset:121,assign:18,associ:[89,121],assumpt:145,async:129,attent:140,attribut:[2,66,87],auc:3,auto:32,averag:78,avg:7,aws:121,background:[32,50,67,93,94,95,96,97],backward:[33,57,61,85],base:[44,75],basepool:7,basic:[66,93,145],batch:84,batch_norm:[4,18],batch_siz:84,beam:[75,92],beam_search:[4,18],beam_search_decod:18,becaus:109,benchmark:[67,68],benefit:[50,85],between:[31,47,83,85,91,93],bidirectional_gru:5,bidirectional_lstm:5,big:109,bilinear_interp:4,binari:34,bla:27,block:[34,58,60,61,81,83,85],block_expand:4,blockdesc:81,blockguard:18,blockguardserv:18,blockguardwithcomplet:18,bootstrap:145,brelu:[1,18],bring:145,bucket:121,build:[33,60,85],can:89,capi:72,capi_priv:72,cast:18,ceil:18,challeng:[33,50,82],chang:75,channel:47,check:[4,32],checkpoint:[36,37,43],choic:65,choos:[35,121],chunk:3,chunk_ev:18,chunkevalu:14,cifar:10,classif:3,classification_error:3,classification_error_print:3,client:41,clip:[4,18,51],clip_by_norm:18,clone:98,close:32,cloudform:121,cluster:[42,121,145],cmake:[35,67,68],code:[44,58,83,135],column_sum:3,commit:98,compar:145,comparis:83,compat:57,compil:[34,56,58,81,85,135],complet:57,compos:84,comput:[34,61,69,85,87],con:145,concat:[4,18],concept:[83,85,121],concern:68,conclus:[43,62,145],concurr:[45,47],condit:60,conditionalblock:18,configur:[24,121],conll05:10,connect:4,constant:16,construct:61,content:[67,68,72,92,104,105,107,108,109,121,134,136],context_project:4,control:[66,85],control_flow:18,contruct:66,conv2d:18,conv2d_transpos:18,conv:4,conv_oper:4,conv_project:4,conv_shift:4,convert:43,convolut:92,core:[32,83,121],corner:33,cos_sim:[4,18],cost:4,cpu:49,creat:[33,47,84,85,89,121],create_arrai:18,create_global_var:18,create_paramet:18,create_tensor:18,createreaderop:46,creation:[40,78,87],creator:84,credenti:121,crf:4,crf_decod:[4,18],cross_channel_norm:4,cross_entropi:18,cross_entropy_cost:4,cross_entropy_with_selfnorm_cost:4,csp:47,ctc:[4,92],ctc_error:3,ctc_greedy_decod:18,cuda:[27,56,104],cuda_profil:22,cudnn:27,cudnnavg:7,cudnnmax:7,current:[56,86],custom:84,data:[4,8,9,18,36,46,48,84,95,121],data_feed:13,datafeed:[9,13],dataflow:66,dataprovid:129,dataset:[8,10,36,40],datatyp:[9,73],decayedadagrad:[6,20],decod:75,decor:84,decoratedread:46,deep:[34,57],deepspeech2:92,defin:121,definit:97,delet:121,demo:[60,121],dens:43,depend:[60,92],deploi:44,describ:[57,76],descript:[55,85],design:[31,32,34,36,40,41,42,43,45,47,48,49,50,52,53,56,57,59,60,61,64,67,68,69,70,73,74,75,76,77,79,81,83,84,85,86,89,90,92,93,94,96],destroi:[89,121],detail:[39,92],detect:3,detection_map:3,detection_output:4,develop:85,devic:[18,93],devicecontext:93,dictionari:84,differ:[85,93],directori:121,discrimin:60,discuss:[50,60],dispatch:[36,40],distribut:[31,36,39,44,48,50,121],dnn:68,doc:[31,34,36,40,41,42,43,45,47,48,49,50,53,56,57,59,61,64,67,68,69,70,73,77,79,81,83,84,85,86,90,92,93,94],docker:[26,44],doe:84,dot_prod:4,dot_product_attent:5,dotmul_oper:4,dotmul_project:4,down:121,download:121,driver:104,drop_out:108,dropout:[4,18],duplic:108,dure:[75,84],dylib:72,dynam:[36,96],dynamic_gru:18,dynamic_lstm:18,dynamic_lstmp:18,dynamicrnn:18,dynet:62,ec2:121,edit_dist:18,eigen:102,elast:121,elect:43,elementwise_add:18,elementwise_div:18,elementwise_max:18,elementwise_min:18,elementwise_mul:18,elementwise_pow:18,elementwise_sub:18,els:34,elu:18,embed:[4,18],engin:60,enough:32,entri:84,environ:44,eos:4,error:51,evalu:[3,14,52],event:[25,31,80],evolut:57,exampl:[31,35,45,47,63,72],except:107,execut:[34,49,57,81,85],executor:[15,53],exp:[1,18],expand:[4,136],expandlevel:4,explan:32,extern:121,factorization_machin:4,faq:106,fault:36,feed:46,file:[34,121],fileread:46,fill_const:18,fill_constant_batch_size_lik:18,find:121,first_seq:[4,136],float16:56,floor:18,flow:66,fluid:[12,45,47,57,58,69,77,135],fork:98,format:[34,36,70],forward:61,frame:34,framework:[32,102],from:[31,43,91],full_matrix_project:4,fulli:4,functor:93,futur:[57,92],gan:60,gate:140,gated_unit:4,gener:[58,60,75,145],get_inference_program:17,get_output:4,get_plac:18,give:84,global:[81,83],global_scop:15,glu:19,gpu:129,grad_op:33,gradient:[32,33,41,68,86],gradient_print:3,graph:[61,62,66,85,87],group:[4,121],gru:[5,129],gru_group:5,gru_step:4,gru_unit:[5,18],grumemori:4,handler:[31,71],happen:43,hard_shrink:18,hard_sigmoid:18,hardwar:56,helper:83,hierarchi:34,high:[76,78,87,91],how:[32,39,78,84,85,93],hsigmoid:4,huber_classification_cost:4,huber_regression_cost:4,iam:121,ident:1,identity_project:4,ifels:[18,63],ifelseop:34,illeg:104,im2sequ:18,imag:[4,5,11,44],imdb:10,img_cmrnorm:4,img_conv:4,img_conv_bn_pool:5,img_conv_group:5,img_pool:4,imikolov:10,implement:[32,33,35,39,49,51,52,56,70,74,77,78,83,84,85,86,87],increment:18,infer:[25,107],infershap:[81,90],infervartyp:64,ingredi:31,ingress:54,initi:[16,41,60,121],insid:89,inspect:121,instal:[121,145],instanc:121,instead:84,instruct:104,insuffici:104,integr:[93,121],intel:[67,68],interact:91,interfac:[8,9,11,32,36,41,42,53,76,84,89],intermedi:85,interpol:4,introduc:[75,96],introduct:[80,87],isn:84,issu:[56,98],job:[36,44,121,122],kei:[67,73,121],kernel:[69,73,85],kmax_sequence_scor:4,kube:121,kubectl:121,kubernet:[44,121,122],l1decai:23,l2_distanc:4,l2_normal:18,l2decai:23,lambda_cost:4,languag:[34,58],larg:39,last_seq:[4,136],layer:[4,18,31,59,67,68,83,108],layer_norm:18,layout:73,leaky_relu:18,learn:[34,57,109],less_than:18,leval:91,level:[76,78,87,91],libpaddle_capi_shar:72,libpaddle_capi_whol:72,librari:[41,56,73,85,93],limit:48,linear:1,linear_chain_crf:18,linear_comb:4,list:[37,84],listenandserv:18,live:66,load:47,load_inference_model:17,load_param:17,load_persist:17,load_var:17,local:[48,89,121],lod:75,lod_rank_t:18,lod_tensor_to_arrai:18,lodtensor:[74,75,96],lodtensordesc:97,log:[1,18],logic:40,logsigmoid:18,low:[78,87,91],lstm:[5,129],lstm_step:4,lstm_unit:18,lstmemori:4,lstmemory_group:5,lstmemory_unit:5,machin:75,macro:85,main:60,make:66,manag:35,map:[84,85],master:[36,40,44,45],math:[4,93],mathemat:32,matmul:18,matrix:68,max:7,max_sequence_len:18,maxframe_print:3,maxid:4,maxid_print:3,maxout:4,mean:18,member:60,memori:[4,66,74,93,108,137,139],merge_lod_tensor:18,messag:[91,109],method:75,might:60,migrat:85,mileston:85,mini:84,minibatch:[9,47],mix:4,mkl:[67,68],mkldnn:69,mkldnn_helper:69,mkldnndevicecontext:69,mnist:10,model:[24,31,39,41,43,47,57,60,70,75,140],modul:[85,93,104],momentum:[6,20],more:60,motiv:[33,47,53,70,77,82],movielen:10,mul:18,multi:[49,58],multi_binary_label_cross_entropy_cost:4,multibox_loss:4,multipl:84,multiplex:[4,18],mxnet:62,name:[89,104,108,121],nativ:58,nccl:77,nce:[4,18],necess:83,necessari:85,need:84,nest:74,net:19,network:[5,85,140],neural:140,nlp:[5,129],norm:[4,87],normal:16,note:32,numer:32,numpi:32,nvprof:134,nvvp:134,object:36,offset:75,ones:18,onli:[84,89],op_mak:85,oper:[46,59,63,66,69,73,78,81,83,85,86,90,96],opinfomap:85,opkernel:[85,93],opproto:91,ops:[18,87],optim:[6,20,36,41,61,66,76,83],option:55,opwithkernel:85,order:55,org:103,origin:85,orthogon:89,other:68,out_prod:4,output:121,overview:[43,51,53,67,68,85,89,92,135],pack:[67,75],packag:35,pad:4,paddl:[39,77,84,89,102,104,108],paddlejob:44,paddlepaddl:[31,34,47,57,58,67,68,78,81,87,88,92,103,104,121,135,141],pair:121,paradigm:57,parallel_nn:131,paralleldo:18,param_attr:21,paramattr:21,paramet:[2,25,31,36,41,42,44,47,50,68,78,79,83,87,121],parameteraverageoptim:78,parent:89,part:61,partit:41,path:[43,55],penalti:87,perform:[78,129],persist:40,pfsclient:[54,55],pfsserver:54,place:[66,73,93],placement:48,platform:104,pnpair:3,point:[67,107,121],polici:66,pool2d:18,pool:[4,7,136],pose:[64,86],potenti:65,pow:18,power:4,pre:98,precision_recal:3,prefetch:84,prelu:4,prepar:121,principl:69,print:[3,18],privat:121,pro:145,problem:[52,64,65,66,73,76,86,95],procedur:145,process:[36,41,44,76,85],profil:22,program:[34,45,47,49,57,58,81,83],programdesc:[58,81],project:35,propos:[64,86,87],protobuf:90,protocol:109,provid:84,prune:82,pserver:43,pull:98,push:98,python:[32,44,48,67,68,74,76,78,83,84,87,91,97,115],qualiti:85,queue:[36,40],rank:3,rank_cost:4,raspberri:144,rate:109,reader:[8,9,31,46,84],readerbas:46,readerhold:46,readop:46,realiz:85,reciproc:18,recoveri:36,recurr:[4,5,108,139,140],recurrent_group:4,recv:47,reduce_max:18,reduce_mean:18,reduce_min:18,reduce_sum:18,refactor:85,refer:[32,48,50,66,67,68,92],region:121,regist:[64,85,91],registr:[85,86],registri:85,regular:[23,41,87],reject:109,rel:75,relat:[46,85,96],relu6:18,relu:[1,18],remot:42,remoteexecutor:48,render:121,reorder_lod_tensor_by_rank:18,repeat:4,represent:[34,85],request:98,requir:[35,60],reset_profil:22,reshap:[4,18],resiz:4,retri:40,reus:83,rmsprop:6,rnn:[74,96,129,137],rnnop:[34,74,85],roi_pool:4,rotat:4,round:18,route53:121,row:[90,92],row_conv:[4,18],row_l2_norm:4,rpc:47,run:[53,135],runtim:44,sampl:4,sampling_id:4,save:43,save_inference_model:17,save_param:17,save_persist:17,save_var:17,scale:[4,18,36],scale_shift:4,scaled_dot_product_attent:19,scaling_project:4,scope:[34,74,85,89],scope_guard:15,search:[75,92],secur:121,select:[41,47,90],selectedrow:90,selective_fc:4,semant:94,send:[18,47],sentiment:10,separ:85,seq_concat:4,seq_reshap:4,seq_slic:4,seqtext_print:3,sequenc:[75,140],sequence_conv:18,sequence_conv_pool:[5,19],sequence_expand:18,sequence_first_step:18,sequence_last_step:18,sequence_pool:18,sequence_reshap:18,sequence_softmax:18,sequencesoftmax:1,server:[36,40,41,44,47,50,121],servic:121,setup:121,sextant:145,sgd:[20,129],shape:75,share:[31,33,66,89],should:89,shrink_memori:18,shuffl:84,sigmoid:[1,18],sigmoid_cross_entropy_with_logit:18,simpl:75,simple_attent:5,simple_gru2:5,simple_gru:5,simple_img_conv_pool:[5,19],simple_lstm:5,singl:84,slice_project:4,slope_intercept:4,small_vgg:5,smooth_l1_cost:4,soft_relu:18,softmax:1,softplu:18,softrelu:1,softshrink:18,softsign:[1,18],solut:[64,65,66,67,73,82,86,95],sourc:135,spars:[41,42,43,90],split:18,split_lod_tensor:18,spp:4,sqrt:18,squar:[1,18],square_error_cost:[4,18],squarerootn:7,stack:34,stanh:[1,18],start:[31,121],statement:52,staticrnn:18,staticrnnmemorylink:18,step2:116,step:[74,116],storag:87,store:36,strategi:66,sub_nested_seq:4,sub_seq:4,subcommond:55,submit:44,suffici:84,suitabl:35,sulut:69,sum:[3,7,18],sum_cost:4,sum_to_one_norm:4,summar:[31,45],summari:70,support:[56,77,93,96,104],survei:[56,62,87,145],swish:18,switch_scop:15,synopsi:55,syntax:47,system:[57,121],tabl:[72,92],table_project:4,tanh:[1,18],tanh_shrink:18,task:[36,40,92],tear:121,tecton:145,templat:121,tensor:[4,18,85,93,102],tensorarrai:[75,96],tensordesc:97,tensorflow:62,test:[67,68,69],text_conv_pool:5,theori:32,thi:[89,104],think:60,three:96,thresholded_relu:18,time:135,timelin:43,todo:[37,38,49],togeth:89,toler:36,too:109,tool:[35,145],topic:93,topk:18,toward:58,train:[25,31,36,39,42,44,48,76,84,121],trainer:[25,36,41,43,44,47,121],tran:4,trans_full_matrix_project:4,transform:95,translat:75,transpil:[48,49,50,58,66,77],transpos:18,tune:129,ture:57,two:32,type:[47,73],uci_h:10,uniform:[16,96],unit:[67,68,69],unpack:75,updat:[31,42,43,121],usag:[33,51,74,75,84],use:[39,84],user:36,util:3,valu:83,value_print:3,vardesc:97,variabl:[33,66,83,85,89,97],verifi:121,version:[45,56,104],vgg_16_network:5,volum:121,vpc:121,warp_ctc:4,warpctc:18,weightnormparamattr:21,what:[39,43],wheel:104,when:[43,89],whileguard:18,whl:104,why:[56,57,78,84,85,96],wmt14:10,wmt16:10,work:92,worker:45,xavier:16,zero:18}})
\ No newline at end of file
+Search.setIndex({docnames:["build_and_install/build_cn","build_and_install/build_from_source_cn","build_and_install/docker_install_cn","build_and_install/index_cn","build_and_install/pip_install_cn","design/api","design/auto_gradient_check","design/backward","design/block","design/build_system/README","design/cluster_train/README","design/cluster_train/checkpointing","design/cluster_train/data_dispatch","design/cluster_train/large_model_dist_train","design/cluster_train/master_server","design/cluster_train/pserver_client","design/cluster_train/remote_parameter_updater","design/cluster_train/save_model","design/cluster_train/submit-job","design/concurrent_programming","design/cpp_data_feeding","design/csp","design/dist_refactor/distributed_architecture","design/dist_refactor/multi_cpu","design/dist_refactor/parameter_server","design/error_clip","design/evaluator","design/executor","design/file_manager/README","design/file_manager/pfs/pfsclient","design/float16","design/fluid","design/fluid_compiler","design/functions_operators_layers","design/gan_api","design/graph","design/graph_survey","design/if_else_op","design/infer_var_type","design/kernel_hint_design","design/memory_optimization","design/mkl/mkl_packed","design/mkl/mkldnn","design/mkl/mkldnn_fluid","design/model_format","design/multi_language_interface/00.why_plain_c","design/multi_language_interface/01.inference_implementation","design/operator_kernel_type","design/ops/rnn","design/ops/sequence_decoder","design/optimizer","design/paddle_nccl","design/parameter_average","design/parameters_in_cpp","design/profiler","design/program","design/prune","design/python_api","design/reader/README","design/refactorization","design/register_grad_op","design/regularization","design/releasing_process","design/scope","design/selected_rows","design/simple_op_design","design/speech/deep_speech_2","design/support_new_device","design/switch","design/switch_kernel","design/tensor_array","design/var_desc","dev/contribute_to_paddle_cn","dev/index_cn","dev/new_layer_cn","dev/new_op_cn","dev/use_eigen_cn","dev/write_docs_cn","faq/build_and_install/index_cn","faq/cluster/index_cn","faq/index_cn","faq/local/index_cn","faq/model/index_cn","faq/parameter/index_cn","getstarted/concepts/use_concepts_cn","getstarted/index_cn","getstarted/quickstart_cn","howto/capi/compile_paddle_lib_cn","howto/capi/index_cn","howto/capi/organization_of_the_inputs_cn","howto/capi/workflow_of_capi_cn","howto/cluster/cmd_argument_cn","howto/cluster/index_cn","howto/cluster/multi_cluster/fabric_cn","howto/cluster/multi_cluster/index_cn","howto/cluster/multi_cluster/k8s_aws_cn","howto/cluster/multi_cluster/k8s_cn","howto/cluster/multi_cluster/k8s_distributed_cn","howto/cluster/multi_cluster/openmpi_cn","howto/cluster/multi_cluster/src/k8s_data/README","howto/cluster/multi_cluster/src/k8s_train/README","howto/cluster/preparations_cn","howto/cmd_parameter/arguments_cn","howto/cmd_parameter/detail_introduction_cn","howto/cmd_parameter/index_cn","howto/cmd_parameter/use_case_cn","howto/index_cn","howto/optimization/cpu_profiling_cn","howto/optimization/gpu_profiling_cn","howto/read_source","howto/rnn/hierarchical_layer_cn","howto/rnn/hrnn_rnn_api_compare_cn","howto/rnn/index_cn","howto/rnn/recurrent_group_cn","howto/rnn/rnn_config_cn","index_cn","mobile/cross_compiling_for_android_cn","mobile/cross_compiling_for_ios_cn","mobile/cross_compiling_for_raspberry_cn","survey/cluster_bootstrapping_tools"],envversion:50,filenames:["build_and_install/build_cn.md","build_and_install/build_from_source_cn.rst","build_and_install/docker_install_cn.rst","build_and_install/index_cn.rst","build_and_install/pip_install_cn.rst","design/api.md","design/auto_gradient_check.md","design/backward.md","design/block.md","design/build_system/README.md","design/cluster_train/README.md","design/cluster_train/checkpointing.md","design/cluster_train/data_dispatch.md","design/cluster_train/large_model_dist_train.md","design/cluster_train/master_server.md","design/cluster_train/pserver_client.md","design/cluster_train/remote_parameter_updater.md","design/cluster_train/save_model.md","design/cluster_train/submit-job.md","design/concurrent_programming.md","design/cpp_data_feeding.md","design/csp.md","design/dist_refactor/distributed_architecture.md","design/dist_refactor/multi_cpu.md","design/dist_refactor/parameter_server.md","design/error_clip.md","design/evaluator.md","design/executor.md","design/file_manager/README.md","design/file_manager/pfs/pfsclient.md","design/float16.md","design/fluid.md","design/fluid_compiler.md","design/functions_operators_layers.md","design/gan_api.md","design/graph.md","design/graph_survey.md","design/if_else_op.md","design/infer_var_type.md","design/kernel_hint_design.md","design/memory_optimization.md","design/mkl/mkl_packed.md","design/mkl/mkldnn.md","design/mkl/mkldnn_fluid.md","design/model_format.md","design/multi_language_interface/00.why_plain_c.md","design/multi_language_interface/01.inference_implementation.md","design/operator_kernel_type.md","design/ops/rnn.md","design/ops/sequence_decoder.md","design/optimizer.md","design/paddle_nccl.md","design/parameter_average.md","design/parameters_in_cpp.md","design/profiler.md","design/program.md","design/prune.md","design/python_api.md","design/reader/README.md","design/refactorization.md","design/register_grad_op.md","design/regularization.md","design/releasing_process.md","design/scope.md","design/selected_rows.md","design/simple_op_design.md","design/speech/deep_speech_2.md","design/support_new_device.md","design/switch.md","design/switch_kernel.md","design/tensor_array.md","design/var_desc.md","dev/contribute_to_paddle_cn.md","dev/index_cn.rst","dev/new_layer_cn.rst","dev/new_op_cn.md","dev/use_eigen_cn.md","dev/write_docs_cn.rst","faq/build_and_install/index_cn.rst","faq/cluster/index_cn.rst","faq/index_cn.rst","faq/local/index_cn.rst","faq/model/index_cn.rst","faq/parameter/index_cn.rst","getstarted/concepts/use_concepts_cn.rst","getstarted/index_cn.rst","getstarted/quickstart_cn.rst","howto/capi/compile_paddle_lib_cn.md","howto/capi/index_cn.rst","howto/capi/organization_of_the_inputs_cn.md","howto/capi/workflow_of_capi_cn.md","howto/cluster/cmd_argument_cn.md","howto/cluster/index_cn.rst","howto/cluster/multi_cluster/fabric_cn.md","howto/cluster/multi_cluster/index_cn.rst","howto/cluster/multi_cluster/k8s_aws_cn.md","howto/cluster/multi_cluster/k8s_cn.md","howto/cluster/multi_cluster/k8s_distributed_cn.md","howto/cluster/multi_cluster/openmpi_cn.md","howto/cluster/multi_cluster/src/k8s_data/README.md","howto/cluster/multi_cluster/src/k8s_train/README.md","howto/cluster/preparations_cn.md","howto/cmd_parameter/arguments_cn.md","howto/cmd_parameter/detail_introduction_cn.md","howto/cmd_parameter/index_cn.rst","howto/cmd_parameter/use_case_cn.md","howto/index_cn.rst","howto/optimization/cpu_profiling_cn.md","howto/optimization/gpu_profiling_cn.rst","howto/read_source.md","howto/rnn/hierarchical_layer_cn.rst","howto/rnn/hrnn_rnn_api_compare_cn.rst","howto/rnn/index_cn.rst","howto/rnn/recurrent_group_cn.md","howto/rnn/rnn_config_cn.rst","index_cn.rst","mobile/cross_compiling_for_android_cn.md","mobile/cross_compiling_for_ios_cn.md","mobile/cross_compiling_for_raspberry_cn.md","survey/cluster_bootstrapping_tools.md"],objects:{},objnames:{},objtypes:{},terms:{"00m":108,"01org":78,"03m":108,"0424m":108,"04\u4ee5\u4e0a":4,"04\u4ee5\u53camaco":86,"055ee37d":95,"0630u":108,"06u":108,"0810u":108,"0957m":108,"0\u53f7\u8bad\u7ec3\u8282\u70b9\u662f\u4e3b\u8bad\u7ec3\u8282\u70b9":103,"0\u5c42\u5e8f\u5217":110,"0_cudnn5":1,"0_cudnn5_avx_mkl":[2,4],"0_cudnn7_avx_mkl":4,"0rc1":62,"0rc2":62,"0x10f256d50":36,"0x7ffe4de00110":36,"100gi":95,"100m":81,"10g":18,"1150u":108,"11\u5b9e\u73b0\u4e86c":46,"11e6":96,"124n":108,"12\u4ee5\u4e0a":4,"12\u64cd\u4f5c\u7cfb\u7edf":78,"12gb":40,"13m":96,"1490u":108,"14\u7248\u672c\u4ee5\u4e0a\u7684":118,"14\u8fd9\u79cd\u5199\u6cd5\u5c06\u4f1a\u6d4b\u8bd5\u6a21\u578b":105,"1550u":108,"15\u884c":111,"16\u5b57\u8282\u8868\u793a\u4fdd\u5b58\u7684\u53c2\u6570\u603b\u4e2a\u6570":83,"16u":108,"173n":108,"1770u":108,"18ad":95,"18e457ce3d362ff5f3febf8e7f85ffec852f70f3b629add10aed84f930a68750":96,"197u":108,"1\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2a\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":1,"1\u7684\u5c42\u4e4b\u5916":105,"1\u7a00\u758f\u6570\u636e":74,"1\u8f6e\u5b58\u50a8\u7684\u6240\u6709\u6a21\u578b":105,"210u":108,"215n":108,"228u":108,"2520u":108,"2680u":108,"26\u884c":111,"279n":108,"27m":108,"285m":108,"2863m":108,"28m":108,"2977m":108,"2\u4e09\u7c7b\u7684\u6bd4\u4f8b\u4e3a":83,"2\u4e2a\u5b50\u5e8f\u5217":89,"2\u5206\u522b\u4ee3\u88683\u4e2a\u8282\u70b9\u7684trainer":97,"2\u610f\u5473\u77400\u53f7\u548c1\u53f7gpu\u5c06\u4f1a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97fc1\u548cfc2\u5c42":105,"2\u8fd9\u51e0\u4e2a\u76ee\u5f55\u8868\u793apaddlepaddle\u8282\u70b9\u4e0etrain":97,"2cbf7385":95,"302n":108,"30u":108,"328n":108,"32u":108,"331n":108,"3320u":108,"365e":95,"36u":108,"3710m":108,"3768m":108,"387u":108,"38u":108,"3920u":108,"39u":108,"3\u4ee5\u4e0a\u7684\u7b26\u53f7":4,"3\u53f7gpu":81,"4035m":108,"4090u":108,"4096mb":103,"4279m":108,"43u":108,"448a5b355b84":96,"4560u":108,"4563m":108,"45u":108,"4650u":108,"4726m":108,"473m":96,"4\u4e2a\u5e8f\u5217\u7684\u957f\u5ea6\u5206\u522b\u4e3a":89,"4\u5b57\u8282\u8868\u793apaddlepaddle\u7248\u672c\u4fe1\u606f":83,"4gb":103,"500m":81,"50bd":95,"50gi":95,"514u":108,"525n":108,"526u":108,"536u":108,"5460u":108,"5470u":108,"54u":108,"5690m":108,"573u":108,"578n":108,"5798m":108,"586u":108,"58s":96,"5969m":108,"5\u4f5c\u4e3a\u7f16\u8bd1\u73af\u5883":4,"5\u5373\u5c06\u505c\u6b62\u7ef4\u62a4":4,"5_cudnn5_avx_mkl":4,"5_cudnn5_avx_openbla":[4,86],"6080u":108,"6140u":108,"6305m":108,"639u":108,"64\u5e73\u53f0\u4e3a\u4f8b":116,"64m":44,"655u":108,"6780u":108,"6810u":108,"682u":108,"6970u":108,"6\u4e07\u4ebf\u6b21\u6d6e\u70b9\u8fd0\u7b97\u6bcf\u79d2":108,"6\u4ee5\u4e0a":[4,86],"6\u4f5c\u4e3a\u6807\u51c6\u7f16\u8bd1\u73af\u5883":4,"6ce9":95,"704u":108,"7090u":108,"72u":108,"73u":108,"75u":108,"760u":108,"767u":108,"783n":108,"784u":108,"78m":108,"7\u4ee5\u4e0a":116,"7\u4ee5\u4e0a\u7684\u7b26\u53f7":4,"7\u4ee5\u4e0b":116,"7\u548cpip":78,"7\u7248\u672c\u5f00\u59cb":116,"7\u7cfb\u5217":4,"7kb":96,"8000\u5c31\u53ef\u4ee5\u5728\u7f51\u9875\u4e0a\u751f\u6210\u9700\u8981\u7684\u6587\u6863":77,"8250u":108,"8300u":108,"830n":108,"849m":108,"861u":108,"8661m":108,"892m":108,"8\u5b57\u8282\u8868\u793a\u6bcf\u4e2a\u53c2\u6570\u5360\u7528\u7684\u5b57\u8282\u6570":83,"901n":108,"90u":108,"918u":108,"9247m":108,"924n":108,"9261m":108,"9330m":108,"94u":108,"9530m":108,"983m":108,"988u":108,"997u":108,"99u":108,"9a235":117,"9f18":96,"\u4e00\u4e2a":89,"\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a0\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u4e0d\u5171\u4eab\u7684\u4f8b\u5b50\u662f":75,"\u4e00\u4e2a\u5178\u578b\u7684chunk\u5982\u4e0b\u6240\u793a":28,"\u4e00\u4e2a\u5206\u5e03\u5f0fpaddlepaddle\u8bad\u7ec3\u4efb\u52a1\u4e2d":96,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u6269\u5c55\u6210\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8fdb\u5165":113,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8fdb\u5165":113,"\u4e00\u4e2a\u53cc\u5c42rnn\u7531\u591a\u4e2a\u5355\u5c42rnn\u7ec4\u6210":113,"\u4e00\u4e2a\u53ef\u8c03\u7528\u7684\u51fd\u6570":113,"\u4e00\u4e2a\u5e38\u7528\u7684cmake\u914d\u7f6e\u5982\u4e0b":118,"\u4e00\u4e2a\u6570\u636e\u96c6\u5927\u90e8\u5206\u5e8f\u5217\u957f\u5ea6\u662f100":81,"\u4e00\u4e2a\u662f\u6d6e\u70b9\u8ba1\u7b97\u91cf":108,"\u4e00\u4e2a\u72ec\u7acb\u7684\u5143\u7d20":110,"\u4e00\u4e2a\u72ec\u7acb\u7684\u8bcd\u8bed":110,"\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u6a21\u578b\u7531\u5927\u91cf\u7684\u53c2\u6570\u7ec4\u6210":11,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad\u90e8\u5206\u628a\u8f93\u5165\u8f6c\u5316\u4e3a\u76f8\u5e94\u7684\u8f93\u51fa":74,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684\u53c2\u6570\u662f\u5728":74,"\u4e00\u4e2a\u7f51\u7edc\u5c42\u7684c":74,"\u4e00\u4e2a\u8f93\u51fa\u6570\u636e\u540c\u6837\u88ab\u7ec4\u7ec7\u4e3a\u4e00\u4e2a":89,"\u4e00\u4e2a\u8f93\u51fa\u7ec4\u6210":75,"\u4e00\u4e2a\u91cd\u8981\u7684\u95ee\u9898\u662f\u9009\u62e9\u6b63\u786e\u7684learning_r":83,"\u4e00\u4e2achunk\u7531\u6240\u5728\u7684\u6587\u4ef6\u504f\u79fb":28,"\u4e00\u4e2agpu\u8bbe\u5907\u4e0a\u4e0d\u5141\u8bb8\u914d\u7f6e\u591a\u4e2a\u6a21\u578b":103,"\u4e00\u4e2agradientmachine\u7c7b\u7684\u5bf9\u8c61\u7ba1\u7406\u7740\u4e00\u7ec4\u8ba1\u7b97\u5c42":90,"\u4e00\u4e2alabel":111,"\u4e00\u4e2amemory\u5305\u542b":114,"\u4e00\u4e2aposix\u517c\u5bb9\u7684\u6587\u4ef6\u7cfb\u7edf":28,"\u4e00\u4e9b\u60c5\u51b5\u4e3a\u4e86\u4fbf\u4e8e\u53d1\u5e03":90,"\u4e00\u53e5\u8bdd\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":113,"\u4e00\u53f0\u7535\u8111":0,"\u4e00\u662fbatch":81,"\u4e00\u6837\u7684\u65b9\u5f0f":0,"\u4e00\u7ef4\u6570\u7ec4":[89,90],"\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":89,"\u4e00\u81f4":[110,111],"\u4e00\u822c\u4e0d\u5141\u8bb8\u518d\u4ece":62,"\u4e00\u822c\u4ece":72,"\u4e00\u822c\u5728paddlepaddle\u4e2d":111,"\u4e00\u822c\u662f\u7531\u4e8e\u76f4\u63a5\u4f20\u9012\u5927\u5b57\u5178\u5bfc\u81f4\u7684":83,"\u4e00\u822c\u6765\u8bf4":114,"\u4e00\u822c\u7531mkl":42,"\u4e00\u822c\u8868\u793a":111,"\u4e00\u822c\u8bbe\u7f6e":83,"\u4e00\u8282":90,"\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":84,"\u4e0a":72,"\u4e0a\u4ea4\u53c9\u7f16\u8bd1raspberri":118,"\u4e0a\u4f20\u5230cloud\u6216\u8005\u4e0b\u8f7d\u5230\u672c\u5730\u7684\u65f6\u95f4\u53ef\u80fd\u6bd4\u8f83\u957f":28,"\u4e0a\u4f20\u65b9\u6cd5":62,"\u4e0a\u4f20\u8ba1\u7b97\u5f97\u51fa\u7684\u68af\u5ea6":92,"\u4e0a\u56fe\u4e2d\u7684":89,"\u4e0a\u56fe\u4e2d\u865a\u7ebf\u7684\u8fde\u63a5":111,"\u4e0a\u56fe\u63cf\u8ff0\u4e86\u4e00\u4e2a3\u8282\u70b9\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3\u573a\u666f":97,"\u4e0a\u6ce8\u518c\u4e00\u4e0b":28,"\u4e0a\u7f16\u8bd1\u5f88\u6162":0,"\u4e0a\u8fd0\u884c":116,"\u4e0a\u8ff0\u4ee3\u7801\u5c06bias\u5168\u90e8\u521d\u59cb\u5316\u4e3a1":83,"\u4e0a\u8ff0\u547d\u4ee4\u4e2d":2,"\u4e0a\u8ff0\u547d\u4ee4\u7f16\u8bd1\u51fa\u4e00\u4e2a":0,"\u4e0a\u8ff0\u7684":82,"\u4e0a\u8ff0\u7684\u4ee3\u7801\u7247\u6bb5\u5305\u542b\u4e86\u4e24\u79cd\u65b9\u6cd5":108,"\u4e0a\u8ff0paddlepaddl":62,"\u4e0a\u9762\u7684\u4ee3\u7801\u5728":75,"\u4e0a\u9762\u7684\u4ee3\u7801\u9996\u5148\u5bfc\u5165\u4f9d\u8d56\u7684\u5305":75,"\u4e0b":[75,77],"\u4e0b\u4e00\u4e2awheel\u5305\u9700\u8981\u66f4\u65b0\u7248\u672c\u53f7\u624d\u53ef\u4ee5\u4e0a\u4f20":62,"\u4e0b\u4f1a\u770b\u5230\u5982\u4e0b\u76ee\u5f55\u7ed3\u6784":87,"\u4e0b\u540c":83,"\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":108,"\u4e0b\u56fe\u662f\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u793a\u610f\u56fe":74,"\u4e0b\u56fe\u662fcsr\u5b58\u50a8\u7a00\u758f\u77e9\u9635\u7684\u793a\u610f\u56fe":89,"\u4e0b\u5b58\u653e\u516c\u5171\u6570\u636e\u96c6\u5408":12,"\u4e0b\u627e\u5230":87,"\u4e0b\u62c9\u6846\u4e2d\u627e\u5230\u751f\u6210\u76843\u4e2a\u4e8c\u8fdb\u5236\u6587\u4ef6":62,"\u4e0b\u6587\u4ee5nlp\u4efb\u52a1\u4e3a\u4f8b":113,"\u4e0b\u6587\u4f1a\u8be6\u7ec6\u8fdb\u884c\u4ecb\u7ecd":89,"\u4e0b\u6587\u4f7f\u7528":97,"\u4e0b\u6587\u5c31\u662f\u7528job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u8fdb\u884c\u8bad\u7ec3":96,"\u4e0b\u6587\u8be6\u7ec6\u89e3\u91ca":89,"\u4e0b\u7684":[90,97],"\u4e0b\u8868\u5217\u51fa\u4e86python\u7aef\u8bad\u7ec3\u63a5\u53e3\u66b4\u9732\u7684\u6570\u636e\u7c7b\u578b":89,"\u4e0b\u8f7d":28,"\u4e0b\u8f7d\u5230\u672c\u5730":28,"\u4e0b\u8f7d\u5b8c\u6570\u636e\u540e":96,"\u4e0b\u8f7d\u5f97\u5230":62,"\u4e0b\u8f7d\u6307\u5b9a\u7248\u672c\u7684docker\u955c\u50cf":2,"\u4e0b\u8f7dgpu\u7248\u672c":2,"\u4e0b\u9762":90,"\u4e0b\u9762\u4e3e\u4e2a\u7b80\u5355\u7684\u4f8b\u5b50":108,"\u4e0b\u9762\u4ecb\u7ecd\u4ecb\u7ecd":75,"\u4e0b\u9762\u4ee5":91,"\u4e0b\u9762\u4ee5\u77e9\u9635\u4e58\u64cd\u4f5c":75,"\u4e0b\u9762\u4ee5addop\u4e3a\u4f8b\u8bf4\u660etensor\u7684\u4f7f\u7528\u8fc7\u7a0b":76,"\u4e0b\u9762\u5206\u522b\u4ecb\u7ecd\u67d0\u4e00\u7c7b\u6587\u4ef6\u7684\u5b9e\u73b0\u65b9\u5f0f":46,"\u4e0b\u9762\u5217\u51fa\u4e86":114,"\u4e0b\u9762\u5217\u51fa\u4e86\u5168\u8fde\u63a5\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":74,"\u4e0b\u9762\u5c31\u6839\u636e\u8fd9\u51e0\u4e2a\u6b65\u9aa4\u5206\u522b\u4ecb\u7ecd":97,"\u4e0b\u9762\u6211\u4eec\u4f7f\u7528\u8fd9\u4e2a\u955c\u50cf\u6765\u4e0b\u8f7d\u6570\u636e\u5230docker":96,"\u4e0b\u9762\u662f":75,"\u4e0b\u9762\u662f\u5bf9":75,"\u4e0b\u9762\u662fc":90,"\u4e0b\u9762\u7684\u4ee3\u7801\u5c06\u968f\u673a\u751f\u6210\u7684\u77e9\u9635\u8f6c\u5316\u4e3a\u53ef\u4ee5\u88abpaddlepaddle\u52a0\u8f7d\u7684\u6a21\u578b\u53c2\u6570":83,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u4ecegithub\u62c9\u53d6\u6700\u65b0\u4ee3\u7801":87,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u521b\u5efa\u4e86\u4e00\u4e2a\u9ad8\u5ea6\u4e3a1":89,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u521b\u5efa\u4e86\u4e00\u4e2acpu\u4e0a\u7684\u4e8c\u503c\u7a00\u758f\u77e9\u9635":89,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u521b\u5efa\u4e86\u542b\u6709\u4e09\u4e2a\u5143\u7d20":89,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u5728\u521b\u5efa\u4e86\u4e00\u4e2acpu\u4e0a\u7684\u5e26\u5143\u7d20\u503c\u7684\u7a00\u758f\u77e9\u9635":89,"\u4e0b\u9762\u7684\u4ee3\u7801\u7247\u6bb5\u5b9e\u73b0\u4e86":74,"\u4e0b\u9762\u7684\u70b9\u5b9e\u73b0\u4e86mulop\u7684\u5b9a\u4e49":75,"\u4e0b\u9762\u7ed9\u51fa\u4e86\u4e00\u4e2a\u4f8b\u5b50":74,"\u4e0b\u9762\u7ed9\u51fa\u5728\u4e09\u7ef4\u7a7a\u95f4\u4e2d\u4f7f\u7528\u7ebf\u6027\u56de\u5f52\u62df\u5408\u4e00\u6761\u76f4\u7ebf\u7684\u4f8b\u5b50":84,"\u4e0b\u9762\u8be6\u7ec6\u89e3\u91ca\u4ec0\u4e48\u662f":89,"\u4e0b\u9762\u8fd9\u4e9blayer\u80fd\u591f\u63a5\u53d7\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165":110,"\u4e0d\u4e00\u81f4\u7684\u7531pfsclient\u4e0b\u8f7d\u6216\u8005\u4f20\u8f93chunk\u5b8c\u6210":28,"\u4e0d\u4ec5\u8981\u63d0\u4f9b\u6bcf\u4e00\u4e2a\u5916\u5c42\u5e8f\u5217\u5728\u6574\u4e2a":89,"\u4e0d\u4f1a\u4fdd\u7559\u5728\u78c1\u76d8\u4e0a":0,"\u4e0d\u4f1a\u518d\u4ece":81,"\u4e0d\u4f1a\u865a\u62df\u4efb\u4f55\u786c\u4ef6":0,"\u4e0d\u4f7f\u7528\u9759\u6001\u5e93":45,"\u4e0d\u4f7f\u7528\u989d\u5916\u7a7a\u95f4":74,"\u4e0d\u4f7f\u7528c":45,"\u4e0d\u4f7f\u7528swig":45,"\u4e0d\u5141\u8bb8\u4e00\u4e2a\u6587\u4ef6\u4e2d\u5305\u542b\u591a\u4e2aop":75,"\u4e0d\u5171\u4eab\u5219\u4e0d\u52a0":75,"\u4e0d\u5171\u4eab\u7684\u4f8b\u5b50\u53ef\u4ee5\u53c2\u8003":75,"\u4e0d\u53ef\u4ee5\u66f4\u6539":62,"\u4e0d\u53ef\u518d\u8fdb\u884c\u62c6\u5206":89,"\u4e0d\u540c":42,"\u4e0d\u540c\u4e8e\u4e0a\u8ff0\u4ecb\u7ecd\u7684recurr":82,"\u4e0d\u540c\u4e8eop\u7684\u7f16\u8bd1\u6d4b\u8bd5":75,"\u4e0d\u540c\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u6570\u636e\u5927\u5c0f\u7684\u6700\u5927\u503c\u4e0e\u6700\u5c0f\u503c\u7684\u6bd4\u7387":103,"\u4e0d\u540c\u5e8f\u5217\u53ef\u80fd\u4f1a\u542b\u6709\u4e0d\u540c\u6570\u76ee\u4e2a\u5143\u7d20":89,"\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u662f\u4e0d\u540c\u7684":114,"\u4e0d\u540c\u7248\u672c\u7684\u7f16\u8bd1\u5668\u4e4b\u95f4":45,"\u4e0d\u540c\u7684\u4f18\u5316\u7b97\u6cd5\u9700\u8981\u4f7f\u7528\u4e0d\u540c\u5927\u5c0f\u7684\u5185\u5b58":81,"\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf":97,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":84,"\u4e0d\u540c\u8ba1\u7b97\u5c42\u5bf9\u7a7a\u8f93\u5165\u7684\u5904\u7406\u7b56\u7565\u6709\u53ef\u80fd\u4e0d\u540c":89,"\u4e0d\u540c\u8bbe\u5907":75,"\u4e0d\u540c\u8bed\u8a00\u7684\u63a5\u53e3\u9002\u5e94\u4e0d\u540c\u8bed\u8a00\u7684\u7279\u6027":45,"\u4e0d\u540c\u8f93\u5165\u542b\u6709\u7684\u5b50\u53e5":113,"\u4e0d\u540c\u8f93\u5165\u5e8f\u5217\u542b\u6709\u7684\u8bcd\u8bed\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":113,"\u4e0d\u540cdataprovider\u5bf9\u6bd4\u5982\u4e0b":111,"\u4e0d\u540crank\u7684tensor\u662f\u4e0d\u540c\u7c7b\u578b":76,"\u4e0d\u5728":46,"\u4e0d\u5bb9\u6613\u51fa\u9519":28,"\u4e0d\u5d4c\u5165\u5176\u4ed6\u8bed\u8a00\u89e3\u91ca\u5668":45,"\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":45,"\u4e0d\u5e94\u8be5\u88ab\u62c6\u89e3":113,"\u4e0d\u6307\u5b9a\u65f6":113,"\u4e0d\u652f\u6301":89,"\u4e0d\u652f\u6301\u5e8f\u5217\u957f\u5ea6\u4e3a":89,"\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":84,"\u4e0d\u662f\u771f\u6b63\u7684layer":82,"\u4e0d\u662f\u901a\u8fc7\u4e00\u822c\u7684\u65b9\u5f0f\u6765\u5b9e\u73b0\u5bf9\u8f93\u51fa\u7684\u6fc0\u6d3b":82,"\u4e0d\u663e\u793a\u7684\u5199\u6bcf\u4e2a\u7c7b\u5177\u4f53\u5305\u542b\u4ec0\u4e48":45,"\u4e0d\u6ee1\u8db3\u94a9\u5b50\u7684":72,"\u4e0d\u7528mount\u7684\u65b9\u5f0f\u6765\u8bbf\u95ee\u6570\u636e":12,"\u4e0d\u80fd\u4fee\u6539op\u7684\u6210\u5458\u53d8\u91cf":75,"\u4e0d\u80fd\u592a\u968f\u610f":72,"\u4e0d\u80fd\u88ab\u63d0\u4ea4\u5230":72,"\u4e0d\u8981\u5728\u6ce8\u91cd\u6027\u80fd\u7684\u8bad\u7ec3\u573a\u666f\u4e0b\u4f7f\u7528":81,"\u4e0d\u8bba\u5e8f\u5217\u4e2d\u7684\u5143\u7d20\u5728\u5185\u5b58\u4e2d\u5360\u7528\u591a\u5c11\u5b9e\u9645\u5b58\u50a8\u7a7a\u95f4":89,"\u4e0d\u8bba\u6570\u636e\u57df\u662f":89,"\u4e0d\u8bba\u662f\u4e00\u7ef4\u6574\u578b\u6570\u7ec4\u8fd8\u662f\u4e8c\u7ef4\u6d6e\u70b9\u6570\u77e9\u9635":89,"\u4e0d\u8bba\u662f\u5355\u5c42\u5e8f\u5217\u8fd8\u662f\u53cc\u5c42\u5e8f\u5217\u7684\u5e8f\u5217\u4fe1\u606f":89,"\u4e0d\u8fc7\u5b9e\u9645\u4e0a\u662f\u8fd0\u884c\u5728\u4e00\u4e2a":0,"\u4e0d\u9700\u5728\u4f7f\u7528c":90,"\u4e0d\u9700\u8981\u4f9d\u8d56\u5176\u4ed6\u4efb\u4f55\u8f6f\u4ef6\u4e86":0,"\u4e0d\u9700\u8981\u63d0\u4f9b\u5143\u7d20\u503c":89,"\u4e0d\u9700\u8981\u8bbe\u7f6e":116,"\u4e0e":[42,75,97,107],"\u4e0e\u4e4b\u76f8\u5bf9\u7684\u662flocal":28,"\u4e0e\u5176\u4ed6\u7b2c\u4e09\u65b9\u5e93\u4e00\u6837":42,"\u4e0e\u5176\u5b83":90,"\u4e0e\u529f\u80fd\u5206\u652f\u4e0d\u540c\u7684\u662f":62,"\u4e0e\u5355\u5c42rnn\u7684\u914d\u7f6e\u7c7b\u4f3c":111,"\u4e0e\u53ef\u80fd\u6709\u7684":62,"\u4e0e\u540c\u6b65sgd\u76f8\u6bd4":92,"\u4e0e\u5bfb\u627epython\u4ee3\u7801\u7684\u6027\u80fd\u74f6\u9888\u7c7b\u4f3c":107,"\u4e0e\u5f53\u524d\u7684\u8870\u51cf\u56e0\u5b50\u7684\u4e58\u79ef":83,"\u4e0e\u672c\u5730\u8bad\u7ec3\u76f8\u540c":93,"\u4e0e\u6b64\u4e0d\u540c\u7684\u662f":97,"\u4e0e\u8c03\u4f18":107,"\u4e0e\u8f93\u5165\u4e0d\u540c\u7684\u662f":90,"\u4e0e\u8fd9\u4e2a\u8bad\u7ec3\u6570\u636e\u4ea4\u4e92\u7684layer":81,"\u4e0ebatch":41,"\u4e0ejob":97,"\u4e0eoperator\u524d\u5411\u8ba1\u7b97\u7684\u8f93\u51fa\u8fdb\u884c\u5bf9\u6bd4":75,"\u4e0eoperator\u6ce8\u518c\u65f6\u6ce8\u518c\u7684\u7c7b\u578b\u4e00\u81f4":75,"\u4e0epython\u4e0d\u540c":107,"\u4e14\u4e0d\u6392\u9664commit\u4e4b\u95f4\u7684\u4fee\u6539\u5b58\u5728\u76f8\u4e92\u8986\u76d6\u7684\u60c5\u51b5":72,"\u4e14\u4f7f\u7528":87,"\u4e14\u589e\u52a0\u4e00\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00":45,"\u4e14\u5c55\u793a\u6548\u679c\u66f4\u597d":107,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":84,"\u4e14\u6bcf\u4e2a\u53e5\u5b50\u8868\u793a\u4e3a\u5bf9\u5e94\u7684\u8bcd\u8868\u7d22\u5f15\u6570\u7ec4":111,"\u4e14\u8c03\u7528\u65f6\u4e0d\u80fd\u629b\u51fa\u5f02\u5e38\u6216\u51fa\u73b0\u8fd0\u884c\u65f6\u9519\u8bef":46,"\u4e14c99\u652f\u6301bool\u7c7b\u578b\u548c\u5b9a\u957f\u6574\u6570":45,"\u4e14c99\u76f8\u5bf9\u4e8ec11\u4f7f\u7528\u66f4\u52a0\u5e7f\u6cdb":45,"\u4e24\u4e2a\u5b50\u76ee\u5f55\u4e0b":77,"\u4e24\u4e2a\u5d4c\u5957\u7684":113,"\u4e24\u4e2a\u64cd\u4f5c":108,"\u4e24\u4e2a\u8f93\u5165\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u4e5f\u5e76\u4e0d\u76f8\u540c":111,"\u4e24\u4e2a\u90e8\u5206":77,"\u4e24\u4e2a\u9690\u5c42\u7684\u7b80\u5355\u5168\u8fde\u63a5\u7f51\u7edc":90,"\u4e24\u6b21":89,"\u4e24\u79cd\u5e38\u7528\u7684\u6a21\u578b\u52a0\u8f7d\u65b9\u5f0f":90,"\u4e24\u79cd\u65b9\u6cd5\u7684\u533a\u522b":81,"\u4e24\u79cdblas\u5e93":1,"\u4e24\u8005\u90fd\u662f\u5bf9\u68af\u5ea6\u7684\u622a\u65ad":81,"\u4e25\u683c\u7684\u547d\u540d\u89c4\u8303pep":62,"\u4e2a\u5185\u5b58\u6c60\u5b9e\u9645\u4e0a\u51b3\u5b9a\u4e86shuffle\u7684\u7c92\u5ea6":81,"\u4e2a\u6027\u5316\u63a8\u8350":62,"\u4e2a\u6279\u6b21\u7684\u53c2\u6570\u5e73\u5747\u503c\u8fdb\u884c\u6d4b\u8bd5":103,"\u4e2a\u6a21\u578b\u6d4b\u8bd5\u6570\u636e":103,"\u4e2d":[41,42,45,46,74,75,76,81,89,97,107],"\u4e2d\u4e0d\u8981\u6dfb\u52a0\u5927\u6587\u4ef6\u7b49":72,"\u4e2d\u4f1a\u4f7f\u7528\u5230\u7684\u5b57\u5178\u6570\u636e\u6587\u4ef6":91,"\u4e2d\u4f1a\u63d0\u4f9b\u4e00\u4e9b\u5fc5\u8981\u7684\u63a5\u53e3\u548c\u51fd\u6570":42,"\u4e2d\u4f20\u5165\u53c2\u6570":91,"\u4e2d\u4f20\u5165\u7684\u53c2\u6570":91,"\u4e2d\u5143\u7d20\u4e2a\u6570\u603b\u662f\u7b49\u4e8e\u884c\u6570":89,"\u4e2d\u5143\u7d20\u7684\u4e2a\u6570\u7b49\u4e8e\u7f51\u7edc\u4e2d\u8f93\u51fa\u5c42\u7684\u4e2a\u6570":81,"\u4e2d\u5173\u4e8e\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc\u7684\u4ecb\u7ecd":111,"\u4e2d\u5199\u5165json\u5185\u5bb9":11,"\u4e2d\u5305\u542b\u4e00\u4e2araspberri":118,"\u4e2d\u5305\u542b\u6240\u4f9d\u8d56\u7684\u6240\u6709\u7b2c\u4e09\u65b9\u5e93":116,"\u4e2d\u5305\u542b\u82e5\u5e72\u4e2a\u4e0d\u540candroid":116,"\u4e2d\u5305\u542bc":[116,118],"\u4e2d\u5355\u5143\u6d4b\u8bd5\u7684\u4e00\u90e8\u5206":72,"\u4e2d\u5355\u5143\u6d4b\u8bd5\u80fd\u987a\u5229\u901a\u8fc7":72,"\u4e2d\u542b\u6709\u591a\u4e2a\u5e8f\u5217":89,"\u4e2d\u5b8c\u5168\u4e00\u81f4":45,"\u4e2d\u5b9a\u4e49":114,"\u4e2d\u5b9a\u4e49\u548c\u4f7f\u7528":113,"\u4e2d\u5b9e\u73b0\u4e86\u4e00\u4e2amerge\u7684\u65b9\u6cd5":42,"\u4e2d\u5b9e\u73b0\u7684\u7ed3\u6784\u4f53":46,"\u4e2d\u5bf9\u5e94\u7684layer\u5904":41,"\u4e2d\u5f15\u5165\u7684":41,"\u4e2d\u6253\u5370\u5176\u503c":81,"\u4e2d\u6307\u5b9a":103,"\u4e2d\u6307\u5b9a\u7684\u540d\u5b57":105,"\u4e2d\u63d0\u4f9b\u4e00\u4e2a\u4e0emkl\u6709\u5173\u7684\u603b\u5f00\u5173":42,"\u4e2d\u63d0\u4f9b\u4e86\u4e00\u4e9b\u5168\u5c40\u51fd\u6570\u7528\u6765\u5b9e\u73b0paddl":76,"\u4e2d\u641c\u7d22\u8fd9\u51e0\u4e2a\u5e93":1,"\u4e2d\u64cd\u4f5c":89,"\u4e2d\u6587\u6587\u6863":77,"\u4e2d\u6587\u6587\u6863\u76ee\u5f55":77,"\u4e2d\u6587\u7ef4\u57fa\u767e\u79d1\u9875\u9762":111,"\u4e2d\u6839\u636e":41,"\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7ed3\u679c\u77e9\u9635":81,"\u4e2d\u6bcf\u4e2apod\u7684ip\u5730\u5740":97,"\u4e2d\u6bcf\u5c42\u7684\u6570\u503c\u7edf\u8ba1":103,"\u4e2d\u6dfb\u52a0":41,"\u4e2d\u6dfb\u52a0\u4e00\u4e2a":42,"\u4e2d\u6dfb\u52a0\u4e24\u4e2a\u8f93\u5165":75,"\u4e2d\u7528\u4e8e\u5b58\u50a8\u6570\u636e\u7684":90,"\u4e2d\u7684":[76,90],"\u4e2d\u7684\u4e00\u884c":72,"\u4e2d\u7684\u4ee3\u7801\u4f5c\u4e3a\u5b9e\u4f8b":91,"\u4e2d\u7684\u504f\u79fb":89,"\u4e2d\u7684\u5bf9\u5e94\u5206\u652f\u5373\u53ef":72,"\u4e2d\u7684\u7248\u672c\u4fe1\u606f":62,"\u4e2d\u7684\u76f8\u5173\u811a\u672c":90,"\u4e2d\u7684\u8d77\u59cb\u504f\u79fb":89,"\u4e2d\u83b7\u53d6":97,"\u4e2d\u8bbe\u7f6e\u7684\u6240\u6709\u8282\u70b9":93,"\u4e2d\u8be6\u7ec6\u4ecb\u7ecd":74,"\u4e2d\u8c03\u7528":75,"\u4e2d\u8fd0\u884c\u4efb\u52a1\u7684\u89d2\u5ea6":12,"\u4e2d\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":114,"\u4e34\u65f6\u53d8\u91cf\u7b49\u7b49":81,"\u4e3a":[41,42,75,89,114,116,117,118],"\u4e3a\u4e86\u4f7f":75,"\u4e3a\u4e86\u4f7f\u8bc4\u5ba1\u4eba\u5728\u8bc4\u5ba1\u4ee3\u7801\u65f6\u66f4\u597d\u5730\u4e13\u6ce8\u4e8e\u4ee3\u7801\u672c\u8eab":72,"\u4e3a\u4e86\u4fdd\u8bc1\u6548\u7387":74,"\u4e3a\u4e86\u4fdd\u8bc1gpu\u9a71\u52a8\u80fd\u591f\u5728\u955c\u50cf\u91cc\u9762\u6b63\u5e38\u8fd0\u884c":2,"\u4e3a\u4e86\u51cf\u5c11\u751f\u6210\u94fe\u63a5\u5e93\u7684\u5927\u5c0f\u628a":87,"\u4e3a\u4e86\u5c01\u88c5\u80fd\u591f\u6b63\u786e\u5de5\u4f5c":74,"\u4e3a\u4e86\u5c3d\u53ef\u80fd\u5c11\u7684\u5728\u7236\u7c7blayer\u4e2d\u6dfb\u52a0\u53d8\u91cf\u6216\u8005\u51fd\u6570":42,"\u4e3a\u4e86\u5e94\u5bf9\u4ee5\u4e0a\u7684\u95ee\u9898":28,"\u4e3a\u4e86\u5f00\u53d1paddlepaddl":0,"\u4e3a\u4e86\u63cf\u8ff0\u65b9\u4fbf":113,"\u4e3a\u4e86\u65b9\u4fbf\u5927\u5bb6":72,"\u4e3a\u4e86\u66b4\u9732\u7684\u63a5\u53e3\u5c3d\u91cf\u7b80\u5355":46,"\u4e3a\u4e86\u66f4\u597d\u7684\u7b26\u5408paddlepaddle\u7684\u4ee3\u7801\u98ce\u683c":42,"\u4e3a\u4e86\u6700\u5927\u7a0b\u5ea6\u51cf\u5c11\u591a\u6b21\u8c03\u7528":41,"\u4e3a\u4e86\u751f\u6210\u66f4\u53ef\u8bfb\u7684\u6027\u80fd\u5206\u6790\u7ed3\u679c":107,"\u4e3a\u4e86\u7b80\u5316cmake\u914d\u7f6e":116,"\u4e3a\u4e86\u8fbe\u5230\u6027\u80fd\u6700\u4f18":108,"\u4e3a\u4e86\u8fbe\u5230\u6700\u5feb\u7684\u8ba1\u7b97\u901f\u5ea6":[116,117],"\u4e3a\u4e86\u8fdb\u4e00\u6b65\u63d0\u5347paddlepaddle\u5728\u57fa\u672c\u6570\u5b66\u8fd0\u7b97\u7684\u8ba1\u7b97\u901f\u5ea6":42,"\u4e3a\u4ec0\u4e48\u7528":0,"\u4e3a\u4f7f\u7528c":90,"\u4e3a\u4f8b":[75,82],"\u4e3a\u4f8b\u6765\u4ecb\u7ecd\u5982\u4f55\u5199\u5e26kernel\u7684oper":75,"\u4e3a\u53c2\u6570\u77e9\u9635\u7684\u5bbd\u5ea6":83,"\u4e3a\u5b83\u4eec\u9644\u52a0\u4e0a\u5e8f\u5217\u4fe1\u606f\u5c06\u53d8\u6210\u5e8f\u5217\u8f93\u5165":89,"\u4e3a\u5bb9\u5668\u5185\u6267\u884c\u7684\u547d\u4ee4":2,"\u4e3a\u60a8\u505a\u6027\u80fd\u8c03\u4f18\u63d0\u4f9b\u4e86\u65b9\u5411":108,"\u4e3a\u65b9\u4fbf\u4f5c\u4e1a\u542f\u52a8\u63d0\u4f9b\u4e86\u4e24\u4e2a\u72ec\u7279\u7684\u547d\u4ee4\u9009\u9879":93,"\u4e3a\u6b64":96,"\u4e3a\u6bcf\u4e00\u4e2a":[89,90],"\u4e3a\u6bcf\u4e00\u4e2a\u8f93\u5165":[89,90],"\u4e3a\u6bcf\u4e2aop\u521b\u5efa\u5355\u72ec\u7684":75,"\u4e3a\u8f93\u51fa\u5206\u914d\u5185\u5b58":74,"\u4e3aconst\u51fd\u6570":75,"\u4e3aoutput_\u7533\u8bf7\u5185\u5b58":74,"\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":103,"\u4e3b\u8981\u529f\u80fd\u5305\u62ec":28,"\u4e3b\u8981\u5305\u62ec":42,"\u4e3b\u8981\u5305\u62ec\u4e86\u6df1\u5ea6\u5b66\u4e60\u76f8\u5173\u7684\u6570\u5b66\u539f\u8bed\u4e0e\u64cd\u4f5c":42,"\u4e3b\u8981\u5305\u62ec\u56db\u79cd\u7c7b\u578b":84,"\u4e3b\u8981\u539f\u56e0\u5305\u62ec\u4e24\u4e2a\u65b9\u9762":81,"\u4e3b\u8981\u7528\u4e8epython":75,"\u4e3b\u8981\u9488\u5bf9paddlepaddle\u5728\u91cd\u6784\u4e4b\u524d\u7684\u4ee3\u7801\u6846\u67b6\u4ee5\u53cav1\u7684api":42,"\u4e3e\u4e00\u4e2a\u4f8b\u5b50":83,"\u4e3e\u4f8b":81,"\u4e3e\u4f8b\u8bf4\u660e":111,"\u4e4b\u524d":72,"\u4e4b\u540e":[74,84],"\u4e4b\u540e\u4f7f\u7528":74,"\u4e4b\u540e\u4f7f\u7528\u77e9\u9635\u8fd0\u7b97\u51fd\u6570\u6765\u8ba1\u7b97":74,"\u4e4b\u540e\u518d\u7528\u7f51\u9875\u8fde\u5230http":77,"\u4e4b\u540e\u521d\u59cb\u5316\u6240\u6709\u7684\u6743\u91cd\u77e9\u9635":74,"\u4e4b\u540e\u624d\u80fd\u5f00\u59cb\u7f16\u8bd1\u7684\u6b65\u9aa4":1,"\u4e4b\u5916\u7684\u6240\u6709\u5934\u6587\u4ef6":46,"\u4e4b\u7c7b\u7684\u7a0b\u5e8f\u6765\u7f16\u8bd1\u6e90\u7801":0,"\u4e4b\u95f4\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":113,"\u4e58\u4e0a\u8f93\u51fa\u7684\u68af\u5ea6":74,"\u4e58\u6cd5\u548c\u4e58\u6cd5\u68af\u5ea6\u7684\u8ba1\u7b97\u5360\u75282":107,"\u4e58\u9664\u7b49\u65f6\u5019":81,"\u4e5f\u4e0d\u4f7f\u7528\u5176\u4ed6\u52a8\u6001\u5e93":45,"\u4e5f\u4e0d\u5b58\u5728\u4e00\u4e2asubseq\u76f4\u63a5\u751f\u6210\u4e0b\u4e00\u4e2asubseq\u7684\u60c5\u51b5":113,"\u4e5f\u4e0d\u5e94\u8be5\u62a5\u9519":46,"\u4e5f\u4e0d\u751f\u6210":46,"\u4e5f\u4e0d\u80fd\u63a5\u6536\u5e8f\u5217\u6570\u636e\u4f5c\u4e3a\u8f93\u5165":82,"\u4e5f\u4f1a\u5360\u7528\u78c1\u76d8":0,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":72,"\u4e5f\u53ef\u4ee5\u4f7f\u7528\u8fd9\u4e9b\u955c\u50cf":62,"\u4e5f\u53ef\u4ee5\u5229\u7528paddlepaddl":77,"\u4e5f\u53ef\u4ee5\u662f\u4e00\u4e2a\u8bcd\u8bed":113,"\u4e5f\u53ef\u4ee5\u662f\u5728\u4efb\u52a1\u542f\u52a8\u524d\u4e0b\u8f7d\u5230\u672c\u5730\u7684":91,"\u4e5f\u53ef\u4ee5\u76f4\u63a5\u5728\u7f51\u9875\u9884\u89c8\u6587\u6863":77,"\u4e5f\u53ef\u4ee5\u8bf4\u662f\u67d0\u4e9b\u7279\u5b9a\u6307\u4ee4\u7684\u4f7f\u7528\u60c5\u51b5":108,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539":97,"\u4e5f\u53ef\u5199\u6210":75,"\u4e5f\u53ef\u81ea\u884c\u524d\u5f80\u5b98\u7f51\u4e0b\u8f7d":117,"\u4e5f\u53ef\u901a\u8fc7\u4ee5\u4e0b\u547d\u4ee4\u83b7\u53d6":116,"\u4e5f\u5c31\u662f":72,"\u4e5f\u5c31\u662f\u672c\u5730\u7684\u6e90\u7801\u6811\u6839\u76ee\u5f55\u91cc\u7684":0,"\u4e5f\u5c31\u662f\u7a7a\u8f93\u5165":89,"\u4e5f\u5c31\u662f\u81ea\u5df1\u7528\u6237\u540d\u4e0b\u7684":72,"\u4e5f\u5c31\u662f\u8bf4":[89,103,105],"\u4e5f\u5c31\u662f\u8bf4\u8f93\u51fa\u7684\u7ed3\u679c\u4e0d\u4f1a\u5728\u539f\u6765\u7684\u6570\u636e\u4e0a\u7d2f\u52a0":42,"\u4e5f\u5c31\u662fpaddlepaddle\u4e2d\u7684\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":89,"\u4e5f\u63cf\u8ff0\u4e86\u5bb9\u5668\u9700\u8981\u4f7f\u7528\u7684\u5b58\u50a8\u5377\u6302\u8f7d\u7684\u60c5\u51b5":97,"\u4e5f\u652f\u6301cpu\u7684\u6027\u80fd\u5206\u6790":108,"\u4e5f\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":111,"\u4e5f\u662fdecoder\u5faa\u73af\u5c55\u5f00\u7684\u4f9d\u636e":113,"\u4e5f\u6ca1\u7528":78,"\u4e66\u5199":45,"\u4e86":0,"\u4e86\u89e3\u5176\u8c03\u7528\u5173\u7cfb":107,"\u4e86\u89e3\u60a8\u7684\u786c\u4ef6":108,"\u4e86\u89e3\u66f4\u591a\u7ec6\u8282":114,"\u4e86\u89e3\u66f4\u591a\u8be6\u7ec6\u4fe1\u606f":114,"\u4e8c\u7ef4\u6d6e\u70b9\u578b\u77e9\u9635":89,"\u4e8c\u7ef4\u6d6e\u70b9\u6570\u77e9\u9635":89,"\u4e8c\u7ef4\u77e9\u9635":90,"\u4e8c\u7ef4\u77e9\u9635\u53ef\u4ee5\u8868\u793a\u884c\u5411\u91cf\u548c\u5217\u5411\u91cf":89,"\u4e8c\u8005\u8bed\u610f\u4e0a\u5b8c\u5168\u4e00\u81f4":111,"\u4e8e\u662f":89,"\u4e8e\u662f\u6211\u4eec\u53ef\u4ee5\u70b9\u51fb":107,"\u4e8e\u662f\u8fd9\u91cc\u4f7f\u7528":107,"\u4ea4\u4e92\u7684\u65b9\u6cd5":107,"\u4ea4\u53c9\u7f16\u8bd1\u5de5\u5177\u94fe\u4e3a":116,"\u4ea4\u53c9\u7f16\u8bd1android\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":116,"\u4ea4\u53c9\u7f16\u8bd1android\u7248\u672c\u7684paddlepaddle\u5e93\u65f6":116,"\u4ea4\u53c9\u7f16\u8bd1ios\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":117,"\u4ea4\u53c9\u7f16\u8bd1ios\u7248\u672c\u7684paddlepaddle\u5e93\u65f6":117,"\u4ea4\u53c9\u7f16\u8bd1raspberri":118,"\u4ea4\u7531cmake\u7cfb\u7edf\u672c\u8eab\u6765\u5904\u7406":116,"\u4ea6\u53ef\u4ee5\u901a\u8fc7\u624b\u52a8\u8bbe\u7f6e":[116,117],"\u4eab\u53d7\u60a8\u7684\u65c5\u7a0b":2,"\u4eba\u8138\u8bc6\u522b":12,"\u4ec0\u4e48\u662f":0,"\u4ec5\u4ec5\u4f7f\u7528":45,"\u4ec5\u4f1a\u5728\u652f\u6301avx2\u6307\u4ee4\u96c6\u53ca\u4ee5\u4e0a\u7684\u673a\u5668\u624d\u4f7f\u7528mkl":42,"\u4ec5\u5728\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u65f6\u6709\u6548":74,"\u4ec5\u5bf9\u7a00\u758f\u6570\u636e\u6709\u6548":74,"\u4ec5\u652f\u6301\u6574\u578b\u503c":89,"\u4ec5\u7528\u4e8e\u5b58\u50a8\u6574\u578b\u503c":90,"\u4ecb\u7ecd\u4e86\u4e00\u79cd\u901a\u8fc7ssh\u8fdc\u7a0b\u5206\u53d1\u4efb\u52a1":97,"\u4ecb\u7ecd\u4ea4\u53c9\u7f16\u8bd1android\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93\u7684\u65b9\u6cd5\u548c\u6b65\u9aa4":116,"\u4ecb\u7ecd\u4f7f\u7528paddlepaddl":91,"\u4ece":[62,79,108],"\u4ece0\u5230num":103,"\u4ece0\u5f00\u59cb\u7684\u6574\u6570":91,"\u4ece\u4e00\u4e2aword\u751f\u6210\u4e0b\u4e00\u4e2aword":113,"\u4ece\u5185\u6838\u51fd\u6570\u7684\u89d2\u5ea6":108,"\u4ece\u6a21\u578b\u6587\u4ef6\u5c06\u9884\u8bad\u7ec3\u53c2\u6570\u8f7d\u5165":83,"\u4ece\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u770b":111,"\u4ece\u6e90\u7801\u4ea4\u53c9\u7f16\u8bd1ios\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":117,"\u4ece\u6e90\u7801\u4ea4\u53c9\u7f16\u8bd1paddlepaddl":116,"\u4ece\u6e90\u7801\u7f16\u8bd1":3,"\u4ece\u78c1\u76d8\u52a0\u8f7d\u9884\u6d4b\u6a21\u578b":90,"\u4ece\u78c1\u76d8\u6587\u4ef6\u4e2d\u52a0\u8f7duuid\u6587\u4ef6\u540d\u7684\u68c0\u67e5\u70b9\u5feb\u7167\u6587\u4ef6":11,"\u4ece\u800c\u53ef\u4ee5\u505a\u4e00\u4e9b\u4e0e\u8ba1\u7b97\u91cd\u53e0\u7684\u5de5\u4f5c":74,"\u4ece\u800c\u5f15\u53d1\u5176\u4ed6\u8282\u70b9\u65e0\u6cd5\u8fde\u63a5\u5bfc\u81f4":79,"\u4ece\u800c\u907f\u514d\u4e86packing\u5197\u4f59":41,"\u4ece\u8bed\u4e49\u4e0a\u770b":113,"\u4ece\u8d77\u59cb\u7aef\u53e3\u76d1\u542c\u591a\u4e2a\u7aef\u53e3\u7528\u4e8e\u901a\u4fe1":91,"\u4ece\u8f93\u5165\u6570\u636e\u4e0a\u770b":111,"\u4ececmake":116,"\u4eceetcd\u4e2d\u8bfb\u53d6\u8282\u70b9":11,"\u4ecestart":103,"\u4ed3\u5e93\u7684\u8fdc\u7a0b\u4e3b\u673a":72,"\u4ed6\u4e3b\u8981\u5305\u542b\u4e86\u5b9e\u9645\u66b4\u9732\u7684\u7c7b\u578b\u7ed3\u6784":46,"\u4ed6\u4eec\u5206\u522b\u662f":111,"\u4ed6\u4eec\u5728\u81ea\u5df1\u7684":0,"\u4ed6\u4eec\u5728paddle\u7684\u6587\u6863\u548capi\u4e2d\u662f\u4e00\u4e2a\u6982\u5ff5":111,"\u4ed6\u662f\u5c06":46,"\u4ed6\u7684\u76ee\u6807\u662f\u4f7f\u7528c":45,"\u4ee3\u66ff":97,"\u4ee3\u7801\u4e2d9":111,"\u4ee3\u7801\u53c2\u8003":91,"\u4ee3\u7801\u5982\u4e0b":[81,82,83,114],"\u4ee3\u7801\u6ce8\u91ca\u8bf7\u9075\u5b88":72,"\u4ee3\u7801\u7247\u6bb5\u5982\u4e0b":89,"\u4ee3\u7801\u751f\u6210\u7684\u7b26\u53f7\u53ef\u80fd\u4e0d\u4e00\u81f4":45,"\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790":107,"\u4ee3\u7801\u793a\u4f8b\u5982\u4e0b":[75,90],"\u4ee3\u8868\u5bbf\u4e3b\u673a\u76ee\u5f55":97,"\u4ee3\u8868\u8fd9\u4e2alayer\u662f\u7528\u4e8e\u8dd1\u5728mkl":42,"\u4ee3\u8868\u8fd9\u4e2ashard\u7684\u6700\u5927index":12,"\u4ee3\u8868shard\u7684index":12,"\u4ee5":82,"\u4ee5\u4e0a":[72,116],"\u4ee5\u4e0a\u4e24\u79cd\u65b9\u5f0f\u53ea\u9700\u9009\u62e9\u5176\u4e00\u5373\u53ef":90,"\u4ee5\u4e0a\u4ee3\u7801\u7684reader\u8f93\u51fa\u7684data":12,"\u4ee5\u4e0a\u547d\u4ee4\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u4e0b\u751f\u6210100\u4e2a\u6587\u4ef6":12,"\u4ee5\u4e0b":12,"\u4ee5\u4e0b\u4ee3\u7801\u7247\u6bb5\u5b9a\u4e49":114,"\u4ee5\u4e0b\u547d\u4ee4\u542f\u52a8\u4e00\u4e2a":0,"\u4ee5\u4e0b\u6307\u4ee4\u80fd\u68c0\u67e5linux\u7535\u8111\u662f\u5426\u652f\u6301avx":2,"\u4ee5\u4e0b\u64cd\u4f5c\u5747\u5728head\u8282\u70b9\u4e2d\u6267\u884c":98,"\u4ee5\u4e0b\u6559\u7a0b\u5c06\u6307\u5bfc\u60a8\u63d0\u4ea4\u4ee3\u7801":72,"\u4ee5\u4e0b\u7b80\u79f0rnn":41,"\u4ee5\u4ea4\u4e92\u5f0f\u7684\u65b9\u5f0f\u6267\u884c\u6216\u8c03\u8bd5\u60a8\u7684\u4ee3\u7801":2,"\u4ee5\u4f7f\u7528":116,"\u4ee5\u4f7f\u7528adam\u7b97\u6cd5\u4e3a\u4f8b":83,"\u4ee5\u4fbf\u6211\u4eec\u53ef\u4ee5\u628a\u66f4\u591a\u7684\u7cbe\u529b\u653e\u5230\u903b\u8f91\u672c\u8eab\u4e0a":28,"\u4ee5\u4fbf\u83b7\u5f97\u8bad\u7ec3\u6570\u636e\u7684\u4f4d\u7f6e\u548c\u83b7\u53d6\u73af\u5883\u53d8\u91cf\u914d\u7f6e":91,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u7684\u6b63\u786e\u8ba1\u7b97":74,"\u4ee5\u4fdd\u8bc1\u68af\u5ea6\u8ba1\u7b97\u7684\u6b63\u786e\u6027":74,"\u4ee5\u4fdd\u8bc1\u7f16\u8bd1\u9ad8\u6548":0,"\u4ee5\u53ca":[0,41,74,89],"\u4ee5\u53ca\u4f7f\u7528\u5b50\u5e8f\u5217\u6765\u5b9a\u4e49\u5206\u7ea7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":114,"\u4ee5\u53ca\u5207\u6362\u673a\u5668\u65f6\u9700\u8981\u65b0\u5b89\u88c5\u7684\u8f9b\u82e6":0,"\u4ee5\u53ca\u53cc\u5c42\u5e8f\u5217":110,"\u4ee5\u53ca\u5982\u4f55\u89e3\u6790\u795e\u7ecf\u7f51\u7edc\u524d\u5411\u8ba1\u7b97\u7684\u8f93\u51fa\u7ed3\u679c":89,"\u4ee5\u53ca\u76ee\u6807\u673a\u7248openblas\u5e93":118,"\u4ee5\u53ca\u76f8\u5173\u7684\u5c5e\u6027\u53c2\u6570":75,"\u4ee5\u53ca\u7b2c\u4e09\u65b9\u4f9d\u8d56\u94fe\u63a5\u5e93\u548c\u5934\u6587\u4ef6":87,"\u4ee5\u53ca\u8ba1\u7b97\u903b\u8f91\u5728\u5e8f\u5217\u4e0a\u7684\u5faa\u73af\u5c55\u5f00":113,"\u4ee5\u53ca\u8f93\u5165\u7684\u68af\u5ea6":74,"\u4ee5\u53caandroid":116,"\u4ee5\u53cagcc":1,"\u4ee5\u53canumpi":12,"\u4ee5\u53carelu":74,"\u4ee5\u63d0\u4f9b\u4e00\u4e9b\u9ed8\u8ba4\u7684\u7f16\u8bd1\u5668\u548c\u7f16\u8bd1\u53c2\u6570\u76f8\u5173\u914d\u7f6e":116,"\u4ee5\u63d0\u4f9b\u4e00\u4e9b\u9ed8\u8ba4\u7684\u7f16\u8bd1\u5668\u548c\u7f16\u8bd1\u53c2\u6570\u914d\u7f6e":117,"\u4ee5\u6b64\u8fbe\u5230\u6700\u597d\u7684\u6027\u80fd":42,"\u4ee5\u786e\u4fdd\u6240\u6709\u7684\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93\u548cpaddlepaddle\u4ee3\u7801\u90fd\u662f\u9488\u5bf9\u65b0\u7684cmake\u914d\u7f6e\u91cd\u65b0\u7f16\u8bd1\u7684":[116,117,118],"\u4ee5\u793a\u533a\u5206":[41,42],"\u4ee5\u8f93\u51fa":81,"\u4ee5\u9017\u53f7\u95f4\u9694":103,"\u4ee5\u907f\u514d\u94fe\u63a5\u4e0d\u5fc5\u8981\u7684\u5e93":87,"\u4ee5eigentensor\u4e3a\u4f8b":76,"\u4ee5embedding\u5c42\u4e3a\u4f8b":83,"\u4ee5lstm\u4e3a\u4f8b":82,"\u4efb\u4f55\u65f6\u5019\u5982\u679c\u9700\u8981\u6d6e\u70b9\u578b\u6570\u7ec4":89,"\u4efb\u52a1\u6765\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":93,"\u4efb\u52a1\u88ab\u8c03\u5ea6\u5728\u96c6\u7fa4\u4e2d\u65f6":94,"\u4efb\u610f\u5c06\u4e00\u4e9b\u6570\u636e\u7ec4\u5408\u6210\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":111,"\u4efb\u610f\u65f6\u523b\u53ea\u53ef\u80fd\u540c\u65f6\u6709\u4e00\u53f0\u670d\u52a1\u5668\u6545\u969c":11,"\u4f18\u5316\u524d":41,"\u4f18\u5316\u540e":41,"\u4f18\u5316\u5668\u5219\u7528\u94fe\u5f0f\u6cd5\u5219\u6765\u5bf9\u6bcf\u4e2a\u53c2\u6570\u8ba1\u7b97\u635f\u5931\u51fd\u6570\u7684\u68af\u5ea6":74,"\u4f1a\u4ea7\u751f\u5f53\u524dpython\u4e8c\u8fdb\u5236\u7684\u5b8c\u6574\u8def\u5f84":107,"\u4f1a\u4ee5":[41,42],"\u4f1a\u4f7f\u7528":90,"\u4f1a\u4f7f\u7528\u76f8\u540c\u7684\u539f\u6570\u636e":41,"\u4f1a\u5148\u4e34\u65f6\u4fdd\u5b58\u5728":42,"\u4f1a\u5148\u8fdb\u884c\u53c2\u6570\u7684\u521d\u59cb\u5316\u4e0e\u89e3\u6790":97,"\u4f1a\u5171\u4eab\u53c2\u6570":83,"\u4f1a\u5173\u8054\u53c2\u6570":82,"\u4f1a\u5206\u522b\u4ecb\u7ecd\u96c6\u7fa4\u4f5c\u4e1a\u7684\u542f\u52a8\u548c\u505c\u6b62\u65b9\u6cd5":94,"\u4f1a\u52a0\u8f7d\u4e0a\u4e00\u8f6e\u7684\u53c2\u6570":103,"\u4f1a\u53d8\u6210\u8bcd\u8868\u4e2d\u7684\u4f4d\u7f6e":111,"\u4f1a\u542f\u52a8pserver\u4e0etrainer\u8fdb\u7a0b":97,"\u4f1a\u5728":[42,77],"\u4f1a\u5728\u5f53\u524d\u76ee\u5f55\u751f\u6210\u4e24\u4e2a\u5b50\u76ee\u5f55":77,"\u4f1a\u5728\u7f16\u8bd1paddlepaddle\u7684\u65f6\u5019\u4e0b\u8f7d\u5e76\u7f16\u8bd1mkl":42,"\u4f1a\u5927\u4e0d\u76f8\u540c":91,"\u4f1a\u5bf9\u6bcf\u4e00\u4e2a\u6fc0\u6d3b\u6682\u5b58\u4e00\u4e9b\u6570\u636e":81,"\u4f1a\u5bf9\u8bad\u7ec3\u6027\u80fd\u9020\u6210\u5f71\u54cd":81,"\u4f1a\u5bf9\u8fd9\u7c7b\u8f93\u5165\u8fdb\u884c\u62c6\u89e3":113,"\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7248\u672cpython\u5728\u4e00\u4e2a\u8fdb\u7a0b\u91cc\u7684bug":45,"\u4f1a\u5c06\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u62fc\u63a5":113,"\u4f1a\u5c06\u7b2c\u4e00\u4e2a":81,"\u4f1a\u5f15\u5165":42,"\u4f1a\u6210\u4e3astep\u51fd\u6570\u7684\u8f93\u5165":113,"\u4f1a\u6253\u5370\u5230\u6807\u51c6\u8f93\u51fa":107,"\u4f1a\u628a\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\u5206\u522b\u5206\u5272\u6210\u591a\u4e2a\u6587\u4ef6":91,"\u4f1a\u628acpu\u7684buffer\u5bf9\u9f50\u4e3a4096":42,"\u4f1a\u62a5\u5982\u4e0b\u7684\u9519\u8bef":81,"\u4f1a\u62a5\u9519":113,"\u4f1a\u6dfb\u52a0\u76f8\u5e94\u7684\u811a\u672c\u5728":42,"\u4f1a\u6dfb\u52a0\u76f8\u5e94\u7684\u811a\u672c\u7528\u4e8e\u6d4b\u8bd5\u548c\u5bf9\u6bd4\u5728\u4f7f\u7528mkl":41,"\u4f1a\u72ec\u7acb\u62e5\u6709\u4e00\u4efd\u8bad\u7ec3\u597d\u7684\u6a21\u578b":90,"\u4f1a\u751f\u6210\u6027\u80fd\u5206\u6790\u7ed3\u679c\u6587\u4ef6":107,"\u4f1a\u76f4\u63a5\u62a5\u9519\u9000\u51fa":45,"\u4f1a\u76f8\u5e94\u5730\u6539\u53d8\u8f93\u51fa\u7684\u5c3a\u5bf8":74,"\u4f1a\u81ea\u52a8\u4f7f\u7528mklml\u5e93\u4f5c\u4e3apaddlepaddle\u7684cblas\u548clapack\u5e93":42,"\u4f1a\u81ea\u52a8\u5173\u95ed\u5bf9\u5e94\u7684issu":72,"\u4f1a\u81ea\u52a8\u5728\u7f16\u8bd1\u65f6\u4e0b\u8f7d":1,"\u4f1a\u81ea\u52a8\u6839\u636e\u786c\u4ef6\u914d\u7f6e":42,"\u4f1a\u83b7\u53d6\u5f53\u524dnamespace\u4e0b\u7684\u6240\u6709pod":97,"\u4f1a\u88ab":91,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":113,"\u4f1a\u88ab\u62c6\u89e3\u4e3a\u975e\u5e8f\u5217":113,"\u4f1a\u88abpickle\u5e8f\u5217\u5316\u6210\u5b57\u7b26\u4e32":12,"\u4f1a\u901a\u8fc7\u5224\u6570\u636e\u662f\u5426\u9644\u5e26\u6709\u5e8f\u5217\u4fe1\u606f\u6765\u5224\u65ad\u4e00\u4e2a\u5411\u91cf":89,"\u4f1a\u9020\u6210\u90ae\u4ef6\u707e\u96be":72,"\u4f20\u5165":12,"\u4f20\u7ed9dataprovider\u7684\u67d0\u4e00\u4e2aargs\u8fc7\u5927":83,"\u4f20\u9012\u7ed9\u914d\u7f6e\u6587\u4ef6\u7684\u53c2\u6570":103,"\u4f46":46,"\u4f46\u4e0d\u66b4\u9732":46,"\u4f46\u4e0d\u7528\u4e8e\u8ba1\u7b97\u68af\u5ea6":74,"\u4f46\u4e0d\u9700\u8981\u63d0\u524d\u521b\u5efa":103,"\u4f46\u4e8e\u53cc\u5c42\u5e8f\u5217\u7684lstm\u6765\u8bf4":111,"\u4f46\u53ef\u4ee5\u83b7\u53d6":81,"\u4f46\u548c\u5355\u5c42rnn\u4e0d\u540c":111,"\u4f46\u5b50\u53e5\u542b\u6709\u7684\u8bcd\u8bed\u6570\u53ef\u4ee5\u4e0d\u76f8\u7b49":113,"\u4f46\u5c3d\u91cf\u8bf7\u4fdd\u6301\u7f16\u8bd1\u548c\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u662f\u540c\u4e00\u4e2a\u7248\u672c":1,"\u4f46\u5e76\u6ca1\u6709\u7ecf\u8fc7\u56de\u5f52\u6d4b\u8bd5":62,"\u4f46\u5e8f\u5217\u8f93\u51fa\u65f6":111,"\u4f46\u622a\u65ad\u65f6\u673a\u4e0d\u540c":81,"\u4f46\u6240\u6709fork\u7684\u7248\u672c\u5e93\u7684\u6240\u6709\u5206\u652f\u90fd\u76f8\u5f53\u4e8e\u7279\u6027\u5206\u652f":62,"\u4f46\u662f":[81,111],"\u4f46\u662f\u53c8\u8fc7\u4e8e\u7410\u788e":46,"\u4f46\u662f\u5728mkl":42,"\u4f46\u662f\u5728paddlepaddle\u4e2d":42,"\u4f46\u662f\u5927\u90e8\u5206\u53c2\u6570\u662f\u4e3a\u5f00\u53d1\u8005\u63d0\u4f9b\u7684":102,"\u4f46\u662f\u5b50\u5e8f\u5217\u7684\u6570\u76ee\u5fc5\u987b\u4e00\u6837":111,"\u4f46\u662f\u5e76\u4e0d\u80fd\u4fdd\u8bc1\u53c2\u6570\u540c\u6b65\u66f4\u65b0":92,"\u4f46\u662f\u652f\u6301avx\u6307\u4ee4\u96c6":72,"\u4f46\u662f\u6574\u4e2a\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4e0d\u9700\u8981\u4efb\u4f55\u8f6c\u6362":42,"\u4f46\u662f\u6bcf\u4e2a\u6837\u672c\u4ec5\u5305\u542b\u51e0\u4e2a\u8bcd":105,"\u4f46\u662f\u6ce8\u610f\u7684\u662f":42,"\u4f46\u662f\u7a81\u7136\u6709\u4e00\u4e2a10000\u957f\u7684\u5e8f\u5217":81,"\u4f46\u662f\u865a\u62df\u7684\u4e0d\u4ec5\u4ec5\u662f":0,"\u4f46\u662f\u89e3\u91ca\u6027\u8bed\u8a00":45,"\u4f46\u662f\u8c03\u8bd5python\u4e2d\u4f7f\u7528\u7684\u52a8\u6001\u94fe\u63a5\u5e93\u4e0e\u76f4\u63a5\u8c03\u8bd5\u539f\u59cb\u4e8c\u8fdb\u5236\u76f8\u6bd4\u589e\u52a0\u4e86\u5f88\u591a\u590d\u6742\u5ea6":107,"\u4f46\u662fbatch":81,"\u4f46\u6709\u503c\u7684\u5730\u65b9\u5fc5\u987b\u4e3a1":84,"\u4f46\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":84,"\u4f46\u7531\u4e8ecuda\u5e93\u901a\u5e38\u9700\u8981cento":4,"\u4f46\u9700\u6ce8\u610f\u53cd\u5411op\u6ca1\u6709":75,"\u4f46eigen":76,"\u4f5c\u4e3a\u4e0b\u4e00\u4e2a\u5b50\u53e5memory\u7684\u521d\u59cb\u72b6\u6001":111,"\u4f5c\u4e3a\u4f8b\u5b50\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":114,"\u4f5c\u4e3a\u53c2\u6570\u5c5e\u6027":75,"\u4f5c\u4e3a\u53c2\u6570\u7684id":83,"\u4f5c\u4e3a\u53e6\u4e00\u4e2a\u7b2c\u4e09\u65b9\u5e93\u96c6\u6210\u8fdbpaddlepaddl":42,"\u4f5c\u4e3a\u5b58\u50a8\u7cfb\u7edf":12,"\u4f5c\u4e3a\u5f53\u524d\u65f6\u523b\u8f93\u5165":113,"\u4f5c\u4e3a\u7c7b\u53e5\u67c4":45,"\u4f5c\u4e3a\u7edf\u8ba1\u7684\u57fa\u672c\u5355\u4f4d":89,"\u4f5c\u4e3a\u7f16\u8bd1\u5de5\u5177":1,"\u4f5c\u4e3a\u8c03\u7528":90,"\u4f5c\u4e3a\u8f93\u5165":89,"\u4f5c\u4e3a\u8f93\u51fa":114,"\u4f5c\u4e3aboot_layer\u4f20\u7ed9\u4e0b\u4e00\u4e2a\u5b50\u53e5\u7684memori":111,"\u4f5c\u7528":110,"\u4f60\u53ef\u4ee5\u5c06\u7f51\u7edc\u914d\u7f6e\u6210\u67d0\u4e9b\u5c42\u4f7f\u7528gpu\u8ba1\u7b97":105,"\u4f60\u8fd8\u53ef\u4ee5\u901a\u8fc7\u8fd0\u884cdjango\u6846\u67b6\u76f4\u63a5\u6fc0\u6d3b\u5de5\u5177\u7684\u670d\u52a1\u5668":77,"\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":74,"\u4f60\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a\u8bbe\u5907\u7684id\u53f7":105,"\u4f60\u9700\u8981\u5728\u914d\u7f6ecmake\u65f6\u5c06":74,"\u4f60\u9700\u8981\u628a\u8be5\u6587\u4ef6\u52a0\u5165":74,"\u4f7f\u4e4b\u53d8\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u8f93\u5165":89,"\u4f7f\u4e4b\u53d8\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":89,"\u4f7f\u5f97\u5355\u5143\u6d4b\u8bd5\u6709\u4e00\u4e2a\u5e72\u51c0\u7684\u73af\u5883":78,"\u4f7f\u5f97\u642d\u6a21\u578b\u65f6\u66f4\u65b9\u4fbf":74,"\u4f7f\u68af\u5ea6\u7684\u63d0\u4ea4\u548c\u53c2\u6570\u7684\u66f4\u65b0\u6309\u7167\u987a\u5e8f\u65b9\u5f0f\u6267\u884c":92,"\u4f7f\u7528":[42,46,62,74,81,82,83,87,89,90,103,107,108,111,113,114,116],"\u4f7f\u75280\u53f7\u548c1\u53f7gpu\u8ba1\u7b97fc2\u5c42":105,"\u4f7f\u75280\u53f7gpu\u8ba1\u7b97fc2\u5c42":105,"\u4f7f\u75281\u53f7gpu\u8ba1\u7b97fc3\u5c42":105,"\u4f7f\u75282\u53f7\u548c3\u53f7gpu\u8ba1\u7b97fc3\u5c42":105,"\u4f7f\u7528\u4e00\u4e2a\u5c3a\u5ea6\u4e3a":74,"\u4f7f\u7528\u4e00\u4e2a\u8bcd\u524d\u4e24\u4e2a\u8bcd\u548c\u540e\u4e24\u4e2a\u8bcd":81,"\u4f7f\u7528\u4e0a\u6587\u521b\u5efa\u7684yaml\u6587\u4ef6\u521b\u5efakubernet":96,"\u4f7f\u7528\u4e0b\u9762\u547d\u4ee4":12,"\u4f7f\u7528\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u8fd0\u884c\u5b83":77,"\u4f7f\u7528\u4e86\u540c\u6837\u7684parameter\u548cbia":83,"\u4f7f\u7528\u4ee5\u4e0a\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u8fdb\u884c\u9884\u6d4b":84,"\u4f7f\u7528\u52a8\u6001\u5e93":45,"\u4f7f\u7528\u53c2\u6570":[1,91],"\u4f7f\u7528\u540c\u6837\u7684\u8bad\u7ec3\u6570\u636eblock":11,"\u4f7f\u7528\u57fa\u4e8edocker\u5bb9\u5668\u7684\u7f16\u8bd1\u65b9\u5f0f":116,"\u4f7f\u7528\u591a\u5757\u663e\u5361\u8bad\u7ec3":81,"\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bad\u7ec3":81,"\u4f7f\u7528\u5b66\u4e60\u5b8c\u6210\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u751f\u6210\u5e8f\u5217":114,"\u4f7f\u7528\u5b83\u4f1a\u5f00\u542f\u4e00\u4e2ahttp\u670d\u52a1":107,"\u4f7f\u7528\u5bb9\u5668\u65b9\u5f0f\u8fd0\u884c\u8bad\u7ec3\u4efb\u52a1\u7684kubernet":97,"\u4f7f\u7528\u6211\u4eec\u4e4b\u524d\u6784\u9020\u7684\u955c\u50cf":96,"\u4f7f\u7528\u6570\u503c\u6cd5\u68c0\u6d4b\u68af\u5ea6\u6b63\u786e\u6027\u548c\u7a33\u5b9a\u6027":75,"\u4f7f\u7528\u6587\u6863":75,"\u4f7f\u7528\u663e\u5361\u8bad\u7ec3":81,"\u4f7f\u7528\u667a\u80fd\u6307\u9488\u7684\u539f\u56e0\u662f":46,"\u4f7f\u7528\u6848\u4f8b":104,"\u4f7f\u7528\u73af\u5883\u53d8\u91cf":91,"\u4f7f\u7528\u7684\u53c2\u6570\u4e0epaddlepaddle\u7533\u8bf7\u7684buffer\u5171\u7528\u4e00\u5757\u5185\u5b58":42,"\u4f7f\u7528\u76f8\u5bf9\u8def\u5f84\u7684\u5f15\u7528\u65b9\u5f0f":46,"\u4f7f\u7528\u8005\u4e0d\u9700\u8981\u5173\u5fc3":103,"\u4f7f\u7528\u8005\u53ea\u9700\u8981\u5173\u6ce8\u4e8e\u8bbe\u8ba1rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":113,"\u4f7f\u7528\u8005\u65e0\u9700\u5173\u5fc3\u8fd9\u4e2a\u53c2\u6570":103,"\u4f7f\u7528\u8005\u901a\u5e38\u65e0\u9700\u5173\u5fc3":103,"\u4f7f\u7528\u8be5learning_rate_schedule\u65f6":83,"\u4f7f\u7528\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u5b8c\u6210\u5bf9\u65b0\u6570\u636e\u7684\u9884\u6d4b":11,"\u4f7f\u7528\u8fd9\u79cd\u65b9\u5f0f":[90,111],"\u4f7f\u7528\u8fdc\u7a0b\u7a00\u758f\u65b9\u5f0f\u8bad\u7ec3\u65f6":74,"\u4f7f\u7528\u9759\u6001\u5e93\u548c\u52a8\u6001\u5e93\u96be\u5ea6\u5dee\u4e0d\u591a":45,"\u4f7f\u7528c":[46,87],"\u4f7f\u7528c99\u505a\u63a5\u53e3":45,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c11\u7684\u539f\u56e0\u662f":45,"\u4f7f\u7528c99\u800c\u4e0d\u4f7f\u7528c89":45,"\u4f7f\u7528checkgrad\u6a21\u5f0f\u65f6\u7684\u53c2\u6570\u53d8\u5316\u5927\u5c0f":103,"\u4f7f\u7528cmake\u7684\u8bdd":107,"\u4f7f\u7528cpu\u4e24\u7ebf\u7a0b\u8ba1\u7b97fc4\u5c42":105,"\u4f7f\u7528cpu\u8ba1\u7b97fc4\u5c42":105,"\u4f7f\u7528docker":2,"\u4f7f\u7528docker\u5b89\u88c5\u548c\u8fd0\u884cpaddlepaddle\u53ef\u4ee5\u65e0\u9700\u8003\u8651":2,"\u4f7f\u7528docker\u5b89\u88c5\u8fd0\u884c":3,"\u4f7f\u7528docker\u5c31\u4e0d\u7528\u914d\u7f6e\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883\u4e86":0,"\u4f7f\u7528docker\u6784\u5efapaddlepaddle\u7684\u6587\u6863":77,"\u4f7f\u7528docker\u7684\u60c5\u51b5\u4e0b":1,"\u4f7f\u7528eigen\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":116,"\u4f7f\u7528fabric\u542f\u52a8\u96c6\u7fa4\u8bad\u7ec3":94,"\u4f7f\u7528init":105,"\u4f7f\u7528lstm\u4f5c\u4e3aencod":111,"\u4f7f\u7528memory\u7684rnn\u5b9e\u73b0\u4fbf\u5982\u4e0b\u56fe\u6240\u793a":111,"\u4f7f\u7528model":105,"\u4f7f\u7528openblas\u7684\u955c\u50cf":2,"\u4f7f\u7528openblas\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":116,"\u4f7f\u7528paddlepaddl":[87,90],"\u4f7f\u7528pip\u5b89\u88c5":3,"\u4f7f\u7528rdma\u8fd8\u662ftcp\u4f20\u8f93\u534f\u8bae":103,"\u4f7f\u7528regress":62,"\u4f7f\u7528swig\u53ea\u652f\u6301cpython\u89e3\u91ca\u5668":45,"\u4f7f\u7528swig\u9700\u8981\u591a\u8bed\u8a00\u7ed1\u5b9a\u7684\u5f00\u53d1\u4eba\u5458\u719f\u7ec3\u638c\u63e1swig\u914d\u7f6e":45,"\u4f7f\u7528void":45,"\u4f7f\u8be5\u5c42\u7684\u53c2\u6570\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4fdd\u6301\u4e0d\u53d8":83,"\u4f8b\u5982":[1,12,45,46,62,74,81,82,84,89,97,102,103,105,108,111,114],"\u4f8b\u5982\u4e0b\u56fe\u4e2d":107,"\u4f8b\u5982\u4e0b\u9762\u4ee3\u7801":81,"\u4f8b\u5982\u4e5f\u53ef\u5728\u7a0b\u5e8f\u8fd0\u884c\u8fc7\u7a0b\u4e2d\u518d\u52a0\u8f7d\u53e6\u5916\u4e00\u4e2a\u6a21\u578b":90,"\u4f8b\u5982\u4f7f\u7528":81,"\u4f8b\u5982\u542b\u6709\u591a\u4e2a\u901a\u9053\u7684\u56fe\u7247":89,"\u4f8b\u5982\u5728deepspeech2":41,"\u4f8b\u5982\u5bf9\u4e8ejava\u6216\u8005python":45,"\u4f8b\u5982\u5bf9\u4e8ejava\u6765\u8bf4":45,"\u4f8b\u5982\u5bf9\u4e8epython":45,"\u4f8b\u5982\u5c06\u7b2c\u4e00\u6761\u6570\u636e\u8f6c\u5316\u4e3a":111,"\u4f8b\u5982\u6587\u672c\u5206\u7c7b\u4e2d":111,"\u4f8b\u5982\u672c\u4f8b\u4e2d\u7684\u4e24\u4e2a\u7279\u5f81":111,"\u4f8b\u5982\u673a\u5668\u4e0a\u67094\u5757gpu":81,"\u4f8b\u5982c":45,"\u4f8b\u5982java\u4e0epython\u7684\u9519\u8bef\u5904\u7406\u662f\u76f4\u63a5\u6254\u51fa\u6765except":45,"\u4f8b\u5982output\u76ee\u5f55\u4e0b\u5c31\u5b58\u653e\u4e86\u8f93\u51fa\u7ed3\u679c":97,"\u4f8b\u5982python\u53ef\u4ee5\u4f7f\u7528":45,"\u4f8b\u5982python\u7684":45,"\u4f8b\u5982rnn":41,"\u4f8b\u5982sigmoid":74,"\u4f8b\u5b50\u4e2d\u4e3a3\u4e2a":91,"\u4f8b\u5b50\u4e2d\u662f":74,"\u4f8b\u5b50\u4e2d\u662f0":74,"\u4f8b\u5b50\u4e2d\u662f100":74,"\u4f8b\u5b50\u4e2d\u662f4096":74,"\u4f8b\u5b50\u4e2d\u662f8192":74,"\u4f8b\u5b50\u4e2d\u662ffc":74,"\u4f8b\u5b50\u4e2d\u662fsoftmax":74,"\u4f9bpaddlepaddle\u52a0\u8f7d":103,"\u4f9d\u636e\u662f\u5426\u5305\u542bkernel":75,"\u4f9d\u6b21\u7c7b\u63a8":62,"\u4f9d\u8d56":[1,4],"\u4f9d\u8d56\u73af\u5883\u5373\u53ef\u8fd0\u884c":2,"\u4f9d\u8d56libpython2":1,"\u4fbf\u548c\u5355\u5c42rnn\u914d\u7f6e\u4e2d\u7684":111,"\u4fbf\u662f\u5c06\u9759\u6001\u5e93\u52a0\u5165jvm\u4e2d":45,"\u4fdd\u5b58\u6a21\u578b\u53c2\u6570\u7684\u76ee\u5f55":103,"\u4fdd\u5b58\u7684\u53c2\u6570\u4e5f\u662ffloat\u7c7b\u578b":83,"\u4fdd\u5b58\u7f51\u7edc\u5c42\u8f93\u51fa\u7ed3\u679c\u7684\u76ee\u5f55":103,"\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c\u7684\u6587\u4ef6\u540d":103,"\u4fdd\u6301\u5c3d\u91cf\u5c11\u7684commit":72,"\u4fdd\u8bc1\u4f7f\u7528gpu\u8bad\u7ec3\u65f6\u4e5f\u53ef\u4ee5\u83b7\u5f97":81,"\u4fe1\u53f7\u6765\u81ea\u52a8\u7ec8\u6b62\u5b83\u542f\u52a8\u7684\u6240\u6709\u8fdb\u7a0b":93,"\u4fe1\u606f":89,"\u4fee\u590d\u6240\u6709bug\u540e":62,"\u4fee\u590ddocker\u7f16\u8bd1\u955c\u50cf\u95ee\u9898":62,"\u4fee\u6539":[42,62,96],"\u4fee\u6539\u542f\u52a8\u811a\u672c\u540e":96,"\u4fee\u6539\u6210":62,"\u4fee\u6539\u6210\u66f4\u5feb\u7684\u7248\u672c":108,"\u503c\u5f97\u6ce8\u610f\u7684\u662f":[72,111],"\u503c\u5f97\u6df1\u5165\u5206\u6790":108,"\u503c\u7c7b\u578b":105,"\u5047\u5982\u6211\u4eec\u662f\u4e09\u5206\u7c7b\u95ee\u9898":83,"\u5047\u8bbe":74,"\u5047\u8bbe\u60a8\u5df2\u7ecf\u5728\u5f53\u524d\u76ee\u5f55":2,"\u5047\u8bbe\u635f\u5931\u51fd\u6570\u662f":74,"\u5047\u8bbe\u7b2c\u4e00\u4e2alayer\u7684\u8f93\u51faa\u662f\u4e00\u4e2a":81,"\u504f\u7f6e\u53c2\u6570\u7684\u5927\u5c0f":74,"\u505a\u4e00\u4e2a\u4ecb\u7ecd":76,"\u505a\u53ea\u8bfb\u6302\u8f7d":12,"\u505a\u5982\u4e0b\u51e0\u4e2a\u64cd\u4f5c":62,"\u505a\u63a5\u53e3":45,"\u505a\u68af\u5ea6\u68c0\u6d4b":75,"\u505a\u68c0\u67e5":75,"\u505c\u6b62\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":11,"\u505c\u6b62\u52a0\u8f7d\u6570\u636e":103,"\u5141\u8bb8\u5916\u7f51\u8bbf\u95ee\u8fd9\u4e2ahttp\u670d\u52a1":107,"\u5143\u7d20":110,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u5e8f\u5217\u6240\u643a\u5e26\u7684\u91cd\u8981\u4fe1\u606f":89,"\u5143\u7d20\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684\u8f93\u5165\u4fe1\u606f":110,"\u5145\u5206\u53d1\u6325\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u4f18\u52bf":41,"\u5145\u5206\u5c55\u73b0\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u4f18\u52bf":42,"\u5148\u4ece\u5355\u7ebf\u7a0b\u5f00\u59cb":107,"\u5148\u5378\u8f7d\u4e4b\u524d\u7684\u7248\u672c":1,"\u5148\u5b8c\u6210\u5bf9\u6743\u91cd\u7684packing\u64cd\u4f5c":41,"\u5148\u5b9e\u73b0\u6a21\u578b\u63a8\u65ad\u7684api":46,"\u5148\u627e\u51fa\u53c2\u6570":82,"\u5148\u67e5\u770b\u4e00\u4e0b\u662f\u5426\u66fe\u7ecf\u5b89\u88c5\u8fc7paddl":78,"\u5148\u68c0\u67e5\u5173\u952e\u8def\u5f84\u7684\u6027\u80fd\u95ee\u9898":107,"\u514b\u9686\u4e0b\u9762":118,"\u5168\u8fde\u63a5\u5c42\u4ee5\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":74,"\u5168\u8fde\u63a5\u5c42\u6ca1\u6709\u7f51\u7edc\u5c42\u914d\u7f6e\u7684\u8d85\u53c2\u6570":74,"\u5168\u8fde\u63a5\u5c42\u7684\u5b9e\u73b0\u4f4d\u4e8e":74,"\u5168\u8fde\u63a5\u5c42\u7684\u6bcf\u4e2a\u8f93\u51fa\u90fd\u8fde\u63a5\u5230\u4e0a\u4e00\u5c42\u7684\u6240\u6709\u7684\u795e\u7ecf\u5143\u4e0a":74,"\u5168\u8fde\u63a5\u5c42python\u5c01\u88c5\u7684\u4f8b\u5b50\u4e2d\u5305\u542b\u4e0b\u9762\u51e0\u6b65":74,"\u516c\u5f0f":2,"\u5171\u4eab\u4e00\u4e2aop\u5b9a\u4e49":75,"\u5171\u4eab\u5185\u5b58":42,"\u5171\u4eab\u540c\u4e00\u4e2a\u6743\u91cd":41,"\u5171\u4eab\u540c\u4e00\u4e2akernel\u65f6":75,"\u5171\u4eab\u5b58\u50a8\u6302\u5728\u7684\u8def\u5f84":97,"\u5173\u4e8e\u4ec0\u4e48\u662f":89,"\u5173\u4e8e\u5728paddlepaddle\u4e2d\u5982\u4f55\u4f7f\u7528eigen\u5e93":75,"\u5173\u4e8e\u65f6\u95f4\u5e8f\u5217":111,"\u5173\u4e8e\u6784\u5efa\u548c\u6d4b\u8bd5\u7684\u66f4\u591a\u4fe1\u606f":72,"\u5173\u4e8eavx":2,"\u5173\u4e8eeigen":76,"\u5173\u4e8elstm":82,"\u5173\u4e8epaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":97,"\u5173\u4e8epaddlepaddle\u7684\u66f4\u591a\u4f7f\u7528\u65b9\u6cd5\u8bf7\u53c2\u8003":84,"\u5173\u4e8eunbound":113,"\u5173\u952e\u8bcd\u5305\u62ec":72,"\u5176\u4e2d":[45,62,74,81,83,84,107,114,116,118],"\u5176\u4e2d\u5305\u542b\u4e86\u7528\u6237\u7684\u8bad\u7ec3\u7a0b\u5e8f":91,"\u5176\u4e2d\u5305\u542b\u6240\u4f9d\u8d56\u7684\u6240\u6709\u7b2c\u4e09\u65b9\u5e93":117,"\u5176\u4e2d\u5305\u542b\u6240\u6709c":117,"\u5176\u4e2d\u5305\u542bpaddlepaddle\u7684c":117,"\u5176\u4e2d\u6bcf\u4e2a\u5143\u7d20\u662f\u53cc\u5c42\u5e8f\u5217\u4e2d\u6bcf\u4e2asubseq\u6700\u540e\u4e00\u4e2a":110,"\u5176\u4e2dcheckgrad\u4e3b\u8981\u4e3a\u5f00\u53d1\u8005\u4f7f\u7528":103,"\u5176\u4e2dmean\u548cstd\u662f\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u53c2\u6570":103,"\u5176\u4e2dx\u8868\u793a\u8f93\u5165\u6570\u636e\u662f\u4e00\u4e2a\u7ef4\u5ea6\u4e3a2\u7684\u7a20\u5bc6\u5411\u91cf":84,"\u5176\u4e3b\u8981\u63a5\u53e3\u5982\u4e0b":76,"\u5176\u4ed6\u4eba\u53ef\u4ee5\u590d\u73b0\u95ee\u9898\u4ee5\u4fbf\u5e2e\u52a9":0,"\u5176\u4ed6\u5185\u5b58\u6742\u9879":81,"\u5176\u4ed6\u5185\u5b58\u6742\u9879\u662f\u6307paddlepaddle\u672c\u8eab\u6240\u7528\u7684\u4e00\u4e9b\u5185\u5b58":81,"\u5176\u4ed6\u51fd\u6570\u5747\u8fd4\u56de":46,"\u5176\u4ed6\u6240\u6709\u5c42\u90fd\u4f1a\u4f7f\u7528gpu\u8ba1\u7b97":105,"\u5176\u4ed6\u7528\u6237\u7684fork\u7248\u672c\u5e93\u5e76\u4e0d\u9700\u8981\u4e25\u683c\u9075\u5b88":62,"\u5176\u4ed6\u7684\u4f9d\u8d56\u8f6f\u4ef6":1,"\u5176\u4ed6\u914d\u7f6e\u53c2\u6570":[116,117],"\u5176\u4ed6\u9ad8\u7ea7\u529f\u80fd\u5305\u62ec\u5b9a\u4e49\u591a\u4e2amemori":114,"\u5176\u4f1a\u81ea\u52a8\u88ab\u52a0\u5165\u7f16\u8bd1\u5217\u8868":74,"\u5176\u547d\u4ee4\u5982\u4e0b":107,"\u5176\u5b83\u53ef\u9009\u7f16\u8bd1\u9009\u9879\u6309\u9700\u8fdb\u884c\u8bbe\u5b9a":87,"\u5176\u5b83layer\u7684\u8f93\u51fa":113,"\u5176\u5b9e\u4e5f\u662f\u548c\u6bcf\u4e2amini":81,"\u5176\u6b21":111,"\u5176\u8bf4\u660e\u5982\u4e0b":111,"\u5176\u8f6c\u6362\u6b21\u6570\u51cf\u5c11\u81f3":41,"\u5176\u8f93\u51fa\u88ab\u7528\u4f5cmemory\u7684\u521d\u59cb\u503c":114,"\u5176name\u7531\u53c2\u6570":82,"\u5177\u4f53\u4f7f\u7528\u65b9\u6cd5\u4e3a":[46,81],"\u5177\u4f53\u505a\u6cd5\u8bf7\u53c2\u8003":0,"\u5177\u4f53\u539f\u56e0\u53c2\u8003":46,"\u5177\u4f53\u53ef\u4ee5\u53c2\u8003":[74,81],"\u5177\u4f53\u53ef\u4ee5\u53c2\u8003mkl":42,"\u5177\u4f53\u53ef\u53c2\u8003\u6587\u6863":113,"\u5177\u4f53\u5b9e\u73b0\u65b9\u5f0f\u6bd4\u5982":[41,42],"\u5177\u4f53\u60c5\u51b5\u56e0\u4eba\u800c\u5f02":108,"\u5177\u4f53\u64cd\u4f5c\u5982\u4e0b":78,"\u5177\u4f53\u6b65\u9aa4\u5982\u4e0b":78,"\u5177\u4f53\u7684\u5b8c\u6210\u72b6\u6001\u53ef\u4ee5\u53c2\u89c1":42,"\u5177\u4f53\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":78,"\u5177\u4f53\u8bf7\u53c2\u8003":[46,72],"\u5177\u4f53\u8bf7\u89c1":72,"\u5177\u6709\u76f8\u540c\u7684\u7ed3\u679c\u4e86":111,"\u5185":114,"\u5185\u5b58":108,"\u5185\u5b58\u4e0d\u8db3":79,"\u5185\u5b58\u5bb9\u9650\u9608\u503c":103,"\u5185\u5bb9":75,"\u5185\u5bb9\u5982\u4e0b":96,"\u5185\u5c42\u5e8f\u5217\u5728":89,"\u5185\u5c42inner_step\u7684recurrent_group\u548c\u5355\u5c42\u5e8f\u5217\u7684\u51e0\u4e4e\u4e00\u6837":111,"\u5185\u5df2\u7ecf\u5305\u542bpaddlepaddle\u7684\u6267\u884c\u7a0b\u5e8f\u4f46\u662f\u8fd8\u6ca1\u4e0a\u8ff0\u529f\u80fd":97,"\u5185\u7f6e\u7684":90,"\u5185\u90e8":[90,97],"\u5185\u90e8\u5b58\u50a8":42,"\u5185\u90e8\u7531":[89,90],"\u5185\u90e8\u9a71\u52a8python\u89e3\u91ca\u5668\u8fdb\u884c\u6a21\u578b\u914d\u7f6e\u89e3\u6790\u548c\u6570\u636e\u8bfb\u53d6":45,"\u518d\u4ee5":75,"\u518d\u505a\u4e00\u5b9a\u7684reshap":82,"\u518d\u5199\u5165\u7f51\u7edc\u53c2\u6570":83,"\u518d\u5728\u6bcf\u4e00\u4e2aapi\u4e2d\u81ea\u5df1\u68c0\u67e5\u7c7b\u578b":45,"\u518d\u57fa\u4e8e":62,"\u518d\u5b89\u88c5":[4,78],"\u518d\u5bf9\u6bcf\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u8fdb\u884c\u5904\u7406":111,"\u518d\u5bf9\u6bcf\u4e00\u53e5\u8bdd\u7684\u7f16\u7801\u5411\u91cf\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u6bb5\u843d\u7684\u5411\u91cf":111,"\u518d\u5bf9\u8fd9\u4e2a\u6bb5\u843d\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":111,"\u518d\u5c06\u66f4\u65b0\u540e\u7684\u53c2\u6570\u4e0b\u53d1\u5230\u6bcf\u4e2a\u8ba1\u7b97\u8282\u70b9":92,"\u518d\u5f00\u542f\u591a\u7ebf\u7a0b":107,"\u518d\u628a\u5df2\u8f6c\u6362\u4e3apacked\u683c\u5f0f\u7684\u6570\u636e\u4f20\u9012\u7ed9\u90a3\u4e9b\u590d\u7528\u540c\u4e00\u6570\u636e\u7684gemm":41,"\u518d\u6307\u5b9a":1,"\u518d\u68c0\u67e5\u5176\u4ed6\u90e8\u5206\u7684\u6027\u80fd\u95ee\u9898":107,"\u518d\u6b21\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":108,"\u518d\u6b21\u8fdb\u884c\u6027\u80fd\u5206\u6790":107,"\u518d\u7528\u8fd9\u4e2a\u68af\u5ea6\u53bb\u548c":74,"\u518d\u901a\u8fc7\u51fd\u6570":97,"\u518d\u91cd\u65b0\u5b89\u88c5":1,"\u5199\u4ee3\u7801":45,"\u5199\u5165\u5feb\u7167\u6570\u636e":11,"\u5199\u5165\u6587\u4ef6\u4e2d":90,"\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":74,"\u5199\u7684":107,"\u51c6\u5907\u60a8\u7684\u8ba1\u7b97\u96c6\u7fa4":101,"\u51c6\u5907\u8bad\u7ec3\u6570\u636e":98,"\u51c6\u5907\u8bad\u7ec3\u6570\u636e\u548c\u9a8c\u8bc1\u6570\u636e\u96c6":91,"\u51c6\u5907\u9884\u6d4b\u6a21\u578b\u548c":90,"\u51c6\u5907\u9884\u6d4b\u6a21\u578b\u90e8\u5206":90,"\u51cf\u5c0f\u5e8f\u5217\u7684\u957f\u5ea6":81,"\u51cf\u5c0f\u8fd9\u4e2a\u5185\u5b58\u6c60\u5373\u53ef\u51cf\u5c0f\u5185\u5b58\u5360\u7528":81,"\u51cf\u5c0fbatch":81,"\u51e0\u53f0\u5230\u51e0\u5343\u53f0\u89c4\u6a21":101,"\u51fa\u73b0":78,"\u51fa\u73b0\u4ee5\u4e0b\u9519\u8bef":83,"\u51fa\u73b0\u8be5\u9519\u8bef\u7684\u539f\u56e0\u4e00\u822c\u662f\u7528\u6237\u5bf9\u4e0d\u540clayer\u7684\u53c2\u6570":82,"\u51fa\u73b0\u8fd9\u4e2a\u95ee\u9898\u7684\u4e3b\u8981\u539f\u56e0\u662f":[4,78],"\u51fd\u6570":[41,42,74,89,107,108,114],"\u51fd\u6570\u4e2d\u64cd\u4f5c\u7684\u91cd\u8981\u53d8\u91cf\u7684\u8be6\u7ec6\u89e3\u91ca":75,"\u51fd\u6570\u5047\u8bbe":114,"\u51fd\u6570\u52a0\u5230\u4ee3\u7801\u4e2d":108,"\u51fd\u6570\u5373\u53ef\u5b8c\u6210\u8f6c\u6362":12,"\u51fd\u6570\u53ea\u5173\u6ce8\u4e8ernn\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8ba1\u7b97":113,"\u51fd\u6570\u540d":107,"\u51fd\u6570\u540d\u4e3a":46,"\u51fd\u6570\u547d\u540d":45,"\u51fd\u6570\u5b9a\u4e49\u8f93\u5165":75,"\u51fd\u6570\u5b9e\u9645\u4f7f\u7528\u7684\u603b\u65f6\u95f4":107,"\u51fd\u6570\u5f97\u5230\u7684\u68af\u5ea6\u53bb\u5bf9\u6bd4":74,"\u51fd\u6570\u5fc5\u987b\u5148\u8c03\u7528\u57fa\u7c7b\u4e2d\u7684\u51fd\u6570":74,"\u51fd\u6570\u5fc5\u987b\u8fd4\u56de\u4e00\u4e2a\u6216\u591a\u4e2alayer\u7684\u8f93\u51fa":113,"\u51fd\u6570\u603b\u65f6\u95f4":107,"\u51fd\u6570\u6307\u51fa\u4e86\u5728\u8bad\u7ec3\u65f6\u9700\u8981\u4ece\u53c2\u6570\u670d\u52a1\u5668\u53d6\u51fa\u7684\u884c":74,"\u51fd\u6570\u6765\u5c06\u4fe1\u606f\u8f93\u51fa\u5230\u754c\u9762\u4e2d":108,"\u51fd\u6570\u7684\u5b9e\u73b0\u662f\u6b63\u786e\u7684":74,"\u51fd\u6570\u7684\u5f00\u5934\u5fc5\u987b\u8c03\u7528":74,"\u51fd\u6570\u7684\u603b\u5171\u8017\u65f6\u5f88\u957f":107,"\u51fd\u6570\u7684\u8c03\u7528\u6b21\u6570":107,"\u51fd\u6570\u80fd\u591f\u5c06\u4f7f\u7528":90,"\u51fd\u6570\u91cc\u5b9e\u73b0":75,"\u5206\u4e3a":90,"\u5206\u522b\u4e3a\u6570\u636e\u8f93\u5165\u6dfb\u52a0\u5916\u5c42\u5e8f\u5217\u548c\u5185\u5c42\u5e8f\u5217\u7684\u5e8f\u5217\u4fe1\u606f":89,"\u5206\u522b\u4ece\u8bcd\u8bed\u548c\u53e5\u5b50\u7ea7\u522b\u7f16\u7801\u8f93\u5165\u6570\u636e":113,"\u5206\u522b\u4ee3\u8868\u8f93\u5165\u6570\u636e":42,"\u5206\u522b\u4f7f\u7528\u5355\u53cc\u5c42rnn\u4f5c\u4e3a\u7f51\u7edc\u914d\u7f6e\u7684\u6a21\u578b":111,"\u5206\u522b\u5b9a\u4e49\u5b50\u53e5\u7ea7\u522b\u548c\u8bcd\u8bed\u7ea7\u522b\u4e0a\u9700\u8981\u5b8c\u6210\u7684\u8fd0\u7b97":113,"\u5206\u522b\u5bf9\u5e94capi":62,"\u5206\u522b\u662f":110,"\u5206\u522b\u662frnn\u72b6\u6001\u548c\u8f93\u5165\u7684\u53d8\u6362\u77e9\u9635":114,"\u5206\u522b\u662fsentences\u548clabel":111,"\u5206\u522b\u662fwords\u548clabel":111,"\u5206\u522b\u8ba1\u7b97\u6bcf\u4e2a\u53c2\u6570\u7684\u68af\u5ea6":74,"\u5206\u522b\u8fdb\u884c\u5e8f\u5217\u64cd\u4f5c":111,"\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":11,"\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u901a\u5e38\u4f1a\u901a\u8fc7api\u6216\u8005\u73af\u5883\u53d8\u91cf\u63d0\u4f9b\u4efb\u52a1\u8fd0\u884c\u9700\u8981\u7684\u53c2\u6570":94,"\u5206\u5e03\u5f0f\u8bad\u7ec3":106,"\u5206\u5e03\u5f0f\u8bad\u7ec3\u67b6\u6784\u5982\u4e0b\u56fe\u6240\u793a":92,"\u5206\u652f":[62,72],"\u5206\u652f\u4e00\u65e6\u5efa\u7acb":62,"\u5206\u652f\u4e0a":72,"\u5206\u652f\u4e0a\u521b\u5efa\u65b0\u5206\u652f":72,"\u5206\u652f\u4e2d":62,"\u5206\u652f\u4e3a\u5f00\u53d1":62,"\u5206\u652f\u4e3a\u6bcf\u4e00\u6b21release\u65f6\u5efa\u7acb\u7684\u4e34\u65f6\u5206\u652f":62,"\u5206\u652f\u4e3a\u7a33\u5b9a":62,"\u5206\u652f\u529f\u80fd\u7684\u5c01\u95ed":62,"\u5206\u652f\u5408\u5165":62,"\u5206\u652f\u5408\u5165master\u5206\u652f":62,"\u5206\u652f\u540c\u6b65\u4e3b\u7248\u672c\u5e93\u7684":62,"\u5206\u652f\u540d":72,"\u5206\u652f\u540d\u4e3a":62,"\u5206\u652f\u5b58\u5728\u7684\u65f6\u5019":62,"\u5206\u652f\u6d3e\u751f\u51fa\u65b0\u7684\u5206\u652f":62,"\u5206\u652f\u7528\u6765\u6d4b\u8bd5\u53ea\u9700\u8981\u8ba1\u7b97\u4e00\u4e2a\u8f93\u5165\u68af\u5ea6\u7684\u60c5\u51b5":75,"\u5206\u652f\u7684\u7248\u672c\u90fd\u662f\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5\u548c\u56de\u5f52\u6d4b\u8bd5\u7684\u7248\u672c":62,"\u5206\u652f\u7684\u7248\u672c\u90fd\u7ecf\u8fc7\u5355\u5143\u6d4b\u8bd5":62,"\u5206\u652f\u89c4\u8303":72,"\u5206\u6790\u5f97\u5230\u7684\u4fe1\u606f\u7528\u4e8e\u534f\u52a9\u8fdb\u884c\u7a0b\u5e8f\u7684\u4f18\u5316":108,"\u5206\u7247":11,"\u5206\u7c7b\u4efb\u52a1\u4e2d\u7c7b\u522b\u6807\u7b7e":89,"\u5206\u914d\u5230\u5f53\u524d\u6570\u636e\u5757\u6837\u672c\u6570\u7684\u56db\u5206\u4e4b\u4e00":103,"\u5207\u6362\u5230":72,"\u5207\u6362\u5230\u6240\u5efa\u5206\u652f":72,"\u5217\u5143\u7d20\u6392\u5217\u6210\u7684\u77e9\u5f62\u9635\u5217":89,"\u5217\u540d":107,"\u5217\u8868\u5982\u4e0b":84,"\u5219\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":74,"\u5219\u4f1a\u4f7f\u7528openblas\u4f5c\u4e3ablas\u5e93":1,"\u5219\u4f7f\u7528":117,"\u5219\u4f7f\u7528\u540c\u6b65\u8bad\u7ec3":103,"\u5219\u4f7f\u7528\u542f\u52a8\u53c2\u6570\u5b9a\u4e49\u7684\u521d\u59cb\u5316\u65b9\u6cd5\u521d\u59cb\u5316\u53c2\u6570":11,"\u5219\u4f7f\u7528\u8be5\u53c2\u6570\u4f5c\u4e3a\u9ed8\u8ba4\u503c":103,"\u5219\u53ef\u8bbe\u7f6e":[116,117,118],"\u5219\u5e76\u4e0d\u4f1a\u7b49\u5f85\u6240\u6709trainer\u63d0\u4ea4\u68af\u5ea6\u624d\u66f4\u65b0\u53c2\u6570":92,"\u5219\u5ffd\u7565":11,"\u5219\u603b\u4f1a\u663e\u793a\u963b\u9694\u6458\u8981\u4fe1\u606f":103,"\u5219\u628a\u53e6\u4e00\u4e2a\u6162\u901f\u7684kill\u6389":11,"\u5219\u662f\u5e26gui\u7684nvidia\u53ef\u89c6\u5316\u6027\u80fd\u5206\u6790\u5de5\u5177":108,"\u5219\u663e\u793a\u963b\u9694\u6027\u80fd\u7684\u6458\u8981\u4fe1\u606f":103,"\u5219\u76f4\u63a5\u5f15\u5165\u53e6\u4e00\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5219\u8bbe\u7f6e\u6210":[116,118],"\u5219\u9700\u8981\u4f7f\u7528\u7b49\u4e8e\u6743\u91cd\u53c2\u6570\u89c4\u6a21\u5927\u7ea65\u500d\u7684\u5185\u5b58":81,"\u5219\u9700\u8981\u5206\u522b\u7f16\u8bd1\u771f\u673a\u548c\u6a21\u62df\u5668\u7248\u672c":117,"\u5219\u9700\u8981\u56de\u6eda\u5230\u4e0a\u4e00\u4e2a\u68c0\u67e5\u70b9":11,"\u5219\u9700\u8981\u5728\u672c\u673a\u5b89\u88c5\u4e0b\u9762\u7ae0\u8282\u5217\u51fa\u7684":1,"\u5219\u9700\u8981\u624b\u52a8\u62f7\u8d1d\u5c5e\u4e8e\u6bcf\u4e2atrainer\u8282\u70b9\u7684\u8bad\u7ec3\u6570\u636e\u5230\u5bf9\u5e94\u7684\u8282\u70b9\u4e0a":91,"\u521b\u5efa":[42,89,90],"\u521b\u5efa\u4e00\u4e2a":86,"\u521b\u5efa\u4e00\u4e2akubernet":97,"\u521b\u5efa\u5e76\u5207\u6362\u5230\u65b0\u5206\u652f":72,"\u521b\u5efa\u6210\u529f\u540e":97,"\u521b\u5efa\u65e5\u5fd7\u76ee\u5f55":98,"\u521b\u5efa\u7a00\u758f\u77e9\u9635\u65f6\u9700\u8981\u663e\u793a\u5730\u6307\u5b9a\u77e9\u9635\u7684":89,"\u521d\u59cb\u5316\u504f\u7f6e\u5411\u91cf":74,"\u521d\u59cb\u5316\u6743\u91cd\u8868":74,"\u521d\u59cb\u5316\u6a21\u578b\u7684\u8def\u5f84":103,"\u521d\u59cb\u5316\u7236\u7c7b":74,"\u521d\u59cb\u5316biases_":74,"\u521d\u59cb\u72b6\u6001":113,"\u5220\u9664":72,"\u5220\u9664\u78c1\u76d8\u76ee\u5f55\u4e2d\u4e0d\u662f\u5f53\u524duuid\u7684\u5feb\u7167\u6587\u4ef6":11,"\u5224\u65ad\u662f\u5426\u5b89\u88c5\u6210\u529f":117,"\u5229\u7528\u5206\u5e03\u5f0f\u8bad\u7ec3\u9a7e\u9a6d\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":81,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90\u53ef\u4ee5\u5206\u4e3a\u4ee5\u4e0b\u51e0\u4e2a\u65b9\u5f0f\u6765\u8fdb\u884c":81,"\u5229\u7528\u8fd9\u79cd\u7279\u6027":113,"\u522b\u4eba\u5e2e\u4e86\u5fd9":72,"\u522b\u5fd8\u4e86":0,"\u5230":[11,78,114],"\u5230\u6307\u5b9a\u6587\u4ef6\u4e2d":90,"\u5230\u672c\u5730":72,"\u5230\u6b64":75,"\u5230\u7b2c\u4e8c\u6b65":62,"\u5236\u4f5c\u65b0\u955c\u50cf\u6765\u5b8c\u6210\u4ee5\u4e0a\u7684\u5de5\u4f5c":97,"\u5236\u4f5cpaddlepaddle\u955c\u50cf":97,"\u524d\u4e00\u7bc7\u6587\u7ae0\u4ecb\u7ecd\u4e86\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673apaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":97,"\u524d\u540e\u7684\u7f51\u7edc\u6027\u80fd":41,"\u524d\u5411\u4f20\u64ad":74,"\u524d\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u5165":74,"\u524d\u5411\u548c\u540e\u5411":74,"\u524d\u5411\u8ba1\u7b97\u4e4b\u540epaddlepaddle\u5185\u90e8\u5df2\u7ecf\u5206\u914d":90,"\u524d\u5411op\u5b9e\u73b0\u5b8c\u6210":75,"\u524d\u8005\u5728":81,"\u524d\u8005\u5b58\u50a8op\u7684\u8f93\u5165\u8f93\u51fa\u548c\u53c2\u6570\u5c5e\u6027":75,"\u524d\u8005\u622a\u65ad\u53ef\u5b66\u4e60\u53c2\u6570\u7684\u68af\u5ea6":81,"\u524d\u8005op\u7684\u5b9a\u4e49\u7ee7\u627f\u81ea":75,"\u524d\u81ea\u52a8\u68c0\u67e5\u4e00\u4e9b\u57fa\u672c\u4e8b\u5b9c":72,"\u524d\u9700\u8981\u5b89\u88c5":107,"\u524d\u9988":92,"\u529f\u80fd":28,"\u529f\u80fd\u7684\u6b63\u786e\u6027\u5305\u62ec\u9a8c\u8bc1paddlepaddle\u76ee\u524d\u7684":62,"\u52a0\u4e0a\u504f\u7f6e\u5411\u91cf":74,"\u52a0\u5165":108,"\u52a0\u6743\u548c\u7528\u6765\u751f\u6210":114,"\u52a0\u6743\u7f16\u7801\u5411\u91cf":114,"\u52a0\u8f7d\u5177\u4f53\u7f51\u7edc\u53c2\u6570":83,"\u52a0\u8f7d\u6a21\u578b\u53ef\u5176\u5b83\u591a\u79cd\u65b9\u5f0f":90,"\u52a0\u8f7d\u6a21\u578b\u9700\u540c\u65f6\u6307\u5b9a":90,"\u52a0\u8f7d\u9884\u6d4b\u6a21\u578b":90,"\u52a0\u8f7d\u9884\u8bad\u7ec3\u53c2\u6570":83,"\u52a0\u8f7dtest":103,"\u52a0\u901f\u7f16\u8bd1":1,"\u52a0\u901fpaddlepaddle\u8bad\u7ec3\u53ef\u4ee5\u8003\u8651\u4ece\u4ee5\u4e0b\u51e0\u4e2a\u65b9\u9762":81,"\u52a8\u6001\u5e93":[45,87],"\u52a9\u624b":74,"\u5305":107,"\u5305\u542b\u4e86\u67d0\u79cd\u7c7b\u578b\u7684\u7c7b\u578b\u5b9a\u4e49\u548c\u66b4\u9732\u7684\u5168\u90e8\u51fd\u6570":46,"\u5305\u542b\u4f46\u4e0d\u9650\u4e8e":1,"\u5305\u542b\u6d4b\u8bd5\u6570\u636e\u96c6\u7684\u76ee\u5f55":91,"\u5305\u542b\u8bad\u7ec3\u6570\u636e\u7684\u76ee\u5f55":91,"\u5305\u542b\u8fd9\u4e2a\u51fd\u6570\u8c03\u7528\u5176\u4ed6\u51fd\u6570\u7684\u65f6\u95f4":107,"\u5305\u542bkernel\u7684op\u548c\u4e0d\u5305\u542bkernel\u7684op":75,"\u5305\u62ec":[12,41,42,87,94,103],"\u5305\u62ec\u4e86\u7f16\u8bd1\u51fa\u7684paddlepaddle\u5934\u6587\u4ef6\u548c\u94fe\u63a5\u5e93":87,"\u5305\u62ec\u5b57\u7b26\u4e32\u5206\u914d":81,"\u5305\u62ec\u6743\u91cdw\u548c\u504f\u7f6eb":11,"\u5305\u62ec\u751f\u6210cpu":1,"\u5305\u62ec\u795e\u7ecf\u7f51\u7edc\u62d3\u6251\u7ed3\u6784":84,"\u5305\u62ecbool":105,"\u5305\u62eclinux":116,"\u5305\u62ecmkl":42,"\u5305\u7684\u65b9\u6cd5\u662f":78,"\u533a\u522b\u662f\u540c\u65f6\u5904\u7406\u4e86\u4e24\u4e2a\u8f93\u5165":111,"\u533a\u522b\u662frnn\u4f7f\u7528\u4e24\u5c42\u5e8f\u5217\u6a21\u578b":111,"\u534f\u540c\u5b8c\u6210releas":62,"\u5355\u4e2a\u503c":12,"\u5355\u4f4d\u662fmb":103,"\u5355\u5143\u6d4b\u8bd5":[0,76],"\u5355\u5143\u6d4b\u8bd5\u4f1a\u5f15\u7528site":78,"\u5355\u5143\u6d4b\u8bd5\u4f1a\u88ab\u81ea\u52a8\u52a0\u5165\u5de5\u7a0b\u8fdb\u884c\u7f16\u8bd1":75,"\u5355\u5143\u6d4b\u8bd5\u5728\u5185\u7684\u6240\u6709\u5355\u5143\u6d4b\u8bd5":0,"\u5355\u5143\u6d4b\u8bd5checkgrad_ep":102,"\u5355\u53cc\u5c42\u5e8f\u5217\u7684\u53e5\u5b50\u662f\u4e00\u6837\u7684":111,"\u5355\u53cc\u5c42rnn":112,"\u5355\u5c42":113,"\u5355\u5c42\u4e0d\u7b49\u957frnn":111,"\u5355\u5c42\u548c\u53cc\u5c42\u5e8f\u5217\u7684\u4f7f\u7528\u548c\u793a\u4f8b2\u4e2d\u7684\u793a\u4f8b\u7c7b\u4f3c":111,"\u5355\u5c42\u5e8f\u5217":[89,110],"\u5355\u5c42\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20":110,"\u5355\u5c42\u5e8f\u5217\u7b2ci\u4e2a\u5143\u7d20":110,"\u5355\u5c42\u6216\u53cc\u5c42":110,"\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":111,"\u5355\u5c42rnn":[111,113],"\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":111,"\u5355\u673acpu\u8bad\u7ec3":81,"\u5355\u673agpu\u8bad\u7ec3":81,"\u5355\u6b65\u51fd\u6570":114,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u5728":114,"\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570\u90fd\u975e\u5e38\u7b80\u5355":114,"\u5355\u6b65\u51fd\u6570\u7684\u5b9e\u73b0\u5982\u4e0b\u6240\u793a":114,"\u5355\u6d4b\u5305\u62ec\u5bf9\u6bd4\u524d\u5411op\u4e0d\u540c\u8bbe\u5907":75,"\u5355\u70b9\u6545\u969c":11,"\u5355\u7eaf\u7684":107,"\u5355\u8fdb\u5355\u51fa":113,"\u5360\u7528\u4e8617":107,"\u5373":[46,75,77,81,82,97],"\u5373\u4e0a\u8ff0\u4ee3\u7801\u4e2d\u7684\u7b2c19\u884c":111,"\u5373\u4e0b\u8f7d\u5931\u8d25":78,"\u5373\u4e0d\u5141\u8bb8\u5728":75,"\u5373\u4e0d\u9700\u8981\u4f7f\u7528memori":111,"\u5373\u4e3a\u4e00\u4e2a\u65f6\u95f4\u6b65":111,"\u5373\u4e3a\u5355\u5c42rnn\u5e8f\u5217\u7684\u4f7f\u7528\u4ee3\u7801":111,"\u5373\u4e3a\u65f6\u95f4\u5e8f\u5217\u7684\u8f93\u5165":111,"\u5373\u4e3a\u8fd9\u4e2a\u53cc\u5c42rnn\u7684\u7f51\u7edc\u7ed3\u6784":111,"\u5373\u4e8c\u7ef4\u6570\u7ec4":111,"\u5373\u4f7f\u7528":[46,82],"\u5373\u4f7f\u7528\u6237\u76f4\u63a5\u5f15\u7528\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5373\u4f7f\u95f4\u9694\u5f88\u5c0f":103,"\u5373\u4f7fc":46,"\u5373\u4f8b\u5982":46,"\u5373\u4fbf\u662f":0,"\u5373\u4fbf\u8bbe\u7f6e":78,"\u5373\u4fbfpaddl":46,"\u5373\u521d\u59cb\u72b6\u6001\u4e3a0":113,"\u5373\u5355\u65f6\u95f4\u6b65\u6267\u884c\u7684\u51fd\u6570":114,"\u5373\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":111,"\u5373\u53cc\u5c42rnn\u7684\u6bcf\u4e2a\u72b6\u6001":113,"\u5373\u53ef":72,"\u5373\u53ef\u4ee5\u6781\u5927\u7684\u52a0\u901f\u6570\u636e\u8f7d\u5165\u6d41\u7a0b":81,"\u5373\u53ef\u4f7f\u7528\u5f00\u53d1\u955c\u50cf\u6765\u7f16\u8bd1android\u7248paddlepaddl":116,"\u5373\u53ef\u5728":118,"\u5373\u53ef\u5f00\u59cb\u4e0b\u8f7d":4,"\u5373\u53ef\u5f00\u59cb\u4e0b\u9762\u7684\u6b65\u9aa4":2,"\u5373\u53ef\u663e\u793a\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c":107,"\u5373\u53ef\u68c0\u67e5\u6211\u4eec\u8c03\u4f18\u540e\u7684\u4fee\u6b63\u662f\u5426\u80fd\u591f\u6539\u5584\u7a0b\u5e8f\u7684\u6027\u80fd":107,"\u5373\u5728\u53cc\u5c42\u5e8f\u5217\u7684\u539f\u59cb\u6570\u636e\u4e2d":111,"\u5373\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":81,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":84,"\u5373\u5b8c\u6210\u67d0\u4e00\u4e2a\u4efb\u52a1\u7684\u6700\u5c11\u51fd\u6570":46,"\u5373\u5c06\u4e00\u6bb5\u8bdd\u8fdb\u884c\u5206\u7c7b":111,"\u5373\u5c06nchw\u8f6c\u6362\u6210nhwc":82,"\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":111,"\u5373\u6211\u4eec\u53ef\u4ee5\u5148\u5b9a\u4e49\u4e00\u4e2atensor":76,"\u5373\u628a\u5355\u5c42rnn\u751f\u6210\u540e\u7684subseq\u7ed9\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u53cc\u5c42seq":113,"\u5373\u6574\u4e2a\u53cc\u5c42group\u662f\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":111,"\u5373\u6574\u4e2a\u8f93\u5165\u5e8f\u5217":110,"\u5373\u6574\u6570\u6570\u7ec4":111,"\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":111,"\u5373\u662f\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684\u7f51\u7edc\u8fde\u63a5":111,"\u5373\u66b4\u9732":46,"\u5373\u7279\u5f81\u7684\u6570\u7ec4":111,"\u5373\u7f51\u5361\u540d":97,"\u5373\u8868\u793a\u4e0d\u9700\u8981\u8f6c\u6362":42,"\u5373\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0nan\u6216\u8005inf":81,"\u5373\u8bbe\u7f6e":81,"\u5373\u8fd0\u884c\u8bad\u7ec3\u7a0b\u5e8f":2,"\u5373\u8fd9\u4e2a\u52a8\u6001\u5e93\u662f\u4e0d\u4f9d\u8d56\u4e8e\u5176\u4ed6\u4efb\u4f55\u6587\u4ef6\u7684":45,"\u5373define_py_data_sources2\u5e94\u6539\u4e3a":83,"\u5373input":113,"\u5373rnn\u4e4b\u95f4\u6709\u4e00\u6b21\u5d4c\u5957\u5173\u7cfb":111,"\u5378\u8f7dpaddlepaddle\u5305":78,"\u538b\u6241\u6210\u4e3aeigen\u7684\u4e00\u7ef4tensor":76,"\u538b\u7f29\u6210\u4e00\u4e2a\u5411\u91cf":111,"\u539f\u56e0":[72,78],"\u539f\u56e0\u5728\u4e8e\u6ca1\u6709\u628a\u673a\u5668\u4e0acuda\u76f8\u5173\u7684\u9a71\u52a8\u548c\u5e93\u6620\u5c04\u5230\u5bb9\u5668\u5185\u90e8":78,"\u539f\u56e0\u662f\u6bcf\u4e2a\u56de\u590d\u90fd\u4f1a\u53d1\u9001\u4e00\u5c01\u90ae\u4ef6":72,"\u539f\u6765\u7684\u65b9\u6848":42,"\u53c2\u6570":[0,45,74,81,90,97,102],"\u53c2\u6570\u4e3a":75,"\u53c2\u6570\u5171\u4eab\u7684\u914d\u7f6e\u793a\u4f8b\u4e3a":83,"\u53c2\u6570\u548c\u73af\u5883\u53d8\u91cf":91,"\u53c2\u6570\u670d\u52a1\u5668":[92,102],"\u53c2\u6570\u670d\u52a1\u5668\u4e4b\u95f4\u4e0d\u76f8\u4e92\u4f9d\u8d56":92,"\u53c2\u6570\u670d\u52a1\u5668\u4e5f\u4e0d\u4f1a\u7b49\u5f85\u8ba1\u7b97\u8282\u70b9\u5168\u90e8\u90fd\u63d0\u4ea4\u68af\u5ea6\u4e4b\u540e\u624d\u5f00\u59cb\u4e0b\u4e00\u6b65":92,"\u53c2\u6570\u670d\u52a1\u5668\u63a5\u6536\u4ece\u8ba1\u7b97\u8282\u70b9\u4e0a\u4f20\u7684\u68af\u5ea6":92,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":103,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u76d1\u542c\u7aef\u53e3":103,"\u53c2\u6570\u670d\u52a1\u5668\u7684\u7f51\u7edc\u8bbe\u5907\u540d\u79f0":103,"\u53c2\u6570\u670d\u52a1\u5668\u7684ip\u5730\u5740":103,"\u53c2\u6570\u670d\u52a1\u5668\u7a00\u758f\u66f4\u65b0\u7684\u53c2\u6570\u5206\u5757\u5927\u5c0f":103,"\u53c2\u6570\u6765\u63a7\u5236\u7f13\u5b58\u65b9\u6cd5":81,"\u53c2\u6570\u6982\u8ff0":104,"\u53c2\u6570\u7684\u4e2a\u6570\u548c\u53c2\u6570\u5217\u8868":90,"\u53c2\u6570\u7684\u89e3\u6790":97,"\u53c2\u6570\u8bbe\u7f6e":80,"\u53c2\u6570\u8bbe\u7f6e\u4e86\u5916\u5c42":111,"\u53c2\u6570\u8bf4\u660e":91,"\u53c2\u6570\u8bf4\u660e\u5bb9\u5668\u5df2\u4ea4\u4e92\u5f0f\u8fd0\u884c":2,"\u53c2\u6570\u8f93\u5165":81,"\u53c2\u6570\u9700\u8981\u5b9e\u73b0":114,"\u53c2\u7167\u4e0a\u8ff0\u6b65\u9aa4\u66f4\u65b0":72,"\u53c2\u8003":[2,28,45],"\u53c2\u8003\u4e0b\u56fe":62,"\u53c2\u8003\u5f3a\u8c03\u90e8\u5206":108,"\u53c2\u8003\u65f6\u95f4\u5e8f\u5217":111,"\u53c2\u8003\u6837\u4f8b\u6570\u636e\u51c6\u5907\u811a\u672c":91,"\u53c2\u8003\u955c\u50cf\u7684":97,"\u53c8\u53ef\u4ee5\u907f\u514d\u4e0d\u5fc5\u8981\u7684\u8f6c\u6362":42,"\u53c8\u662f\u4e00\u4e2a\u5355\u5c42\u7684\u5e8f\u5217":110,"\u53c8\u8981\u4fdd\u8bc1\u6570\u636e\u662f\u968f\u673a\u7684":81,"\u53ca":74,"\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u9690\u85cf\u72b6\u6001":114,"\u53cc\u5411\u9a8c\u8bc1":28,"\u53cc\u5c42":113,"\u53cc\u5c42\u4e0d\u7b49\u957frnn":111,"\u53cc\u5c42\u5e8f\u5217":[89,110],"\u53cc\u5c42\u5e8f\u5217\u5728\u5904\u7406\u957f\u5e8f\u5217\u7684\u4efb\u52a1\u6216\u662f\u6784\u5efa\u5c42\u7ea7\u6a21\u578b\u65f6\u4f1a\u53d1\u6325\u4f5c\u7528":89,"\u53cc\u5c42\u5e8f\u5217\u6216\u5355\u5c42\u5e8f\u5217":110,"\u53cc\u5c42\u5e8f\u5217\u6570\u636e\u4e00\u5171\u67094\u4e2a\u6837\u672c":111,"\u53cc\u5c42\u5e8f\u5217\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u5e8f\u5217":110,"\u53cc\u5c42\u5e8f\u5217\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":113,"\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u4e2d\u6bcf\u4e2a\u5143\u7d20":110,"\u53cc\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"\u53cc\u5c42\u5e8f\u5217\u9700\u8981\u8bbe\u7f6e\u5206\u522b\u4e3a\u5916\u5c42\u5e8f\u5217\u548c\u5185\u5c42\u5e8f\u5217\u5206\u522b\u8bbe\u7f6e":89,"\u53cc\u5c42\u6216\u8005\u5355\u5c42":110,"\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u7684\u4ee3\u7801":111,"\u53cc\u5c42rnn":113,"\u53cc\u5c42rnn\u6570\u636e\u968f\u610f\u52a0\u4e86\u4e00\u4e9b\u9694\u65ad":111,"\u53cc\u5c42rnn\u987e\u540d\u601d\u4e49":111,"\u53cc\u8fdb\u5355\u51fa":113,"\u53cc\u8fdb\u53cc\u51fa":113,"\u53cd\u5411\u4f20\u64ad":74,"\u53cd\u5411\u4f20\u64ad\u6839\u636e\u8f93\u51fa\u7684\u68af\u5ea6":74,"\u53cd\u5411\u8ba1\u7b97\u5df2\u7ecf\u81ea\u52a8\u96c6\u6210\u8fdb\u6d4b\u8bd5\u6846\u67b6":75,"\u53cd\u5411op\u7684\u68af\u5ea6\u6d4b\u8bd5":75,"\u53cd\u5411op\u7c7b":75,"\u53cd\u5411op\u7c7b\u7684\u5b9a\u4e49":75,"\u53cd\u5411opkernel\u7684\u5b9a\u4e49\u4e0e\u524d\u5411op\u7c7b\u4f3c":75,"\u53d1\u578b\u7248":62,"\u53d1\u5e03\u5230dockerhub":62,"\u53d1\u5e03docker\u955c\u50cf\u53ea\u9700\u8981\u5bf9\u81ea\u52a8push\u7684\u955c\u50cf\u6253\u4e0a":62,"\u53d1\u6563\u5230\u4e86\u4e00\u4e2a\u6570\u503c\u7279\u522b\u5927\u7684\u5730\u65b9":81,"\u53d1\u884c\u548c\u7ef4\u62a4":72,"\u53d1\u9001\u53c2\u6570\u7684\u7aef\u53e3\u53f7":103,"\u53d6\u503c\u76f8\u540c\u7684layer":82,"\u53d6\u5176\u4e2d\u4e00\u4e2a\u6a21\u578bparams_pass_90":84,"\u53d6\u51b3\u4e8e":75,"\u53d8\u6362\u77e9\u9635":74,"\u53d8\u91cf\u6765\u533a\u5206layer\u7684\u5c5e\u6027":42,"\u53e5\u5b50\u662f\u7531\u8bcd\u8bed\u6784\u6210\u7684\u5e8f\u5217":89,"\u53e6\u4e00\u4e2a\u65b9\u6cd5\u662f\u4ea4\u53c9\u7f16\u8bd1":118,"\u53e6\u4e00\u4e2a\u662f\u5185\u5b58\u64cd\u4f5c\u91cf":108,"\u53e6\u4e00\u4e2a\u662f\u6bcf\u6761\u5e8f\u5217":81,"\u53e6\u4e00\u79cd\u65b9\u5f0f\u662f\u5c06\u7f51\u7edc\u5c42\u5212\u5206\u5230\u4e0d\u540c\u7684gpu\u4e0a\u53bb\u8ba1\u7b97":105,"\u53e6\u5916":[0,111],"\u53e6\u5916\u6700\u65b0\u7684pip\u5b98\u65b9\u6e90\u4e2d\u7684\u5b89\u88c5\u5305\u9ed8\u8ba4\u662fmanylinux1\u6807\u51c6":4,"\u53ea\u4f5c\u4e3aread":113,"\u53ea\u4fdd\u5b58\u6700\u540e\u4e00\u8f6e\u7684\u53c2\u6570":103,"\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":1,"\u53ea\u5bf9\u7279\u6b8a\u5728\u7ebf\u7cfb\u7edf\u8003\u8651\u4e24\u53f0\u4ee5\u4e0a\u540c\u65f6\u6545\u969c\u7684\u5bb9\u707e":11,"\u53ea\u5bf9\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u8fdb\u884c\u5e8f\u5217\u5316":90,"\u53ea\u5c06\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u8fdb\u884c\u5e8f\u5217\u5316":90,"\u53ea\u662f\u53cc\u5c42\u5e8f\u5217\u5c06\u5176\u53c8\u505a\u4e86\u5b50\u5e8f\u5217\u5212\u5206":111,"\u53ea\u66b4\u9732\u6982\u5ff5\u7684\u63a5\u53e3":46,"\u53ea\u6709\u5728\u9047\u5230\u9700\u8981":3,"\u53ea\u6709\u5f53\u8bbe\u7f6e\u4e86spars":103,"\u53ea\u7528\u4e8e\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d\u6307\u5b9a\u8f93\u5165\u6570\u636e":113,"\u53ea\u7559\u4e0b\u6838\u5fc3\u8ba1\u7b97\u5c42":90,"\u53ea\u80fd\u5728recurrent_group\u4e2d\u4f5c\u4e3astep":82,"\u53ea\u80fd\u6309\u884c\u8ba1\u7b97":82,"\u53ea\u80fd\u6d4b\u8bd5\u5355\u4e2a\u6a21\u578b":105,"\u53ea\u80fd\u8bbf\u95ee\u5b83\u4eec\u7684\u8f93\u51fa\u503c":82,"\u53ea\u80fd\u8c03\u7528paddle\u7684\u52a8\u6001\u5e93":45,"\u53ea\u8981\u4e00\u7cfb\u5217\u7279\u5f81\u6570\u636e\u4e2d\u7684":111,"\u53ea\u8981\u51fa\u73b0\u6d6e\u70b9\u6570\u5f02\u5e38":81,"\u53ea\u8bfbmemory\u8f93\u5165":113,"\u53ea\u9488\u5bf9\u5185\u5b58":81,"\u53ea\u9700\u4e2d\u65ad":93,"\u53ea\u9700\u5728\u7f16\u8bd1\u65f6\u9700\u914d\u5236\u4e0b\u9762\u8fd9\u4e9b\u7f16\u8bd1\u9009\u9879":87,"\u53ea\u9700\u7528\u60a8\u5b9a\u4e49\u7684\u76ee\u5f55\u4fee\u6539":93,"\u53ea\u9700\u8981":114,"\u53ea\u9700\u8981\u6062\u590d\u8fd9\u53f0\u8282\u70b9":11,"\u53ea\u9700\u8981\u8bbe\u7f6e\u884c\u504f\u79fb":89,"\u53ea\u9700\u8981\u94fe\u63a5":87,"\u53ea\u9700\u8fdb\u884c\u524d\u5411\u8ba1\u7b97\u800c\u65e0\u9700\u8c03\u7528\u53cd\u5411\u8ba1\u7b97":90,"\u53ef\u4ee5":[2,62,72,77],"\u53ef\u4ee5\u4ece":2,"\u53ef\u4ee5\u4ece\u6211\u4eec\u7684ci\u7cfb\u7edf\u4e2d\u4e0b\u8f7d\u6700\u65b0\u7684whl\u5b89\u88c5\u5305\u548cc":4,"\u53ef\u4ee5\u4f30\u8ba1\u51fa\u5982\u679c\u6a21\u578b\u91c7\u7528\u4e0d\u53d8\u7684\u8f93\u51fa\u6700\u5c0f\u7684cost0\u662f\u591a\u5c11":83,"\u53ef\u4ee5\u4f7f\u7528":[83,90,94],"\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u547d\u4ee4\u66f4\u65b0\u60a8\u7684pip":4,"\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u4ee3\u7801":83,"\u53ef\u4ee5\u4f7f\u7528\u76f8\u5e94\u6570\u636e\u7c7b\u578b\u7684":83,"\u53ef\u4ee5\u4f7f\u7528\u8be5\u53c2\u6570":103,"\u53ef\u4ee5\u4f7f\u7528kubernetes\u7684\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":97,"\u53ef\u4ee5\u5148\u4f7f\u7528":82,"\u53ef\u4ee5\u51cf\u5c0f\u7cfb\u7edf\u590d\u6742\u6027":11,"\u53ef\u4ee5\u51cf\u5c11\u7f13\u5b58\u6c60\u7684\u5927\u5c0f":81,"\u53ef\u4ee5\u521b\u5efa\u4e00\u4e2a":96,"\u53ef\u4ee5\u521b\u5efa\u975e":75,"\u53ef\u4ee5\u52a0\u901fpaddlepaddle\u7684\u8ba1\u7b97":2,"\u53ef\u4ee5\u53c2\u8003":[0,2,72,111,114],"\u53ef\u4ee5\u53c2\u8003\u4e0b\u9762\u7684\u6b65\u9aa4\u6392\u67e5":79,"\u53ef\u4ee5\u53c2\u8003paddlepaddl":84,"\u53ef\u4ee5\u540c\u65f6\u5728cpu":76,"\u53ef\u4ee5\u542b\u6709\u4e00\u6761\u6216\u591a\u6761\u6837\u672c":89,"\u53ef\u4ee5\u544a\u8bc9\u60a8\u67d0\u4e2a\u64cd\u4f5c\u5230\u5e95\u82b1\u4e86\u591a\u957f\u65f6\u95f4":108,"\u53ef\u4ee5\u5728":[1,93],"\u53ef\u4ee5\u5728\u4efb\u4f55\u673a\u5668\u4e0a\u6267\u884c\u7684":45,"\u53ef\u4ee5\u5728\u5171\u4eab\u5b58\u50a8\u4e0a\u67e5\u770b\u8f93\u51fa\u7684\u65e5\u5fd7\u548c\u6a21\u578b":97,"\u53ef\u4ee5\u5728\u6b64\u9875\u9762\u7684":62,"\u53ef\u4ee5\u5728\u8fd9\u4e2a":72,"\u53ef\u4ee5\u5728event_handler\u4e2d":81,"\u53ef\u4ee5\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684sgd\u65b9\u6cd5\u7684\u8bad\u7ec3":92,"\u53ef\u4ee5\u5b9e\u73b0\u4ecepaddl":76,"\u53ef\u4ee5\u5c06cmake":107,"\u53ef\u4ee5\u5c06memory\u7406\u89e3\u4e3a\u4e00\u4e2a\u65f6\u5ef6\u64cd\u4f5c":113,"\u53ef\u4ee5\u5c06op\u5206\u4e3a\u4e24\u79cd":75,"\u53ef\u4ee5\u5c1d\u8bd5\u4ee5\u4e0b\u7684\u65b9\u6cd5":2,"\u53ef\u4ee5\u5e2e\u60a8\u63d0\u4f9b\u4e00\u4e9b\u5b9a\u4f4d\u6027\u80fd\u74f6\u9888\u7684\u5efa\u8bae":108,"\u53ef\u4ee5\u5e76\u884c\u7f16\u8bd1\u5417":0,"\u53ef\u4ee5\u5feb\u901f\u5728\u672c\u5730\u542f\u52a8\u4e00\u4e2a\u5305\u542b\u4e86paddlepaddle\u5b98\u65b9book\u6559\u7a0b\u7684jupyt":2,"\u53ef\u4ee5\u6267\u884c":[4,78],"\u53ef\u4ee5\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u7f16\u8bd1\u751f\u6210\u6587\u6863":77,"\u53ef\u4ee5\u6267\u884cctest\u547d\u4ee4\u5373\u53ef":1,"\u53ef\u4ee5\u628a\u5b83\u60f3\u8c61\u4e3a\u4e00\u4e2a\u7c7b\u4f3c":0,"\u53ef\u4ee5\u628a\u672c\u5730\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5b58\u50a8\u96c6\u7fa4\u4e2d":12,"\u53ef\u4ee5\u6307\u5b9a\u540c\u65f6\u6267\u884cgpu\u4e0a\u7684\u5355\u5143\u6d4b\u8bd5":1,"\u53ef\u4ee5\u6307\u5b9a\u54ea\u4e00\u4e2a\u8f93\u5165\u548c\u8f93\u51fa\u5e8f\u5217\u4fe1\u606f\u4e00\u81f4":111,"\u53ef\u4ee5\u6307\u5b9a\u5f00\u542f\u81ea\u52a8\u68c0\u6d4bsm\u67b6\u6784":1,"\u53ef\u4ee5\u6309\u7167\u4e0b\u9762\u7684\u65b9\u6cd5":1,"\u53ef\u4ee5\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":[110,113],"\u53ef\u4ee5\u662f\u4e00\u4e2a\u975e\u5e8f\u5217":113,"\u53ef\u4ee5\u662f\u4ece\u5206\u5e03\u5f0f\u5b58\u50a8\u6302\u8f7d\u8fc7\u6765\u7684":91,"\u53ef\u4ee5\u662f\u4ee5\u4e0b\u51e0\u79cd":74,"\u53ef\u4ee5\u662f\u6574\u578b":89,"\u53ef\u4ee5\u663e\u793a\u5730\u6307\u5b9a\u4e00\u4e2alayer\u7684\u8f93\u51fa\u7528\u4e8e\u521d\u59cb\u5316memori":113,"\u53ef\u4ee5\u66f4\u6709\u6b21\u5e8f\u7684\u5b8c\u6210\u6027\u80fd\u7684\u4f18\u5316":107,"\u53ef\u4ee5\u6709\u4ee5\u4e0b\u4e24\u79cd":113,"\u53ef\u4ee5\u6709\u6548\u51cf\u5c0f\u7f51\u7edc\u7684\u963b\u585e":103,"\u53ef\u4ee5\u6709\u6548\u7684\u907f\u514dparamet":11,"\u53ef\u4ee5\u67e5\u770b":97,"\u53ef\u4ee5\u67e5\u770b\u6b64pod\u8fd0\u884c\u7684\u5bbf\u4e3b\u673a":96,"\u53ef\u4ee5\u6d4b\u8bd5\u591a\u4e2a\u6a21\u578b":105,"\u53ef\u4ee5\u7528":[0,28],"\u53ef\u4ee5\u7528\u4ee5\u4e0b\u6307\u4ee4":12,"\u53ef\u4ee5\u7528\u5982\u4e0b\u547d\u4ee4":72,"\u53ef\u4ee5\u7528\u6765\u8ba1\u7b97cpu\u51fd\u6570\u6216cuda\u5185\u6838\u7684\u65f6\u95f4\u6d88\u8017":108,"\u53ef\u4ee5\u76f4\u63a5\u8fd0\u884c":90,"\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u975e\u5e8f\u5217\u8f93\u5165":110,"\u53ef\u4ee5\u770b\u51fa":92,"\u53ef\u4ee5\u770b\u5230\u6700\u8017\u65f6\u7684\u51fd\u6570\u662fc":107,"\u53ef\u4ee5\u7cbe\u786e\u8bf4\u660e\u4e00\u4e2a\u957f\u8017\u65f6\u64cd\u4f5c\u7684\u5177\u4f53\u539f\u56e0":108,"\u53ef\u4ee5\u7ee7\u7eed\u5728\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f\u63d0\u4ea4\u4ee3\u7801":62,"\u53ef\u4ee5\u8003\u8651\u4f7f\u7528\u4e00\u4e9b\u4f18\u5316\u7b97\u6cd5":81,"\u53ef\u4ee5\u8054\u7cfbop":79,"\u53ef\u4ee5\u8054\u7cfbop\u662f\u5426\u53ef\u4ee5\u66f4\u6362\u96c6\u7fa4\u6216\u5347\u7ea7\u5f53\u524d\u96c6\u7fa4":79,"\u53ef\u4ee5\u83b7\u53d6\u7f51\u7edc\u4e2d\u5b9a\u4e49\u7684\u4efb\u610f\u591a\u4e2a":90,"\u53ef\u4ee5\u88c5\u7684\u662f":0,"\u53ef\u4ee5\u8bbe\u7f6e":[107,116,117,118],"\u53ef\u4ee5\u8bbf\u95ee\u7531recurr":82,"\u53ef\u4ee5\u8c03\u7528resize\u63a5\u53e3\u8fdb\u884c\u6539\u53d8":76,"\u53ef\u4ee5\u8f7b\u677e\u5730\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":84,"\u53ef\u4ee5\u8fd0\u884c":91,"\u53ef\u4ee5\u9009\u5728\u5728\u5f53\u524d\u673a\u5668\u5b89\u88c5\u4e5f\u53ef\u4ee5\u62f7\u8d1d\u5230\u76ee\u6807\u673a\u5668\u5b89\u88c5":1,"\u53ef\u4ee5\u9009\u62e9\u662f\u5426\u4f7f\u7528\u53c2\u6570":105,"\u53ef\u4ee5\u901a\u8fc7":72,"\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539\u8fd9\u4e24\u4e2a\u51fd\u6570\u6765\u5b9e\u73b0\u590d\u6742\u7684\u7f51\u7edc\u914d\u7f6e":114,"\u53ef\u4ee5\u901a\u8fc7\u5728":81,"\u53ef\u4ee5\u901a\u8fc7\u7f51\u9875\u6d4f\u89c8":2,"\u53ef\u4ee5\u901a\u8fc7\u8fd9\u4e2a\u8f93\u51fa\u6765\u5b8c\u6210\u81ea\u5b9a\u4e49\u7684\u8bc4\u4f30\u6307\u6807\u8ba1\u7b97\u7b49\u529f\u80fd":81,"\u53ef\u4ee5\u901a\u8fc7\u9636\u6bb5\u6027\u7684\u4fdd\u5b58\u6bcf\u4e2aparamet":11,"\u53ef\u4ee5\u91c7\u53d6\u4e0b\u9762\u51e0\u70b9\u63aa\u65bd":107,"\u53ef\u4ee5\u91cd\u547d\u540d\u8fd9\u4e2awhl\u5305\u4e3a":[4,78],"\u53ef\u53c2\u8003":90,"\u53ef\u5728":87,"\u53ef\u5728\u547d\u4ee4\u884c\u6267\u884c":117,"\u53ef\u663e\u5f0f\u6307\u5b9a\u4e3a":117,"\u53ef\u7528\u4e8e\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u89e3\u6790\u8fd9\u4e9b\u53c2\u6570":105,"\u53ef\u76f4\u63a5\u8fd0\u884c":90,"\u53ef\u80fd\u4f1a\u5bfc\u81f4\u51fa\u9519":97,"\u53ef\u80fd\u4f1a\u9020\u6210\u7f51\u7edc\u62e5\u585e":11,"\u53ef\u80fd\u7684\u4ee3\u7801\u4e3a":81,"\u53ef\u80fd\u7684\u539f\u56e0\u662f":83,"\u53ef\u80fd\u7684\u60c5\u51b5\u4e0b":108,"\u53ef\u80fd\u9700\u8981\u6ce8\u610f\u7ed9\u8fd9\u4e2a\u865a\u62df\u673a\u591a\u5206\u914d\u4e00\u4e9b":0,"\u53ef\u89c1\u8be5\u8ba1\u7b97\u7531\u4e24\u4e2a\u8f93\u5165":75,"\u53ef\u8bbe\u7f6e":[116,117],"\u53ef\u8bbe\u7f6e\u4e3a":117,"\u53ef\u8bbe\u7f6e\u7684\u76ee\u6807\u67b6\u6784\u5982\u4e0b\u8868\u6240\u793a":117,"\u53ef\u9009":[1,74,90,91],"\u53ef\u9009\u7684\u4e0d\u540c\u7f16\u8bd1\u73af\u5883docker\u955c\u50cf":1,"\u53ef\u9009\u914d\u7f6e\u9009\u9879":87,"\u53ef\u914d\u7f6e\u4e3a":87,"\u53ef\u91c7\u7528\u7b2c\u4e8c\u79cd\u65b9\u5f0f":82,"\u53f3\u4fa7\u7684":62,"\u5404\u6b21\u524d\u5411\u4e4b\u95f4\u4e5f\u90fd\u4f7f\u7528\u4e86\u76f8\u540c\u7684\u6743\u91cd":41,"\u5404\u9879\u66f4\u52a0\u5177\u4f53\u7684\u5355\u5143\u6d4b\u8bd5\u5728":75,"\u5408\u5e76\u5165\u4e00\u4e2a\u6587\u4ef6":90,"\u5408\u5e76\u6a21\u578b\u6587\u4ef6":90,"\u540c\u4e00\u6b21\u524d\u5411":41,"\u540c\u65f6":[41,42,78,81,108],"\u540c\u65f6\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":1,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u52a0\u901f\u5f00\u59cb\u8bad\u7ec3\u524d\u6570\u636e\u8f7d\u5165\u7684\u8fc7\u7a0b":81,"\u540c\u65f6\u4e5f\u53ef\u4ee5\u901a\u8fc7":72,"\u540c\u65f6\u4e5f\u80fd\u591f\u5f15\u5165\u66f4\u52a0\u590d\u6742\u7684\u8bb0\u5fc6\u673a\u5236":113,"\u540c\u65f6\u4f1a\u5f00\u542fintel":42,"\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":108,"\u540c\u65f6\u518d\u5c06":62,"\u540c\u65f6\u53c8\u5c3d\u53ef\u80fd\u5c11\u7684\u727a\u7272mkl":42,"\u540c\u65f6\u5728\u5185\u5b58\u91cc\u76f4\u63a5\u968f\u5373\u9009\u53d6\u6570\u636e\u6765\u505ashuffl":81,"\u540c\u65f6\u5c06\u53c2\u6570\u521d\u59cb\u5316\u4e3a":83,"\u540c\u65f6\u628a\u5f53\u524d\u76ee\u5f55":0,"\u540c\u65f6\u63d0\u8d77":62,"\u540c\u65f6\u6570\u636e\u683c\u5f0f\u5c31\u662f":42,"\u540c\u65f6\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":105,"\u540c\u65f6\u8bbe\u7f6e\u5185\u5b58\u7f13\u5b58\u529f\u80fd":81,"\u540c\u65f6\u8f93\u51fa\u5e8f\u5217\u5c42\u548c\u975e\u5e8f\u5217\u5c42":81,"\u540c\u6837":84,"\u540c\u6837\u4e5f\u53ef\u4ee5\u5728\u6d4b\u8bd5\u6a21\u5f0f\u4e2d\u6307\u5b9a\u6a21\u578b\u8def\u5f84":103,"\u540c\u6837\u53ef\u4ee5\u6269\u5c55\u5230\u53cc\u5c42\u5e8f\u5217\u7684\u5904\u7406\u4e0a":113,"\u540c\u6837\u53ef\u83b7\u53d6\u5230\u8f93\u5165\u8f93\u51fa\u548c\u5c5e\u6027\u53c2\u6570":75,"\u540c\u6b65\u6267\u884c\u64cd\u4f5c\u7684\u7ebf\u7a0b\u6570":103,"\u540c\u7406":75,"\u540d\u5b57\u4fee\u9970":45,"\u540e":[1,72,83,97,116,117,118],"\u540e\u5411":41,"\u540e\u5411\u4f20\u64ad":74,"\u540e\u5411\u4f20\u64ad\u7ed9\u5b9a\u8f93\u51fa\u7684\u68af\u5ea6":74,"\u540e\u5411\u65f6\u590d\u7528\u5df2\u7ecf\u8f6c\u6362\u8fc7\u7684\u6743\u91cd":41,"\u540e\u7f00\u4e3a":91,"\u540e\u8005\u5728\u6fc0\u6d3b\u51fd\u6570\u53cd\u5411\u8ba1\u7b97\u65f6\u88ab\u8c03\u7528":81,"\u540e\u8005\u622a\u65ad\u56de\u4f20\u7ed9\u524d\u5c42\u7684\u68af\u5ea6":81,"\u540e\u8005\u7528\u4e8e\u68c0\u67e5\u53c2\u6570\u5c5e\u6027\u7684\u5408\u6cd5\u6027":75,"\u540e\u8005\u7ee7\u627f\u81ea":75,"\u540e\u9988":92,"\u5411\u6307\u5b9a\u7684\u76ee\u5f55\u4e2d\u4e00\u4e2a\u65b0\u7684\u6587\u4ef6":11,"\u5411\u91cf":89,"\u5411\u91cfenable_parallel_vector":102,"\u5411paddlepaddle\u7684\u4e3b\u7248\u672c\u5e93\u63d0\u4ea4":62,"\u5417":0,"\u5426\u5219":[75,116,118],"\u5426\u5219\u4f1a\u628a":72,"\u5426\u5219\u4f7f\u7528\u591a\u673a\u8bad\u7ec3":103,"\u5426\u5219\u4f7f\u7528cpu\u6a21\u5f0f":103,"\u5426\u5219\u4f7f\u7528gpu":105,"\u5426\u5219\u5b83\u4ee5\u4e00\u4e2a\u5e8f\u5217\u8f93\u5165":114,"\u5426\u5219\u5f97\u628apaddle\u9759\u6001\u5e93\u94fe\u63a5\u5230\u89e3\u91ca\u5668\u91cc":45,"\u5426\u5219\u9891\u7e41\u7684\u591a\u8282\u70b9\u5de5\u4f5c\u7a7a\u95f4\u90e8\u7f72\u53ef\u80fd\u4f1a\u5f88\u9ebb\u70e6":93,"\u542b\u4e49":107,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u548c\u5b50\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":74,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u6574\u6570":74,"\u542b\u6709\u5e8f\u5217\u4fe1\u606f\u7684\u7a20\u5bc6\u5411\u91cf":74,"\u542f\u52a8\u4e00\u4e2a\u65b0\u7684\u7ebf\u7a0b\u5f00\u59cb\u4fdd\u5b58\u68c0\u67e5\u70b9":11,"\u542f\u52a8\u4e00\u4e2a\u6d4b\u8bd5\u96c6\u7fa4":93,"\u542f\u52a8\u53c2\u6570\u8bf4\u660e":92,"\u542f\u52a8\u5bb9\u5668\u5f00\u59cb\u8bad\u7ec3":97,"\u542f\u52a8\u5e76\u884c\u5411\u91cf\u7684\u9608\u503c":103,"\u542f\u52a8\u5feb\u901f\u5e94\u7b54":103,"\u542f\u52a8\u8bad\u7ec3\u4efb\u52a1":98,"\u542f\u7528\u68af\u5ea6\u53c2\u6570\u7684\u9608\u503c":103,"\u547d\u4ee4\u4e3a":[78,96],"\u547d\u4ee4\u521b\u5efa\u65b0\u955c\u50cf":96,"\u547d\u4ee4\u5220\u9664":[116,117,118],"\u547d\u4ee4\u53ef\u4ee5\u8bbe\u7f6e":1,"\u547d\u4ee4\u65f6":116,"\u547d\u4ee4\u6709\u65f6\u5019\u4f1a\u4ea7\u751f\u4e00\u4e9b\u4e2d\u95f4\u7ed3\u679c":0,"\u547d\u4ee4\u770b\u5230\u505c\u6b62\u540e\u4f46\u662f\u6ca1\u6709\u5220\u9664\u7684":0,"\u547d\u4ee4\u7f16\u8bd1\u6e90\u7801\u5373\u53ef":0,"\u547d\u4ee4\u884c\u4e2d\u7684":107,"\u547d\u4ee4\u884c\u53c2\u6570\u8bbe\u7f6e":106,"\u547d\u4ee4\u8bbe\u7f6e\u8be5\u7c7b\u7f16\u8bd1\u9009\u9879":1,"\u547d\u4ee4\u9009\u9879\u5e76\u4e14":93,"\u547d\u4ee4\u91cc\u90fd\u7528\u4e86":0,"\u547d\u540d\u4e3a":72,"\u547d\u540d\u89c4\u8303":75,"\u547d\u540d\u8bf7\u9075\u5b88":75,"\u548c":[0,1,12,41,42,45,46,62,72,74,75,76,77,81,82,83,87,89,91,92,93,105,107,108,111,114,116,118],"\u548c\u4e00\u4e2a\u5df2\u7ecf\u5206\u8bcd\u540e\u7684\u53e5\u5b50":111,"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":84,"\u548c\u4e0b\u9762\u5c06\u8981\u4ecb\u7ecd\u7684\u6ce8\u518c\u51fd\u6570\u4e00\u8d77\u653e\u5728":75,"\u548c\u4e2d\u6587\u6587\u6863":77,"\u548c\u4e4b\u524d\u51cf\u5c0f\u901a\u8fc7\u51cf\u5c0f\u7f13\u5b58\u6c60\u6765\u51cf\u5c0f\u5185\u5b58\u5360\u7528\u7684\u539f\u7406\u4e00\u81f4":81,"\u548c\u504f\u7f6e\u5411\u91cf":74,"\u548c\u5185\u5b58":0,"\u548c\u5217\u53f7":89,"\u548c\u53cc\u5c42\u5e8f\u5217\u542b\u6709subseq":110,"\u548c\u5bf9\u5e94\u884c\u7684\u4ee3\u7801":107,"\u548c\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":110,"\u548c\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":92,"\u548c\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165":114,"\u548c\u64cd\u4f5c\u7cfb\u7edf\u4e0a\u76f4\u63a5\u8fd0\u884c\u7684":0,"\u548c\u672a\u6765\u53ef\u80fd\u8fd8\u4f1a\u7528\u5230":42,"\u548c\u793a\u4f8b2\u4e2d\u7684\u914d\u7f6e\u7c7b\u4f3c":111,"\u548c\u79bb\u7ebf\u6570\u636e\u7684\u65b9\u5f0f":12,"\u548c\u90e8\u5206layer":113,"\u548cpool":110,"\u548cpserver\u4e4b\u95f4\u7528\u4e8e\u7a00\u758f\u7c7b\u578b\u53c2\u6570\u901a\u4fe1\u7684\u7aef\u53e3\u4e2a\u6570":91,"\u54ea\u4e2atrainer\u5148\u5b8c\u6210block\u7684\u8bad\u7ec3":11,"\u54ea\u4e9b\u4e0d\u662f":111,"\u56db\u79cd\u6570\u636e\u7c7b\u578b":84,"\u56e0\u4e3a\u5168\u8fde\u63a5\u5c42\u7684\u6fc0\u6d3b\u53ef\u4ee5\u662fsoftmax":74,"\u56e0\u4e3a\u53c2\u6570":105,"\u56e0\u4e3a\u5b83\u4eec\u7684\u8ba1\u7b97\u6548\u7387\u6bd4":114,"\u56e0\u4e3a\u5b83\u6bd4":114,"\u56e0\u4e3a\u5b98\u65b9\u955c\u50cf":97,"\u56e0\u4e3a\u6211\u4eec\u4f1a\u628a\u6240\u6709\u7f16\u8bd1\u5de5\u5177\u90fd\u5b89\u88c5\u8fdb\u4e00\u4e2a":0,"\u56e0\u4e3a\u6e90\u7801\u5c31\u5728\u672c\u673a\u4e0a":0,"\u56e0\u4e3a\u8fd9\u4e2a\u5de5\u5177\u5177\u6709web\u670d\u52a1\u754c\u9762":107,"\u56e0\u4e3a\u8fd9\u6837\u505a\u4e5f\u6ca1\u6cd5\u4fdd\u8bc1\u6d88\u9664\u968f\u673a\u6027":11,"\u56e0\u4e3ac":107,"\u56e0\u4e3apython\u7684\u641c\u7d22\u8def\u5f84\u662f\u4f18\u5148\u5df2\u7ecf\u5b89\u88c5\u7684python\u5305":78,"\u56e0\u4e3aswig\u5728\u7b2c\u4e09\u65b9\u8bed\u8a00\u4e2d\u66b4\u9732\u7684\u51fd\u6570\u540d":45,"\u56e0\u6b64":[41,74,111,113,116],"\u56e0\u6b64\u53cc\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":111,"\u56e0\u6b64\u53ef\u80fd\u6d4b\u8bd5\u4e0d\u591f\u5b8c\u5907":76,"\u56e0\u6b64\u5728\u8f6c\u6362\u65f6\u9700\u8981\u663e\u793a\u7684\u6307\u5b9a":76,"\u56e0\u6b64\u5b83\u662finteger_value_sub_sequ":111,"\u56e0\u6b64\u5f53":116,"\u56e0\u6b64\u6211\u4eec\u91c7\u7528\u8f93\u51fa\u7684\u52a0\u6743\u548c":74,"\u56e0\u6b64\u7528\u6237\u5e76\u4e0d\u9700\u8981\u5173\u5fc3\u5b83\u4eec":102,"\u56e0\u6b64\u9519\u8bef\u7684\u4f7f\u7528\u4e8c\u8fdb\u5236\u53d1\u884c\u7248\u53ef\u80fd\u4f1a\u5bfc\u81f4\u8fd9\u79cd\u9519\u8bef":78,"\u56fd\u5185\u7528\u6237\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u955c\u50cf\u6e90\u6765\u52a0\u901f\u8bbf\u95ee":2,"\u56fe1":[89,90],"\u56fe2":89,"\u56fe\u50cf\u5206\u7c7b":62,"\u56fe\u8868":2,"\u5728":[0,41,42,46,62,72,75,89,90,91,107,110,114,117],"\u5728\u4e00\u4e2a\u4e0d\u53ef\u4e2d\u65ad\u5e76\u7f3a\u5c11\u5907\u4efd\u7684\u8bad\u7ec3\u4efb\u52a1\u4e2d":11,"\u5728\u4e00\u4e2a\u529f\u80fd\u9f50\u5168\u7684kubernetes\u673a\u7fa4\u91cc":96,"\u5728\u4e00\u4e2a\u53c2\u6570\u7684\u68af\u5ea6\u88ab\u66f4\u65b0\u540e":74,"\u5728\u4e00\u4e9b\u5206\u5e03\u5f0f\u7cfb\u7edf\u4e2d":91,"\u5728\u4e00\u6b21\u8bad\u7ec3\u4e2d":107,"\u5728\u4e00\u8f6e\u4e2d\u6bcfsave":103,"\u5728\u4e0a\u56fe\u4e2d\u663e\u793a\u4e86\u5728\u4e00\u4e2a\u5b9e\u9645\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u5e94\u7528":12,"\u5728\u4e0a\u9762\u4ee3\u7801\u4e2d":111,"\u5728\u4e0a\u9762\u7684\u4ee3\u7801\u4e2d":75,"\u5728\u4e0b\u4e00\u7bc7\u4e2d":96,"\u5728\u4e0d\u540c\u96c6\u7fa4\u4e2d\u8fd0\u884c":92,"\u5728\u4e4b\u540e\u7684":81,"\u5728\u4e86\u89e3docker\u7684\u57fa\u672c\u4f7f\u7528\u65b9\u6cd5\u4e4b\u540e":2,"\u5728\u4efb\u610f\u65f6\u95f4\u67d0\u4e00\u53f0\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u4fdd\u5b58\u7684\u53c2\u6570\u53ef\u80fd\u6bd4\u53e6\u4e00\u53f0\u8981\u66f4\u65b0":92,"\u5728\u4f7f\u7528\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u65f6":91,"\u5728\u4f7f\u7528\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u8fdb\u884c\u8bad\u7ec3\u65f6":94,"\u5728\u4f7f\u7528\u540c\u6b65sgd\u8bad\u7ec3\u795e\u7ecf\u7f51\u7edc\u65f6":92,"\u5728\u4f7f\u7528\u65f6":82,"\u5728\u4f7f\u7528\u8be5\u6587\u6863\u4e4b\u524d":84,"\u5728\u4f7f\u7528c":87,"\u5728\u4f7f\u7528paddlepaddl":87,"\u5728\u4f7f\u7528twine\u4e0a\u4f20\u4e4b\u524d":62,"\u5728\u5168\u8fde\u63a5\u5c42\u4e2d":74,"\u5728\u5177\u4f53\u7684\u8ba1\u7b97\u4e2d":76,"\u5728\u51c6\u5907\u53d1\u8d77":72,"\u5728\u51fa\u73b0\u5355\u70b9\u6545\u969c\u65f6":11,"\u5728\u51fd\u6570":97,"\u5728\u5206\u5e03\u5f0f\u73af\u5883\u4e2d\u6d4b\u8bd5":103,"\u5728\u5206\u5e03\u5f0f\u8bad\u7ec3\u4e2d":103,"\u5728\u521b\u5efaparameters\u540e":83,"\u5728\u5355\u5c42\u6570\u636e\u7684\u57fa\u7840\u4e0a":111,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u52a0\u8f7d\u548c\u4fdd\u5b58\u53c2\u6570":103,"\u5728\u53c2\u6570\u670d\u52a1\u5668\u7ec8\u7aef\u6bcflog":103,"\u5728\u53cc\u5c42rnn\u4e2d\u7684\u7ecf\u5178\u60c5\u51b5\u662f\u5c06\u5185\u5c42\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u6570\u636e":111,"\u5728\u53cd\u5411\u4f20\u9012\u7684\u65f6\u5019":81,"\u5728\u53d8\u6362\u65f6\u9700\u8981\u5c06\u8f93\u5165\u5e8f\u5217\u4f20\u5165":111,"\u5728\u542f\u52a8job\u4e4b\u524d":97,"\u5728\u547d\u4ee4\u884c\u663e\u793a\u5206\u6790\u7ed3\u679c":107,"\u5728\u56de\u590d\u8bc4\u5ba1\u4eba\u610f\u89c1\u65f6":72,"\u5728\u56fe\u50cf\u4efb\u52a1\u4e2d":82,"\u5728\u591acpu\u8bad\u7ec3\u65f6\u5171\u4eab\u8be5\u53c2\u6570":103,"\u5728\u5b8c\u6210\u4e00\u5b9a\u91cf\u6570\u636e\u7684\u8bad\u7ec3\u540e":92,"\u5728\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684\u642d\u5efa\u4e4b\u540e":84,"\u5728\u5b9a\u4e49\u8f93\u5165layer\u4e4b\u540e":84,"\u5728\u5b9e\u73b0\u6bcf\u4e2a\u5b50\u7c7b\u7684\u65f6\u5019\u5c31\u4e0d\u9700\u8981\u5173\u5fc3\u5206\u652f\u7684\u4e8b\u60c5\u4e86":42,"\u5728\u5b9e\u73b0\u8fc7\u7a0b\u4e2d":46,"\u5728\u5b9e\u9645\u5e94\u7528\u4e2d":82,"\u5728\u5bb9\u5668\u4e2d\u7f16\u8f91\u4ee3\u7801":2,"\u5728\u5bb9\u5668\u521b\u5efa\u540e":97,"\u5728\u5bf9\u5bb9\u5668\u7684\u63cf\u8ff0":97,"\u5728\u5bf9\u5e94\u7684":41,"\u5728\u5c42\u4e2d\u6307\u5b9a":105,"\u5728\u5c42\u521d\u59cb\u5316\u7684\u65f6\u5019":41,"\u5728\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u4e2d":113,"\u5728\u5f00\u59cb\u8bad\u7ec3\u4e4b\u524d":12,"\u5728\u5f02\u6784\u96c6\u7fa4\u4e2d":11,"\u5728\u5f02\u6b65sgd\u4e2d":92,"\u5728\u5f15\u5165\u5176\u4ed6\u7c7b\u578b\u7684\u5934\u6587\u4ef6\u65f6":46,"\u5728\u5f53\u524d":81,"\u5728\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":74,"\u5728\u5f97\u5230":97,"\u5728\u5feb\u7167\u5199\u5165\u5b8c\u6210\u540e":11,"\u5728\u60a8\u7684\u5b9e\u9645\u73af\u5883\u4e2d":11,"\u5728\u6211\u4eec\u7684\u4f8b\u5b50\u4e2d":114,"\u5728\u6267\u884c\u65f6":76,"\u5728\u63d0\u4ea4":72,"\u5728\u642d\u5efa\u795e\u7ecf\u7f51\u7edc\u7684\u8fc7\u7a0b\u4e2d":84,"\u5728\u65e0\u7279\u6b8a\u9700\u6c42\u60c5\u51b5\u4e0b":87,"\u5728\u6709\u666e\u901a\u7684cpu":42,"\u5728\u672c\u4f8b\u4e2d":[72,105,111],"\u5728\u672c\u6559\u7a0b\u4e2d":114,"\u5728\u672c\u6587\u6863\u4e2d":28,"\u5728\u672c\u793a\u4f8b\u4e2d":111,"\u5728\u672c\u8282\u4e2d":114,"\u5728\u673a\u7fa4\u4e0a\u8fd0\u884c\u8f6c\u6362\u7a0b\u5e8f":12,"\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b":81,"\u5728\u6811\u7684\u6bcf\u4e00\u5c42\u4e0a":103,"\u5728\u6837\u4f8b\u4e2d":46,"\u5728\u6b64":[102,105],"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":114,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u76f8\u7b49":111,"\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u957f":114,"\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":97,"\u5728\u6d4b\u8bd5\u9636\u6bb5":103,"\u5728\u6e90\u7801\u76ee\u5f55\u6811\u7684\u6839\u76ee\u5f55\u4e2d\u8fd0\u884c":72,"\u5728\u751f\u6210\u65f6":114,"\u5728\u7528\u6237\u4f7f\u7528c":46,"\u5728\u76f8\u5e94\u7684\u4f18\u5316\u7b97\u6cd5\u91cc\u8bbe\u7f6elearning_rate_schedule\u53ca\u76f8\u5173\u53c2\u6570":83,"\u5728\u76f8\u5e94layer\u7684":82,"\u5728\u795e\u7ecf\u7f51\u7edc\u4e2d\u7b49\u4e8e\u4e00\u6b21\u9884\u6d4b\u5904\u7406\u7684\u6837\u672c\u6570":89,"\u5728\u7a0b\u5e8f\u5b9e\u73b0\u4e2d\u90fd\u4f1a\u8f6c\u5316\u4e3a\u4e8c\u7ef4\u77e9\u9635":89,"\u5728\u7b2c\u4e8c\u4e2atab":62,"\u5728\u7ebf\u4e0a\u7cfb\u7edf\u4e2d":91,"\u5728\u7ebf\u6a21\u578b\u9884\u6d4b\u670d\u52a1":12,"\u5728\u7ec4\u5408\u65f6":84,"\u5728\u7ec4\u7ec7\u795e\u7ecf\u7f51\u7edc\u8f93\u5165":90,"\u5728\u7ec4\u7ec7\u795e\u7ecf\u7f51\u7edc\u8f93\u5165\u65f6":89,"\u5728\u7ec8\u7aef\u6267\u884c":90,"\u5728\u7f16\u8bd1\u5bbf\u4e3b\u673a\u7248protoc\u53ef\u6267\u884c\u6587\u4ef6\u548c\u76ee\u6807\u673a\u7248openblas\u5e93\u65f6\u9700\u8981\u7528\u5230":[116,118],"\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d":74,"\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d":110,"\u5728\u8bad\u7ec3\u4e2d":114,"\u5728\u8bad\u7ec3\u4e4b\u524d":97,"\u5728\u8bad\u7ec3\u65f6":96,"\u5728\u8bad\u7ec3\u7ed3\u675f\u7684\u65f6\u5019\u518d\u4fdd\u5b58\u4e3apaddlepaddle\u7684\u683c\u5f0f":42,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":97,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6bcfshow":103,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u53c2\u6570\u7684\u6743\u91cd\u548c\u68af\u5ea6":81,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u67d0\u4e00\u4e2alayer\u7684output":81,"\u5728\u8bbe\u7f6e":[116,117,118],"\u5728\u8bc4\u5ba1\u8fc7\u7a0b\u4e2d":62,"\u5728\u8be5\u793a\u4f8b\u4e2d":83,"\u5728\u8be5\u914d\u7f6e\u76847":111,"\u5728\u8c03\u7528":90,"\u5728\u8c03\u7528c":90,"\u5728\u8f6f\u4ef6\u5de5\u7a0b\u7684\u8303\u7574\u91cc":108,"\u5728\u8f93\u51fa\u7684\u8fc7\u7a0b\u4e2d":113,"\u5728\u8fd0\u884c\u5b8c\u6027\u80fd\u5206\u6790\u540e":107,"\u5728\u8fd0\u884c\u65f6\u5c06\u795e\u7ecf\u7f51\u7edc\u7684\u591a\u4e2a\u53ef\u5b66\u4e60\u53c2\u6570\u653e\u5728\u540c\u4e00\u4e2a\u76ee\u5f55\u4e2d":90,"\u5728\u8fd0\u884c\u795e\u7ecf\u7f51\u7edc\u8ba1\u7b97\u56fe\u65f6":76,"\u5728\u8fd9\u4e2a":62,"\u5728\u8fd9\u4e2a\u4f8b\u5b50\u91cc":[74,96],"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":111,"\u5728\u8fd9\u4e2a\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":45,"\u5728\u8fd9\u4e2a\u6559\u7a0b\u4e2d":108,"\u5728\u8fd9\u4e2a\u6a21\u578b\u4e2d":114,"\u5728\u8fd9\u4e2a\u9636\u6bb5\u7684\u4ee3\u7801\u6b63\u5728\u7ecf\u5386\u56de\u5f52\u6d4b\u8bd5":62,"\u5728\u8fd9\u4e9b\u5934\u6587\u4ef6\u4e2d":46,"\u5728\u8fd9\u4e9b\u6587\u4ef6\u4e2d":46,"\u5728\u8fd9\u4e9blayer\u4e2d":111,"\u5728\u8fd9\u65f6\u771f\u6b63\u7684\u5206\u914d\u5185\u5b58":76,"\u5728\u8fd9\u6bb5\u4ee3\u7801\u4e2d":76,"\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b":[74,114],"\u5728\u8fd9\u79cd\u7ed3\u6784\u4e2d":113,"\u5728\u8fd9\u7bc7\u6587\u6863\u91cc":96,"\u5728\u8fd9\u7bc7\u6587\u7ae0\u91cc":97,"\u5728\u8fd9\u91cc":113,"\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":74,"\u5728\u8fd9\u91cc\u7528eigenvector\u6765\u8868\u793a":76,"\u5728\u8fd9\u91cc\u9700\u8981\u6ce8\u610f\u7684\u662f":76,"\u5728\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u65f6":91,"\u5728\u8fdb\u884c\u7f51\u7edc\u914d\u7f6e\u4e4b\u524d":84,"\u5728\u91c7\u7528sgd":83,"\u5728\u91cd\u6784\u524d\u7684paddlepaddle\u4e2d":42,"\u5728\u95ee\u9898\u672c\u8eab\u7684\u8ba1\u7b97\u91cf\u6bd4\u8f83\u5c0f\u7684\u65f6\u5019":41,"\u5728\u96c6\u7fa4\u4e0a\u8bad\u7ec3\u4e00\u4e2a\u7a00\u758f\u6a21\u578b\u9700\u8981\u52a0\u4e0a\u4e0b\u9762\u7684\u53c2\u6570":105,"\u5728\u975e\u5e8f\u5217\u8f93\u5165\u65f6":81,"\u5728android\u5e73\u53f0\u4e0a\u4e0d\u652f\u6301\u901a\u8fc7swig\u8c03\u7528\u6765\u8bad\u7ec3\u6216\u8005\u9884\u6d4b":116,"\u5728android\u5e73\u53f0\u4e0a\u53ea\u652f\u6301\u4f7f\u7528c":116,"\u5728batch":41,"\u5728build\u76ee\u5f55\u4e0b\u6267\u884c":78,"\u5728c":[45,89],"\u5728c\u7684\u5934\u6587\u4ef6":45,"\u5728cmake\u53c2\u6570\u914d\u7f6e\u4e0a":[116,117],"\u5728cmake\u7684\u547d\u4ee4\u884c\u4e2d":1,"\u5728eigen\u4e2d":76,"\u5728ios\u5e73\u53f0\u4e0a\u4e0d\u652f\u6301\u901a\u8fc7swig\u8c03\u7528\u6765\u8bad\u7ec3\u6216\u8005\u9884\u6d4b":117,"\u5728ios\u5e73\u53f0\u4e0a\u53ea\u652f\u6301\u4f7f\u7528c":117,"\u5728main":107,"\u5728openmpi\u96c6\u7fa4\u4e2d\u63d0\u4ea4\u8bad\u7ec3\u4f5c\u4e1a":94,"\u5728packing\u4e0a\u7684\u8017\u65f6":41,"\u5728paddl":97,"\u5728paddle\u4e2d":105,"\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":12,"\u5728paddlepaddl":89,"\u5728paddlepaddle\u4e2d":[84,113],"\u5728paddlepaddle\u4e2d\u4f7f\u7528dropout\u6709\u4e24\u79cd\u65b9\u5f0f":82,"\u5728paddlepaddle\u4e2d\u5305\u542b\u4ee5\u4e0b":82,"\u5728paddlepaddle\u5185\u90e8":[89,90],"\u5728paddlepaddle\u7684\u6587\u6863\u4e2d":111,"\u5728paramet":11,"\u5728python\u811a\u672c\u4e2d\u5b9e\u73b0\u4e0e\u524d\u5411operator\u76f8\u540c\u7684\u8ba1\u7b97\u903b\u8f91":75,"\u5728rnn\u7684\u60c5\u51b5\u4e0b":41,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49":113,"\u5728step\u51fd\u6570\u4e2d\u5b9a\u4e49memori":113,"\u5728trainer":105,"\u5728trainer\u4e2d\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u53d6\u6a21\u7684\u65b9\u6cd5\u4e3a\u6bcf\u4e2atrainer\u5206\u914d\u8bad\u7ec3\u6570\u636e\u6587\u4ef6":91,"\u5747\u4f1a\u5b58\u653e\u4e8e":87,"\u5747\u4f1a\u88ab\u5b89\u88c5\u5230includ":46,"\u5747\u5300\u5206\u5e03":83,"\u5747\u5300\u5206\u5e03\u7684\u8303\u56f4\u662f":103,"\u5747\u662f\u5728":46,"\u5747\u6709\u4e09\u4e2a\u5b50\u5e8f\u5217":111,"\u5747\u6709\u4e24\u7ec4\u7279\u5f81":111,"\u57fa\u4e8e\u53cc\u5c42\u5e8f\u5217\u8f93\u5165":113,"\u57fa\u4e8e\u7c98\u6027\u4f1a\u8bdd\u7684\u8d1f\u8f7d\u5747\u8861\u529f\u80fd":28,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":[85,90],"\u57fa\u7840\u4e0a":89,"\u57fa\u7c7b":75,"\u586b\u5199":72,"\u589e\u52a0":75,"\u589e\u52a0\u4e86\u4e00\u6761cd\u547d\u4ee4":96,"\u589e\u52a0\u4e86\u8bbe\u5907\u7c7b\u578b":75,"\u589e\u52a0\u5982\u4e0b\u53c2\u6570":105,"\u589e\u52a0\u68af\u5ea6\u68c0\u6d4b\u7684\u5355\u5143\u6d4b\u8bd5":74,"\u5904\u7406\u5668\u6709\u4e24\u4e2a\u5173\u952e\u6027\u80fd\u9650\u5236":108,"\u5904\u7406\u7684\u8f93\u5165\u5e8f\u5217\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u4e09\u79cd\u7c7b\u578b":113,"\u5907\u6ce8":108,"\u590d\u6742\u5ea6\u6216\u65f6\u95f4\u590d\u6742\u5ea6":108,"\u5916\u5c42\u5e8f\u5217\u5728":89,"\u5916\u5c42memory\u662f\u4e00\u4e2a\u5143\u7d20":111,"\u5916\u5c42outer_step\u4e2d":111,"\u5916\u90e8\u5b58\u50a8":42,"\u591a\u4e2a":90,"\u591a\u4e2a\u503c":12,"\u591a\u4e2a\u5c42\u7684\u8f93\u51fa\u77e9\u9635\u7684\u9ad8\u5ea6\u4e0d\u4e00\u81f4\u5bfc\u81f4\u62fc\u63a5\u5931\u8d25":81,"\u591a\u4e2a\u6392\u6210\u4e00\u5217\u7684\u5143\u7d20":89,"\u591a\u4e2a\u8f93\u51fa\u5c42\u5904\u7406\u591a\u4e2a\u4e0d\u540c\u957f\u5ea6\u7684\u5e8f\u5217":81,"\u591a\u4e2aip\u4f7f\u7528":91,"\u591a\u4e2aparamet":11,"\u591a\u53e5\u8bdd\u8fdb\u4e00\u6b65\u6784\u6210\u4e86\u6bb5\u843d":113,"\u591a\u673a\u8bad\u7ec3":81,"\u591a\u6b21\u8c03\u7528":41,"\u591a\u7528\u4e8e\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1":90,"\u591a\u8f6e\u5bf9\u8bdd\u7b49\u66f4\u4e3a\u590d\u6742\u7684\u8bed\u8a00\u6570\u636e":113,"\u5927\u4e8e\u7b49\u4e8e\u4e00\u4e2a":90,"\u5927\u591a\u6570\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u8bad\u7ec3\u51fd\u6570":74,"\u5927\u591a\u6570\u5c42\u9700\u8981\u8bbe\u7f6e\u4e3a":74,"\u5927\u591a\u6570\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u652f\u6301\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":74,"\u5927\u591a\u6570\u8bed\u8a00\u90fd\u652f\u6301\u4f7f\u7528c\u8bed\u8a00api":45,"\u5927\u5bb6\u53ef\u4ee5\u7528\u628a\u5f00\u53d1\u5de5\u5177\u5b89\u88c5\u8fdb\u5165":0,"\u5927\u5bb6\u53ef\u4ee5\u901a\u8fc7\u5b83\u9605\u8bfb\u6559\u7a0b":2,"\u5927\u5c0f\u4e0d\u4e00\u6837\u65f6":81,"\u5927\u6982\u82b1\u5341\u5206\u949f\u770b\u4e00\u4e0b":0,"\u5927\u90e8\u5206":107,"\u5934\u4fe1\u606f\u4e2d":83,"\u5934\u6587\u4ef6\u4e2d\u628a\u53c2\u6570\u5b9a\u4e49\u4e3a\u7c7b\u7684\u6210\u5458\u53d8\u91cf":74,"\u5934\u6587\u4ef6\u5982\u4e0b":74,"\u5982":[72,75,105,114],"\u5982\u4e0a\u4e00\u5c0f\u8282\u6240\u793a":76,"\u5982\u4e0b":91,"\u5982\u4e0b\u56fe\u6240\u793a":[108,111],"\u5982\u4e0b\u6240\u793a":105,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":73,"\u5982\u4f55\u8d21\u732e\u6587\u6863":73,"\u5982\u56fe\u4e2dtrainer":11,"\u5982\u6709":75,"\u5982\u679c\u4e00\u4e2a\u7f51\u7edc\u5c42\u9700\u8981\u914d\u7f6e\u7684\u8bdd":74,"\u5982\u679c\u4e0a\u9762\u4e24\u6b65\u51fa\u73b0\u9519\u8bef":11,"\u5982\u679c\u4e0d\u4e3a0":103,"\u5982\u679c\u4e0d\u4f7f\u7528\u5206\u5e03\u5f0f\u5b58\u50a8":91,"\u5982\u679c\u4e0d\u4f7f\u7528docker":1,"\u5982\u679c\u4e0d\u4f7f\u7528docker\u7f16\u8bd1\u73af\u5883":1,"\u5982\u679c\u4e0d\u60f3\u4f7f\u7528":77,"\u5982\u679c\u4e0d\u6307\u5b9a\u8fd9\u4e2a\u6587\u4ef6":107,"\u5982\u679c\u4e0d\u6536\u655b":83,"\u5982\u679c\u4e0d\u9700\u8981\u5916\u90e8\u5b58\u50a8\u7528\u4e8e\u8f6c\u6362":42,"\u5982\u679c\u4e3a0":103,"\u5982\u679c\u4e3a\u5426\u5219\u662f\u7528openbla":1,"\u5982\u679c\u4e3afals":103,"\u5982\u679c\u4e3atrue":103,"\u5982\u679c\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe\u7f6e":1,"\u5982\u679c\u4ec5\u4ec5\u4fee\u6539\u4e00\u4e2a\u6587\u4ef6\u4f46\u63d0\u4ea4\u4e86\u5341\u51e0\u4e2acommit":72,"\u5982\u679c\u4ecd\u7136\u5b58\u5728\u95ee\u9898":4,"\u5982\u679c\u4ed4\u7ec6\u8bbe\u7f6e\u7684\u8bdd":103,"\u5982\u679c\u4f60\u53ea\u9700\u8981\u4f7f\u7528\u7b80\u5355\u7684rnn":114,"\u5982\u679c\u4f60\u60f3\u4f7f\u7528\u8fd9\u4e9b\u7279\u6027":105,"\u5982\u679c\u4f60\u60f3\u8981\u4fdd\u5b58\u67d0\u4e9b\u5c42\u7684\u7279\u5f81\u56fe":103,"\u5982\u679c\u4f60\u66fe\u5728\u6e90\u7801\u76ee\u5f55\u4e0b\u7f16\u8bd1\u8fc7\u5176\u4ed6\u5e73\u53f0\u7684paddlepaddle\u5e93":117,"\u5982\u679c\u4f60\u66fe\u7ecf\u5728\u6e90\u7801\u76ee\u5f55\u4e0b\u7f16\u8bd1\u8fc7\u5176\u4ed6\u5e73\u53f0\u7684paddlepaddle\u5e93":[116,118],"\u5982\u679c\u4f60\u6b63\u5728\u5904\u7406\u5e8f\u5217\u6807\u8bb0\u4efb\u52a1":114,"\u5982\u679c\u4f60\u8981\u4e3a\u4e86\u6d4b\u8bd5\u800c\u589e\u52a0\u65b0\u7684\u6587\u4ef6":74,"\u5982\u679c\u4f7f\u7528":90,"\u5982\u679c\u4f7f\u7528docker\u7f16\u8bd1\u73af\u5883":1,"\u5982\u679c\u4f7f\u7528mkl\u5e76\u4e14\u673a\u5668\u542b\u6709avx2\u6307\u4ee4\u96c6":1,"\u5982\u679c\u4f7f\u7528swig\u6211\u4eec\u9700\u8981\u5c06\u5728interface\u6587\u4ef6\u91cc":45,"\u5982\u679c\u5173\u95edmkl":1,"\u5982\u679c\u51fa\u73b0\u4ee5\u4e0bpython\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86\u7684\u60c5\u51b5":78,"\u5982\u679c\u53c2\u6570\u4fdd\u5b58\u4e0b\u6765\u7684\u6a21\u578b\u76ee\u5f55":81,"\u5982\u679c\u53d1\u73b0\u6700\u65e9\u7684\u62a5\u9519\u5c31\u662f\u7f51\u7edc\u901a\u4fe1\u7684\u95ee\u9898":79,"\u5982\u679c\u540c\u65f6\u4f7f\u7528":91,"\u5982\u679c\u5728\u4f7f\u7528mkl":42,"\u5982\u679c\u5728\u70b9\u51fb\u4e0b\u9762\u94fe\u63a5\u65f6\u51fa\u73b0\u5982\u4e0b\u767b\u9646\u754c\u9762":4,"\u5982\u679c\u5728\u7f16\u8bd1":87,"\u5982\u679c\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u672a\u8bbe\u7f6easync":103,"\u5982\u679c\u5728\u8bad\u7ec3\u671f\u95f4\u540c\u65f6\u53d1\u8d77\u53e6\u5916\u4e00\u4e2a\u8fdb\u7a0b\u8fdb\u884c\u6d4b\u8bd5":103,"\u5982\u679c\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6ebatch":103,"\u5982\u679c\u5728\u8bad\u7ec3nlp\u76f8\u5173\u6a21\u578b\u65f6":83,"\u5982\u679c\u591a\u4e2aop\u4f9d\u8d56\u4e00\u4e9b\u5171\u7528\u7684\u51fd\u6570":75,"\u5982\u679c\u5931\u8d25":62,"\u5982\u679c\u5b58\u5728\u6570\u636e\u6392\u5217\u683c\u5f0f\u4e0d\u4e00\u6837\u7684\u60c5\u51b5\u65f6":42,"\u5982\u679c\u5b58\u5728\u67d0\u4e9btrainer\u6267\u884c\u901f\u5ea6\u8fc7\u6162\u4f1a\u5f71\u54cd\u6574\u4f53\u96c6\u7fa4\u7684\u901f\u5ea6":11,"\u5982\u679c\u5c06\u8fd9\u4e2a\u5185\u5b58\u6c60\u51cf\u5c0f":81,"\u5982\u679c\u5c0f\u4e8e75m":78,"\u5982\u679c\u5df2\u7ecf\u6709pod\u8fd0\u884c":97,"\u5982\u679c\u5df2\u7ecf\u6b63\u5728\u6267\u884c\u4fdd\u5b58\u68c0\u67e5\u70b9\u7684\u7ebf\u7a0b":11,"\u5982\u679c\u5e0c\u671b\u53ef\u4ee5\u5728\u540e\u53f0\u8fd0\u884cpserver\u7a0b\u5e8f":91,"\u5982\u679c\u5f53\u524dmpi\u96c6\u7fa4\u5e76\u4e0d\u652f\u6301\u4efb\u52a1\u72ec\u5360\u6a21\u5f0f":79,"\u5982\u679c\u60a8\u5728\u4f7f\u7528window":2,"\u5982\u679c\u60a8\u60f3\u8981\u66f4\u6df1\u5165\u4e86\u89e3deep":2,"\u5982\u679c\u60a8\u671f\u671b\u5728\u7f16\u8bd1\u5b8c\u6210\u540e\u7acb\u5373\u6267\u884c\u6240\u6709\u7684\u5355\u5143\u6d4b\u8bd5":1,"\u5982\u679c\u60a8\u6ca1\u6709\u542c\u8bf4":0,"\u5982\u679c\u60a8\u7684\u7535\u8111\u4e0d\u652f\u6301avx":2,"\u5982\u679c\u60a8\u7684gpu\u7406\u8bba\u53ef\u4ee5\u8fbe\u52306":108,"\u5982\u679c\u60a8\u9009\u62e9\u4e0d\u4f7f\u7528docker\u955c\u50cf":1,"\u5982\u679c\u60f3\u4f7f\u7528\u53ef\u89c6\u5316\u7684\u5206\u6790\u5668":108,"\u5982\u679c\u60f3\u5f88\u597d\u7684\u7406\u89e3\u7a0b\u5e8f\u7684\u884c\u4e3a":108,"\u5982\u679c\u60f3\u6539\u53d8\u539f\u6709tensor\u7684shape\u4fe1\u606f":76,"\u5982\u679c\u60f3\u8981\u4e86\u89e3\u53cc\u5c42rnn\u5728\u5177\u4f53\u95ee\u9898\u4e2d\u7684\u4f7f\u7528":111,"\u5982\u679c\u60f3\u8981\u542f\u7528paddlepaddle\u7684\u5185\u7f6e\u5b9a\u65f6\u5668":108,"\u5982\u679c\u6211\u4eec\u53ea\u9700\u8981\u7f16\u8bd1\u4e00\u4e2a\u53ea\u652f\u6301":0,"\u5982\u679c\u6211\u77e5\u9053\u5185\u6838\u82b1\u4e8610ms\u6765\u79fb\u52a81gb\u6570\u636e":108,"\u5982\u679c\u6307\u5b9a\u4e862\u4e2alayer\u4f5c\u4e3a\u8f93\u51fa\u5c42":81,"\u5982\u679c\u63d0\u793a\u6b63\u786e":77,"\u5982\u679c\u652f\u6301\u589e\u52a0\u6b64\u53c2\u6570\u63d0\u4ea4":79,"\u5982\u679c\u662f\u4e00\u4e2a\u5e8f\u5217\u8f93\u5165":89,"\u5982\u679c\u662f\u5176\u5b83\u7c7b\u578b":12,"\u5982\u679c\u662f\u7528\u7f16\u8bd1\u65f6\u6307\u5b9acpu\u7248\u672c":87,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165":113,"\u5982\u679c\u6709\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":113,"\u5982\u679c\u6709\u9700\u8981\u4fee\u6539\u7684\u5730\u65b9":72,"\u5982\u679c\u6709bugfix\u7684\u884c\u4e3a":62,"\u5982\u679c\u672a\u8bbe\u7f6e":103,"\u5982\u679c\u672a\u8bbe\u7f6egpu":105,"\u5982\u679c\u673a\u5668\u4e2d\u5df2\u7ecf\u5b89\u88c5\u8fc7paddlepaddl":1,"\u5982\u679c\u67d0\u4e00\u4e2a\u7c7b\u578b\u9700\u8981\u5f15\u7528\u53e6\u4e00\u4e2a\u7c7b\u578b":46,"\u5982\u679c\u67d0\u4e00\u4e2apaddl":46,"\u5982\u679c\u67d0\u4e00\u4e2apaddle\u6982\u5ff5\u5fc5\u987b\u8981\u66b4\u9732":46,"\u5982\u679c\u67d0\u4e00\u5757\u6839\u672c\u5c31\u4e0d\u600e\u4e48\u8017\u65f6":108,"\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":103,"\u5982\u679c\u6ca1\u6709\u5b89\u88c5nvidia":2,"\u5982\u679c\u6ca1\u6709\u5b9a\u4e49memori":113,"\u5982\u679c\u6ca1\u8fc7":72,"\u5982\u679c\u6d88\u606f\u6570\u636e\u592a\u5c0f":103,"\u5982\u679c\u6ee1\u8db3\u6761\u4ef6":11,"\u5982\u679c\u7528\u516c\u7528\u7684\u7535\u8111\u5f00\u53d1":0,"\u5982\u679c\u7528\u6237\u4e0d\u9700\u8981\u8bbf\u95eelstm\u7684\u4e2d\u95f4\u53d8\u91cf":82,"\u5982\u679c\u7528\u6237\u60f3\u8981\u81ea\u5b9a\u4e49\u521d\u59cb\u5316\u65b9\u5f0f":83,"\u5982\u679c\u7528\u6237\u8981\u628apaddle\u7684\u9759\u6001\u5e93":45,"\u5982\u679c\u7528\u81ea\u5df1\u7684\u7535\u8111\u5f00\u53d1":0,"\u5982\u679c\u771f\u60f3\u6316\u6398\u5185\u6838\u6df1\u5904\u7684\u67d0\u4e2a\u79d8\u5bc6":108,"\u5982\u679c\u795e\u7ecf\u7f51\u7edc\u6709\u591a\u4e2a\u8f93\u5165\u6216\u8005\u591a\u4e2a\u8f93\u51fa":[89,90],"\u5982\u679c\u7a0b\u5e8f\u5d29\u6e83\u4f60\u4e5f\u53ef\u4ee5\u624b\u52a8\u7ec8\u6b62":93,"\u5982\u679c\u7cfb\u7edf\u5b89\u88c5\u4e86\u591a\u4e2apython\u7248\u672c":78,"\u5982\u679c\u7cfb\u7edf\u652f\u6301":[4,78],"\u5982\u679c\u7cfb\u7edf\u652f\u6301\u7684\u662f":[4,78],"\u5982\u679c\u7f16\u8bd1\u65f6\u6307\u5b9a\u7f16\u8bd1cpu\u7248\u672c":87,"\u5982\u679c\u7f16\u8bd1\u65f6\u6307\u5b9a\u7f16\u8bd1gpu\u7248\u672c":87,"\u5982\u679c\u7f16\u8bd1\u7684\u65f6\u5019\u6211\u4eec\u7528\u4e86":0,"\u5982\u679c\u7f51\u7edc\u5c42\u4e0d\u9700\u8981\u8fdc\u7a0b\u7a00\u758f\u66f4\u65b0":74,"\u5982\u679c\u7f51\u7edc\u67b6\u6784\u7b80\u5355":114,"\u5982\u679c\u8981\u4e0a\u4f20gpu\u7248\u672c\u7684\u5305":62,"\u5982\u679c\u8981\u542f\u7528gpu":101,"\u5982\u679c\u8981\u8fd0\u884c\u6240\u6709\u7684\u5355\u5143\u6d4b\u8bd5":72,"\u5982\u679c\u89e3\u51b3\u4e86\u67d0\u4e2aissue\u7684\u95ee\u9898":72,"\u5982\u679c\u8bad\u7ec3\u4e00\u4e2apass":83,"\u5982\u679c\u8bad\u7ec3\u8fc7\u7a0b\u7684\u7684cost\u660e\u663e\u9ad8\u4e8e\u8fd9\u4e2a\u5e38\u6570\u8f93\u51fa\u7684cost":83,"\u5982\u679c\u8bbe\u7f6e\u8be5\u53c2\u6570":103,"\u5982\u679c\u8bc4\u5ba1\u610f\u89c1\u6bd4\u8f83\u591a":72,"\u5982\u679c\u8c03\u7528\u9759\u6001\u5e93\u53ea\u80fd\u5c06\u9759\u6001\u5e93\u4e0e\u89e3\u91ca\u5668\u94fe\u63a5":45,"\u5982\u679c\u8f93\u5165\u662f\u5e8f\u5217\u6570\u636e":89,"\u5982\u679c\u8f93\u51fa\u662f\u4e00\u4e2a\u5e8f\u5217":89,"\u5982\u679c\u8f93\u51fa\u662fno":2,"\u5982\u679c\u8fd0\u884c":78,"\u5982\u679c\u8fd8\u4e0d\u884c":78,"\u5982\u679c\u9700\u8981":87,"\u5982\u679c\u9700\u8981\u5728c\u7ef4\u5ea6\u8ba1\u7b97softmax":82,"\u5982\u679c\u9700\u8981\u5b89\u88c5\u652f\u6301gpu\u7684\u7248\u672c":[4,86],"\u5982\u679c\u9700\u8981\u624b\u52a8\u7f16\u8bd1":62,"\u5982\u679c\u9700\u8981\u6269\u5927\u77e9\u9635":74,"\u5982\u679c\u9700\u8981\u7f29\u51cf\u77e9\u9635":74,"\u5982\u679c\u9700\u8981\u83b7\u53d6":4,"\u5982\u679ccuda":75,"\u5982\u679clearning_rate\u592a\u5927":83,"\u5982\u679clearning_rate\u592a\u5c0f":83,"\u5982\u679cmkl":42,"\u5982\u679cop\u6ca1\u6709\u5b9e\u73b0cuda":75,"\u5982\u679cop\u7684\u67d0\u4e2a\u8f93\u5165\u4e0d\u53c2\u4e0e\u53cd\u5411\u68af\u5ea6\u7684\u8ba1\u7b97":75,"\u5982\u679cpaddlepaddle\u5305\u5df2\u7ecf\u5728python\u7684sit":78,"\u5982\u679cpaddlepaddle\u5e93\u9700\u8981\u540c\u65f6\u652f\u6301\u771f\u673a\u548c\u6a21\u62df\u5668":117,"\u5982\u679cparamet":11,"\u5982\u6bcf\u4e2a\u6587\u4ef6\u53ea\u6709\u4e00\u4e2a":72,"\u5982\u795e\u7ecf\u5143\u6fc0\u6d3b\u503c\u7b49":81,"\u5982\u8981build\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf":72,"\u5982\u9ad8\u4eae\u90e8\u5206":108,"\u5982train":91,"\u5b50\u53e5":113,"\u5b50\u53e5\u7684\u5355\u8bcd\u6570\u548c\u6307\u5b9a\u7684\u4e00\u4e2a\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":113,"\u5b50\u76ee\u5f55":0,"\u5b50\u7c7b\u53ea\u9700\u8981\u4f7f\u7528\u5b9a\u4e49\u597d\u7684\u63a5\u53e3":42,"\u5b57\u6bb5\u4e2d":97,"\u5b57\u6bb5\u4e3a\u4f8b":81,"\u5b57\u6bb5\u7684\u53d6\u503c":89,"\u5b57\u6bb5\u8868\u793a\u5bb9\u5668\u7684\u73af\u5883\u53d8\u91cf":97,"\u5b57\u6bb5\u8868\u793a\u8fd9\u4e2ajob\u4f1a\u540c\u65f6\u5f00\u542f3\u4e2apaddlepaddle\u8282\u70b9":97,"\u5b57\u6bb5\u8bbe\u4e3a":62,"\u5b57\u7b26\u4e32":12,"\u5b58\u50a8":[12,89,90],"\u5b58\u50a8\u6d6e\u70b9\u7c7b\u578b\u8f93\u5165":90,"\u5b58\u50a8\u7684\u538b\u7f29\u6587\u4ef6":90,"\u5b58\u6570\u6570\u636e":89,"\u5b66\u4e60":0,"\u5b66\u4e60\u6210\u672c\u9ad8":45,"\u5b66\u4e60\u7387\u4e3a":83,"\u5b83\u4eec\u4e3b\u8981\u662f\u7528\u4e8e":42,"\u5b83\u4eec\u7684\u6587\u4ef6\u540d\u662f":12,"\u5b83\u5305\u542b\u4ee5\u4e0b\u51e0\u6b65":74,"\u5b83\u5305\u542b\u4ee5\u4e0b\u53c2\u6570":74,"\u5b83\u53ea\u4f1a\u5305\u62ec\u751f\u6210\u597d\u7684\u52a8\u6001\u5e93\u548c\u5934\u6587\u4ef6":42,"\u5b83\u53eb\u505a":114,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u51cf\u5c11\u5206\u53d1\u5ef6\u8fdf":93,"\u5b83\u53ef\u4ee5\u5e2e\u52a9\u6211\u4eec\u683c\u5f0f\u5316\u6e90\u4ee3\u7801":72,"\u5b83\u53ef\u4ee5\u6307\u6d4b\u91cf\u4e00\u4e2a\u7a0b\u5e8f\u7684\u7a7a\u95f4":108,"\u5b83\u53ef\u80fd\u6709\u4e0d\u6b62\u4e00\u4e2a\u6743\u91cd":74,"\u5b83\u5b9a\u4e49\u4e86":114,"\u5b83\u5b9a\u4e49\u89e3\u7801\u7f51\u7edc\u7684":114,"\u5b83\u5c06\u88ab\u5206\u53d1\u5230":93,"\u5b83\u5e76\u4e0d\u662f\u4e00\u4e2a\u5b8c\u6574\u7684recurr":82,"\u5b83\u5e94\u8be5\u6253\u5370\u51fa\u9884\u6d4b\u4f4f\u623f\u6570\u636e\u7684\u6e05\u5355":86,"\u5b83\u652f\u6301\u591a\u7ebf\u7a0b\u66f4\u65b0":74,"\u5b83\u662finteger_value\u7c7b\u578b\u7684":111,"\u5b83\u662finteger_value_sequence\u7c7b\u578b\u7684":111,"\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":93,"\u5b83\u7684":114,"\u5b83\u7684\u529f\u80fd\u662f":75,"\u5b83\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20":110,"\u5b83\u7684\u8f93\u5165\u4e0e\u7ecf\u8fc7\u5b66\u4e60\u7684\u53c2\u6570\u505a\u5185\u79ef\u5e76\u52a0\u4e0a\u504f\u7f6e":74,"\u5b83\u8868\u793a":0,"\u5b83\u8d1f\u8d23\u51b3\u5b9a\u7f16\u8bd1\u65f6\u662f\u5426\u4f7f\u7528mklml\u548cmkl":42,"\u5b83\u9996\u5148\u8c03\u7528\u57fa\u6784\u9020\u51fd\u6570":74,"\u5b89\u88c5":107,"\u5b89\u88c5\u4e0e\u7f16\u8bd1":115,"\u5b89\u88c5\u4e0e\u7f16\u8bd1c":88,"\u5b89\u88c5\u540e":2,"\u5b89\u88c5\u540e\u7684\u76ee\u5f55\u7ed3\u6784\u4e3a":46,"\u5b89\u88c5\u597ddocker\u4e4b\u540e\u53ca\u53ef\u7528\u4ee5\u4e0b\u547d\u4ee4\u542f\u52a8\u5de5\u5177":77,"\u5b89\u88c5\u597ddocker\u4e4b\u540e\u53ef\u4ee5\u4f7f\u7528\u6e90\u7801\u76ee\u5f55\u4e0b\u7684\u811a\u672c\u6784\u5efa\u6587\u6863":77,"\u5b89\u88c5\u5b8c\u6210\u4e4b\u540e":[101,117],"\u5b89\u88c5\u5b8c\u6bd5\u540e":107,"\u5b89\u88c5\u5f00\u53d1\u5de5\u5177\u5230":0,"\u5b89\u88c5\u6587\u6863":84,"\u5b89\u88c5\u65b9\u5f0f\u6765\u5feb\u901f\u5b89\u88c5paddlepaddl":101,"\u5b89\u88c5\u6d41\u7a0b":86,"\u5b8c\u6210":72,"\u5b8c\u6210\u4e00\u4e2a\u4f20\u8f93\u52a8\u4f5c\u5b8c\u6210\u7684\u65f6\u95f4\u4e5f\u6bd4\u8f83\u77ed":28,"\u5b8c\u6210\u4e0a\u8ff0\u51c6\u5907\u4e4b\u540e":90,"\u5b8c\u6210\u4efb\u610f\u7684\u8fd0\u7b97\u903b\u8f91":113,"\u5b8c\u6210\u540evolume\u4e2d\u7684\u6587\u4ef6\u5185\u5bb9\u5927\u81f4\u5982\u4e0b":97,"\u5b8c\u6210\u5728windows\u4e0a\u5b89\u88c5\u548c\u4f7f\u7528dock":2,"\u5b8c\u6210\u5b89\u88c5":4,"\u5b8c\u6210\u5e38\u7528layer\u7684mkl":42,"\u5b8c\u6210\u5e38\u89c1\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edcvgg":42,"\u5b8c\u6210\u6570\u636e\u7684\u9884\u5904\u7406":12,"\u5b8c\u6210\u76f8\u5e94\u7684\u8ba1\u7b97":110,"\u5b8c\u6210\u81ea\u52a8\u5316\u4e8c\u8fdb\u5236\u7f16\u8bd1":62,"\u5b8c\u6210paddlepaddle\u7684\u5b89\u88c5":84,"\u5b8c\u6574\u4ee3\u7801\u53ef\u4ee5\u53c2\u8003\u793a\u4f8b":81,"\u5b8c\u6574\u4ee3\u7801\u53ef\u4ee5\u67e5\u770b":90,"\u5b8c\u6574\u6e90\u7801\u53ef\u53c2\u8003":83,"\u5b8c\u6574\u7684\u53c2\u6570\u77e9\u9635\u88ab\u5206\u5e03\u5728\u4e0d\u540c\u7684\u53c2\u6570\u670d\u52a1\u5668\u4e0a":74,"\u5b8c\u6574\u7684\u914d\u7f6e\u6587\u4ef6\u5728":114,"\u5b98\u65b9\u6587\u6863":1,"\u5b9a\u4e49":42,"\u5b9a\u4e49\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185rnn\u5355\u5143\u5b8c\u6210\u7684\u8ba1\u7b97":113,"\u5b9a\u4e49\u4e00\u4e9b\u9664\u4e86layer\u548cmemory\u76f8\u5173\u7684\u7c7b\u548c\u51fd\u6570":42,"\u5b9a\u4e49\u4e86\u4e00\u4e2a\u53ea\u8bfb\u7684memori":113,"\u5b9a\u4e49\u4e86lstm\u5355\u5143\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u7684\u8ba1\u7b97\u8fc7\u7a0b":82,"\u5b9a\u4e49\u4f4d\u7f6e":75,"\u5b9a\u4e49\u5728\u5916\u5c42":113,"\u5b9a\u4e49\u5f02\u6b65\u8bad\u7ec3\u7684\u957f\u5ea6":103,"\u5b9a\u4e49\u6e90\u8bed\u53e5\u7684\u6570\u636e\u5c42":114,"\u5b9a\u4e49\u7684\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784":90,"\u5b9a\u4e49\u7c7b\u578b":75,"\u5b9a\u4e49\u89e3\u7801\u5668\u7684memori":114,"\u5b9a\u4e49\u8f93\u5165":75,"\u5b9a\u4e49\u8f93\u51fa":75,"\u5b9a\u4e49\u8f93\u51fa\u51fd\u6570":114,"\u5b9a\u4e49\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5355\u6b65\u51fd\u6570":114,"\u5b9d\u5854\u7684\u5e95\u7aef\u9700\u8981\u575a\u5b9e\u7684\u57fa\u5ea7\u6765\u652f\u6491":84,"\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":111,"\u5b9e\u73b0\u5177\u4f53\u7684\u51fd\u6570\u529f\u80fd\u5373\u53ef":42,"\u5b9e\u73b0\u524d\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":74,"\u5b9e\u73b0\u5355\u6b65\u51fd\u6570":114,"\u5b9e\u73b0\u540e\u5411\u4f20\u64ad\u7684\u90e8\u5206\u6709\u4e0b\u9762\u51e0\u4e2a\u6b65\u9aa4":74,"\u5b9e\u73b0\u5728":75,"\u5b9e\u73b0\u5bf9":76,"\u5b9e\u73b0\u65b0\u7684op\u90fd\u6dfb\u52a0\u81f3\u76ee\u5f55":75,"\u5b9e\u73b0\u6784\u9020\u51fd\u6570":74,"\u5b9e\u73b0\u7684":82,"\u5b9e\u73b0\u7b80\u5355":45,"\u5b9e\u73b0\u7ec6\u8282":74,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u524d\u5411\u4f20\u64ad":74,"\u5b9e\u73b0\u7f51\u7edc\u5c42\u7684\u540e\u5411\u4f20\u64ad":74,"\u5b9e\u73b0\u8bcd\u8bed\u548c\u53e5\u5b50\u4e24\u4e2a\u7ea7\u522b\u7684\u53cc\u5c42rnn\u7ed3\u6784":113,"\u5b9e\u73b0\u8be5\u5c42\u7684c":74,"\u5b9e\u9645\u4e0a\u4f7f\u7528\u4e86":82,"\u5b9e\u9645\u4e0a\u9700\u8981\u7684\u8f93\u51fa\u7ed3\u679c\u662f\u4e24\u4e2a\u77e9\u9635":81,"\u5bb9\u5668\u8fd0\u884c\u90fd\u8fd0\u884c":97,"\u5bb9\u5668\u9ed8\u8ba4\u6267\u884c":116,"\u5bbd\u5ea6":89,"\u5bbd\u5ea6\u4e3a":89,"\u5bbd\u5ea6\u7b49\u4e8e\u914d\u7f6e\u4e2dlayer\u7684s":81,"\u5bbf\u4e3b\u673a\u7684c":[116,117,118],"\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":108,"\u5bf9":90,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f01\u5411\u91cf":84,"\u5bf9\u4e00\u4e2a5\u7ef4\u975e\u5e8f\u5217\u7684\u7a00\u758f\u6d6e\u70b9\u5411\u91cf":84,"\u5bf9\u4e8e":114,"\u5bf9\u4e8e\u4e0d\u540c\u7684\u8bad\u7ec3\u4efb\u52a1":91,"\u5bf9\u4e8e\u4e0d\u540c\u7684\u96c6\u7fa4\u5e73\u53f0":94,"\u5bf9\u4e8e\u4e0d\u540c\u8bed\u8a00":45,"\u5bf9\u4e8e\u4e24\u79cd\u4e0d\u540c\u7684\u8f93\u5165\u6570\u636e\u7c7b\u578b":111,"\u5bf9\u4e8e\u4e60\u60ef\u4f7f\u7528windows\u548cmacos\u7684\u5f00\u53d1\u8005\u6765\u8bf4":0,"\u5bf9\u4e8e\u5355\u5c42rnn":111,"\u5bf9\u4e8e\u5355\u5c42rnn\u7684\u6570\u636e\u4e00\u5171\u6709\u4e24\u4e2a\u6837\u672c":111,"\u5bf9\u4e8e\u53cc\u5c42rnn":111,"\u5bf9\u4e8e\u540c\u4e00\u6bb5c":45,"\u5bf9\u4e8e\u540c\u6837\u7684\u6570\u636e":111,"\u5bf9\u4e8e\u540c\u6837\u8bbe\u7f6e\u7684\u7f51\u7edc\u6a21\u578b":41,"\u5bf9\u4e8e\u56fd\u5185\u7528\u6237":2,"\u5bf9\u4e8e\u591a\u8bed\u8a00\u63a5\u53e3":45,"\u5bf9\u4e8e\u5927\u591a\u6570\u8bed\u8a00":45,"\u5bf9\u4e8e\u5e8f\u5217\u957f\u5ea6":41,"\u5bf9\u4e8e\u6027\u80fd\u7684\u5173\u952e\u8def\u5f84\u90fd\u505a\u51fa\u4e86\u7ea2\u8272\u6807\u8bb0":107,"\u5bf9\u4e8e\u6211\u4eec\u652f\u6301\u7684\u5168\u90e8\u77e9\u9635\u64cd\u4f5c":74,"\u5bf9\u4e8e\u6709\u53c2\u6570\u7684\u5c42":42,"\u5bf9\u4e8e\u672c\u6837\u4f8b\u4ee3\u7801":91,"\u5bf9\u4e8e\u6bb5\u843d\u7684\u6587\u672c\u5206\u7c7b":111,"\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5355\u5c42rnn\u7684\u6570\u636e":111,"\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u65b0\u52a0\u7684rnn":41,"\u5bf9\u4e8e\u6bcf\u79cd\u7c7b\u578b":46,"\u5bf9\u4e8e\u6bcf\u79cdc":46,"\u5bf9\u4e8e\u8fd9\u6837\u7684\u9700\u6c42":90,"\u5bf9\u4e8e\u914d\u5907\u6709\u6ce8\u610f\u529b\u673a\u5236\u7684\u89e3\u7801\u5668":114,"\u5bf9\u4e8enchw":82,"\u5bf9\u4ee3\u7801\u8fdb\u884c\u6027\u80fd\u5206\u6790":108,"\u5bf9\u4f7f\u7528\u7684\u4e2d\u95f4\u53d8\u91cf\u548c\u8d44\u6e90\u8fdb\u884c\u6e05\u7406\u548c\u91ca\u653e":90,"\u5bf9\u5168\u8fde\u63a5\u5c42\u6765\u8bf4":74,"\u5bf9\u52a0\u8f7d\u9884\u8bad\u7ec3\u53c2\u6570\u7684\u5c42":83,"\u5bf9\u53cc\u5c42\u5e8f\u5217\u6765\u8bb2":89,"\u5bf9\u5df2\u7ecfpush\u5230\u8fdc\u7a0b\u4ed3\u5e93\u7684\u591a\u4e2acommit":72,"\u5bf9\u5e94":117,"\u5bf9\u5e94\u4e00\u4e2a\u5b50\u53e5":113,"\u5bf9\u5e94\u4e00\u4e2a\u8bcd":113,"\u5bf9\u5e94\u4e8e\u8c03\u7528c":89,"\u5bf9\u5e94\u7684\u68af\u5ea6op\u8ba1\u7b97\u4e4b\u4e2d":75,"\u5bf9\u5e94\u7740\u4e0a\u6587\u63d0\u5230\u7684\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":89,"\u5bf9\u5e94\u7740\u4e0a\u6587\u63d0\u5230\u7684\u4e8c\u7ef4\u6d6e\u70b9\u578b\u77e9\u9635":89,"\u5bf9\u63a8\u8350\u914d\u7f6e\u4e2d\u7684\u9009\u9879\u5efa\u8bae\u6309\u7167\u8bbe\u7f6e":87,"\u5bf9\u65b0\u7684\u6743\u91cd\u8fdb\u884c\u8f6c\u6362\u7528\u4e8e\u4e0b\u6b21\u8fed\u4ee3":41,"\u5bf9\u6bcf\u4e2a\u8f93\u5165":74,"\u5bf9\u6bcf\u4e2a\u8f93\u5165\u4e58\u4e0a\u53d8\u6362\u77e9\u9635":74,"\u5bf9\u6bd4":45,"\u5bf9\u6bd4\u4f18\u5316\u540elayer\u4e0e\u76f8\u5bf9\u5e94\u7684paddlepaddle\u539f\u6709lay":41,"\u5bf9\u6bd4\u4f18\u5316\u540elayer\u81ea\u8eab":41,"\u5bf9\u6bd4\u53cd\u5411op\u4e0d\u540c\u8bbe\u5907":75,"\u5bf9\u6fc0\u6d3b\u6c42\u5bfc":74,"\u5bf9\u795e\u7ecf\u7f51\u7edc\u6765\u8bf4":89,"\u5bf9\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u8fdb\u884c\u5e8f\u5217\u5316":90,"\u5bf9\u8bc4\u5ba1\u610f\u89c1\u4e0d\u540c\u610f\u7684":72,"\u5bf9\u8bc4\u5ba1\u610f\u89c1\u540c\u610f\u4e14\u6309\u5176\u4fee\u6539\u5b8c\u7684":72,"\u5bf9\u8c61":83,"\u5bf9\u8c61\u5206\u914d\u7a7a\u95f4":90,"\u5bf9\u8f93\u5165\u53c2\u6570\u7684\u5b89\u5168\u6027\u8fdb\u884c\u4e86\u5fc5\u8981\u7684\u5224\u65ad":46,"\u5bf9\u8f93\u51fa\u7684\u5408\u5e76":113,"\u5bf9\u8fd9\u4e2a\u7248\u672c\u7684\u63d0\u4ea4":62,"\u5bf9sparse_binary_vector\u548csparse_float_vector":84,"\u5bfb\u627e\u6709\u6ca1\u6709\u5176\u4ed6\u53ef\u4ee5\u4f18\u5316\u7684\u53ef\u80fd":42,"\u5bfb\u627epython\u4e0ec":107,"\u5bfc\u51fa\u8fd9\u4e9b\u63a5\u53e3":46,"\u5bfc\u81f4\u4e86\u6d6e\u70b9\u6570\u6ea2\u51fa":81,"\u5bfc\u81f4\u53c2\u6570\u6536\u655b\u5230\u4e86\u4e00\u4e9b\u5947\u5f02\u7684\u60c5\u51b5":81,"\u5bfc\u81f4\u53c2\u6570\u7d2f\u52a0":81,"\u5bfc\u81f4\u7f16\u8bd1paddlepaddle\u5931\u8d25":78,"\u5bfc\u81f4\u8bad\u7ec3\u65f6\u95f4\u8fc7\u957f":83,"\u5bfc\u81f4mklml\u5e93\u4e0b\u8f7d\u4e0d\u6210\u529f":78,"\u5c01\u88c5\u4e86":108,"\u5c01\u88c5\u8be5\u5c42\u7684python\u63a5\u53e3":74,"\u5c06":[62,83,108],"\u5c06\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u62c6\u5206\u6210\u591a\u4efd":11,"\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":114,"\u5c06\u4f1a\u4f18\u5148\u4f7f\u7528":91,"\u5c06\u4f1a\u59cb\u7ec8\u4f7f\u7528":116,"\u5c06\u4f1a\u5c06\u7528\u6237\u4f20\u8fdb\u6765\u7684\u914d\u7f6e\u53c2\u6570\u4f20\u9012cmake\u7cfb\u7edf":116,"\u5c06\u4f1a\u81ea\u52a8\u8ba1\u7b97\u51fa\u4e00\u4e2a\u5408\u9002\u7684\u503c":103,"\u5c06\u5176\u8bbe\u7f6e\u6210":81,"\u5c06\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e\u5148\u53d8\u6362\u6210\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u6570\u636e":111,"\u5c06\u542b\u6709\u5b50\u53e5":113,"\u5c06\u542b\u6709\u8bcd\u8bed\u7684\u53e5\u5b50\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":113,"\u5c06\u56fe\u7247\u5206\u7c7b\u5230":90,"\u5c06\u591a\u53e5\u8bdd\u770b\u6210\u4e00\u4e2a\u6574\u4f53\u540c\u65f6\u4f7f\u7528encoder\u538b\u7f29":111,"\u5c06\u591a\u53f0\u673a\u5668\u7684\u6d4b\u8bd5\u7ed3\u679c\u5408\u5e76":103,"\u5c06\u5927\u91cf\u7684":45,"\u5c06\u5b57\u5178\u7684\u5730\u5740\u4f5c\u4e3aargs\u4f20\u7ed9dataprovid":83,"\u5c06\u5b83\u4eec\u653e\u5728\u540c\u4e00\u76ee\u5f55\u4e2d":90,"\u5c06\u5bf9\u5e94\u6570\u636e\u5c42\u7684\u7ef4\u6570\u8bbe\u7f6e\u6210\u4e00\u4e2a\u5927\u4e8e\u8f93\u5165\u6570\u636e\u7ef4\u6570\u7684\u503c\u7528\u4e8e\u5360\u4f4d\u5373\u53ef":82,"\u5c06\u5e8f\u5217\u5316\u7ed3\u679c\u5199\u5165\u4e00\u4e2a\u6587\u4ef6\u5185":90,"\u5c06\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u524d\u5411\u548c\u53cd\u5411\u90e8\u5206\u6df7\u5408\u5728\u4e00\u8d77":114,"\u5c06\u6027\u80fd\u5206\u6790\u7ed3\u679c\u4ee5\u7f51\u9875\u7684\u5f62\u5f0f\u5c55\u793a\u51fa\u6765":107,"\u5c06\u6027\u80fd\u5206\u6790\u7ed3\u679c\u6309\u7167tottime\u6392\u5e8f":107,"\u5c06\u6570\u636e\u5207\u5206\u6210\u591a\u4efd":91,"\u5c06\u65b0\u5206\u652f\u7684\u7248\u672c\u6253\u4e0atag":62,"\u5c06\u65b0\u5efa\u7684\u6743\u91cd\u52a0\u5165\u6743\u91cd\u8868":74,"\u5c06\u660e\u6587\u53c2\u6570\u8f6c\u5316\u4e3apaddlepaddle\u53ef\u52a0\u8f7d\u7684\u6a21\u578b\u53c2\u6570\u65f6":83,"\u5c06\u672c\u5730\u7684\u4fee\u6539\u63a8\u9001\u5230":72,"\u5c06\u6b64\u76ee\u5f55\u6302\u8f7d\u4e3a\u5bb9\u5668\u7684":97,"\u5c06\u73af\u5883\u53d8\u91cf\u8f6c\u6362\u6210paddle\u7684\u547d\u4ee4\u884c\u53c2\u6570":97,"\u5c06\u7528\u4e8epython":75,"\u5c06\u7ed3\u679c\u4fdd\u5b58\u5230\u6b64\u76ee\u5f55\u91cc":97,"\u5c06\u7f51\u7edc\u7ed3\u6784\u5b9a\u4e49\u548c\u8bad\u7ec3\u7ed3\u675f\u5b58\u50a8\u4e0b\u6765\u7684\u6a21\u578b\u53c2\u6570\u6587\u4ef6":90,"\u5c06\u8bad\u7ec3\u6587\u4ef6\u4e0e\u5207\u5206\u597d\u7684\u6570\u636e\u4e0a\u4f20\u5230\u5171\u4eab\u5b58\u50a8":97,"\u5c06\u8df3\u8fc7\u5206\u53d1\u9636\u6bb5\u76f4\u63a5\u542f\u52a8\u6240\u6709\u8282\u70b9\u7684\u96c6\u7fa4\u4f5c\u4e1a":93,"\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":111,"\u5c06\u8fdc\u7a0b\u4ed3\u5e93":72,"\u5c06\u900f\u660e":93,"\u5c06\u9700\u8981\u8f93\u51fa\u7684\u5c42\u4f5c\u4e3a":81,"\u5c06cuda\u5e93\u548clinux\u8bbe\u5907\u6302\u8f7d\u5230docker\u5bb9\u5668\u5185":2,"\u5c06ip\u6392\u5e8f\u751f\u6210\u7684\u5e8f\u53f7\u4f5c\u4e3atrain":97,"\u5c06master\u5206\u652f\u7684\u5408\u5165commit\u6253\u4e0atag":62,"\u5c06node\u8282\u70b9\u7684ip\u5730\u5740\u4fdd\u5b58\u5230machines\u6587\u4ef6\u4e2d":98,"\u5c06paddlepaddle\u4fdd\u5b58\u7684\u6a21\u578b\u53c2\u6570\u8fd8\u539f\u56de\u660e\u6587\u65f6":83,"\u5c06recurr":82,"\u5c0f\u4e8e\u67d0\u4e2a\u6bd4\u8f83\u5c0f\u7684\u9608\u503c\u8ba4\u4e3a\u901a\u8fc7":42,"\u5c31\u4f1a\u5728\u5b8c\u6210\u7f16\u8bd1\u4e4b\u540e":1,"\u5c31\u53ef\u4ee5\u4e86\u89e3\u5230\u95ee\u9898\u4ee3\u7801\u5728\u54ea\u91cc":107,"\u5c31\u53ef\u4ee5\u4f7f\u7528\u4e0b\u9762\u7684\u547d\u4ee4\u5f00\u59cb\u6267\u884c\u8bad\u7ec3":2,"\u5c31\u53ef\u4ee5\u6309":0,"\u5c31\u5c06\u8fd9\u4e9b\u5c42\u52a0\u5165\u4e00\u4e2apython":90,"\u5c31\u5f88\u5bb9\u6613\u5bfc\u81f4\u5185\u5b58\u8d85\u9650":81,"\u5c31\u662f\u7528\u4e8e\u5c55\u793a\u4e0a\u8ff0\u5206\u6790\u5de5\u5177\u7684\u7528\u6cd5":108,"\u5c31\u662fpaddlepaddle\u4e2d\u6240\u6307\u7684":89,"\u5c31\u8fd9\u4e48\u7b80\u5355":2,"\u5c31\u901a\u5e38\u7684gpu\u6027\u80fd\u5206\u6790\u6765\u8bf4":108,"\u5c31\u9700\u8981\u5bf9\u8fd9\u4e2a\u7b2c\u4e09\u65b9\u8bed\u8a00\u589e\u52a0\u4e00\u4e9b\u5b9a\u4e49":45,"\u5c31\u9700\u8981\u9009\u62e9\u4f7f\u7528no":2,"\u5c3d\u65e9\u62a5\u9519":75,"\u5c42\u524d\u5411\u8ba1\u7b97\u7684\u7ed3\u679c":90,"\u5c42\u548c\u8f93\u5165\u7684\u914d\u7f6e":74,"\u5c42\u6b21\u5316\u7684rnn":113,"\u5c42\u7684\u540d\u79f0\u4e0e":114,"\u5c42\u7684\u5927\u5c0f":74,"\u5c42\u7684\u7c7b\u578b":74,"\u5c42\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684":114,"\u5c55\u793a\u4e86\u4e00\u4e2a\u542b\u67094\u4e2a\u5e8f\u5217\u7684":89,"\u5c55\u793a\u7684\u8c03\u7528\u56fe\u4e5f\u53ef\u4ee5\u5e2e\u52a9\u6211\u4eec\u53d1\u73b0\u6027\u80fd\u4e2d\u7684\u95ee\u9898":107,"\u5c5e\u4e8e\u8fd9\u4e00\u7c7b\u7684\u5b9e\u73b0":82,"\u5c5e\u6027":75,"\u5de5\u4f5c\u6a21\u5f0f":103,"\u5de5\u4f5c\u7a7a\u95f4\u4e2d\u7684":93,"\u5de5\u4f5c\u7a7a\u95f4\u5e94\u5982\u4e0b\u6240\u793a":91,"\u5de5\u5177\u4e0a\u4f20\u5373\u53ef":62,"\u5de5\u5177\u5408\u5e76fat\u5e93":117,"\u5de5\u5177\u670d\u52a1\u5668\u5c06\u8bfb\u53d6\u73af\u5883\u53d8\u91cf":77,"\u5de5\u5177\u6765\u7ba1\u7406":72,"\u5de5\u5177\u6765\u7f16\u8bd1\u6587\u6863":77,"\u5de5\u5177\u94fe":116,"\u5de5\u5177\u94fe\u7684android":116,"\u5de6\u53f3\u7684\u8ba1\u7b97\u65f6\u95f4":107,"\u5df2\u6253\u5f00":72,"\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":103,"\u5e02\u9762\u4e0a\u5df2\u7ecf\u6709nvidia\u6216\u7b2c\u4e09\u65b9\u63d0\u4f9b\u7684\u4f17\u591a\u5de5\u5177":108,"\u5e0c\u671b\u80fd\u591f\u5c06\u5e8f\u5217\u5316\u540e\u7684\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u53c2\u6570\u6253\u5305\u8fdb\u4e00\u4e2a\u6587\u4ef6":90,"\u5e26\u6709\u4e0b\u9762\u4e24\u4e2a\u6a21\u677f\u53c2\u6570":75,"\u5e2e\u52a9\u6211\u4eec\u5b8c\u6210\u5bf9\u8f93\u5165\u5e8f\u5217\u7684\u62c6\u5206":113,"\u5e2e\u52a9\u6211\u4eec\u66f4\u597d\u5730\u63cf\u8ff0\u6bb5\u843d":113,"\u5e2e\u52a9\u6211\u4eec\u6784\u9020\u4e00\u4e9b\u590d\u6742\u7684\u8f93\u5165\u4fe1\u606f":110,"\u5e38\u5e38\u51fa\u73b0":78,"\u5e38\u7528\u4e8e\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1":89,"\u5e38\u7528\u7684cmake\u914d\u7f6e\u5982\u4e0b":[116,117],"\u5e38\u89c1\u7684\u5305\u62ec":107,"\u5e73\u5747\u6545\u969c\u4fee\u590d\u65f6\u95f4":11,"\u5e73\u5747\u6545\u969c\u7387":11,"\u5e76\u4e0d\u4fdd\u8bc1":74,"\u5e76\u4e0d\u662f\u4f7f\u7528\u53cc\u5c42rnn\u89e3\u51b3\u5b9e\u9645\u7684\u95ee\u9898":111,"\u5e76\u4e0d\u662fkubernetes\u4e2d\u7684node\u6982\u5ff5":97,"\u5e76\u4e0d\u771f\u6b63":[89,90],"\u5e76\u4e0d\u771f\u6b63\u7684\u548c":111,"\u5e76\u4e0d\u96be":0,"\u5e76\u4e14":114,"\u5e76\u4e14\u4e5f\u53ef\u4ee5\u5728windows\u7684docker\u4e2d\u8fd0\u884c":2,"\u5e76\u4e14\u4e66\u5199\u4e00\u4efd\u4ee3\u7801":76,"\u5e76\u4e14\u4f1a\u5199\u597d":42,"\u5e76\u4e14\u4f1a\u6839\u636e":116,"\u5e76\u4e14\u4f7f\u7528":46,"\u5e76\u4e14\u5185\u5c42\u7684\u5e8f\u5217\u64cd\u4f5c\u4e4b\u95f4\u72ec\u7acb\u65e0\u4f9d\u8d56":111,"\u5e76\u4e14\u52a0\u4e0a\u4e0b\u9762\u7684\u547d\u4ee4\u884c\u53c2\u6570":105,"\u5e76\u4e14\u5305\u62ecunit":72,"\u5e76\u4e14\u53ea\u9700\u8981\u5728\u5fc5\u8981\u7684\u65f6\u5019\u8f6c\u6362\u8fd9\u79cd\u683c\u5f0f":42,"\u5e76\u4e14\u53ef\u80fd\u4f1a\u52a0\u901f\u8bad\u7ec3\u8fc7\u7a0b":81,"\u5e76\u4e14\u542f\u52a8\u8bad\u7ec3":97,"\u5e76\u4e14\u5728\u5e38\u89c1\u7684\u5e73\u53f0\u4e0a":45,"\u5e76\u4e14\u5c55\u793a\u4e86\u5982\u4f55\u5229\u7528paddlepaddle\u6765\u89e3\u51b3\u4e00\u4e2a\u7ecf\u5178\u7684\u7ebf\u6027\u56de\u5f52\u95ee\u9898":84,"\u5e76\u4e14\u5f3a\u5236\u8bbe\u7f6e\u4e00\u4e9bpaddlepaddle\u53c2\u6570\u7684\u503c":117,"\u5e76\u4e14\u5f53\u7f16\u8bd1\u65f6":41,"\u5e76\u4e14\u628a\u5404\u79cd\u5f00\u53d1\u5de5\u5177\u5b89\u88c5\u8fdb\u53bb":0,"\u5e76\u4e14\u628a\u7cfb\u7edf\u751f\u6210\u7684ca":28,"\u5e76\u4e14\u628a\u7ed3\u679c\u8fd4\u56depfsclient\u7aef":28,"\u5e76\u4e14\u67e5\u8be2paddlepaddle\u5355\u5143\u6d4b\u8bd5\u7684\u65e5\u5fd7":78,"\u5e76\u4e14\u7f16\u8bd1\u65f6\u9700\u8981\u6253\u5f00":75,"\u5e76\u4e14\u7f16\u8bd1\u80fd\u901a\u8fc7\u4ee3\u7801\u6837\u5f0f\u68c0\u67e5":72,"\u5e76\u4e14\u8ba9\u63a5\u53e3\u8131\u79bb\u5b9e\u73b0\u7ec6\u8282":45,"\u5e76\u4e14\u8bbe\u7f6e\u9ed8\u8ba4\u503c\u4e3a1":75,"\u5e76\u4e14\u8f93\u5165\u8f93\u51fa\u90fd\u662f\u5171\u7528\u4e00\u5757\u5185\u5b58":42,"\u5e76\u4e14\u8f93\u51fa\u4e00\u4e2a":72,"\u5e76\u4e14\u8fd0\u884c":0,"\u5e76\u4e14\u9700\u8981\u91cd\u5199\u57fa\u7c7b\u4e2d\u7684\u4ee5\u4e0b\u51e0\u4e2a\u865a\u51fd\u6570":74,"\u5e76\u4e14cpu":75,"\u5e76\u4e14softmax\u5c42\u7684\u4e24\u4e2a\u8f93\u5165\u4e5f\u4f7f\u7528\u4e86\u540c\u6837\u7684\u53c2\u6570":83,"\u5e76\u4f7f\u7528":93,"\u5e76\u4fdd\u5b58\u8f93\u51fa\u5230\u4e00\u4e2a\u65e5\u5fd7\u6587\u4ef6":91,"\u5e76\u521b\u5efa\u4e86\u4e00\u4e2a\u65b0\u6587\u4ef6":72,"\u5e76\u521b\u5efaoptim":84,"\u5e76\u521d\u59cb\u5316":75,"\u5e76\u5220\u9664":62,"\u5e76\u5220\u9664\u66f4\u65e9\u7684\u5feb\u7167":11,"\u5e76\u52a0\u8f7d\u5176\u4e2d\u7684\u53c2\u6570":11,"\u5e76\u53d1\u5e03\u5230pypi":62,"\u5e76\u53ef\u4ee5\u5728\u5927\u591a\u6570\u4e3b\u6d41\u7684linux\u64cd\u4f5c\u7cfb\u7edf\u4ee5\u53camacos\u4e0a\u6267\u884c":4,"\u5e76\u548c\u53c2\u6570\u670d\u52a1\u5668\u901a\u4fe1":92,"\u5e76\u5728\u4e58\u79ef\u7ed3\u679c\u4e0a\u518d\u52a0\u4e0a\u7ef4\u5ea6\u4e3a":74,"\u5e76\u5728\u6700\u5f00\u59cb\u521d\u59cb\u5316\u4e3a\u8d77\u59cb\u8bcd":114,"\u5e76\u5728\u6bcf\u6b21\u6743\u91cd\u66f4\u65b0\u540e":41,"\u5e76\u5728\u7c7b\u6784\u5efa\u51fd\u6570\u4e2d\u628a\u5b83\u653e\u5165\u4e00\u4e2a\u7c7b\u6210\u5458\u53d8\u91cf\u91cc":74,"\u5e76\u5728\u8be5layer\u91cc\u91c7\u7528\u7b2c\u4e00\u79cd\u65b9\u5f0f\u8bbe\u7f6e":82,"\u5e76\u5728\u96c6\u7fa4\u4e2d\u8fd0\u884c\u591a\u4e2a\u5206\u5e03\u5f0f\u6570\u636e\u5904\u7406\u4efb\u52a1":12,"\u5e76\u5728python\u811a\u672c\u4e2d\u5b8c\u6210\u4e0eoperator\u540c\u6837\u7684\u8ba1\u7b97\u903b\u8f91":75,"\u5e76\u5904\u7406\u4e0e\u4e4b\u76f8\u5173\u7684\u6240\u6709\u7ec6\u8282":90,"\u5e76\u5b89\u88c5\u4e86python":78,"\u5e76\u5b89\u88c5\u6700\u65b0":4,"\u5e76\u5b89\u88c5\u6709python2":86,"\u5e76\u5b8c\u6210\u53c2\u6570\u4f18\u5316\u66f4\u65b0":92,"\u5e76\u5bf9\u6bd4\u662f\u5426\u548c\u6b63\u5728\u5b89\u88c5\u7684\u540e\u7f00\u4e00\u81f4":78,"\u5e76\u5bf9\u76f8\u5e94\u7684\u53c2\u6570\u8c03\u7528":74,"\u5e76\u5c06":62,"\u5e76\u5c06\u5176\u6295\u5c04\u5230":114,"\u5e76\u5c06\u8be5layer\u4e0a\u4e00\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u4f5c\u4e3a\u81ea\u8eab\u5f53\u524d\u65f6\u95f4\u6b65\u7684\u8f93\u51fa":82,"\u5e76\u5c06c":46,"\u5e76\u624b\u52a8\u751f\u6210download\u6210\u529f\u6807\u7b7e":78,"\u5e76\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4":2,"\u5e76\u628a\u5feb\u7167\u4fdd\u5b58\u5230\u8fd9\u4e2a\u76ee\u5f55\u4e0b":11,"\u5e76\u628a\u7ed3\u679c\u653e\u5230\u5f53\u524d\u5c42\u7684":42,"\u5e76\u628a\u8fd9\u4e2a\u5305\u542b\u4e86\u8bad\u7ec3\u6570\u636e\u7684container\u4fdd\u5b58\u4e3a\u4e00\u4e2a\u65b0\u7684\u955c\u50cf":96,"\u5e76\u66f4\u6362job":79,"\u5e76\u6839\u636e\u5206\u5e03\u5f0f\u8bad\u7ec3\u5e76\u53d1\u6570":91,"\u5e76\u68c0\u67e5\u548c\u9700\u5b89\u88c5\u7684\u5305\u662f\u5426\u5339\u914d":4,"\u5e76\u6ca1\u6709paddle\u7279\u522b\u9700\u8981\u7684\u7279\u6027":45,"\u5e76\u6dfb\u52a0\u5934\u6587\u4ef6":41,"\u5e76\u6dfb\u52a0\u6ce8\u91ca":75,"\u5e76\u7279\u5316\u6a21\u677f\u53c2\u6570\u4e3a":75,"\u5e76\u7c98\u8d34\u6b64python\u4ee3\u7801":86,"\u5e76\u81ea\u52a8\u4e0b\u8f7d\u5b89\u88c5\u4f9d\u8d56\u8f6f\u4ef6":4,"\u5e76\u81ea\u52a8\u7f16\u8bd1\u5bbf\u4e3b\u673a\u7248protoc\u53ef\u6267\u884c\u6587\u4ef6":118,"\u5e76\u81ea\u52a8\u7f16\u8bd1paddlepaddle\u6240\u9700\u7684\u6240\u6709\u7b2c\u4e09\u65b9\u5e93":116,"\u5e76\u884c\u5730\u6267\u884c\u6a21\u578b\u7684\u8bad\u7ec3":92,"\u5e76\u884c\u5730\u63a5\u6536\u68af\u5ea6\u548c\u66f4\u65b0\u53c2\u6570":92,"\u5e76\u88ab\u5b58\u50a8\u5728\u8bf8\u5982hadoop":12,"\u5e76\u89c2\u5bdf\u7ed3\u679c":108,"\u5e76\u89e3\u91ca\u4e86\u5404\u81ea\u542b\u4e49":75,"\u5e76\u8bb0\u5f55\u5b83\u7684\u7f16\u53f7":72,"\u5e76\u8fdb\u884c\u521d\u59cb\u5316\u64cd\u4f5c":84,"\u5e76\u9002\u5e94github\u7684\u7279\u6027\u505a\u4e86\u4e00\u4e9b\u533a\u522b":62,"\u5e76\u91cd\u65b0\u6253\u5305wheel\u5305":62,"\u5e76\u94fe\u63a5\u5230\u751f\u6210\u7684lib\u5e93\u4e2d":75,"\u5e78\u800cpython\u7684\u4e00\u4e2a\u7b2c\u4e09\u65b9\u5e93":107,"\u5e8f\u5217\u4e2d\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee\u540c":110,"\u5e8f\u5217\u4e2d\u7684\u4e00\u4e2a\u5143\u7d20":89,"\u5e8f\u5217\u4e2d\u7684\u5143\u7d20\u662f\u8bcd\u8bed":89,"\u5e8f\u5217\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":89,"\u5e8f\u5217\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u662f\u975e\u5e8f\u5217":89,"\u5e8f\u5217\u4fe1\u606f":89,"\u5e8f\u5217\u5316\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u914d\u7f6e":90,"\u5e8f\u5217\u5316\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u65f6":90,"\u5e8f\u5217\u5316\u7ed3\u679c\u4f1a\u5199\u5165\u5f53\u524d\u8fd0\u884c\u76ee\u5f55\u4e0b\u7684":90,"\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":113,"\u5e8f\u5217\u662f\u4e00\u79cd\u5e38\u89c1\u7684\u6570\u636e\u7c7b\u578b":110,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u5927\u591a\u9075\u5faaencod":113,"\u5e8f\u5217\u751f\u6210\u4efb\u52a1\u7684\u8f93\u5165":113,"\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u539f\u6765\u53cc\u5c42\u5e8f\u5217\u6bcf\u4e2asubseq\u5143\u7d20\u7684\u5e73\u5747\u503c":110,"\u5e8f\u5217\u8f93\u5165":89,"\u5e8f\u5217\u8f93\u5165\u65f6\u7b49\u4e8e":81,"\u5e8f\u5217\u8f93\u5165\u793a\u610f\u56fe":89,"\u5e93\u6709\u81ea\u5df1\u72ec\u7acb\u7684\u52a8\u6001\u5e93\u6587\u4ef6":87,"\u5e94\u7528\u524d\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":114,"\u5e94\u7528\u53cd\u5411\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":114,"\u5e94\u80fd\u53cd\u6620\u5f53\u524dcommit\u7684\u5185\u5bb9":72,"\u5e94\u8be5\u4e0e\u5b83\u7684memory\u540d\u5b57\u76f8\u540c":114,"\u5e94\u8be5\u8bf4\u8c22\u8c22":72,"\u5e94\u8be5\u8bfb\u53d6\u5f53\u524d\u76ee\u5f55\u4e0b\u7684":0,"\u5e94\u8be5\u964d\u4f4e\u5b66\u4e60\u7387":83,"\u5e95\u5c42\u8fdb\u7a0b":93,"\u5efa\u7acb\u4e00\u4e2a":72,"\u5efa\u8bae":[3,62,72],"\u5efa\u8bae\u5c06\u8be5\u53c2\u6570\u8bbe\u4e3atrue":103,"\u5f00\u53d1\u4e86\u6a21\u578b\u9884\u6d4b\u7684\u6837\u4f8b\u4ee3\u7801":46,"\u5f00\u53d1\u4eba\u5458\u4f7f\u7528":72,"\u5f00\u53d1\u5206\u652f":4,"\u5f00\u53d1\u6807\u51c6":115,"\u5f00\u53d1\u8005\u4f7f\u7528":0,"\u5f00\u53d1\u8005\u4fee\u6539\u81ea\u5df1\u7684\u4ee3\u7801":62,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4e2d":62,"\u5f00\u53d1\u8005fork\u7684\u7248\u672c\u5e93\u4f7f\u7528":62,"\u5f00\u53d1\u955c\u50cf":72,"\u5f00\u53d1\u9884\u6d4b\u5e8f":90,"\u5f00\u53d1\u9884\u6d4b\u7a0b\u5e8f\u94fe\u63a5":87,"\u5f00\u542f":1,"\u5f00\u5934":[41,42],"\u5f00\u5934\u7684\u90e8\u5206":91,"\u5f00\u5934\u90e8\u5206\u6307\u5b9a":91,"\u5f00\u59cb\u63d0\u4f9b\u670d\u52a1":11,"\u5f00\u59cb\u6807\u8bb0":114,"\u5f00\u59cb\u795e\u7ecf\u7f51\u7edc\u7684":92,"\u5f00\u59cb\u9636\u6bb5":108,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":102,"\u5f15\u5165\u4e86\u4ee5\u4e0b\u56db\u4e2aapi":41,"\u5f15\u5165\u4e86\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5f15\u5bfc\u5c42":114,"\u5f15\u7528memory\u5f97\u5230\u8fd9layer\u4e0a\u4e00\u65f6\u523b\u8f93\u51fa":113,"\u5f39\u51fa\u4e0b\u9762\u7684\u9009\u62e9\u6846":62,"\u5f52\u4e00\u5316\u6982\u7387\u5411\u91cf":114,"\u5f53":105,"\u5f53\u4e00\u4e2a":89,"\u5f53\u4e0a\u8ff0\u63a5\u53e3\u7b2c4\u4e2a\u53c2\u6570":89,"\u5f53\u4f60\u6267\u884c\u547d\u4ee4":74,"\u5f53\u4fdd\u5b58\u7684\u7f51\u7edc\u53c2\u6570\u4e3afloat\u7c7b\u578b\u65f6\u4e3a4":83,"\u5f53\u524d\u65f6\u95f4\u6b65\u5904\u7684memory\u7684\u8f93\u51fa\u4f5c\u4e3a\u4e0b\u4e00\u65f6\u95f4\u6b65memory\u7684\u8f93\u5165":114,"\u5f53\u524d\u7684\u5b66\u4e60\u7387\u4e3a\u6240\u8bbe\u7f6e":83,"\u5f53\u524d\u7684\u5b9e\u73b0\u65b9\u5f0f\u4e0b":74,"\u5f53\u524d\u7684\u8f93\u5165y\u548c\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51farnn_state\u505a\u4e86\u4e00\u4e2a\u5168\u94fe\u63a5":111,"\u5f53\u524d\u8bad\u7ec3\u4efb\u52a1\u542f\u52a8\u7684pserver\u7684ip\u5217\u8868":91,"\u5f53\u524d\u8bad\u7ec3\u4efb\u52a1pserver\u603b\u6570":91,"\u5f53\u524d\u8bad\u7ec3\u4efb\u52a1trainer\u603b\u6570":91,"\u5f53\u524dtrainer\u7684\u7ebf\u7a0b\u6570\u76ee":91,"\u5f53\u529f\u80fd\u5206\u652f\u5f00\u53d1\u5b8c\u6bd5\u540e":62,"\u5f53\u53ea\u505a\u63a8\u65ad":41,"\u5f53\u5728\u7f51\u7edc\u5c42\u914d\u7f6e\u4e2d\u8bbe\u7f6e":103,"\u5f53\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":103,"\u5f53\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5927\u4e8e1000\u5c0f\u4e8e\u7b49\u4e8e2000\u65f6":83,"\u5f53\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5927\u4e8e2000\u65f6":83,"\u5f53\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5c0f\u4e8e\u7b49\u4e8e1000\u65f6":83,"\u5f53\u5df2\u8bad\u7ec3pass\u6570\u5927\u4e8e1\u5c0f\u4e8e\u7b49\u4e8e2\u65f6":83,"\u5f53\u5df2\u8bad\u7ec3pass\u6570\u5927\u4e8e2\u65f6":83,"\u5f53\u5df2\u8bad\u7ec3pass\u6570\u5c0f\u4e8e\u7b49\u4e8e1\u65f6":83,"\u5f53\u5f00\u542f":42,"\u5f53\u6211\u4eec\u505a\u51fa\u6027\u80fd\u4fee\u6b63\u540e":107,"\u5f53\u6240\u6709pod\u90fd\u5904\u4e8erunning\u72b6\u6001":97,"\u5f53\u6253\u5f00":42,"\u5f53\u6570\u636e\u683c\u5f0f\u4e0epaddlepaddle\u9ed8\u8ba4\u7684":42,"\u5f53\u6a21\u578b\u53c2\u6570\u4e0d\u5b58\u5728\u65f6":103,"\u5f53\u6a21\u5f0f\u4e3a":103,"\u5f53\u7136":[2,108],"\u5f53\u7136\u53ef\u4ee5":0,"\u5f53\u7136\u8fd9\u4e24\u8005\u4e5f\u53ef\u4ee5\u76f8\u7b49":42,"\u5f53\u7528\u6237\u4f7f\u7528\u5b8c\u8fd9\u4e2a\u53c2\u6570\u540e":46,"\u5f53\u7528\u6237\u6ca1\u6709\u663e\u5f0f\u8bbe\u5b9a\u65f6":82,"\u5f53\u7f51\u7edc\u51fa\u73b0\u5206\u652f\u4e14\u5728":42,"\u5f53\u7f51\u7edc\u5c42\u7528\u4e00\u4e2a\u6279\u6b21\u505a\u8bad\u7ec3\u65f6":74,"\u5f53\u89e3\u8bfb\u6bcf\u4e00\u4e2a":114,"\u5f53\u8d85\u8fc7\u8be5\u9608\u503c\u65f6":103,"\u5f53\u8f93\u5165\u662f\u7ef4\u5ea6\u5f88\u9ad8\u7684\u7a00\u758f\u6570\u636e\u65f6":105,"\u5f53\u9700\u8981\u5b8c\u6210\u8ba1\u7b97\u65f6":76,"\u5f53\u975e\u5e8f\u5217\u8f93\u5165\u65f6":89,"\u5f53destination\u6587\u4ef6\u4e0d\u5b58\u5728\u6216\u8005\u5927\u5c0f\u548csource\u6587\u4ef6\u4e0d\u4e00\u81f4\u65f6":28,"\u5f53n1":81,"\u5f62\u6210recurr":113,"\u5f62\u6210recurrent\u8fde\u63a5":113,"\u5f88\u591a":0,"\u5f88\u6709\u53ef\u80fd\u5b9e\u9645\u5e94\u7528\u5c31\u662f\u6ca1\u6709\u6309\u7167\u60a8\u7684\u9884\u671f\u60c5\u51b5\u8fd0\u884c":108,"\u5f88\u6709\u53ef\u80fd\u662f\u975e\u72ec\u5360\u65b9\u5f0f\u6267\u884c\u5bfc\u81f4\u7684\u7aef\u53e3\u51b2\u7a81":79,"\u5f88\u96be\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":45,"\u5f97\u4f7f\u7528":45,"\u5f97\u5230\u8f93\u51fa\u503c":75,"\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e2a\u65f6\u95f4\u6b65\u603b\u662f\u80fd\u591f\u5f15\u7528\u6240\u6709\u8f93\u5165":113,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4e2d":114,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u4f5c\u4e3a\u4f7f\u7528":114,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u548c":114,"\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u9aa4\u987a\u5e8f\u5730\u5904\u7406\u5e8f\u5217":114,"\u5faa\u73af\u7f51\u7edc\u4ece":114,"\u5fc5\u8981":46,"\u5fc5\u9009":91,"\u5fc5\u987b":74,"\u5fc5\u987b\u5206\u522b\u4e0e":42,"\u5fc5\u987b\u5c06\u524d\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5143\u7d20":111,"\u5fc5\u987b\u6307\u5411\u4e00\u4e2apaddlepaddle\u5b9a\u4e49\u7684lay":113,"\u5fc5\u987b\u6307\u5b9a\u4e3a":90,"\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":113,"\u5fc5\u987b\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":110,"\u5fc5\u987b\u7531\u53ea\u8bfbmemory\u7684":114,"\u5fc5\u987b\u8bbe\u7f6e\u4e3a":[116,117],"\u5fc5\u987b\u8bbe\u7f6e\u4e3aon":117,"\u5fc5\u987b\u914d\u7f6e\u4e3a":[87,118],"\u5fc5\u987b\u914d\u7f6e\u9009\u9879":87,"\u5feb\u901f\u5f00\u59cb":85,"\u6027\u80fd\u4f18\u5316\u7684\u8fc7\u7a0b\u901a\u5e38\u662f\u4e0d\u65ad\u91cd\u590d\u5730":107,"\u6027\u80fd\u5206\u6790":108,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u662f\u7528\u4e8e\u7ed9\u5e94\u7528\u7a0b\u5e8f\u7684\u6027\u80fd\u505a\u5b9a\u91cf\u5206\u6790\u7684":108,"\u6027\u80fd\u5206\u6790\u662f\u6027\u80fd\u4f18\u5316\u7684\u5173\u952e\u4e00\u6b65":108,"\u6027\u80fd\u548c\u628a\u7f16\u8bd1\u5de5\u5177\u5b89\u88c5\u5728\u672c\u673a\u8fd0\u884c\u4e00\u6837":0,"\u6027\u80fd\u8c03\u4f18":102,"\u6027\u80fdtip":[116,117],"\u60a8\u4e5f\u53ef\u4ee5\u8fdb\u5165\u5230docker\u5bb9\u5668\u4e2d":2,"\u60a8\u4f1a\u5728\u63a5\u4e0b\u6765\u7684\u90e8\u5206\u4e2d\u83b7\u5f97\u66f4\u591a\u7684\u7ec6\u8282\u4ecb\u7ecd":108,"\u60a8\u53ef\u4ee5\u4ece\u4e0b\u9762\u7684\u8868\u683c\u4e2d\u627e\u5230\u9700\u8981\u7684\u7248\u672c":4,"\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":108,"\u60a8\u53ef\u4ee5\u5728":2,"\u60a8\u53ef\u4ee5\u5728\u5bb9\u5668\u4e2d\u6267\u884c":2,"\u60a8\u53ef\u4ee5\u5bfc\u5165":108,"\u60a8\u53ef\u4ee5\u6309\u7167\u4e0b\u9762\u7684\u6b65\u9aa4\u5728openmpi\u96c6\u7fa4\u4e2d\u63d0\u4ea4paddle\u8bad\u7ec3\u4efb\u52a1":98,"\u60a8\u53ef\u4ee5\u91c7\u7528\u4e0b\u9762\u4e94\u4e2a\u6b65\u9aa4":108,"\u60a8\u53ef\u80fd\u9700\u8981\u4fee\u6539":91,"\u60a8\u5c06\u4e86\u89e3\u5982\u4f55":114,"\u60a8\u5c31\u80fd\u83b7\u5f97\u5982\u4e0b\u7684\u5206\u6790\u7ed3\u679c":108,"\u60a8\u6309\u5982\u4e0b\u6b65\u9aa4\u64cd\u4f5c\u5373\u53ef":108,"\u60a8\u6700\u597d\u5148\u786e\u8ba4":108,"\u60a8\u9996\u5148\u9700\u8981\u5728\u76f8\u5173\u4ee3\u7801\u6bb5\u4e2d\u52a0\u5165":108,"\u60c5\u611f\u5206\u6790":62,"\u60f3\u4e86\u89e3\u66f4\u591apaddlepaddl":77,"\u610f\u5473\u7740\u4e0d\u540c\u65f6\u95f4\u6b65\u7684\u8f93\u5165\u90fd\u662f\u76f8\u540c\u7684\u503c":114,"\u610f\u601d\u662f\u4e0d\u4f7f\u7528\u5e73\u5747\u53c2\u6570\u6267\u884c\u6d4b\u8bd5":103,"\u610f\u601d\u662f\u4e0d\u4fdd\u5b58\u7ed3\u679c":103,"\u610f\u601d\u662f\u4f7f\u7528\u7b2ctest":103,"\u610f\u601d\u662f\u5728gpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2agpu":103,"\u6210\u529f\u7f16\u8bd1\u540e":87,"\u6210\u529f\u8bad\u7ec3\u4e14\u9000\u51fa\u7684pod\u6570\u76ee\u4e3a3\u65f6":97,"\u6210\u5458":75,"\u6210\u719f\u7684\u9ad8\u6027\u80fd\u5e76\u884c\u8ba1\u7b97\u6846\u67b6":94,"\u6211\u4eec\u4e0d\u80fd\u901a\u8fc7\u5e38\u89c4\u7684\u68af\u5ea6\u68c0\u67e5\u7684\u65b9\u5f0f\u6765\u8ba1\u7b97\u68af\u5ea6":74,"\u6211\u4eec\u4e3b\u8981\u4f1a\u4ecb\u7ecdnvprof\u548cnvvp":108,"\u6211\u4eec\u4e5f\u53ef\u4ee5\u786e\u5b9a\u6bcf\u4e00\u4e2a\u53c2\u6570\u7684\u7c7b\u578b":46,"\u6211\u4eec\u4e5f\u5c06mklml\u5373":42,"\u6211\u4eec\u4ec5\u4ec5\u5bf9\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165\u8fdb\u884c\u4e86\u63cf\u8ff0":84,"\u6211\u4eec\u4ec5\u6709\u4e00\u4e2a\u8f93\u5165":74,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728":96,"\u6211\u4eec\u4ecb\u7ecd\u5982\u4f55\u5728kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884c\u5206\u5e03\u5f0fpaddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":97,"\u6211\u4eec\u4ee5\u624b\u5199\u6570\u5b57\u8bc6\u522b\u4efb\u52a1\u4e3a\u4f8b\u8fdb\u884c\u4ecb\u7ecd":90,"\u6211\u4eec\u4f1a\u4fdd\u8bc1":42,"\u6211\u4eec\u4f1a\u5728\u7f51\u7edc\u8bad\u7ec3\u4e4b\u524d\u628a\u683c\u5f0f\u8f6c\u6362\u4e3amkl":42,"\u6211\u4eec\u4f1a\u5bf9\u6bcf\u4e2a\u8bad\u7ec3\u4efb\u52a1\u90fd\u4f1a\u5728\u6bcf\u4e2a\u8282\u70b9\u4e0a\u521b\u5efa\u4e00\u4e2a\u5de5\u4f5c\u7a7a\u95f4":91,"\u6211\u4eec\u4f1a\u5bf9\u6bd4\u5982\u4e0b2\u4e2a\u65b9\u9762":41,"\u6211\u4eec\u4f1a\u628amkl":42,"\u6211\u4eec\u4f1a\u6dfb\u52a0":[41,42],"\u6211\u4eec\u4f1a\u7ee7\u7eed\u4f7f\u7528\u73b0\u6709\u7684\u5185\u5b58\u5757":74,"\u6211\u4eec\u4f1a\u91cd\u65b0\u5206\u914d\u5185\u5b58":74,"\u6211\u4eec\u4f7f\u7528":74,"\u6211\u4eec\u4f7f\u7528\u4e0d\u540c\u7684layer\u8fdb\u884c\u7ec4\u5408":84,"\u6211\u4eec\u4f7f\u7528\u4e86":111,"\u6211\u4eec\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":45,"\u6211\u4eec\u4f7f\u7528paddl":91,"\u6211\u4eec\u5047\u8bbe\u4e00\u53f0\u673a\u5668\u4e0a\u67094\u4e2agpu":105,"\u6211\u4eec\u5148\u8c03\u7528\u6bcf\u4e2a":76,"\u6211\u4eec\u51b3\u5b9a\u4f7f\u7528\u5df2\u6709\u7684":42,"\u6211\u4eec\u5373\u53ef\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684\u642d\u5efa":84,"\u6211\u4eec\u53ea\u6f14\u793a\u4e00\u4e2a\u5355\u673a\u4f5c\u4e1a":96,"\u6211\u4eec\u53ea\u9700\u8981":0,"\u6211\u4eec\u53ea\u9700\u8981\u4f7f\u7528lstm":111,"\u6211\u4eec\u53ea\u9700\u8981\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u628a\u7f16\u8bd1\u597d\u7684paddlepaddle\u6253\u5305\u6210\u4e00\u4e2a":72,"\u6211\u4eec\u53ea\u9700\u8981\u914d\u7f6e":0,"\u6211\u4eec\u53ef\u4ee5":0,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528":107,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5176\u4ed6layer\u8fdb\u884c\u7ec4\u5408":84,"\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5b83\u6765\u751f\u6210\u5e8f\u5217":114,"\u6211\u4eec\u53ef\u4ee5\u5148\u5b8c\u6210\u5bf9\u539f\u6570\u636e\u7684packing\u64cd\u4f5c":41,"\u6211\u4eec\u53ef\u4ee5\u521b\u5efatrainer\u6765\u5bf9\u7f51\u7edc\u8fdb\u884c\u8bad\u7ec3":84,"\u6211\u4eec\u53ef\u4ee5\u53c2\u8003tensorflow\u7684":76,"\u6211\u4eec\u53ef\u4ee5\u5728":72,"\u6211\u4eec\u53ef\u4ee5\u5728\u547d\u4ee4\u884c\u4e2d\u7b80\u5355\u7684\u770b\u4e00\u4e0b\u751f\u6210\u6548\u679c":107,"\u6211\u4eec\u53ef\u4ee5\u5b9a\u4e49\u5982\u4e0b\u7684layer\u7ec4\u5408":84,"\u6211\u4eec\u53ef\u4ee5\u5b9a\u4e49\u5982\u4e0blayer\u6765\u63cf\u8ff0\u795e\u7ecf\u7f51\u7edc\u7684\u8f93\u5165":84,"\u6211\u4eec\u53ef\u4ee5\u6309\u7167\u5982\u4e0b\u5c42\u6b21\u5b9a\u4e49\u975e\u5e8f\u5217":110,"\u6211\u4eec\u53ef\u4ee5\u67e5\u770b\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c":107,"\u6211\u4eec\u53ef\u4ee5\u8bbe\u8ba1\u642d\u5efa\u4e00\u4e2a\u7075\u6d3b\u7684":113,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7":107,"\u6211\u4eec\u53ef\u4ee5\u901a\u8fc7\u65e5\u5fd7\u67e5\u770b\u5bb9\u5668\u8bad\u7ec3\u7684\u60c5\u51b5":97,"\u6211\u4eec\u5728":76,"\u6211\u4eec\u5728\u51fd\u6570\u7684\u7ed3\u5c3e\u8fd4\u56de":114,"\u6211\u4eec\u5bf9\u6a21\u578b\u8fdb\u884c\u4e86\u4ee5\u4e0b\u66f4\u6539":114,"\u6211\u4eec\u5c06":97,"\u6211\u4eec\u5c06\u4e00\u6bb5\u8bdd\u770b\u6210\u53e5\u5b50\u7684\u6570\u7ec4":111,"\u6211\u4eec\u5c06\u4ecb\u7ecd\u5982\u4f55\u542f\u52a8\u5206\u5e03\u5f0f\u8bad\u7ec3\u4f5c\u4e1a":96,"\u6211\u4eec\u5c06\u4f7f\u7528":114,"\u6211\u4eec\u5c06\u4f7f\u7528\u7b80\u5355\u7684":114,"\u6211\u4eec\u5c06\u539f\u59cb\u6570\u636e\u7684\u6bcf\u4e00\u7ec4":111,"\u6211\u4eec\u5c06\u5b83\u4eec\u5212\u5206\u4e3a\u4e0d\u540c\u7684\u7c7b\u522b":102,"\u6211\u4eec\u5c06\u795e\u7ecf\u7f51\u7edc\u4e00\u6b21\u8ba1\u7b97\u63a5\u53d7\u7684\u6240\u6709\u8f93\u5165\u6837\u672c\u79f0\u4e4b\u4e3a\u4e00\u4e2a":89,"\u6211\u4eec\u5c06\u8bad\u7ec3\u7ed3\u675f\u540e\u5b58\u50a8\u4e0b\u6765\u7684\u6a21\u578b\u8f6c\u6362\u6210\u9884\u6d4b\u6a21\u578b":90,"\u6211\u4eec\u5c31\u5b8c\u6210\u4e86\u4e00\u6b21\u4ee3\u7801\u8d21\u732e\u7684\u8fc7\u7a0b":72,"\u6211\u4eec\u5df2\u7ecf\u5b9e\u73b0\u4e86\u5927\u591a\u6570\u5e38\u7528\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":114,"\u6211\u4eec\u5efa\u8bae\u4f60\u4e3a\u4f60\u7684python\u5c01\u88c5\u5b9e\u73b0\u4e00\u4e2a":74,"\u6211\u4eec\u5efa\u8bae\u4f60\u5728\u5199\u65b0\u7f51\u7edc\u5c42\u65f6\u628a\u6d4b\u8bd5\u4ee3\u7801\u653e\u5165\u65b0\u7684\u6587\u4ef6\u4e2d":74,"\u6211\u4eec\u5efa\u8bae\u4f7f\u7528\u7b2c\u4e8c\u7c7b\u5b9e\u73b0":82,"\u6211\u4eec\u603b\u7ed3\u51fa\u4e00\u4e9b\u7279\u522b\u9700\u8981\u6ce8\u610f\u7684\u70b9":42,"\u6211\u4eec\u628apaddlepaddle\u7684\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883\u6253\u5305\u6210\u4e00\u4e2a\u955c\u50cf":116,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528":[2,101],"\u6211\u4eec\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":1,"\u6211\u4eec\u63a8\u8350\u60a8\u4f7f\u7528paddlepaddl":1,"\u6211\u4eec\u63d0\u4f9b\u4e24\u4e2a\u8f6c\u6362\u65b9\u5f0f":12,"\u6211\u4eec\u63d0\u4f9b\u4e86\u52a0\u901f\u8bbf\u95ee\u7684\u955c\u50cf\u6e90":2,"\u6211\u4eec\u63d0\u4f9b\u53ef\u4ee5\u76f4\u63a5\u8fd0\u884cpaddlepaddl":2,"\u6211\u4eec\u63d0\u51fa\u4e86chunk\u7684\u6982\u5ff5":28,"\u6211\u4eec\u662f\u5bf9\u6bcf\u4e00\u4e2a\u5b50\u5e8f\u5217\u53d6\u6700\u540e\u4e00\u4e2a\u5143\u7d20":111,"\u6211\u4eec\u6700\u7ec8\u7684\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165python\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u8bed\u8a00\u7684\u89e3\u91ca\u5668":45,"\u6211\u4eec\u6709\u4e00\u4e2a\u5e8f\u5217\u4f5c\u4e3a\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u72b6\u6001":114,"\u6211\u4eec\u7684":0,"\u6211\u4eec\u7684\u6807\u51c6\u5f00\u53d1\u6d41\u7a0b\u662f\u628a\u8fd9\u4e9b\u5de5\u5177\u90fd\u88c5\u8fdb\u4e00\u4e2adocker":72,"\u6211\u4eec\u770b\u4e00\u4e0b\u5355\u5c42rnn\u7684\u914d\u7f6e":111,"\u6211\u4eec\u770b\u4e00\u4e0b\u8bed\u4e49\u76f8\u540c\u7684\u53cc\u5c42rnn\u7684\u7f51\u7edc\u914d\u7f6e":111,"\u6211\u4eec\u771f\u8bda\u5730\u611f\u8c22\u60a8\u7684\u8d21\u732e":72,"\u6211\u4eec\u79f0\u4e4b\u4e3a\u4e00\u4e2a0\u5c42\u7684\u5e8f\u5217":110,"\u6211\u4eec\u8ba1\u5212\u5c06":41,"\u6211\u4eec\u8ba1\u5212\u5c06\u82f1\u7279\u5c14\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u6570\u5b66\u5e93":42,"\u6211\u4eec\u8bbe\u8ba1\u8bf4\u660e\u4e86\u540d\u4e3afilemanager\u7cfb\u7edf":28,"\u6211\u4eec\u8c03\u7528\u4e86eigenvector\u7684flatten\u63a5\u53e3":76,"\u6211\u4eec\u8fd8\u53ef\u4ee5\u767b\u5f55\u5230\u5bbf\u4e3b\u673a\u4e0a\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":96,"\u6211\u4eec\u8fd8\u5c06\u7f16\u7801\u5411\u91cf\u6295\u5c04\u5230":114,"\u6211\u4eec\u9009\u53d6\u5355\u53cc\u5c42\u5e8f\u5217\u914d\u7f6e\u4e2d\u7684\u4e0d\u540c\u90e8\u5206":111,"\u6211\u4eec\u9009\u62e9":12,"\u6211\u4eec\u901a\u5e38\u501f\u52a9":75,"\u6211\u4eec\u901a\u5e38\u5c06\u4e00\u53e5\u8bdd\u7406\u89e3\u6210\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":111,"\u6211\u4eec\u901a\u8fc7\u4f7f\u7528\u65b0\u5f15\u5165\u7684gemm":41,"\u6211\u4eec\u901a\u8fc7\u8bfb\u53d6":97,"\u6211\u4eec\u90fd\u63d0\u4f9bpython\u7684\u8f6c\u6362\u5e93":12,"\u6211\u4eec\u9700\u8981":0,"\u6211\u4eec\u9700\u8981\u5148\u628a\u8f93\u5165tensor\u548c\u8f93\u51fatensor\u8f6c\u6362\u4e3aeigen\u652f\u6301\u7684\u683c\u5f0f":76,"\u6211\u4eec\u9700\u8981\u5236\u4f5c\u4e00\u4e2a\u5305\u542b\u8bad\u7ec3\u6570\u636e\u7684paddlepaddle\u955c\u50cf":96,"\u6211\u4eec\u9700\u8981\u5728\u96c6\u7fa4\u7684\u6240\u6709\u8282\u70b9\u4e0a\u5b89\u88c5":101,"\u6211\u4eec\u9700\u8981\u7b49\u5f0f\u5de6\u8fb9\u7684eigentensor\u8c03\u7528device\u63a5\u53e3":76,"\u6211\u4eec\u9700\u8981\u8ba1\u7b97":74,"\u6211\u4eec\u9996\u5148\u9700\u8981\u6839\u636e\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u6765\u521b\u5efa\u6240\u9700\u8981\u4f18\u5316\u7684paramet":84,"\u6211\u5220\u9664\u4e86":72,"\u6211\u53ef\u4ee5\u7528":0,"\u6211\u53ef\u4ee5\u9009\u62e9\u4e0d\u7528docker\u5417":0,"\u6216":[89,108,117],"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u6216\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"\u6216\u4e00\u4e2a\u5411\u91cf":113,"\u6216\u5355\u5c42\u5e8f\u5217\u7ecf\u8fc7\u8fd0\u7b97\u53d8\u6210\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"\u6216\u662f\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u4e2d\u8868\u793a\u8bcd\u8bed\u5728\u5b57\u5178\u4e2d\u7684\u5e8f\u53f7":89,"\u6216\u6700\u5927\u503c":110,"\u6216\u79f0\u4f5cweight":81,"\u6216\u7b2c\u4e00\u4e2a":110,"\u6216\u7b2c\u4e00\u4e2a\u5143\u7d20":110,"\u6216\u7f16\u5199\u7a0b\u5e8f\u65f6":91,"\u6216\u8005":[0,42,45,46,72,75,81,89,107,108,110,111],"\u6216\u8005\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":113,"\u6216\u8005\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":[110,113],"\u6216\u8005\u4ece\u5de5\u5177\u7684\u754c\u9762\u91cc\u8fd0\u884c\u60a8\u7684\u5e94\u7528":108,"\u6216\u8005\u5236\u4f5c\u548c\u5206\u4eab\u5e26\u6709\u4ee3\u7801":2,"\u6216\u8005\u53cd\u5411\u5730\u4ece":114,"\u6216\u8005\u53ef\u88abdns\u89e3\u6790\u7684\u4e3b\u673a\u540d":101,"\u6216\u8005\u5728cpu\u6a21\u5f0f\u4e0b\u4f7f\u75284\u4e2a\u7ebf\u7a0b":103,"\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":11,"\u6216\u8005\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":102,"\u6216\u8005\u5f15\u8d77\u884c\u65f6\u9519\u8bef":89,"\u6216\u8005\u6570\u7ec4\u7684\u6570\u7ec4\u8fd9\u4e2a\u6982\u5ff5":111,"\u6216\u8005\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":110,"\u6216\u8005\u662f\u51fd\u6570\u8c03\u7528\u7684\u9891\u7387\u548c\u8017\u65f6\u7b49":108,"\u6216\u8005\u66f4\u65e9":83,"\u6216\u8005\u6bcf\u4e00\u4e2a\u7cfb\u5217\u91cc\u7684\u7279\u5f81\u6570\u636e":111,"\u6216\u8005\u7528tuple\u8868\u793a\u7684\u591a\u4e2a\u503c":12,"\u6216\u8005\u7531\u5b83\u4eec\u7ec4\u6210\u7684list":12,"\u6216\u8005\u76f4\u63a5\u6254\u6389\u975e\u5e38\u957f\u7684\u5e8f\u5217":81,"\u6216\u8005\u76f8\u5bf9\u4e8e\u6784\u5efa\u76ee\u5f55\u7684\u76f8\u5bf9\u8def\u5f84":[116,118],"\u6216\u8005\u8f93\u5165\u6570\u636e\u5c3a\u5ea6\u8fc7\u5927":81,"\u6216\u8005\u8fd0\u884c":78,"\u6216\u8005\u91c7\u7528\u5e76\u884c\u8ba1\u7b97\u6765\u52a0\u901f\u67d0\u4e9b\u5c42\u7684\u66f4\u65b0":105,"\u6216activ":42,"\u6216gpu":103,"\u622a\u65ad\u5bf9\u8c61\u4e0d\u540c":81,"\u6240\u4ee5":[2,42,62,81,107],"\u6240\u4ee5\u4e00\u4e2a\u7248\u672c\u53f7\u7684wheel\u5305\u53d1\u5e03\u4e4b\u540e":62,"\u6240\u4ee5\u4e0d\u5b58\u5728\u8fd9\u4e2a\u95ee\u9898":42,"\u6240\u4ee5\u4e0d\u80fd\u91c7\u7528\u7b2c\u4e00\u79cd\u65b9\u5f0f\u5728\u8fd9\u51e0\u4e2alayer\u91cc\u8bbe\u7f6e":82,"\u6240\u4ee5\u505a\u6cd5\u53ef\u4ee5\u6709\u4e24\u79cd":81,"\u6240\u4ee5\u53ef\u4ee5\u7b80\u5316\u5bf9\u73af\u5883\u7684\u8981\u6c42":96,"\u6240\u4ee5\u5728":42,"\u6240\u4ee5\u5728\u5199\u5165\u5feb\u7167\u7684\u8fc7\u7a0b\u4e2d":11,"\u6240\u4ee5\u5916\u5c42\u8f93\u51fa\u7684\u5e8f\u5217\u5f62\u72b6":111,"\u6240\u4ee5\u5bf9":111,"\u6240\u4ee5\u5f00\u53d1\u8005\u9700\u8981\u6839\u636e\u81ea\u5df1\u8bad\u7ec3\u4efb\u52a1\u7684\u5b9e\u9645\u573a\u666f\u5b8c\u6210\u8bad\u7ec3\u6570\u636e\u7684\u5206\u5272\u548c":91,"\u6240\u4ee5\u6027\u80fd\u4e5f\u5c31\u9010\u6b65\u53d8\u6210\u4e86\u6df1\u5ea6\u5b66\u4e60\u9886\u57df\u6700\u91cd\u8981\u7684\u6307\u6807":108,"\u6240\u4ee5\u6211\u4eec\u53ef\u4ee5\u5728\u8fd9\u4e2a\u57fa\u7840\u4e0a":97,"\u6240\u4ee5\u6211\u4eec\u5b9a\u4e49\u4e86\u4e00\u4e2a":42,"\u6240\u4ee5\u6211\u4eec\u786e\u4fdd\u53d1\u5e03\u7684\u4e8c\u8fdb\u5236\u5305\u53ef\u4ee5\u652f\u6301\u4e3b\u6d41\u7684linux\u64cd\u4f5c\u7cfb\u7edf":4,"\u6240\u4ee5\u6211\u4eec\u9700\u8981\u5c06\u8f93\u5165\u6570\u636e\u6807\u8bb0\u6210":111,"\u6240\u4ee5\u6211\u4eec\u9ed8\u8ba4\u4f7f\u7528cento":4,"\u6240\u4ee5\u6574\u4f53\u4e0a":42,"\u6240\u4ee5\u6dfb\u52a0\u4e86\u5bf9\u5e94\u7684":42,"\u6240\u4ee5\u7528\u6237\u9700\u8981\u9996\u5148\u5728":28,"\u6240\u4ee5\u76f8\u6bd4\u4e8erecurr":82,"\u6240\u4ee5\u8fd9\u4e00\u6b65\u662f\u5fc5\u8981\u7684":74,"\u6240\u4ee5\u9700\u8981\u5f15\u5165\u4e00\u4e2a\u8f6c\u6362\u65b9\u6cd5":42,"\u6240\u4f7f\u7528":117,"\u6240\u4f9d\u8d56\u7684\u7b2c\u4e09\u65b9\u5e93\u540c\u65f6\u4e5f\u88ab\u5b89\u88c5\u5230":116,"\u6240\u5bf9\u5e94\u7684\u8bcd\u8868index\u6570\u7ec4":111,"\u6240\u6709\u4e0e\u7c7b\u578b\u76f8\u5173\u7684\u51fd\u6570":46,"\u6240\u6709\u4ee3\u7801\u5fc5\u987b\u5177\u6709\u5355\u5143\u6d4b\u8bd5":72,"\u6240\u6709\u53c2\u6570\u7f6e\u4e3a\u96f6":103,"\u6240\u6709\u547d\u4ee4\u884c\u9009\u9879\u53ef\u4ee5\u8bbe\u7f6e\u4e3a":93,"\u6240\u6709\u5916\u90e8\u7684\u8f6c\u6362\u5de5\u4f5c\u90fd\u4f1a\u5728reset\u7cfb\u5217\u51fd\u6570\u4e2d\u90fd\u51c6\u5907\u597d":42,"\u6240\u6709\u67b6\u6784":116,"\u6240\u6709\u7684":[41,72,74],"\u6240\u6709\u7684\u5355\u6d4b\u90fd\u4f1a\u88ab\u6267\u884c\u4e00\u6b21":74,"\u6240\u6709\u7684\u63a5\u53e3\u5747\u4e3ac\u63a5\u53e3":46,"\u6240\u6709\u7684\u64cd\u4f5c\u90fd\u662f\u9488\u5bf9\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u6765\u8fdb\u884c\u7684":111,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u4f7f\u7528":74,"\u6240\u6709\u7684python\u5c01\u88c5\u90fd\u5728":74,"\u6240\u6709\u76f8\u5173\u7684":41,"\u6240\u6709\u7c7b\u578b\u540d\u4e3a":46,"\u6240\u6709\u7f51\u7edc\u5c42\u7684\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u90fd\u4f4d\u4e8e":74,"\u6240\u6709\u8f93\u5165\u5e8f\u5217\u5e94\u8be5\u6709\u76f8\u540c\u7684\u957f\u5ea6":114,"\u6240\u6709mkl":42,"\u6240\u9700\u652f\u6301\u7684\u6700\u4f4eandroid":116,"\u6240\u9700\u7684\u5f00\u53d1\u5de5\u5177\u548c\u7b2c\u4e09\u65b9\u5e93\u53ef\u4ee5\u53c2\u8003":118,"\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":45,"\u624b\u5199\u6570\u5b57\u8bc6\u522b":90,"\u624b\u5199\u6570\u5b57\u8bc6\u522b\u4efb\u52a1\u5b9a\u4e49\u4e86\u4e00\u4e2a\u542b\u6709":90,"\u624b\u52a8\u4e0b\u8f7d\u4e14\u89e3\u538b\u7f29":78,"\u624b\u52a8\u4e0b\u8f7d\u5e76\u5b89\u88c5":78,"\u624d\u53ef\u4ee5\u5b89\u88c5":4,"\u624d\u80fd\u4fdd\u8bc1\u548c\u5355\u5c42\u5e8f\u5217\u7684\u914d\u7f6e\u4e2d":111,"\u624d\u80fd\u53d1\u6325\u5176\u5168\u90e8\u80fd\u529b":108,"\u624d\u80fd\u66f4\u597d\u7684\u53d1\u6325mkl":42,"\u6253\u5f00":108,"\u6253\u5f00\u6d4f\u89c8\u5668\u8bbf\u95ee\u5bf9\u5e94\u76ee\u5f55\u4e0b\u7684index":77,"\u6253\u5f00\u8fd9\u4e2a\u7f16\u8bd1\u9009\u9879":46,"\u6267\u884c":[62,86,87,93],"\u6267\u884c\u4e0a\u8ff0":116,"\u6267\u884c\u4e0a\u8ff0\u4ee3\u7801\u751f\u6210makefile\u6587\u4ef6\u540e":87,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4":1,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u4ee5\u542f\u52a83\u4e2a\u8282\u70b9\u7684openmpi\u96c6\u7fa4\u548c\u4e00\u4e2a":98,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5373\u53ef\u5728\u5f53\u524d\u673a\u5668\u4e0a\u5b89\u88c5paddlepaddle\u7684\u8fd0\u884c\u65f6\u73af\u5883":4,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u53ef\u4ee5\u67e5\u770b\u5df2\u7ecf\u5b89\u88c5\u7684\u7248\u672c":101,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u5b8c\u6210\u5feb\u901f\u5b89\u88c5":86,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u6765\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5":75,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u7f16\u8bd1cpu":1,"\u6267\u884c\u4e0b\u9762\u7684\u547d\u4ee4\u83b7\u53d6\u6700\u65b0\u7684paddlepaddl":2,"\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4":[116,117,118],"\u6267\u884c\u4ee5\u4e0b\u547d\u4ee4\u542f\u52a8\u4f7f\u7528python\u7f16\u5199\u7684trainer\u7a0b\u5e8f":91,"\u6267\u884c\u4ee5\u4e0b\u64cd\u4f5c":114,"\u6267\u884c\u4ee5\u4e0b\u7684\u547d\u4ee4\u542f\u52a8\u4e00\u4e2a\u53c2\u6570\u670d\u52a1\u5668\u5e76\u7b49\u5f85\u548c\u8ba1\u7b97\u8282\u70b9\u7684\u6570\u636e\u4ea4\u4e92":91,"\u6267\u884c\u5b8c\u5b89\u88c5\u547d\u4ee4\u540e":[116,117,118],"\u6267\u884c\u60a8\u7684\u4ee3\u7801":108,"\u627e\u5230":[1,94,114],"\u627e\u5230\u6700\u65e9\u62a5\u9519\u7684\u5730\u65b9":79,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u539f\u56e0":108,"\u627e\u5230\u8fd0\u884c\u6162\u7684\u90e8\u5206":108,"\u628a":[12,74],"\u628a\u4e4b\u524d\u793a\u4f8b\u4e2d\u8f6c\u6362\u5b8c\u6bd5\u7684random":12,"\u628a\u4efb\u610f\u7ef4\u5ea6\u7684tensor\u8f6c\u4e3a\u4e86\u4e00\u7ef4\u7684eigenvector":76,"\u628a\u5de5\u5177\u548c\u914d\u7f6e\u90fd\u5b89\u88c5\u5728\u4e00\u4e2a":0,"\u628a\u8bad\u7ec3\u6570\u636e\u76f4\u63a5\u653e\u5728":96,"\u628a\u8fd9\u4e9b\u5de5\u5177\u5b89\u88c5\u5230\u672c\u673a":0,"\u6295\u5c04\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b\u5230":114,"\u6295\u5c04\u7f16\u7801\u5411\u91cf\u5230":114,"\u62c6\u6210\u4ee5\u4e0a\u4e24\u4e2a\u9759\u6001\u94fe\u63a5\u5e93":87,"\u62c6\u89e3":113,"\u62c6\u89e3\u6210\u7684\u6bcf\u4e00\u53e5\u8bdd\u518d\u901a\u8fc7\u4e00\u4e2alstm\u7f51\u7edc":111,"\u62f7\u8d1d\u5230numpi":81,"\u62f7\u8d1d\u5fc5\u8981\u7684\u6587\u4ef6\u5230head\u8282\u70b9":98,"\u62f7\u8d1d\u8bad\u7ec3\u6570\u636e\u5230\u5404\u81ea\u7684\u8282\u70b9":98,"\u62f7\u8d1d\u8bad\u7ec3\u6587\u4ef6\u5230\u5bb9\u5668\u5185":97,"\u62f7\u8d1d\u8bad\u7ec3\u7a0b\u5e8f\u548c\u5b57\u5178\u6587\u4ef6\u5230\u6bcf\u53f0mpi\u8282\u70b9":98,"\u62fc\u63a5":81,"\u6302\u8f7d\u5230\u5bb9\u5668\u5185\u90e8\u7684":2,"\u6302\u8f7d\u6216\u4e0b\u8f7d\u7684\u8bad\u7ec3\u6570\u636e\u5206\u7247":91,"\u6307\u53d1\u73b0\u6027\u80fd\u74f6\u9888":107,"\u6307\u5411\u4e00\u4e2alayer":113,"\u6307\u5b9a":[81,82,113,114],"\u6307\u5b9a\u4e00\u53f0\u673a\u5668\u4e0a\u4f7f\u7528\u7684\u7ebf\u7a0b\u6570":103,"\u6307\u5b9a\u4e3a":89,"\u6307\u5b9a\u4f7f\u75282":81,"\u6307\u5b9a\u524d\u5411\u7f51\u7edc\u6700\u7ec8\u7684\u8f93\u51fa\u76ee\u6807\u53d8\u91cf":75,"\u6307\u5b9a\u52a0\u8f7d\u7684\u65b9\u5f0f":103,"\u6307\u5b9a\u5728\u751f\u6210\u6027\u80fd\u5206\u6790\u6587\u4ef6\u4e4b\u540e":107,"\u6307\u5b9a\u5bf9\u8f93\u5165\u53d8\u91cf":75,"\u6307\u5b9a\u5c06\u5f53\u524d\u8def\u5f84":2,"\u6307\u5b9a\u6267\u884c\u5176\u4e2d\u4e00\u4e2a\u5355\u5143\u6d4b\u8bd5":1,"\u6307\u5b9a\u68c0\u6d4b\u68af\u5ea6\u65f6\u80fd\u5bb9\u5fcd\u7684\u6700\u5927\u9519\u8bef\u503c":75,"\u6307\u5b9a\u7684\u5185\u5bb9\u5b58\u50a8\u5e93\u8fd0\u884c\u547d\u4ee4":77,"\u6307\u5b9a\u7684\u8f93\u5165\u4e0d\u4f1a\u88ab":113,"\u6307\u5b9a\u8981\u8f93\u51fa\u7684\u5b57\u6bb5\u8fdb\u884c\u8f93\u51fa":81,"\u6307\u5b9a\u9700\u8981\u4f7f\u7528\u7684\u5bb9\u5668":2,"\u6307\u5b9acudnn\u7684\u6700\u5927\u5de5\u4f5c\u7a7a\u95f4\u5bb9\u9650":103,"\u6307\u5bf9\u4e8e\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217\u8f93\u5165\u6570\u636e":111,"\u6307\u5f00\u542fhttp\u670d\u52a1":107,"\u6307\u6d88\u9664\u74f6\u9888":107,"\u6307\u6df1\u5ea6\u5b66\u4e60\u8bad\u7ec3\u4e4b\u540e\u5f97\u5230\u7684\u6240\u6709\u53c2\u6570":11,"\u6307\u793a\u4f7f\u7528\u54ea\u4e2agpu\u6838":103,"\u6307\u793a\u5728\u7b80\u5355\u7684recurrentlayer\u5c42\u7684\u8ba1\u7b97\u4e2d\u662f\u5426\u4f7f\u7528\u6279\u5904\u7406\u65b9\u6cd5":103,"\u6307\u793a\u5f53\u6307\u5b9a\u8f6e\u7684\u6d4b\u8bd5\u6a21\u578b\u4e0d\u5b58\u5728\u65f6":103,"\u6307\u793a\u662f\u5426\u4f7f\u7528\u591a\u7ebf\u7a0b\u6765\u8ba1\u7b97\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc":103,"\u6307\u793a\u662f\u5426\u5f00\u542f\u53c2\u6570\u670d\u52a1\u5668":103,"\u6307\u793a\u662f\u5426\u663e\u793a\u53c2\u6570\u670d\u52a1\u5668\u4e0a\u7684\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u65e5\u5fd7\u7ec6\u8282":103,"\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":103,"\u6309\u542f\u53d1\u5f0f\u635f\u5931\u7684\u5927\u5c0f\u9012\u589e\u6392\u5e8f":103,"\u6309\u7167\u4e0b\u9762\u6b65\u9aa4\u5373\u53ef":97,"\u6309\u7167\u5176\u5185\u5bb9\u521b\u5efa\u4e00\u4e2a\u540d\u4e3a":0,"\u6309\u7167\u5177\u4f53\u5b9e\u73b0\u65b9\u5f0f\u53ef\u4ee5\u5f52\u7eb3\u4e3a2\u7c7b":82,"\u6309\u7167\u57fa\u672c\u6570\u636e\u7c7b\u578b\u5728paddlepaddle\u5185\u90e8\u7684\u5b9a\u4e49\u548c\u5b9e\u73b0":89,"\u6309\u94ae":[62,72],"\u635f\u5931\u51fd\u6570\u5c42":90,"\u6392\u6210\u4e00\u5217\u7684\u591a\u4e2a\u5143\u7d20":110,"\u63a5\u4e0a\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42":84,"\u63a5\u4e0a\u5e73\u65b9\u8bef\u5dee\u5c42":84,"\u63a5\u4e0b\u6765":75,"\u63a5\u4e0b\u6765\u53ef\u4ee5\u8003\u8651\u4e0b\u65f6\u95f4\u7ebf\u7684\u5206\u6790":108,"\u63a5\u4e0b\u6765\u5c31\u53ef\u4ee5\u4f7f\u7528":108,"\u63a5\u4e0b\u6765\u6211\u4eec\u521b\u5efa\u4e00\u4e2a\u539f\u59cb":72,"\u63a5\u4e0b\u6765\u6211\u4eec\u53d6\u6d88\u5bf9":72,"\u63a5\u4e0b\u6765\u7b49\u5f85":72,"\u63a5\u53d7\u4e00\u4e2a\u8f93\u5165\u53c2\u6570":75,"\u63a5\u53e3":[45,46,75,76,90],"\u63a5\u53e3\u4f1a\u88ab\u8c03\u7528":76,"\u63a5\u53e3\u5b8c\u6210\u795e\u7ecf\u7f51\u7edc\u7684\u524d\u5411\u8ba1\u7b97":90,"\u63a5\u53e3\u5bf9\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u53c2\u6570\u8fdb\u884c\u5e8f\u5217\u5316":90,"\u63a5\u53e3\u5c42\u505a\u8fc7\u591a\u5c01\u88c5":46,"\u63a5\u53e3\u662f":12,"\u63a5\u53e3\u6700\u7ec8\u4f1a\u8c03\u7528\u5bf9\u5e94":76,"\u63a5\u53e3\u6709\u4e00\u4e2a":81,"\u63a5\u53e3\u7684":81,"\u63a5\u53e3\u8bf4\u660e\u8bf7\u67e5\u770b":89,"\u63a5\u6536\u5904\u7406pfsclient\u7aef\u7684\u6587\u4ef6\u7ba1\u7406\u8bf7\u6c42":28,"\u63a5\u7740\u7f16\u8bd1\u5373\u53ef":78,"\u63a7\u5236":103,"\u63a7\u5236\u662f\u5426\u4f7f\u7528mkl":42,"\u63a7\u5236\u662f\u5426\u4f7f\u7528mklml\u5e93":42,"\u63a7\u5236\u7528\u6237\u6743\u9650":12,"\u63a8\u5bfc\u8be5\u5c42\u524d\u5411\u548c\u540e\u5411\u4f20\u9012\u7684\u65b9\u7a0b":74,"\u63a8\u8350\u4f7f\u7528\u6b64\u65b9\u5f0f":87,"\u63a8\u8350\u4f7f\u7528centos\u7684devtools2":1,"\u63a8\u8350\u6e05\u7406\u6574\u4e2a\u7f16\u8bd1\u76ee\u5f55":1,"\u63a8\u8350\u8bbe\u7f6e\u4e3a":87,"\u63a8\u8350\u914d\u7f6e\u4e3a":87,"\u63a8\u8350\u914d\u7f6e\u9009\u9879":87,"\u63a8\u9001\u5230\u8fdc\u7a0b\u4ed3\u5e93":72,"\u63cf\u8ff0\u7684\u9ed8\u8ba4\u5165\u53e3\u7a0b\u5e8f":0,"\u63cf\u8ff0\u8be5op\u7684\u8f93\u5165":75,"\u63cf\u8ff0\u95ee\u9898":72,"\u63d0\u4ea4\u65b9\u5f0f\u53c2\u89c1":77,"\u63d0\u4ea4pull":72,"\u63d0\u4f9b":93,"\u63d0\u4f9b\u4e03\u5c42\u534f\u8bae\u7684\u53cd\u5411\u4ee3\u7406":28,"\u63d0\u4f9b\u4e86\u4e00\u4e2a\u542f\u52a8\u811a\u672c":97,"\u63d0\u4f9b\u4e86\u547d\u4ee4\u6837\u4f8b\u6765\u8fd0\u884c":93,"\u63d0\u4f9b\u4e86\u65b9\u4fbf\u7684\u548c":107,"\u63d0\u4f9b\u4e86\u81ea\u52a8\u5316\u811a\u672c\u6765\u542f\u52a8\u4e0d\u540c\u8282\u70b9\u4e2d\u7684\u6240\u6709":93,"\u63d0\u4f9b\u51e0\u4e4e\u6240\u6709\u8bad\u7ec3\u7684\u5185\u90e8\u8f93\u51fa\u65e5\u5fd7":93,"\u63d0\u4f9b\u5e38\u7528\u7684\u547d\u4ee4\u884c\u7ba1\u7406\u547d\u4ee4\u7ba1\u7406\u6587\u4ef6\u548c\u76ee\u5f55":28,"\u63d0\u4f9b\u6269\u5c55\u7684\u957f\u5ea6\u4fe1\u606f":110,"\u63d0\u4f9b\u7528\u6237\u7ba1\u7406\u6587\u4ef6\u7684\u547d\u4ee4":28,"\u63d0\u4f9b\u7ed9paddle\u4f5c\u4e3a\u8bad\u7ec3\u6570\u636e":12,"\u63d0\u4f9b\u8bad\u7ec3\u8fc7\u7a0b\u7684":93,"\u63d0\u793a":78,"\u641c\u7d22\u4ee3\u7801\u5e93":77,"\u642d\u5efa\u795e\u7ecf\u7f51\u7edc\u5c31\u50cf\u4f7f\u7528\u79ef\u6728\u642d\u5efa\u5b9d\u5854\u4e00\u6837":84,"\u64cd\u4f5c":111,"\u64cd\u4f5c\u7cfb\u7edf":[0,4],"\u652f\u6301\u4e24\u79cd\u5e8f\u5217\u7c7b\u578b":89,"\u652f\u6301\u4ea4\u53c9\u7f16\u8bd1":118,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":[112,113],"\u652f\u6301\u5927\u6587\u4ef6\u7684\u65ad\u70b9\u4e0a\u4f20":28,"\u652f\u6301\u5927\u89c4\u6a21\u96c6\u7fa4\u751f\u4ea7\u73af\u5883\u7684\u5b8c\u6574\u96c6\u7fa4\u65b9\u6848":94,"\u652f\u6301\u7684\u6700\u5c0f\u7684android":116,"\u652f\u6301\u7684\u6700\u5c0fandroid":116,"\u652f\u6301\u7ef4\u6570\u53ef\u53d8\u7684\u6570\u636e\u8f93\u5165":82,"\u652f\u6301\u7f16\u8bd1\u5668":116,"\u6539\u53d8\u7ef4\u5ea6\u987a\u5e8f":82,"\u653e\u5728\u8fd9\u4e2a\u76ee\u5f55\u91cc\u7684\u6587\u4ef6\u5176\u5b9e\u662f\u4fdd\u5b58\u5230\u4e86mfs\u4e0a":97,"\u6545\u800c\u662f\u4e00\u4e2a\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":111,"\u6548\u679c\u5982\u4e0b":107,"\u6559\u7a0b":2,"\u6570":113,"\u6570\u5b66\u5e93":87,"\u6570\u5fc5\u987b\u4e25\u683c\u76f8\u7b49":113,"\u6570\u636e":[28,89,90],"\u6570\u636e\u4e2d0":83,"\u6570\u636e\u5206\u7247":92,"\u6570\u636e\u63d0\u4f9b\u5668":102,"\u6570\u636e\u8bfb\u53d6\u5747\u4ea4\u7531\u5176\u4ed6\u8bed\u8a00\u5b8c\u6210":45,"\u6570\u636e\u8f93\u5165":[89,113],"\u6570\u636e\u957f\u5ea6\u53ca\u6821\u9a8c\u503c\u7ec4\u6210":28,"\u6570\u636e\u96c6":90,"\u6570\u636e\u96c6\u9700\u8981\u9884\u5148\u88ab\u8f6c\u6362\u6210paddlepaddle\u5206\u5e03\u5f0f\u8bad\u7ec3\u4f7f\u7528\u7684\u5b58\u50a8\u683c":12,"\u6570\u636e\u9884\u5904\u7406\u4efb\u52a1":12,"\u6570\u76ee":105,"\u6574\u4f53\u4f7f\u7528\u6d41\u7a0b":90,"\u6574\u4f53\u6570\u636e\u548c\u539f\u59cb\u6570\u636e\u5b8c\u5168\u4e00\u6837":111,"\u6574\u4f53\u7684\u7ed3\u6784\u56fe\u5982\u4e0b":97,"\u6574\u578b\u6570\u7ec4":89,"\u6574\u6570":74,"\u6574\u6570\u6807\u7b7e":84,"\u6587\u4ef6":[0,45,72,75,89,96],"\u6587\u4ef6\u4e2d":[75,90,97],"\u6587\u4ef6\u4e2d\u6ce8\u518c\u524d\u5411":75,"\u6587\u4ef6\u4e2d\u6ce8\u518c\u8be5op\u548ckernel":75,"\u6587\u4ef6\u4e2d\u6ce8\u518ccuda":75,"\u6587\u4ef6\u4e3a":81,"\u6587\u4ef6\u4e4b\u5916":72,"\u6587\u4ef6\u4f20\u8f93\u7684\u7684\u5173\u952e\u5728\u4e8e\u9700\u8981pfsclient\u7aef\u5bf9\u6bd4source\u548cdestination\u7684\u6587\u4ef6chunks\u7684checksum\u662f\u5426\u4fdd\u6301\u4e00\u81f4":28,"\u6587\u4ef6\u5185\u5bb9\u4e3a":45,"\u6587\u4ef6\u540d":107,"\u6587\u4ef6\u540d\u4e3a\u4efb\u610f\u6587\u4ef6\u540d":91,"\u6587\u4ef6\u540d\u4e3a\u6b64uuid":11,"\u6587\u4ef6\u547d\u540d\u4ee5":75,"\u6587\u4ef6\u59390":97,"\u6587\u4ef6\u5bf9\u5e94\u7684data":12,"\u6587\u4ef6\u5de5\u5177\u662f\u4f7f\u7528docker":77,"\u6587\u4ef6\u7684\u4e0a\u4f20\u548c\u4e0b\u8f7d\u90fd\u662f\u901a\u8fc7\u5bf9chunk\u7684\u64cd\u4f5c\u6765\u5b9e\u73b0\u7684":28,"\u6587\u4ef6\u7684\u6539\u53d8":72,"\u6587\u4ef6\u7684\u8def\u5f84\u6765\u52a0\u8f7d\u9884\u6d4b\u6a21\u578b":90,"\u6587\u4ef6model":105,"\u6587\u5b57\u7684\u4ea4\u4e92\u5f0f\u6587\u6863":2,"\u6587\u6863":78,"\u6587\u68631":76,"\u6587\u68632":76,"\u6587\u6863\u8f83\u5c11":76,"\u6587\u6863\u90fd\u662f\u901a\u8fc7":77,"\u6587\u7ae0":97,"\u65b0\u5efa\u4e00\u4e2a\u6743\u91cd":74,"\u65b0\u624b\u5165\u95e8":115,"\u65b0\u624b\u5165\u95e8\u7ae0\u8282":62,"\u65b0\u7248\u672c":42,"\u65b9\u4fbf\u5feb\u901f\u5b89\u88c5":3,"\u65b9\u4fbf\u6392\u67e5\u4ee5\u53ca\u5feb\u901f\u5b9a\u4f4d\u95ee\u9898":81,"\u65b9\u4fbf\u6d4b\u8bd5\u4eba\u5458\u6d4b\u8bd5paddlepaddle\u7684\u884c\u4e3a":62,"\u65b9\u4fbf\u7528\u6237\u4e0a\u4f20\u81ea\u5df1\u7684\u8bad\u7ec3\u6570\u636e\u4ee5\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3":28,"\u65b9\u4fbf\u7528\u6237\u5728python\u7aef\u9009\u62e9\u662f\u5426\u542f\u7528\u8fd9\u4e2a\u529f\u80fd":41,"\u65b9\u4fbf\u7528\u6237\u9009\u62e9\u4f7f\u7528mkl":42,"\u65b9\u5f0f1":81,"\u65b9\u5f0f2":81,"\u65b9\u5f0f\u5c06\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u53c2\u6570\u5e8f\u5217\u5316\u5230\u4e00\u4e2a\u6587\u4ef6":90,"\u65b9\u5f0f\u7c7b\u4f3c\u4e8e":42,"\u65b9\u6cd5\u4e00":105,"\u65b9\u6cd5\u4e09":105,"\u65b9\u6cd5\u4e8c":105,"\u65e0\u5ef6\u8fdf":103,"\u65e0\u6cd5\u505a\u5230\u5bf9\u4e8e\u5404\u79cd\u8bed\u8a00\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u7684\u9002\u914d":45,"\u65e0\u8bba\u5728\u672c\u5730\u8fd8\u662f\u5728\u4e91\u7aef":12,"\u65e0\u8bba\u662f\u4ece":12,"\u65e0\u8bba\u662f\u5728\u672c\u5730\u6216\u662f\u4e91\u7aef\u8f6c\u6362":12,"\u65e0\u8bba\u662f\u91cd\u6784\u524d\u7684layer\u8fd8\u662f\u91cd\u6784\u540e\u7684op":42,"\u65e0\u9700\u5173\u5fc3\u548c\u5904\u7406\u5e8f\u5217\u4fe1\u606f":89,"\u65e0\u9700\u63d0\u4f9b\u975e\u96f6\u5143\u7684\u503c":89,"\u65e0\u9700\u9644\u52a0\u5e8f\u5217\u4fe1\u606f":89,"\u65e0\u9ed8\u8ba4\u503c":[116,118],"\u65e5\u5fd7\u62a5\u9519\u4e3a\u7f51\u7edc\u901a\u4fe1\u7c7b\u9519\u8bef":79,"\u65f6":[11,41,42,74,81,83,87,89,97,103,110,114,116],"\u65f6\u4e00\u8d77\u66f4\u65b0":42,"\u65f6\u4f7f\u7528openblas\u6570\u5b66\u5e93":87,"\u65f6\u5982\u4f55\u7ec4\u7ec7\u8f93\u5165\u6570\u636e":89,"\u65f6\u6709\u6548":117,"\u65f6\u88ab\u8bad\u7ec3\u7684":74,"\u65f6\u8bbe\u5907id\u53f7\u7684\u5206\u914d":105,"\u65f6\u95f4":111,"\u65f6\u95f4\u6b65\u7684\u6982\u5ff5":111,"\u65f6\u987b\u4ece\u7b2c17\u5b57\u8282\u5f00\u59cb":83,"\u6620\u5c04\u4e3a":0,"\u6620\u5c04\u5230\u4e00\u4e2a\u7ef4\u5ea6\u4e3a":74,"\u662f":[4,28,42,78],"\u662f\u4e00\u4e2a\u51681\u7684\u5411\u91cf":74,"\u662f\u4e00\u4e2a\u5185\u7f6e\u7684\u5b9a\u65f6\u5668\u5c01\u88c5":108,"\u662f\u4e00\u4e2a\u52a8\u6001\u7a0b\u5e8f\u5206\u6790\u7684\u672f\u8bed":108,"\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u662f\u4e00\u4e2a\u53cc\u5c42\u7684\u5e8f\u5217":110,"\u662f\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3\u7684\u4ee3\u7801\u751f\u6210\u5668":45,"\u662f\u4e00\u4e2a\u5c01\u88c5\u5bf9\u8c61":108,"\u662f\u4e00\u4e2a\u5f88\u6709\u7528\u7684\u53c2\u6570":105,"\u662f\u4e00\u4e2a\u7c7b\u578b\u7684\u6807\u5fd7":46,"\u662f\u4e00\u4e2a\u975e\u7ebf\u6027\u7684":74,"\u662f\u4e00\u4e2apython\u7684\u7b2c\u4e09\u65b9\u5e93":107,"\u662f\u4e00\u4e2aunbound":113,"\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":84,"\u662f\u4e00\u6b21\u9884\u6d4b\u63a5\u53d7\u7684\u6837\u672c\u6570\u76ee":89,"\u662f\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":113,"\u662f\u4e0d\u5305\u62ec\u6e90\u7801\u7684":96,"\u662f\u4e0d\u5e38\u89c1\u7684\u505a\u6cd5":45,"\u662f\u4f7f\u5f97\u8981\u5171\u4eab\u7684\u53c2\u6570\u4f7f\u7528\u540c\u6837\u7684":83,"\u662f\u4f7f\u7528mkl\u6570\u5b66\u5e93":87,"\u662f\u504f\u5dee":114,"\u662f\u5404\u4e2a\u5b9e\u73b0\u4e2d\u5171\u4eab\u7684\u5934\u6587\u4ef6":46,"\u662f\u5411\u91cf":74,"\u662f\u5426\u4ec5\u7f16\u8bd1capi":1,"\u662f\u5426\u4ee5\u9006\u5e8f\u5904\u7406\u8f93\u5165\u5e8f\u5217":113,"\u662f\u5426\u4f7f\u7528":117,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":1,"\u662f\u5426\u4f7f\u7528\u65e7\u7684remoteparameterupdat":103,"\u662f\u5426\u4f7f\u7528\u6743\u91cd":74,"\u662f\u5426\u4f7f\u7528arm\u6a21\u5f0f":116,"\u662f\u5426\u4f7f\u7528eigen\u5e93\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":[116,117],"\u662f\u5426\u4f7f\u7528mkl\u6570\u5b66\u5e93":1,"\u662f\u5426\u4f7f\u7528neon\u6307\u4ee4":[116,118],"\u662f\u5426\u4f7f\u80fd":117,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":1,"\u662f\u5426\u5219\u5171\u4eab\u540c\u4e00\u4e2a":75,"\u662f\u5426\u542f\u7528gpu\u8bad\u7ec3":91,"\u662f\u5426\u5c06\u5168\u5c40\u79cd\u5b50\u5e94\u7528\u4e8e\u672c\u5730\u7ebf\u7a0b\u7684\u968f\u673a\u6570":103,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":1,"\u662f\u5426\u6253\u5370\u7248\u672c\u4fe1\u606f":103,"\u662f\u5426\u6253\u5f00":41,"\u662f\u5426\u652f\u6301gpu":1,"\u662f\u5426\u663e\u793a":103,"\u662f\u5426\u7a00\u758f":74,"\u662f\u5426\u7f16\u8bd1\u4e2d\u82f1\u6587\u6587\u6863":1,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236\u6587\u4ef6":1,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":1,"\u662f\u5426\u7f16\u8bd1c":117,"\u662f\u5426\u7f16\u8bd1go\u8bed\u8a00\u7684\u53ef\u5bb9\u9519paramet":1,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":1,"\u662f\u5426\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":1,"\u662f\u5426\u9700\u8981\u7b49\u5f85\u8be5\u8f6e\u6a21\u578b\u53c2\u6570":103,"\u662f\u56e0\u4e3a\u8fd9\u4e2a\u6d41\u7a0b\u6bd4\u5176\u4ed6\u65b9\u6cd5\u90fd\u66f4\u7b80\u4fbf":0,"\u662f\u56e0\u4e3ac99\u652f\u6301":45,"\u662f\u5728paddlepaddle\u4e2d\u6784\u9020\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u6700\u91cd\u8981\u7684\u6982\u5ff5":114,"\u662f\u5b58\u6709\u4e00\u7cfb\u5217\u53d8\u6362\u77e9\u9635\u7684\u6743\u91cd":74,"\u662f\u5b58\u6709\u504f\u7f6e\u5411\u91cf\u7684\u6743\u91cd":74,"\u662f\u5bf9\u7528\u6237\u6587\u4ef6\u5b58\u50a8\u7a7a\u95f4\u7684\u62bd\u8c61":28,"\u662f\u5bfb\u627e\u74f6\u9888\u7684\u5173\u952e\u6307\u6807":107,"\u662f\u5f00\u542favx\u7f16\u8bd1\u7684":2,"\u662f\u5f85\u6269\u5c55\u7684\u6570\u636e":110,"\u662f\u6211\u4eec":72,"\u662f\u6211\u4eec\u8981\u5206\u6790\u7684\u7a0b\u5e8f":107,"\u662f\u6307":46,"\u662f\u6307\u4e00\u7cfb\u5217\u7684\u7279\u5f81\u6570\u636e":111,"\u662f\u6307recurrent_group\u7684\u591a\u4e2a\u8f93\u5165\u5e8f\u5217":111,"\u662f\u6570\u636e\u8f93\u5165":114,"\u662f\u6709\u610f\u4e49\u7684":111,"\u662f\u6784\u5efa\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u6700\u91cd\u8981\u7684\u5de5\u5177":114,"\u662f\u6ca1\u6709\u540d\u5b57\u7684":0,"\u662f\u7528\u6237\u4f7f\u7528c":46,"\u662f\u7684":0,"\u662f\u77e9\u9635":74,"\u662f\u795e\u7ecf\u7f51\u7edc\u5b9a\u4e49\u65f6":89,"\u662f\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u6807\u8bc6\u7b26":74,"\u662f\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26":74,"\u662f\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":74,"\u662f\u7f51\u7edc\u5c42\u8f93\u51fa\u7684\u5927\u5c0f":74,"\u662f\u8be5\u5c42\u7684\u6807\u8bc6\u7b26":74,"\u662f\u8be5\u5c42\u7684\u7c7b\u540d":74,"\u662f\u8be5\u7f51\u7edc\u5c42\u7684":74,"\u662f\u8f93\u5165":114,"\u662f\u8fd9\u4e00\u7c7b\u7684":82,"\u662f\u8fdb\u884c\u8ba1\u7b97\u7684\u57fa\u672c\u5355\u4f4d":89,"\u662f\u9700\u8981\u4e86\u89e3\u54ea\u4e9b\u6b65\u9aa4\u62d6\u6162\u4e86\u6574\u4f53":108,"\u662fc":46,"\u662fdecoder\u7684\u6570\u636e\u8f93\u5165":113,"\u662fnvidia\u6027\u80fd\u5206\u6790\u5de5\u5177":108,"\u662fpaddlepaddle\u4e2d\u5355\u5c42\u5e8f\u5217\u548c\u53cc\u5c42\u5e8f\u5217\u5b58\u50a8\u793a\u610f\u56fe":89,"\u662fpaddlepaddle\u652f\u6301\u7684\u4e00\u79cd\u4efb\u610f\u590d\u6742\u7684rnn\u5355\u5143":113,"\u662fpython\u5c01\u88c5\u7684\u7c7b\u540d":74,"\u662frnn\u72b6\u6001":114,"\u663e\u5f97\u76f8\u5bf9\u6765\u8bf4\u8f83\u4e3a\u8017\u65f6":41,"\u663e\u7136":107,"\u6682\u4e0d\u8003\u8651\u5728\u5185":81,"\u6682\u65e0":4,"\u6682\u65f6\u4e0d\u652f\u6301python3":4,"\u6682\u65f6\u4e0d\u8003\u8651\u591a\u4e2aparamet":11,"\u66b4\u9732\u8fd9\u4e2a\u6982\u5ff5\u5fc5\u8981\u51fd\u6570":46,"\u66f4\u522b\u63d0\u7b80\u5316\u95ee\u9898\u590d\u73b0\u5e26\u6765\u7684\u597d\u5904\u4e86":0,"\u66f4\u591a\u5173\u4e8edocker\u7684\u5b89\u88c5\u4e0e\u4f7f\u7528":78,"\u66f4\u591a\u7684\u8f6c\u6362\u65b9\u6cd5\u8bf7\u53c2\u8003eigen":76,"\u66f4\u597d\u5730\u5b8c\u6210\u4e00\u4e9b\u590d\u6742\u7684\u8bed\u8a00\u7406\u89e3\u4efb\u52a1":113,"\u66f4\u5feb":114,"\u66f4\u65b0":78,"\u66f4\u65b0\u53ef\u80fd\u5bfc\u81f4\u9700\u8981\u65b0\u7684\u5f00\u53d1\u5de5\u5177":0,"\u66f4\u65b0\u6a21\u5f0f":81,"\u66f4\u65b0\u7684\u6587\u6863\u4ee5pr\u7684\u5f62\u5f0f\u63d0\u4ea4\u5230github\u4e2d":77,"\u66f4\u65b0\u7f51\u7edc\u53c2\u6570\u65f6\u5e94\u7528":81,"\u66f4\u65b9\u4fbf\u7684\u8bbe\u7f6e\u65b9\u5f0f":83,"\u66f4\u8be6\u7ec6\u7684\u5b89\u88c5\u548c\u7f16\u8bd1\u65b9\u6cd5\u53c2\u8003":86,"\u66f4\u8fdb\u4e00\u6b65":113,"\u66f4\u9ad8":114,"\u66ff\u6211\u4eec\u5b8c\u6210\u4e86\u539f\u59cb\u8f93\u5165\u6570\u636e\u7684\u62c6\u5206":113,"\u6700\u4e3b\u8981\u7684\u5de5\u4f5c\u5c31\u662f\u89e3\u6790\u51fa":97,"\u6700\u540e":[2,72,74,91],"\u6700\u540e\u4e00\u4e2a":110,"\u6700\u540e\u4e00\u5c42cost\u4e2d\u8bb0\u5f55\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u6240\u6709\u62d3\u6251\u7ed3\u6784":84,"\u6700\u540e\u518d\u8c03\u7528mutabl":76,"\u6700\u540e\u5220\u9664":62,"\u6700\u540e\u6211\u4eec\u4f7f\u7528\u94fe\u5f0f\u6cd5\u5219\u8ba1\u7b97":74,"\u6700\u540e\u8ba1\u7b97softmax":82,"\u6700\u5c0f\u7684ios\u90e8\u7f72\u7248\u672c":117,"\u6700\u5c11\u663e\u793a\u591a\u5c11\u4e2a\u8282\u70b9":103,"\u6700\u5e38\u89c1\u7684\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662fexcept":45,"\u6700\u65b0\u7684\u4ee3\u7801":72,"\u6700\u65b0\u7684paddlepaddl":[2,78],"\u6700\u7ec8":74,"\u6700\u7ec8\u5b9e\u73b0\u4e00\u4e2a\u5c42\u6b21\u5316\u7684\u590d\u6742rnn":113,"\u6700\u7ec8\u6211\u4eec\u53ef\u4ee5\u8c03\u7528trainer\u7684train\u65b9\u6cd5\u542f\u52a8\u8bad\u7ec3":84,"\u6700\u7ec8\u7684\u8f93\u51fa\u7ed3\u679c":113,"\u6709\u4e00\u4e9b\u5fc5\u987b\u914d\u7f6e\u7684\u53c2\u6570":[116,117,118],"\u6709\u4e24\u79cd\u65b9\u6cd5":1,"\u6709\u4e9b\u5c42\u53ef\u80fd\u9700\u8981\u9ad8\u7cbe\u5ea6\u6765\u4fdd\u8bc1\u68af\u5ea6\u68c0\u67e5\u5355\u6d4b\u6b63\u786e\u6267\u884c":74,"\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":74,"\u6709\u4e9b\u7279\u5f81\u7684\u53d6\u503c\u8fbe\u5230\u6570\u767e\u4e07":81,"\u6709\u4eba\u7528\u865a\u62df\u673a\u6765\u7c7b\u6bd4":0,"\u6709\u4ee5\u4e0b\u5efa\u8bae":[116,117],"\u6709\u5173":111,"\u6709\u5173\u53c2\u6570\u914d\u7f6e\u7684\u8be6\u7ec6\u8bf4\u660e\u89c1":116,"\u6709\u5173\u7ebf\u6027\u56de\u5f52\u7684\u5b9e\u9645\u5e94\u7528":84,"\u6709\u52a9\u4e8e\u5728\u8bad\u7ec3\u65f6\u89c2\u5bdf\u5177\u4f53\u6570\u503c":81,"\u6709\u52a9\u4e8e\u8bca\u65ad\u5206\u5e03\u5f0f\u9519\u8bef":93,"\u6709\u591a\u96be":0,"\u6709\u6548\u63d0\u5347paddlepaddle\u5728\u82f1\u7279\u5c14\u67b6\u6784\u4e0a\u7684\u6027\u80fd":[41,42],"\u6709\u65f6\u5019\u6211\u4eec\u4f1a\u5e0c\u671b\u6e05\u7406\u6389\u5df2\u7ecf\u4e0b\u8f7d\u7684\u7b2c\u4e09\u65b9\u4f9d\u8d56\u4ee5\u53ca\u5df2\u7ecf\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6":0,"\u6709\u65f6\u5019\u6211\u4eec\u53ea\u60f3\u8fd0\u884c\u4e00\u4e2a\u7279\u5b9a\u7684\u5355\u5143\u6d4b\u8bd5":0,"\u6709\u6807\u51c6\u7684":45,"\u6709\u7684\u65f6\u5019":45,"\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":108,"\u6709\u7684\u8bdd\u9700\u8981\u5148\u5378\u8f7d":78,"\u6709\u975e\u5e38\u5927\u7684\u5dee\u522b":107,"\u670d\u52a1\u5668\u4e4b\u95f4\u53ef\u4ee5\u901a\u8fc7\u5c40\u57df\u7f51":101,"\u672a\u6307\u5b9a\u6309\u7167double\u7cbe\u5ea6\u7f16\u8bd1":83,"\u672a\u8bbe\u7f6e":117,"\u672c\u4f8b\u4e2d\u7684\u539f\u59cb\u6570\u636e\u4e00\u5171\u670910\u4e2a\u6837\u672c":111,"\u672c\u5217\u8868\u8bf4\u660epaddlepaddle\u53d1\u7248\u4e4b\u524d\u9700\u8981\u6d4b\u8bd5\u7684\u529f\u80fd\u70b9":62,"\u672c\u5730":[4,78],"\u672c\u5730\u6d4b\u8bd5":102,"\u672c\u5730\u8bad\u7ec3":[89,102],"\u672c\u5730\u8bad\u7ec3\u4e0e\u9884\u6d4b":80,"\u672c\u5730\u8bad\u7ec3\u7684\u5b9e\u9a8c":105,"\u672c\u6559\u7a0b\u4e3b\u8981\u4ecb\u7ecd\u5e26kernel\u7684op\u5982\u4f55\u5199":75,"\u672c\u6559\u7a0b\u5c06\u6307\u5bfc\u4f60\u5982\u4f55\u5728":114,"\u672c\u6587\u4e2d\u6240\u6709\u7684\u4f8b\u5b50":111,"\u672c\u6587\u4e2d\u7684\u4f8b\u5b50\u91cc":0,"\u672c\u6587\u4e2d\u793a\u4f8b\u6240\u4f7f\u7528\u7684\u5355\u5143\u6d4b\u8bd5\u6587\u4ef6\u662f":111,"\u672c\u6587\u4ee5paddlepaddle\u7684\u53cc\u5c42rnn\u5355\u5143\u6d4b\u8bd5\u4e3a\u793a\u4f8b":111,"\u672c\u6587\u5c06\u4ecb\u7ecd\u5728kubernetes\u5bb9\u5668\u7ba1\u7406\u5e73\u53f0\u4e0a\u5feb\u901f\u6784\u5efapaddlepaddle\u5bb9\u5668\u96c6\u7fa4":97,"\u672c\u6587\u6863\u5c06\u4ee5linux":116,"\u672c\u6587\u6863\u63cf\u8ff0paddl":46,"\u672c\u6587\u7684\u5c06\u4ecb\u7ecd\u5728macos\u4e0a":117,"\u672c\u6b21\u8bad\u7ec3\u6587\u4ef6\u6240\u5728\u76ee\u5f55":97,"\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6\u53ef\u4ee5\u5199\u6210":97,"\u672c\u6b21\u8bad\u7ec3\u8981\u6c42\u67093\u4e2apaddlepaddle\u8282\u70b9":97,"\u672c\u793a\u4f8b\u4e2d\u4f7f\u7528\u7684\u539f\u59cb\u6570\u636e\u5982\u4e0b":111,"\u672c\u793a\u4f8b\u610f\u56fe\u4f7f\u7528\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u5b9e\u73b0\u4e24\u4e2a\u5b8c\u5168\u7b49\u4ef7\u7684\u5168\u8fde\u63a5rnn":111,"\u672c\u8282\u5c06\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528paddlepaddle\u5728\u4e0d\u540c\u7684\u96c6\u7fa4\u6846\u67b6\u4e0b\u5b8c\u6210\u5206\u5e03\u5f0f\u8bad\u7ec3":92,"\u673a\u5668\u4e0a\u4ee5\u53ca":118,"\u673a\u5668\u7684\u8bbe\u5907":105,"\u673a\u5668\u7ffb\u8bd1":62,"\u6743\u91cd\u66f4\u65b0\u7684\u68af\u5ea6":103,"\u6765\u4e3a\u4e00\u4e2a":89,"\u6765\u4ee3\u66ff":72,"\u6765\u4f20\u8f93\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6\u4e2d\u5b9a\u4e49\u7684\u7f51\u7edc\u7ed3\u6784\u548c\u76f8\u5173\u53c2\u6570":90,"\u6765\u4f7f\u7528dropout":82,"\u6765\u4f7f\u7528dropout\u7684":82,"\u6765\u4fdd\u8bc1\u8bad\u7ec3\u8fc7\u7a0b\u53ef\u4ee5\u4ece\u4e2d\u95f4\u72b6\u6001\u91cd\u65b0\u542f\u52a8":11,"\u6765\u505a\u68af\u5ea6\u68c0\u67e5":74,"\u6765\u51b3\u5b9a\u662f\u5426\u5f00\u542fmkl":41,"\u6765\u5206\u6790\u6267\u884c\u6587\u4ef6":108,"\u6765\u521d\u59cb\u5316\u53c2\u6570":83,"\u6765\u542f\u52a8\u548c":0,"\u6765\u5b58\u50a8":89,"\u6765\u5b58\u50a8\u6570\u636e":[89,90],"\u6765\u5b8c\u6210\u524d\u5411\u548c\u53cd\u5411\u8ba1\u7b97":90,"\u6765\u5b8c\u6210\u7f51\u7edc\u7684\u8bad\u7ec3":84,"\u6765\u5b9a\u4e49\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":114,"\u6765\u5b9e\u73b0":42,"\u6765\u5b9e\u9645\u5b58\u50a8\u6570\u636e":[89,90],"\u6765\u5bf9\u6bd4\u5206\u6790\u4e24\u8005\u8bed\u4e49\u76f8\u540c\u7684\u539f\u56e0":111,"\u6765\u5f71\u54cdpaddlepaddle\u7684\u7f16\u8bd1\u8fc7\u7a0b":[116,117],"\u6765\u5f97\u5230\u67d0\u4e2a\u7279\u5b9a\u53c2\u6570\u7684\u68af\u5ea6\u77e9\u9635":74,"\u6765\u6267\u884c":0,"\u6765\u63cf\u8ff0\u7684":76,"\u6765\u63cf\u8ff0\u8be5op\u7684\u8f93\u5165":75,"\u6765\u63cf\u8ff0\u8f93\u5165":89,"\u6765\u642d\u5efa\u795e\u7ecf\u7f51\u7edc":84,"\u6765\u663e\u793a\u6027\u80fd\u5206\u6790\u7ed3\u679c":107,"\u6765\u67e5\u770b\u6027\u80fd\u5206\u6790\u7ed3\u679c":107,"\u6765\u6ce8\u518c\u8be5\u5c42":74,"\u6765\u6df7\u5408\u4f7f\u7528gpu\u548ccpu\u8ba1\u7b97\u7f51\u7edc\u5c42\u7684\u53c2\u6570":105,"\u6765\u6e05\u7406\u8fd9\u4e9b\u5185\u5bb9":0,"\u6765\u7279\u6307":89,"\u6765\u7279\u6307\u8c03\u7528paddlepaddl":90,"\u6765\u7279\u6307paddlepaddl":90,"\u6765\u7279\u6307paddlepaddle\u4e2d\u7684\u4e00\u7ef4\u6574\u578b\u6570\u7ec4":89,"\u6765\u7279\u6307paddlepaddle\u4e2d\u7684\u4e8c\u7ef4\u6d6e\u70b9\u578b\u77e9\u9635":89,"\u6765\u7279\u6307paddlepaddle\u4e2d\u795e\u7ecf\u7f51\u7edc\u8ba1\u7b97\u5c42\u4e00\u4e2a\u8f93\u5165":89,"\u6765\u786e\u4fdd\u628a":45,"\u6765\u786e\u5b9a\u7a00\u758f\u77e9\u9635\u7684\u5185\u5bb9":89,"\u6765\u7f16\u8bd1":0,"\u6765\u83b7\u5f97\u8f93\u51fa\u7684\u68af\u5ea6":74,"\u6765\u8868\u793a":114,"\u6765\u8868\u793a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":114,"\u6765\u8868\u793apaddle\u5185\u90e8\u7c7b":45,"\u6765\u89e3\u51b3\u4e0a\u9762\u7684\u95ee\u9898":81,"\u6765\u8ba1\u7b97\u68af\u5ea6":74,"\u6765\u8bb0\u5f55\u8f93\u5165":89,"\u6765\u8bb2\u89e3\u5982\u4f55\u4f7f\u7528\u53cc\u5c42rnn":111,"\u6765\u8bbe\u7f6e":83,"\u6765\u8bbf\u95ee\u7528\u6237\u81ea\u5df1\u7684\u6570\u636e":12,"\u6765\u8bfb\u53d6\u4e00\u4e2a":89,"\u6765\u8c03\u6574":72,"\u6765\u8c03\u7528":0,"\u6765\u8fd0\u884c\u5305\u62ec":0,"\u6765\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5\u4e86":0,"\u6765\u8fd0\u884c\u6027\u80fd\u5206\u6790\u548c\u8c03\u4f18":108,"\u6765\u8fd0\u884c\u955c\u50cf":2,"\u6765\u8fdb\u884c\u8ba8\u8bba":46,"\u6765\u9884\u6d4b\u8fd9\u4e2a\u4e2d\u95f4\u7684\u8bcd":81,"\u6784\u5efa":116,"\u6784\u5efa\u597d\u5f00\u53d1\u955c\u50cf\u540e":116,"\u6784\u5efa\u76ee\u6807\u4e3a":117,"\u6784\u6210\u4e00\u4e2a\u5e8f\u5217":89,"\u6784\u6210\u4e86\u8f93\u51fa\u53cc\u5c42\u5e8f\u5217\u7684\u7b2ci\u4e2a":110,"\u6784\u9020":97,"\u6784\u9020\u51fd\u6570\u542b\u67092\u4e2a\u53c2\u6570":75,"\u6784\u9020\u51fd\u6570\u91cc\u901a\u8fc7":75,"\u67b6\u6784\u7684\u6a21\u62df\u5668\u5e73\u53f0":117,"\u67b6\u6784\u7684iphone\u6216\u8005ipad\u7b49\u7269\u7406\u8bbe\u5907":117,"\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":111,"\u67d0\u4e9b\u53c2\u6570\u53ea\u53ef\u7528\u4e8e\u7279\u5b9a\u7684\u5c42\u4e2d":102,"\u67e5\u770b":72,"\u67e5\u770b\u5305\u7684\u5927\u5c0f":78,"\u67e5\u770b\u5f53\u524d\u72b6\u6001":72,"\u67e5\u770b\u5f53\u524d\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":72,"\u67e5\u770b\u6587\u4ef6\u5177\u4f53\u88ab\u4fee\u6539\u7684\u5185\u5bb9":72,"\u67e5\u770b\u662f\u5426\u662f\u5176\u4ed6\u9519\u8bef\u5f15\u53d1\u7684\u62a5\u9519":79,"\u67e5\u770bjob\u7684\u8be6\u7ec6\u60c5\u51b5":96,"\u67e5\u770blatest":62,"\u6807\u51c6":4,"\u6807\u51c6\u5dee\u4e3a":83,"\u6807\u51c6\u8868\u793apaddlepaddle\u7248\u672c\u53f7":62,"\u6807\u8bc6\u4e86\u4e00\u4e2a\u8f93\u51fa\u7684\u6587\u4ef6\u540d":107,"\u6807\u8bc6\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c\u6587\u4ef6":107,"\u6807\u8bc6\u662f\u5426\u4e3a\u8fde\u7eed\u7684batch\u8ba1\u7b97":103,"\u6807\u8bc6\u88ab\u6027\u80fd\u5206\u6790\u7684\u6e90\u6587\u4ef6":107,"\u6807\u8bc6http\u670d\u52a1\u7684\u7aef\u53e3":107,"\u6807\u8bc6http\u670d\u52a1\u7ed1\u5b9a\u7684ip":107,"\u6838\u4e00\u6837\u591a\u7684\u8fdb\u7a0b\u6765\u5e76\u884c\u7f16\u8bd1":0,"\u6838\u5fc3\u4ee3\u7801\u7f16\u8bd1\u6210\u94fe\u63a5\u5e93":87,"\u6839\u636e\u4e2a\u4eba\u7684\u9700\u6c42\u4fee\u6539\u5b9a\u5236docker\u5bb9\u5668\u6240\u6267\u884c\u7684\u811a\u672c":116,"\u6839\u636e\u4f60\u7684\u4efb\u52a1":105,"\u6839\u636e\u524d\u6587\u7684\u63cf\u8ff0":97,"\u6839\u636e\u7f51\u7edc\u914d\u7f6e\u4e2d\u7684":103,"\u6839\u636e\u8f93\u5165tensor\u7684\u5927\u5c0f\u6765\u8bbe\u7f6e\u8f93\u51fatensor\u7684\u5927\u5c0f":76,"\u6839\u636e\u8fd9\u4e9b\u53c2\u6570\u7684\u4f7f\u7528\u573a\u5408":102,"\u6839\u636e\u9ed8\u8ba4\u503c\u9012\u589e":103,"\u6839\u636e\u9ed8\u8ba4\u7aef\u53e3\u53f7\u9012\u589e":103,"\u6839\u636ejob\u5bf9\u5e94\u7684pod\u4fe1\u606f":96,"\u6839\u636eport":91,"\u683c\u5f0f":103,"\u683c\u5f0f\u4e0d\u5339\u914d\u65f6":42,"\u683c\u5f0f\u5b58\u50a8":89,"\u683c\u5f0f\u7684\u6587\u4ef6\u6765\u5b58\u653e":75,"\u6846\u67b6\u63d0\u4f9b\u7684blas\u51fd\u6570\u8fdb\u884c\u77e9\u9635\u8ba1\u7b97":117,"\u6846\u67b6\u8fdb\u884cblas\u77e9\u9635\u8ba1\u7b97":117,"\u68af\u5ea6\u4f1a\u5c31\u5730":74,"\u68af\u5ea6\u4f1a\u6709\u566a\u58f0":92,"\u68af\u5ea6\u53c2\u6570\u7684\u5206\u5757\u6570\u76ee":103,"\u68af\u5ea6\u5c31\u53ef\u4ee5\u901a\u8fc7\u8fd9\u4e2a\u65b9\u7a0b\u8ba1\u7b97\u5f97\u5230":74,"\u68af\u5ea6\u670d\u52a1\u5668\u7684\u6570\u91cf":103,"\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5\u901a\u8fc7\u6709\u9650\u5dee\u5206\u6cd5\u6765\u9a8c\u8bc1\u4e00\u4e2a\u5c42\u7684\u68af\u5ea6":74,"\u68af\u5ea6\u68c0\u67e5\u7684\u8f93\u5165\u6570\u636e\u7684\u6279\u6b21\u5927\u5c0f":74,"\u68c0\u67e5\u70b9\u4fdd\u5b58\u7a0b\u5e8f\u6d41\u7a0b":11,"\u68c0\u67e5\u8f93\u5165\u6570\u636e\u7ef4\u5ea6":75,"\u6982\u5ff5\u4e0a":90,"\u6982\u5ff5\u4e0a\u53ef\u4ee5\u5c06":89,"\u6a21\u5757\u4e0b\u7684\u76f8\u5173":76,"\u6a21\u578b\u4e00\u76f4\u4e0d\u6536\u655b":81,"\u6a21\u578b\u4e2d\u6240\u6709\u53ef\u5b66\u4e60\u53c2\u6570\u4f1a\u88ab\u5b58\u4e3a\u4e00\u4e2a\u538b\u7f29\u6587\u4ef6":90,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\u901a\u8fc7\u5b9a\u671f\u5411\u78c1\u76d8\u4e0a\u4fdd\u5b58\u4e00\u4efd\u5b58\u50a8\u5728paramet":11,"\u6a21\u578b\u6570\u636e\u68c0\u67e5\u70b9\u7684\u5b9e\u73b0":11,"\u6a21\u578b\u6587\u4ef6\u5c06\u88ab\u5199\u5165\u8282\u70b9":93,"\u6a21\u578b\u6765\u6307\u5bfc\u4f60\u5b8c\u6210\u8fd9\u4e9b\u6b65\u9aa4":114,"\u6a21\u578b\u6f14\u793a\u5982\u4f55\u914d\u7f6e\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u6a21\u578b":114,"\u6a21\u578b\u7684\u4ee3\u7801\u53ef\u4ee5\u5728":114,"\u6a21\u578b\u7684\u7f16\u7801\u5668\u90e8\u5206\u5982\u4e0b\u6240\u793a":114,"\u6a21\u578b\u8bad\u7ec3\u7b49\u4efb\u52a1":84,"\u6a21\u578b\u914d\u7f6e":80,"\u6a21\u578b\u914d\u7f6e\u89e3\u6790":45,"\u6a21\u5f0f\u4e0b\u7684\u6027\u80fd\u6d4b\u8bd5\u662f\u6ca1\u6709\u610f\u4e49\u7684":107,"\u6a2a\u5411\u62fc\u63a5":81,"\u6b21\u8fed\u4ee3\u6267\u884c\u7684\u8f6c\u6362\u6b21\u6570\u4e3a":41,"\u6b22\u8fce\u901a\u8fc7":72,"\u6b63\u5728\u7b49\u5f85\u672a\u5b8c\u6210\u7684\u4efb\u52a1":78,"\u6b63\u5e38\u60c5\u51b5\u4e0b\u662f75m":78,"\u6b63\u786e\u7684\u89e3\u51b3\u65b9\u6cd5\u662f":78,"\u6b63\u8d1f\u5bf9\u9a8c\u8bc1":102,"\u6b64\u547d\u4ee4\u5c06\u5728":116,"\u6b64\u5904\u90fd\u4e3a2":111,"\u6b64\u5916":[0,72,82,116],"\u6b64\u6559\u7a0b\u4f1a\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528python\u7684cprofile\u5305":107,"\u6b64\u6559\u7a0b\u5c06\u5411\u60a8\u5206\u6b65\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528\u5185\u7f6e\u7684\u5b9a\u65f6\u5de5\u5177":108,"\u6b64\u65b9\u6cd5\u4e0d\u80fd\u83b7\u53d6":81,"\u6b64\u65f6\u53ea\u9700\u8981":0,"\u6b64\u65f6\u53ef\u4ee5\u5728\u8c03\u7528infer\u63a5\u53e3\u65f6\u901a\u8fc7\u8bbe\u7f6e":81,"\u6b64\u65f6\u53ef\u4ee5\u8df3\u8fc7paddlepaddle\u6a21\u578b\u53c2\u6570\u6587\u4ef6\u7684\u5934\u4fe1\u606f":83,"\u6b64\u65f6\u6bcf\u4e2a\u5c0f\u5206\u652f\u7684":42,"\u6b64\u65f6master\u5c06\u8d1f\u8d23\u542f\u52a8\u4e00\u4e2a\u65b0\u7684train":11,"\u6b64\u76ee\u5f55":90,"\u6b64\u793a\u4f8b":90,"\u6b64\u7c7b\u62a5\u9519\u901a\u5e38\u662f\u7531\u4e8e\u67d0\u4e00\u4e2a\u8282\u70b9\u7684\u9519\u8bef\u5bfc\u81f4\u8fd9\u4e2a\u8282\u70b9\u7684\u8bad\u7ec3\u8fdb\u7a0b\u9000\u51fa":79,"\u6b65\u9aa4":81,"\u6bb5\u843d\u53ef\u4ee5\u770b\u4f5c\u662f\u4e00\u4e2a\u5d4c\u5957\u7684\u53cc\u5c42\u7684\u5e8f\u5217":113,"\u6bb5\u843d\u662f\u7531\u53e5\u5b50\u6784\u6210\u7684\u5e8f\u5217":89,"\u6bcf\u4e00\u4e2a":[62,90],"\u6bcf\u4e00\u4e2a\u5916\u5c42\u5e8f\u5217\u53c8\u542b\u6709\u82e5\u5e72\u4e2a\u5185\u5c42\u5e8f\u5217":89,"\u6bcf\u4e00\u4e2a\u5e8f\u5217\u5728\u6574\u4e2a":89,"\u6bcf\u4e00\u4e2a\u6587\u4ef6\u662f\u6570\u636e\u96c6\u7684\u4e00\u4e2ashard":12,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65":111,"\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4e4b\u95f4\u7684\u795e\u7ecf\u7f51\u7edc\u5177\u6709\u4e00\u5b9a\u7684\u76f8\u5173\u6027":111,"\u6bcf\u4e00\u4e2a\u8282\u70b9\u90fd\u6709\u76f8\u540c\u7684\u65e5\u5fd7\u7ed3\u6784":93,"\u6bcf\u4e00\u4e2a\u8f93\u5165":[89,90],"\u6bcf\u4e00\u4e2alayer\u8f93\u51fa\u77e9\u9635\u7684\u9ad8\u5ea6":81,"\u6bcf\u4e00\u5217\u7684\u542b\u4e49\u662f":107,"\u6bcf\u4e00\u7ec4\u5185\u7684\u6240\u6709\u53e5\u5b50\u548clabel":111,"\u6bcf\u4e00\u884c\u5143\u7d20\u5728":89,"\u6bcf\u4e2a":42,"\u6bcf\u4e2a\u503c\u7684\u7c7b\u578b\u53ef\u4ee5\u662f\u6574\u5f62":12,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u6bcf\u4e2a\u5355\u5c42rnn":113,"\u6bcf\u4e2a\u53c2\u6570\u670d\u52a1\u5668\u53ea\u4fdd\u5b58\u6574\u4e2a\u795e\u7ecf\u7f51\u7edc\u6240\u6709\u53c2\u6570\u7684\u4e00\u90e8\u5206":92,"\u6bcf\u4e2a\u53e5\u5b50\u53c8\u662f\u5355\u8bcd\u7684\u6570\u7ec4":111,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u5f00\u59cb\u6807\u8bb0\u5f00\u5934":114,"\u6bcf\u4e2a\u53e5\u5b50\u90fd\u4ee5\u7ed3\u675f\u6807\u8bb0\u7ed3\u5c3e":114,"\u6bcf\u4e2a\u5b50\u5e8f\u5217\u957f\u5ea6\u53ef\u4ee5\u4e0d\u4e00\u81f4":111,"\u6bcf\u4e2a\u5c42\u5728\u5176":74,"\u6bcf\u4e2a\u6279\u6b21\u6570\u636e":103,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e4b\u5185\u7684\u8fd0\u7b97\u662f\u72ec\u7acb\u7684":113,"\u6bcf\u4e2a\u65f6\u95f4\u6b65\u90fd\u7528\u4e86\u4e0a\u4e00\u4e2a\u65f6\u95f4\u6b65\u7684\u8f93\u51fa\u7ed3\u679c":111,"\u6bcf\u4e2a\u6743\u91cd\u5bf9\u5e94\u4e00\u4e2a\u8f93\u5165":74,"\u6bcf\u4e2a\u6837\u672c\u7531\u4e24\u90e8\u5206\u7ec4\u6210":111,"\u6bcf\u4e2a\u6837\u672c\u95f4\u7528\u7a7a\u884c\u5206\u5f00":111,"\u6bcf\u4e2a\u6d4b\u8bd5\u4f1a\u5bf9\u6bd4paddlepaddle\u4e2dcpu\u7b97\u51fa\u7684\u7ed3\u679c\u4e0emkl":42,"\u6bcf\u4e2a\u72b6\u6001":113,"\u6bcf\u4e2a\u7ebf\u7a0b":103,"\u6bcf\u4e2a\u7ebf\u7a0b\u5206\u914d\u5230128\u4e2a\u6837\u672c\u7528\u4e8e\u8bad\u7ec3":103,"\u6bcf\u4e2a\u8bad\u7ec3\u8282\u70b9\u5fc5\u987b\u6307\u5b9a\u4e00\u4e2a\u552f\u4e00\u7684id\u53f7":103,"\u6bcf\u4e2a\u8f93\u5165\u90fd\u662f\u4e00\u4e2a":74,"\u6bcf\u4e2a\u8f93\u51fa\u8282\u70b9\u90fd\u8fde\u63a5\u5230\u6240\u6709\u7684\u8f93\u5165\u8282\u70b9\u4e0a":74,"\u6bcf\u4e2a\u90e8\u5206\u5206\u522b\u7ed9\u6bcf\u4e2atrainer\u4f7f\u7528":92,"\u6bcf\u4e2acommit\u53ea\u505a\u4e86\u5c11\u91cf\u7684\u4fee\u6539":72,"\u6bcf\u4e2adata":12,"\u6bcf\u4e2amkldnnlayer\u90fd\u5305\u542b\u7528\u4e8e\u5185\u90e8\u5b58\u50a8\u548c\u5916\u90e8\u5b58\u50a8\u7684\u4e00\u7cfb\u5217mkldnnmatrix":42,"\u6bcf\u4e2aparamet":11,"\u6bcf\u4e2apod\u5305\u542b\u4e00\u4e2apaddlepaddle\u5bb9\u5668":97,"\u6bcf\u4e2ashard\u5206\u522b\u5b58\u50a8\u5728\u5176\u4e2d\u4e00\u53f0paramet":11,"\u6bcf\u4e2atrainer\u542f\u52a8\u540e\u8bfb\u53d6\u5207\u5206\u597d\u7684\u4e00\u90e8\u5206\u6570\u636e":92,"\u6bcf\u4e2atrainer\u7684\u552f\u4e00id":91,"\u6bcf\u4e2atrainer\u8fdb\u7a0b\u9700\u8981\u80fd\u591f\u8bfb\u53d6\u5c5e\u4e8e\u81ea\u5df1\u7684\u4e00\u4efd\u6570\u636e":91,"\u6bcf\u53f0\u670d\u52a1\u5668\u5177\u6709\u96c6\u7fa4\u4e2d\u552f\u4e00\u7684ip\u5730\u5740":101,"\u6bcf\u5c42\u4e0a\u53ea\u80fd\u4fdd\u5b58\u56fa\u5b9a\u6570\u76ee\u4e2a\u6700\u597d\u7684\u72b6\u6001":103,"\u6bcf\u5c42\u4f7f\u7528\u7684gpu\u53f7\u4f9d\u8d56\u4e8e\u53c2\u6570train":105,"\u6bcf\u6279\u6b21":103,"\u6bcf\u6b21\u63d0\u4ea4\u4ee3\u7801":72,"\u6bcf\u6b21\u63d0\u4ea4\u65f6":72,"\u6bcf\u6b21\u8c03\u7528\u65f6\u5bf9\u539f\u6570\u636e\u7684\u91cd\u590dpacking\u4fbf\u6210\u4e3a\u4e86\u5197\u4f59":41,"\u6bcf\u6b21\u8c03\u7528\u7684\u8017\u65f6\u4e5f\u5f88\u957f":107,"\u6bcf\u6b21\u8f93\u51fa\u4e00\u4e2adata":12,"\u6bcf\u884c\u8868\u793a\u4e00\u4e2a\u6279\u6b21\u4e2d\u7684\u5355\u4e2a\u8f93\u5165":74,"\u6bcf\u8f6e\u4f1a\u5c06\u6570\u636e\u96c6\u4e2d\u7684\u6240\u6709\u8bad\u7ec3\u6837\u672c\u4f7f\u7528\u4e00\u6b21":103,"\u6bcf\u8f6e\u7ed3\u675f\u65f6\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":103,"\u6bcf\u8f6e\u90fd\u4f1a\u4fdd\u5b58\u9884\u6d4b\u7ed3\u679c":103,"\u6bcf\u8fd0\u884c\u591a\u5c11\u4e2a\u6279\u6b21\u6267\u884c\u4e00\u6b21\u7a00\u758f\u53c2\u6570\u5206\u5e03\u7684\u68c0\u67e5":103,"\u6bcf\u969410\u5206\u949f":11,"\u6bcfdot":103,"\u6bcflog":103,"\u6bcfsave":103,"\u6bcftest":103,"\u6bd4\u5982":[0,2,12,42,72,79,81],"\u6bd4\u5982\u4e00\u53e5\u8bdd\u4e2d\u7684\u6bcf\u4e00\u4e2a\u5355\u8bcd":111,"\u6bd4\u5982\u53ef\u80fd\u4f1a\u7528openmp\u6539\u8fdbsgd\u7684\u66f4\u65b0\u6027\u80fd":42,"\u6bd4\u5982\u5728":2,"\u6bd4\u5982\u5982\u679c\u8981build\u4e00\u4e2a\u4e0d\u4f9d\u8d56gpu":72,"\u6bd4\u5982\u5c06":62,"\u6bd4\u5982\u5e0c\u671b\u6700\u5c0f\u5316\u751f\u6210\u5e93\u7684\u5927\u5c0f":117,"\u6bd4\u5982\u5e0c\u671b\u6700\u5c0f\u5316\u751f\u6210\u7684\u5e93\u7684\u5927\u5c0f":[116,118],"\u6bd4\u5982\u6bcf\u969410\u5206\u949f\u6700\u65b0\u7684\u5feb\u7167":11,"\u6bd4\u5982\u6d41\u5f0f\u6570\u636e\u5904\u7406":12,"\u6bd4\u5982\u8282\u70b9\u7684id":94,"\u6bd4\u5982\u8bbe\u7f6e\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u53c2\u6570\u521d\u59cb\u5316\u65b9\u5f0f\u548cbias\u521d\u59cb\u5316\u65b9\u5f0f":83,"\u6bd4\u5982cento":4,"\u6bd4\u5982fpe":79,"\u6bd4\u5982ide\u914d\u7f6e\u91cc":72,"\u6bd4\u5982imagenet\u8fd9\u4e2a\u6570\u636e\u96c6\u53ef\u80fd\u88ab\u5206\u62101000\u4e2ashard":12,"\u6bd4\u5982pil\u5e93\u7b49":91,"\u6bd5\u7adf\u5355\u7ebf\u7a0b\u8c03\u8bd5\u66f4\u5bb9\u6613":107,"\u6ca1\u6709\u5fc5\u8981\u5728\u6bcf\u6b21\u524d\u5411\u4e2d\u6bcf\u4e2a\u65f6\u95f4\u6b65\u7684\u8ba1\u7b97\u65f6\u5bf9\u6743\u91cd\u8fdb\u884c\u91cd\u590d\u7684packing\u64cd\u4f5c":41,"\u6ca1\u6709\u627e\u5230\u548c\u5f53\u524d\u7cfb\u7edf\u5339\u914d\u7684paddlepaddle\u5b89\u88c5\u5305":[4,78],"\u6ca1\u6709\u8bbe\u7f6e":[116,118],"\u6ce8":[2,11,62,89],"\u6ce8\u518c":75,"\u6ce8\u518ccpu":75,"\u6ce8\u518clayer\u7684\u65f6\u5019\u4fdd\u8bc1":[41,42],"\u6ce8\u518cop":75,"\u6ce8\u518cop\u65f6\u7684\u7c7b\u578b\u540d":75,"\u6ce8\u610f":[1,42,74,77,81,84,90,91,97,107,114,116,117,118],"\u6ce8\u610f\u4e0a\u8ff0\u547d\u4ee4\u4e2d":97,"\u6ce8\u610f\u4e8b\u9879":89,"\u6ce8\u610f\u5230\u6211\u4eec\u5df2\u7ecf\u5047\u8bbe\u673a\u5668\u4e0a\u67094\u4e2agpu":105,"\u6ce8\u610fnode":97,"\u6ce8\u91ca":75,"\u6d41\u7a0b\u6765\u63d0\u4ea4\u4ee3\u7801":72,"\u6d4b\u8bd5":72,"\u6d4b\u8bd5\u5206\u4e3a\u6bcf\u4e2alayer":42,"\u6d4b\u8bd5\u65f6\u6307\u5b9a\u7684\u5b58\u50a8\u6a21\u578b\u5217\u8868\u7684\u6587\u4ef6":103,"\u6d4b\u8bd5\u662f":72,"\u6d4b\u8bd5\u672c\u6b21release\u7684\u6b63\u786e\u6027":62,"\u6d4b\u8bd5\u7684\u6027\u80fd\u5bf9\u6bd4\u7ed3\u679c\u4f1a\u5728":42,"\u6d4b\u8bd5\u7684\u6a21\u578b\u5305\u62ec\u4ece\u7b2cm\u8f6e\u5230\u7b2cn":105,"\u6d4b\u8bd5model_list":102,"\u6d4b\u8bd5oper":75,"\u6d4b\u8bd5save_dir":102,"\u6d6e\u70b9\u578b\u6570\u636e":12,"\u6d6e\u70b9\u578b\u7a00\u758f\u77e9\u9635":89,"\u6d6e\u70b9\u578b\u7a20\u5bc6\u77e9\u9635":89,"\u6d6e\u70b9\u5f02\u5e38\u901a\u5e38\u7684\u539f\u56e0\u662f\u6d6e\u70b9\u6570\u6ea2\u51fa":81,"\u6d6e\u70b9\u6570":89,"\u6d6e\u70b9\u6570\u5411\u91cf\u7b49":89,"\u6d6e\u70b9\u7a00\u758f\u6570\u636e":74,"\u6df1\u5165paddlepaddl":42,"\u6df7\u5408\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790\u6765\u8fdb\u884c\u8c03\u4f18":107,"\u6df7\u5408\u4ee3\u7801\u7684\u6027\u80fd\u74f6\u9888\u4e5f\u662f\u8981\u770b":107,"\u6df7\u5408\u5f53\u524d\u8bcd\u5411\u91cf\u548cattention\u52a0\u6743\u7f16\u7801\u5411\u91cf":114,"\u6dfb\u52a0":41,"\u6dfb\u52a0\u4e86\u4e00\u4e2a\u8f93\u51fa":75,"\u6dfb\u52a0\u542f\u52a8\u811a\u672c":97,"\u6dfb\u52a0\u5e8f\u5217\u4fe1\u606f":89,"\u6dfb\u52a0\u7684\u76f8\u5173\u6587\u4ef6\u548c\u76ee\u5f55\u7ed3\u6784\u5982\u4e0b":[41,42],"\u6dfb\u52a0\u8f93\u5165\u53c2\u6570":75,"\u6dfb\u52a0\u8f93\u51fa\u53c2\u6570":75,"\u6dfb\u52a0op\u7684\u6ce8\u91ca":75,"\u6e05\u7406":0,"\u6e05\u7406\u548c\u7ed3\u675f":90,"\u6e05\u7406\u6389\u8001\u65e7\u7684paddlepaddle\u5b89\u88c5\u5305":78,"\u6e90\u4ee3\u7801\u683c\u5f0f":72,"\u6e90\u5e8f\u5217":114,"\u6e90\u7801\u6811\u6839\u76ee\u5f55":0,"\u6f5c\u5728\u4f1a\u5f15\u8d77\u672a\u5b9a\u4e49\u884c\u4e3a":89,"\u6fc0\u6d3b":74,"\u6fc0\u6d3b\u51fd\u6570\u662f\u72ec\u7acb\u4e8e":42,"\u6fc0\u6d3b\u65b9\u7a0b":74,"\u6fc0\u6d3b\u7684\u7c7b\u578b":74,"\u70b9\u51fb":[4,62],"\u70b9\u51fb\u8fd9\u91cc":77,"\u7136\u540e":[93,108],"\u7136\u540e\u4e0b\u8f7d\u4f18\u5316\u66f4\u65b0\u540e\u7684\u795e\u7ecf\u7f51\u7edc\u53c2\u6570":92,"\u7136\u540e\u4ea4\u7ed9step\u51fd\u6570":113,"\u7136\u540e\u4f7f\u7528":117,"\u7136\u540e\u4f7f\u7528resize\u63a5\u53e3\u8bbe\u7f6etensor\u7684\u5927\u5c0f":76,"\u7136\u540e\u5355\u51fb":72,"\u7136\u540e\u53ef\u4ee5\u4ecehead\u8282\u70b9ssh\u65e0\u5bc6\u7801\u767b\u5f55\u5230openmpi\u7684\u6bcf\u4e2a\u8282\u70b9\u4e0a":98,"\u7136\u540e\u53ef\u4ee5\u4f7f\u7528\u547d\u4ee4\u884c\u5de5\u5177\u521b\u5efajob":97,"\u7136\u540e\u5728\u4e0b\u4e00\u4e2a\u65f6\u95f4\u6b65\u8f93\u5165\u7ed9\u53e6\u4e00\u4e2a\u795e\u7ecf\u5143":111,"\u7136\u540e\u5728\u524d\u5411":41,"\u7136\u540e\u5728\u6d4f\u89c8\u5668\u4e2d\u8f93\u5165\u4ee5\u4e0b\u7f51\u5740":2,"\u7136\u540e\u5728dataprovider\u91cc\u9762\u6839\u636e\u8be5\u5730\u5740\u52a0\u8f7d\u5b57\u5178":83,"\u7136\u540e\u5728etcd\u7684":11,"\u7136\u540e\u5b89\u88c5paddle\u7684python\u73af\u5883":78,"\u7136\u540e\u5b9a\u4e49":114,"\u7136\u540e\u5c06\u6784\u5efa\u6210\u529f\u7684\u955c\u50cf\u4e0a\u4f20\u5230\u955c\u50cf\u4ed3\u5e93":97,"\u7136\u540e\u5c06\u8fd9\u4e9blayer\u7684\u53c2\u6570":82,"\u7136\u540e\u5c31\u53ef\u4ee5\u5e76\u53d1\u5199\u5165\u591a\u4e2achunk":28,"\u7136\u540e\u6240\u6709\u7528":72,"\u7136\u540e\u624d\u80fd\u4f7f\u7528pfsclient":28,"\u7136\u540e\u6253\u5370\u8f93\u51fa":84,"\u7136\u540e\u6309\u7167\u4e0a\u8ff0\u7684\u65b9\u6cd5":62,"\u7136\u540e\u63d0\u4ea4\u65b0\u6dfb\u52a0\u7684":72,"\u7136\u540e\u70b9\u51fb":[62,72],"\u7136\u540e\u7533\u660e\u4e00\u4e2a\u5b58\u50a8\u5377":97,"\u7136\u540e\u89c2\u5bdf\u5230\u8f93\u51fa\u7684\u53d8\u5316\u4e3a":74,"\u7136\u540e\u901a\u8fc7\u51fd\u6570":97,"\u7136\u540e\u901a\u8fc7\u81ea\u8eab\u7684ip\u5730\u5740\u5728":97,"\u7136\u540e\u91cd\u65b0cmake\u5373\u53ef":78,"\u7136\u800c":[103,114],"\u7248\u672c":[0,1,4,117],"\u7248\u672c\u4e3acpu_avx_mkl":2,"\u7248\u672c\u4e3acpu_avx_openbla":[4,86],"\u7248\u672c\u5206\u652f":62,"\u7248\u672c\u53f7":62,"\u7248\u672c\u53f7\u5bf9\u5e94\u7684tag\u5373\u53ef":62,"\u7248\u672c\u53f7rc":62,"\u7248\u672c\u5728":72,"\u7248\u672c\u8bf4\u660e":4,"\u7248\u672cfork\u51fa\u81ea\u5df1\u7684\u529f\u80fd\u5206\u652f":62,"\u7279\u522b\u662f\u5728lstm\u7b49rnn\u4e2d":81,"\u7279\u6307":90,"\u7279\u6709\u7684\u8bbe\u5907id":42,"\u72ec\u7acb\u5b9a\u5236\u7684\u4e8c\u8fdb\u5236\u65f6\u624d\u9700\u8981\u7f16\u8bd1":3,"\u72ec\u7acb\u5de5\u5177\u94fe":116,"\u72ec\u7acb\u5de5\u5177\u94fe\u6240\u5728\u7684\u7edd\u5bf9\u8def\u5f84":116,"\u73af\u5883\u51c6\u5907":92,"\u73af\u5883\u53d8\u91cf":91,"\u73af\u5883\u53d8\u91cf\u4e2d":87,"\u73af\u5883\u53d8\u91cf\u6765\u6307\u5b9a\u7279\u5b9a\u7684gpu":81,"\u73b0\u9636\u6bb5\u7684\u4f18\u5316\u4e3b\u8981\u9488\u5bf9":41,"\u73b0\u9636\u6bb5paddle\u6709\u4e00\u4e2a\u95ee\u9898\u662f":45,"\u7406\u89e3":0,"\u7406\u89e3\u4e3a\u4e00\u4e2a\u4e00\u7ef4\u7684\u6574\u578b\u6570\u7ec4":89,"\u751a\u81f3\u80fd\u89e3\u91ca\u4e3a\u4ec0\u4e48\u67d0\u4e2a\u64cd\u4f5c\u82b1\u4e86\u5f88\u957f\u65f6\u95f4":108,"\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u8bad\u7ec3\u6570\u636e\u96c6\u901a\u5e38\u4f53\u79ef\u5f88\u5927":12,"\u751f\u4ea7\u73af\u5883\u7684\u65e5\u5fd7\u6570\u636e\u4f1a\u901a\u8fc7\u5b9e\u65f6\u6d41\u7684\u65b9\u5f0f":12,"\u751f\u4ea7\u955c\u50cf":72,"\u751f\u6210":97,"\u751f\u6210\u5404\u79cd\u8bed\u8a00\u7684\u7ed1\u5b9a\u4ee3\u7801":45,"\u751f\u6210\u540e\u7684\u6587\u6863\u5206\u522b\u5b58\u50a8\u5728\u7f16\u8bd1\u76ee\u5f55\u7684":77,"\u751f\u6210\u5e8f\u5217\u7684\u6700\u5927\u957f\u5ea6":114,"\u751f\u6210\u6587\u6863":45,"\u751f\u6210\u7684":12,"\u751f\u6210\u7684\u6027\u80fd\u5206\u6790\u6587\u4ef6\u4e3a":107,"\u751f\u6210\u7684\u6570\u636e\u5c06\u4f1a\u5b58\u50a8\u5728\u8fd9\u4e2avolume\u4e0b":97,"\u751f\u6210\u7684\u6570\u636e\u7f13\u5b58\u5728\u5185\u5b58\u91cc":81,"\u751f\u6210\u7ed9\u5b9a":12,"\u751f\u6210\u7f51\u7edc\u5c42\u914d\u7f6e":74,"\u751f\u6210\u81ea\u5df1\u76ee\u5f55\u4e0b\u7684\u4ed3\u5e93":72,"\u751f\u6210\u8c03\u8bd5\u4fe1\u606f":107,"\u751f\u6210\u968f\u673a\u7684\u8f93\u5165\u6570\u636e":75,"\u751f\u6210api\u6587\u6863":45,"\u751f\u6210pfsclient\u548cpfsserver\u7684\u6846\u67b6\u90e8\u5206":28,"\u751f\u6210python\u6027\u80fd\u5206\u6790\u7684\u547d\u4ee4\u5982\u4e0b":107,"\u7528":28,"\u7528\u4e8e\u521d\u59cb\u5316\u53c2\u6570\u548c\u8bbe\u7f6e":74,"\u7528\u4e8e\u5c06\u53c2\u6570\u4f20\u9012\u7ed9\u7f51\u7edc\u914d\u7f6e":105,"\u7528\u4e8e\u6307\u5b9a\u5176\u8981\u5173\u8054\u7684layer":82,"\u7528\u4e8e\u6307\u5b9a\u7f51\u7edc\u914d\u7f6e\u6587\u4ef6":103,"\u7528\u4e8e\u6ce8\u518c\u6ca1\u6709\u53cd\u5411\u7684op":75,"\u7528\u4e8e\u6d4b\u8bd5\u548c\u5bf9\u6bd4\u5728\u4f7f\u7528mkl":42,"\u7528\u4e8e\u7a00\u758f\u7c7b\u578b\u53c2\u6570\u901a\u4fe1\u7684\u7aef\u53e3\u4e2a\u6570":91,"\u7528\u4e8e\u7a00\u758f\u8bad\u7ec3\u4e2d":103,"\u7528\u4e8e\u7ba1\u7406mkl":42,"\u7528\u4e8e\u83b7\u53d6\u7279\u5b9alayer\u4e0a\u4e00\u65f6\u95f4\u6b65\u7684\u8f93\u51fa":82,"\u7528\u4e8e\u8ba1\u7b97\u7f16\u7801\u5411\u91cf\u7684\u52a0\u6743\u548c":114,"\u7528\u4e8e\u8bad\u7ec3\u795e\u7ecf\u7f51\u7edc\u7684\u6570\u636e":92,"\u7528\u4e8e\u9009\u62e9\u662f\u5426\u4f7f\u7528\u76f8\u5173\u529f\u80fd":41,"\u7528\u4e8e\u9009\u62e9\u662f\u5426\u4f7f\u7528mkl":42,"\u7528\u4e8emkl":[41,42],"\u7528\u53cc\u5411\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7f16\u7801":114,"\u7528\u591a\u5bf9\u6548\u679c\u5b8c\u5168\u76f8\u540c\u7684":111,"\u7528\u6237\u4e00\u822c\u901a\u8fc7\u8c03\u7528":107,"\u7528\u6237\u4e0a\u4f20\u6570\u636e\u540e":12,"\u7528\u6237\u4e5f\u53ef\u4ee5\u4e0a\u4f20label":12,"\u7528\u6237\u4e5f\u53ef\u4ee5\u4f7f\u7528paddlepaddle\u63d0\u4f9b\u7684\u5b98\u65b9\u5f00\u53d1\u955c\u50cf":116,"\u7528\u6237\u4eceapp":117,"\u7528\u6237\u53ea\u9700\u5b9a\u4e49rnn\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u5b8c\u6210\u7684\u8ba1\u7b97":113,"\u7528\u6237\u53ef\u4ee5\u5206\u522b\u67e5\u770b\u6700\u65b0\u7684":77,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003\u4e0b\u6587":116,"\u7528\u6237\u53ef\u4ee5\u53c2\u8003sphinx\u6559\u7a0b\u8fdb\u884c\u4e66\u5199":77,"\u7528\u6237\u53ef\u4ee5\u5b89\u5168\u7684\u91ca\u653e\u67d0\u4e2ac":46,"\u7528\u6237\u53ef\u4ee5\u628a\u81ea\u5df1\u7684\u6570\u636e\u5206\u4eab\u7ed9\u522b\u4eba":12,"\u7528\u6237\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528\u8fd9\u4e2a\u52a8\u6001\u5e93\u6765\u5f15\u5165paddl":46,"\u7528\u6237\u53ef\u4ee5\u81ea\u5b9a\u4e49beam":103,"\u7528\u6237\u53ef\u4ee5\u8bbe\u7f6e":105,"\u7528\u6237\u53ef\u5728\u81ea\u5df1\u719f\u6089\u7684\u5f00\u53d1\u5e73\u53f0\u4e0a\u7f16\u8bd1android\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u5e93":116,"\u7528\u6237\u53ef\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e\u5b83\u4eec":1,"\u7528\u6237\u53ef\u5c06":116,"\u7528\u6237\u53ef\u5c06\u5408\u6210\u7684fat\u5e93\u7528\u4e8e\u6df1\u5ea6\u5b66\u4e60\u76f8\u5173\u7684io":117,"\u7528\u6237\u53ef\u6839\u636e\u81ea\u5df1\u7684\u7f16\u8bd1\u76ee\u6807\u67b6\u6784":116,"\u7528\u6237\u53ef\u81ea\u884c\u524d\u5f80\u4e0b\u8f7d\u9884\u7f16\u8bd1\u597d\u7684\u7248\u672c":116,"\u7528\u6237\u53ef\u901a\u8fc7\u5982\u4e0b\u4e24\u79cd\u65b9\u5f0f":116,"\u7528\u6237\u5728\u4f7f\u7528\u8fd9\u4e00\u7c7brecurr":82,"\u7528\u6237\u5728\u4f7f\u7528paddlepaddl":78,"\u7528\u6237\u5728\u672c\u5730\u8f6c\u6362\u597d\u518d\u4e0a\u4f20":12,"\u7528\u6237\u5c06\u53c2\u6570\u8f7d\u5165":83,"\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":97,"\u7528\u6237\u5f3a\u5236\u6307\u5b9a\u7279\u5b9a\u7684python\u7248\u672c":78,"\u7528\u6237\u6587\u4ef6\u53ef\u80fd\u662f\u6bd4\u8f83\u5927\u7684":28,"\u7528\u6237\u8fd8\u53ef\u6839\u636e\u81ea\u5df1\u7684\u9700\u6c42\u8bbe\u7f6e\u5176\u4ed6\u7f16\u8bd1\u53c2\u6570":[116,117,118],"\u7528\u6237\u901a\u8fc7\u53c2\u6570":[82,83],"\u7528\u6237\u901a\u8fc7c":46,"\u7528\u6237\u9700\u8981\u5728\u7f51\u7edc\u914d\u7f6e\u4e2d\u6307\u5b9a":105,"\u7528\u6237\u9700\u8981\u5728cmake\u65f6\u624b\u52a8\u8bbe\u7f6e\u8fd9\u4e9b\u503c":[116,118],"\u7528\u6237\u9700\u8981\u6307\u5b9a\u672c\u673a\u4e0apython\u7684\u8def\u5f84":78,"\u7528\u6237\u9700\u8981\u63d0\u524d\u51c6\u5907\u597d\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883":116,"\u7528\u672c\u673a\u7684\u7b2c\u4e00\u4e2a":0,"\u7528\u6765\u4ece\u53c2\u6570\u670d\u52a1\u5668\u9884\u53d6\u53c2\u6570\u77e9\u9635\u76f8\u5e94\u7684\u884c":74,"\u7528\u6765\u5b58\u50a8\u672c\u6b21\u6027\u80fd\u5206\u6790\u7684\u7ed3\u679c":107,"\u7528\u8fd9\u4e2a\u955c\u50cf\u521b\u5efa\u7684\u5bb9\u5668\u9700\u8981\u6709\u4ee5\u4e0b\u4e24\u4e2a\u529f\u80fd":97,"\u7528docker\u7f16\u8bd1\u548c\u6d4b\u8bd5paddlepaddl":3,"\u7528web\u6d4f\u89c8\u5668\u8bbf\u95ee\u5bf9\u5e94\u7f51\u5740":107,"\u7531":[82,89,113],"\u7531\u4e8e":87,"\u7531\u4e8e\u5728\u73b0\u6709\u7684\u67d0\u4e9b\u60c5\u51b5\u4e0b":41,"\u7531\u4e8e\u5b83\u5185\u90e8\u5305\u542b\u4e86\u6bcf\u7ec4\u6570\u636e\u4e2d\u7684\u6240\u6709\u53e5\u5b50":111,"\u7531\u4e8e\u5bf9parameters\u7684\u66f4\u65b0\u9700\u8981\u83b7\u53d6parameters\u5185\u5b58\u7684":11,"\u7531\u4e8e\u6211\u4eec\u60f3\u8981\u7684\u53d8\u6362\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217":111,"\u7531\u4e8e\u6211\u4eec\u652f\u6301\u8bad\u7ec3\u6570\u636e\u6709\u4e0d\u540c\u7684\u6279\u6b21\u5927\u5c0f":74,"\u7531\u4e8e\u96c6\u7fa4\u4e2d\u540c\u65f6\u5b58\u5728\u4e24\u53f0\u673a\u5668\u6545\u969c\u7684\u6982\u7387\u6781\u4f4e":11,"\u7531\u4e8earm64\u67b6\u6784\u8981\u6c42android":116,"\u7531\u4e8ec":45,"\u7531\u4e8echunk\u6bd4\u8f83\u5c0f":28,"\u7531\u4e8eeigen":76,"\u7531\u4e8emkl":42,"\u7531\u4e8epypi":62,"\u7531\u4e8estep":113,"\u7531\u4e8etensor\u7684rank\u662f\u6a21\u677f\u53c2\u6570":76,"\u7531\u5206\u652f\u5904\u7684layer\u8d1f\u8d23\u6c42\u548c":42,"\u7531\u8bcd\u8bed\u6784\u6210\u7684\u53e5\u5b50":110,"\u7531\u94fe\u63a5\u65b9\u5f0f\u51b3\u5b9a":87,"\u7533\u8bf7\u7528\u6237\u7a7a\u95f4":28,"\u767b\u5f55\u5230head\u8282\u70b9":98,"\u7684":[0,4,72,76,89,90,93,97,117],"\u7684\u4e00\u4e2a\u5b50\u96c6":42,"\u7684\u4e00\u4e2a\u7b80\u5355\u8c03\u7528\u5982\u4e0b":113,"\u7684\u4e3a0":103,"\u7684\u4efb\u4e00\u4e00\u79cd":81,"\u7684\u4f5c\u7528\u662f\u5ef6\u8fdf\u5206\u914d\u5185\u5b58":76,"\u7684\u4f7f\u7528\u793a\u4f8b\u5982\u4e0b":110,"\u7684\u4fe1\u606f":42,"\u7684\u503c":[116,117,118],"\u7684\u503c\u81ea\u52a8\u63a8\u5bfc\u5f97\u5230":116,"\u7684\u504f\u7f6e\u5411\u91cf":74,"\u7684\u5171\u4eab\u5df2\u7ecf\u52a0\u8f7d\u7684\u9884\u6d4b\u6a21\u578b":90,"\u7684\u5177\u4f53\u8ba1\u7b97\u903b\u8f91":75,"\u7684\u5185\u5b58":81,"\u7684\u5185\u5bb9\u6765\u5b9a\u5236imag":97,"\u7684\u5185\u6838block\u4f7f\u7528\u60c5\u51b5":108,"\u7684\u5355\u5143\u6d4b\u8bd5":75,"\u7684\u5355\u5143\u6d4b\u8bd5\u548c\u7b80\u5355\u7f51\u7edc\u7684\u6574\u4f53\u6d4b\u8bd5":42,"\u7684\u53cd\u5411\u4f20\u64ad\u5c06\u4f1a\u6253\u5370\u65e5\u5fd7\u4fe1\u606f":103,"\u7684\u53d8\u6362\u77e9\u9635":74,"\u7684\u540d\u79f0\u76f8\u540c":114,"\u7684\u5411\u91cf":74,"\u7684\u542f\u52a8\u53c2\u6570":97,"\u7684\u542f\u52a8\u53c2\u6570\u5e76\u6267\u884c\u8fdb\u7a0b":97,"\u7684\u547d\u4ee4\u548c\u4e00\u822c\u7684":107,"\u7684\u547d\u540d\u98ce\u683c\u5e76\u4e0d\u80fd\u9002\u5e94\u5176\u4ed6\u7b2c\u4e09\u65b9\u8bed\u8a00":45,"\u7684\u5730\u65b9":72,"\u7684\u5747\u5300\u5206\u5e03":83,"\u7684\u57fa\u672c\u903b\u8f91":42,"\u7684\u591a\u79cd\u5b89\u88c5\u65b9\u5f0f":101,"\u7684\u5934\u6587\u4ef6":45,"\u7684\u5b50\u7c7b\u53ea\u9700\u8981\u4f7f\u7528\u5185\u90e8\u5b58\u50a8\u5c31\u53ef\u4ee5\u4e86":42,"\u7684\u5b9e\u73b0":75,"\u7684\u5de5\u4f5c\u6d41\u7a0b\u5982\u56fe1\u6240\u793a":90,"\u7684\u5e73\u5747\u503c":110,"\u7684\u5e8f\u5217":89,"\u7684\u5e8f\u5217\u5f62\u72b6\u4e00\u81f4":111,"\u7684\u5f00\u53d1\u5de5\u4f5c\u90fd\u5e94\u8be5\u5728\u4e00\u4e2a\u65b0\u7684\u5206\u652f\u4e0a\u5b8c\u6210":72,"\u7684\u5f00\u53d1\u6d41\u7a0b":0,"\u7684\u5f00\u59cb\u8bf7\u52a0\u4e0a\u5b8f\u5b9a\u4e49":75,"\u7684\u6027\u80fd\u5206\u6790\u4e0e\u8c03\u4f18\u5206\u4e3a\u4e24\u4e2a\u90e8\u5206":107,"\u7684\u6027\u80fd\u5206\u6790\u5de5\u5177\u975e\u5e38\u591a":107,"\u7684\u6027\u80fd\u6709\u95ee\u9898":107,"\u7684\u60c5\u51b5\u4e0b":41,"\u7684\u63a5\u53e3\u6837\u5f0f":45,"\u7684\u63a5\u53e3\u8bf7\u67e5\u770b":89,"\u7684\u63cf\u8ff0\u8bf4\u660e\u4e2d":72,"\u7684\u64cd\u4f5c":76,"\u7684\u6570\u636e\u6d41\u56fe":12,"\u7684\u6570\u76ee\u4e00\u81f4":110,"\u7684\u6587\u4ef6\u4e5f\u5e26\u5230\u65b0\u5206\u652f\u4e0a":72,"\u7684\u65b9\u7a0b":74,"\u7684\u65f6\u5019":42,"\u7684\u65f6\u95f4\u6b65\u4fe1\u606f\u6210\u6b63\u6bd4":81,"\u7684\u66f4\u8be6\u7ec6\u51c6\u786e\u7684\u5b9a\u4e49":111,"\u7684\u6700\u5c0f\u503c":103,"\u7684\u6700\u65b0\u4ee3\u7801\u5e76\u66f4\u65b0\u5f53\u524d\u5206\u652f":72,"\u7684\u6784\u9020\u51fd\u6570":75,"\u7684\u67b6\u6784\u7684\u793a\u4f8b":114,"\u7684\u6837\u5f0f":72,"\u7684\u6838\u5fc3\u662f\u8bbe\u8ba1step\u51fd\u6570\u7684\u8ba1\u7b97\u903b\u8f91":113,"\u7684\u6839\u76ee\u5f55":117,"\u7684\u683c\u5f0f\u59cb\u7ec8\u662f":42,"\u7684\u683c\u5f0f\u5b58\u50a8":42,"\u7684\u6982\u5ff5":42,"\u7684\u6bb5\u843d\u5b9a\u4e49\u4e3a\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":113,"\u7684\u6bcf\u4e2a\u8fdb\u7a0b\u90fd\u53ef\u4ee5\u4ececeph\u8bfb\u53d6\u6570\u636e":96,"\u7684\u6e90\u7801\u4ee5\u53ca\u751f\u6210\u6587\u6863\u9700\u8981\u591a\u79cd\u5f00\u53d1\u5de5\u5177":72,"\u7684\u6e90\u7801\u91cc\u4f7f\u7528\u4e86":45,"\u7684\u7248\u672c":[0,62,118],"\u7684\u72b6\u6001":113,"\u7684\u72ec\u7acb\u5de5\u5177\u94fe":116,"\u7684\u77e9\u9635":[74,81],"\u7684\u7a20\u5bc6\u5411\u91cf\u4f5c\u4e3a\u8f93\u5165":74,"\u7684\u7a20\u5bc6\u77e9\u9635":89,"\u7684\u7a20\u5bc6\u77e9\u9635\u662f\u4e00\u4e2a\u7531":89,"\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570":90,"\u7684\u7b2ci\u4e2a\u503c":74,"\u7684\u7b2cj\u4e2a\u503c":74,"\u7684\u7cfb\u7edf":0,"\u7684\u7ed3\u679c":41,"\u7684\u7f16\u5199":91,"\u7684\u7f16\u8bd1\u5de5\u5177\u94fe":116,"\u7684\u7f29\u5199":28,"\u7684\u7f51\u7edc\u6a21\u578b":41,"\u7684\u89c4\u8303":45,"\u7684\u89d2\u5ea6":12,"\u7684\u8ba1\u7b97\u4ee3\u7801":76,"\u7684\u8ba1\u7b97\u8fc7\u7a0b\u4e66\u5199\u66f4\u52a0\u7b80\u5355":75,"\u7684\u8bdd":81,"\u7684\u8be6\u7ec6\u4fe1\u606f":107,"\u7684\u8f93\u5165":113,"\u7684\u8f93\u51fa":[81,108],"\u7684\u8f93\u51fa\u4fe1\u606f\u5165\u624b\u662f\u4e2a\u4e0d\u9519\u7684\u9009\u62e9":108,"\u7684\u8f93\u51fa\u51fd\u6570\u8fd4\u56de\u7684\u662f\u4e0b\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa\u8bcd\u7684":114,"\u7684\u8f93\u51fa\u683c\u5f0f":111,"\u7684\u8f93\u51fa\u88ab\u7528\u4f5c":114,"\u7684\u8f93\u51fab\u662f\u4e00\u4e2a":81,"\u7684\u8fd0\u884c\u73af\u5883":0,"\u7684\u8fdc\u7a0b\u4ed3\u5e93\u7684\u540d\u5b57":72,"\u7684\u914d\u7f6e\u5199\u5230\u914d\u7f6e\u6587\u4ef6\u4e2d":12,"\u7684\u96c6\u88c5\u7bb1\u6280\u672f":0,"\u7684\u9875\u9762\u5220\u9664\u8fdc\u7a0b\u4ed3\u5e93\u7684\u5206\u652f":72,"\u7684cpu":75,"\u7684docker\u955c\u50cf":2,"\u7684flag":[41,42],"\u7684linux\u670d\u52a1\u5668\u7ec4\u6210":101,"\u7684paddlepaddle\u5e93":116,"\u7684vanilla":41,"\u76d1\u542c\u7684\u7aef\u53e3\u4e2a\u6570":91,"\u76ee\u524d":113,"\u76ee\u524d\u4f7f\u7528":72,"\u76ee\u524d\u53ea\u8003\u8651":42,"\u76ee\u524d\u53ea\u8003\u8651\u52a8\u6001\u6269\u5bb9trainer\u6570\u91cf":11,"\u76ee\u524d\u5728paddlepaddle\u4e2d":42,"\u76ee\u524d\u5728paddlepaddle\u4e2d\u6570\u636e\u90fd\u662f\u4ee5":42,"\u76ee\u524d\u5d4c\u5165python\u89e3\u91ca\u5668":45,"\u76ee\u524d\u5fc5\u987b\u8bbe\u7f6e\u6210":118,"\u76ee\u524d\u6211\u4eec\u7528cephfs\u6765\u642d\u5efa":28,"\u76ee\u524d\u63d0\u4f9b\u4e09\u79cd\u94fe\u63a5\u65b9\u5f0f":87,"\u76ee\u524d\u652f\u6301":116,"\u76ee\u524d\u652f\u6301\u4e24\u79cd":110,"\u76ee\u524d\u652f\u6301cento":86,"\u76ee\u524d\u652f\u6301fail":103,"\u76ee\u524d\u7684\u4f18\u5316":42,"\u76ee\u524d\u8be5\u53c2\u6570\u4ec5\u7528\u4e8eaucvalidationlayer\u548cpnpairvalidationlayer\u5c42":103,"\u76ee\u524d\u8fd8\u672a\u652f\u6301":113,"\u76ee\u524dpaddle\u7684\u8fdb\u7a0b\u6a21\u578b\u662fc":45,"\u76ee\u524dpaddlepaddle\u7684develop\u5206\u652f\u7684\u6587\u6863\u662f\u81ea\u52a8\u89e6\u53d1\u66f4\u65b0\u7684":77,"\u76ee\u524dpaddlepaddle\u91c7\u7528\u4e86":41,"\u76ee\u5f55":[0,2,93,96,97,116,117,118],"\u76ee\u5f55\u4e0b":[46,74,93],"\u76ee\u5f55\u4e0b\u5bf9\u5e94\u7684\u5730\u65b9":42,"\u76ee\u5f55\u4e0b\u65b0\u589e\u7684":75,"\u76ee\u5f55\u4e0b\u6700\u65b0\u7684":117,"\u76ee\u5f55\u4e0b\u7684\u4ee3\u7801\u793a\u4f8b":90,"\u76ee\u5f55\u4e0b\u7684\u751f\u6210\u6587\u4ef6\u7528\u4e8e\u6df1\u5ea6\u5b66\u4e60\u76f8\u5173android":116,"\u76ee\u5f55\u4e0b\u7684python\u5305":78,"\u76ee\u5f55\u4e2d":[87,90,93],"\u76ee\u5f55\u4e2d\u4f1a\u5305\u542b":[116,118],"\u76ee\u5f55\u4e2d\u4f1a\u5305\u542b\u4ee5\u4e0b\u5185\u5bb9":117,"\u76ee\u5f55\u4e2d\u7684":108,"\u76ee\u5f55\u4e2dpaddl":97,"\u76ee\u5f55\u548c":[116,117,118],"\u76ee\u5f55\u5c31\u6210\u4e3a\u4e86\u5171\u4eab\u5b58\u50a8":97,"\u76ee\u5f55\u751f\u6210\u4e00\u5957\u72ec\u7acb\u7f16\u8bd1\u5de5\u5177\u94fe":116,"\u76ee\u5f55\u91cc\u627e\u5230\u4ea4\u53c9\u7f16\u8bd1\u5668":118,"\u76ee\u6807\u5411\u91cf":114,"\u76ee\u6807\u5de5\u5177\u94fe":116,"\u76ee\u6807\u673a\u7248protobuf\u5e93":118,"\u76ee\u6807\u67b6\u6784":117,"\u76ee\u6807\u67b6\u6784abi":116,"\u76f4\u5230\u8bad\u7ec3\u6536\u655b\u4e3a\u6b62":83,"\u76f4\u63a5\u4f7f\u7528\u4e0a\u8ff0\u5b89\u88c5\u6d41\u7a0b":3,"\u76f4\u63a5\u4f7f\u7528c\u8bed\u8a00\u7684":45,"\u76f4\u63a5\u5220\u9664\u8fd9\u4e2a\u53c2\u6570\u5373\u53ef":46,"\u76f4\u63a5\u5347\u7ea7\u5230\u66f4\u65b0\u7684\u7248\u672c":1,"\u76f4\u63a5\u5bfc\u51fa\u5230c\u7684\u63a5\u53e3\u6bd4\u8f83\u56f0\u96be":45,"\u76f4\u63a5\u8c03\u7528\u76f8\u5e94\u63a5\u53e3\u5373\u53ef":75,"\u76f4\u63a5\u8fd0\u884c":2,"\u76f8\u5173\u5c42":41,"\u76f8\u540c\u540d\u5b57\u7684\u53c2\u6570":83,"\u76f8\u6bd4":75,"\u770b\u5f53\u524dmpi\u96c6\u7fa4\u662f\u5426\u652f\u6301resourc":79,"\u77a7":86,"\u77e9\u9635":102,"\u77e9\u9635\u4e2d\u6bcf\u4e2a\u5143\u7d20\u7684\u503c\u968f\u673a\u751f\u6210":89,"\u77e9\u9635\u4e58\u6cd5\u7684\u516c\u5f0f":75,"\u77e9\u9635\u5927\u5c0f\u662f":41,"\u77e9\u9635\u662f\u5426\u662f\u4e00\u4e2a\u5e8f\u5217":89,"\u77e9\u9635\u7684\u9ad8\u5ea6":89,"\u77e9\u9635\u91cc\u7684\u5143\u7d20\u662f\u6d6e\u70b9\u6570":89,"\u786e\u4fdd\u7f16\u8bd1\u5668\u9009\u9879":72,"\u78c1\u76d8\u4e0d\u591f":0,"\u78c1\u76d8\u7a7a\u95f4\u4e0d\u8db3\u7b49":79,"\u793a\u4f8b":[81,83,90],"\u793a\u4f8b3\u5bf9\u4e8e\u5355\u5c42rnn\u548c\u53cc\u5c42rnn\u6570\u636e\u5b8c\u5168\u76f8\u540c":111,"\u793a\u4f8b3\u7684\u914d\u7f6e\u4f7f\u7528\u4e86\u5355\u5c42rnn\u548c\u53cc\u5c42rnn":111,"\u793a\u4f8b3\u7684\u914d\u7f6e\u5206\u522b\u4e3a":111,"\u793a\u4f8b\u4ee3\u7801\u5982\u4e0b":[81,90],"\u793a\u4f8b\u5982\u4e0b":83,"\u793a\u4f8b\u7a0b\u5e8f":91,"\u793e\u533a\u53c2\u4e0e\u56f0\u96be":45,"\u793e\u533a\u8d21\u732e\u4ee3\u7801\u5b66\u4e60\u6210\u672c\u9ad8":45,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u4e00\u4e2a\u8ba1\u7b97\u5c42\u7684\u8f93\u5165":89,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u4e00\u4e2a\u8ba1\u7b97\u5c42\u7684\u8f93\u5165\u8f93\u51fa\u88ab\u7ec4\u7ec7\u4e3a\u4e00\u4e2a":90,"\u795e\u7ecf\u7f51\u7edc\u4e2d\u7684\u53c2\u6570":11,"\u795e\u7ecf\u7f51\u7edc\u4e5f\u9700\u8981\u4e00\u4e9b\u7279\u5b9a\u7684layer\u4f5c\u4e3a\u8f93\u5165\u63a5\u53e3":84,"\u795e\u7ecf\u7f51\u7edc\u53c2\u6570\u4ee5\u53ca\u8fed\u4ee3\u65b9\u7a0b":84,"\u795e\u7ecf\u7f51\u7edc\u5728\u8bad\u7ec3\u7684\u65f6\u5019":81,"\u795e\u7ecf\u7f51\u7edc\u672c\u8d28\u4e0a\u662f\u4e00\u4e2a\u8ba1\u7b97\u56fe":76,"\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u7ed3\u6784\u548c\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u5c06\u88ab\u5e8f\u5217\u5316\u5408\u5e76\u5165\u4e00\u4e2a\u6587\u4ef6":90,"\u795e\u7ecf\u7f51\u7edc\u7684\u7f51\u7edc\u7ed3\u6784\u4e2d\u5177\u6709\u6709\u5411\u73af\u7ed3\u6784":111,"\u795e\u7ecf\u7f51\u7edc\u7684\u8bad\u7ec3\u672c\u8eab\u662f\u4e00\u4e2a\u975e\u5e38\u6d88\u8017\u5185\u5b58\u548c\u663e\u5b58\u7684\u5de5\u4f5c":81,"\u79bb\u7ebf\u6279\u5904\u7406":12,"\u79f0\u4e3a":[72,114],"\u79f0\u4e3a\u5f00\u53d1\u955c\u50cf":116,"\u79f0\u4e4b\u4e3a":89,"\u79f0\u4e4b\u4e3a\u53cc\u5c42\u5e8f\u5217\u7684\u4e00\u4e2a\u5b50\u5e8f\u5217":110,"\u79f0\u4e4b\u4e3a\u96c6\u675f\u5927\u5c0f":103,"\u79f0\u4f5c\u6709kernel":75,"\u79f0\u4f5ckernel":75,"\u79fb\u52a8\u7aef\u9884\u6d4b":89,"\u7a00\u758f\u6570\u636e\u7684\u683c\u5f0f":74,"\u7a00\u758f\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":97,"\u7a00\u758f\u768401\u5411\u91cf":84,"\u7a00\u758f\u7684\u5411\u91cf":84,"\u7a00\u758f\u77e9\u9635":89,"\u7a00\u758f\u77e9\u9635\u4f7f\u7528":89,"\u7a00\u758f\u77e9\u9635\u53ca\u76f8\u5173\u7684\u63a5\u53e3":89,"\u7a00\u758f\u77e9\u9635\u5b58\u50a8\u793a\u610f\u56fe":89,"\u7a00\u758f\u77e9\u9635\u7684\u4e58\u79ef\u5e94\u7528\u4e8e\u524d\u5411\u4f20\u64ad\u8fc7\u7a0b":105,"\u7a0b\u5e8f\u4ece\u6b64\u76ee\u5f55\u62f7\u8d1d\u6587\u4ef6\u5230\u5bb9\u5668\u5185\u8fdb\u884c\u8bad\u7ec3":97,"\u7a0b\u5e8f\u4f9d\u8d56":91,"\u7a0b\u5e8f\u505c\u6b62":103,"\u7a0b\u5e8f\u662f\u4e00\u6837\u7684":107,"\u7a0b\u5e8f\u76f4\u63a5\u9000\u51fa":103,"\u7a20\u5bc6\u5411\u91cf":74,"\u7a20\u5bc6\u66f4\u65b0\u7684\u7aef\u53e3\u6570\u91cf":97,"\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":84,"\u7a20\u5bc6\u77e9\u9635":89,"\u7acb\u523b\u9000\u51fa":81,"\u7acb\u5373\u6267\u884c\u5355\u5143\u6d4b\u8bd5":1,"\u7ae0\u8282":116,"\u7aef\u53e3":79,"\u7aef\u6570\u636e\u7c7b\u578b":89,"\u7aef\u7684":107,"\u7aef\u8bfb\u53d6\u6570\u636e":81,"\u7b2c\u4e00\u4e2a":72,"\u7b2c\u4e00\u4e2a\u53c2\u6570":75,"\u7b2c\u4e00\u4e2a\u6837\u672c\u540c\u65f6encode\u4e24\u6761\u6570\u636e\u6210\u4e24\u4e2a\u5411\u91cf":111,"\u7b2c\u4e00\u4e2atag\u4e3a":62,"\u7b2c\u4e00\u6b65\u9700\u8c03\u7528":90,"\u7b2c\u4e00\u7ae0\u8282":84,"\u7b2c\u4e09\u4e2a\u53c2\u6570":75,"\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93\u9700\u8981\u6309\u7167\u4e0e\u65b9\u5f0f2\u540c\u6837\u65b9\u6cd5\u663e\u793a\u5730\u8fdb\u884c\u94fe\u63a5":87,"\u7b2c\u4e09\u65b9\u94fe\u63a5\u5e93\u548c\u5934\u6587\u4ef6":87,"\u7b2c\u4e09\u6b65\u5b8c\u6210\u540e":62,"\u7b2c\u4e8c\u4e2a":81,"\u7b2c\u4e8c\u4e2a\u4e3a":62,"\u7b2c\u4e8c\u4e2a\u53c2\u6570":75,"\u7b2c\u4e8c\u7c7b":82,"\u7b2ci\u884c\u7b2cj\u5217\u7684\u6570\u503c":74,"\u7b49":[42,46,75,79,90],"\u7b49\u4e8e\u6837\u672c\u6570":81,"\u7b49\u5168\u90e8\u9759\u6001\u5e93\u4e2d\u7684\u76ee\u6807\u6587\u4ef6\u5168\u90e8\u6253\u5305\u540e\u4ea7\u751f\u7684\u6587\u4ef6":46,"\u7b49\u53c2\u6570":97,"\u7b49\u5f85\u7f16\u8bd1\u5b8c\u6210\u540e":62,"\u7b49\u6587\u4ef6":46,"\u7b49\u7b2c\u4e09\u65b9\u5e93":87,"\u7b80\u5199":75,"\u7b80\u5355\u4ecb\u7ecd\u9700\u8981\u7528\u5230\u57fa\u7c7b":75,"\u7b80\u5355\u603b\u7ed3op\u9700\u8981\u5305\u542b\u7684\u5185\u5bb9\u5982\u4e0b":75,"\u7b80\u5355\u6765\u8bf4":108,"\u7b80\u5355\u7684\u5168\u8fde\u63a5\u7f51\u7edc":83,"\u7b80\u5355\u7684\u6027\u80fd\u5206\u6790":108,"\u7b80\u5355\u7684yaml\u6587\u4ef6\u5982\u4e0b":96,"\u7b97\u6cd5":[81,114],"\u7b97\u6cd5\u4e2d\u7684beam\u5927\u5c0f":114,"\u7ba1\u7406\u4e86\u6bcf\u4e2a\u8ba1\u7b97\u5c42\u8f93\u51fa\u7684\u5b58\u50a8\u7a7a\u95f4":90,"\u7c7b\u4f3c":[46,110],"\u7c7b\u4f5c\u4e3a\u53c2\u6570\u7684\u62bd\u8c61":74,"\u7c7b\u522b\u4e2d\u7684\u53c2\u6570\u53ef\u7528\u4e8e\u6240\u6709\u573a\u5408":102,"\u7c7b\u522b\u6807\u7b7e\u4e4b\u4e00":90,"\u7c7b\u522b\u6807\u7b7e\u5c42":90,"\u7c7b\u540d\u548cc":45,"\u7c7b\u578b":[45,75,89,103],"\u7c7b\u578b\u4e3a":75,"\u7c7b\u578b\u4ecd\u7136\u4e3aeigenvector":76,"\u7c7b\u578b\u53ef\u4ee5\u662fpaddlepaddle\u652f\u6301\u7684\u4efb\u610f\u8f93\u5165\u6570\u636e\u7c7b\u578b":110,"\u7c7b\u578b\u540d\u4e3a":75,"\u7c7b\u578b\u662fnumpy\u7684ndarrai":81,"\u7c7b\u578b\u662fsparse_binary_vector":84,"\u7c7b\u578b\u662fsparse_float_vector":84,"\u7c7b\u578b\u7684":111,"\u7c7b\u578b\u7b49\u662f\u5426\u5408\u6cd5":75,"\u7c7b\u578b\u8fd8\u662f":89,"\u7c7b\u7684\u5b9a\u4e49\u5199\u5728":75,"\u7c7b\u7684\u5bf9\u8c61":90,"\u7c7b\u7684\u6784\u9020\u51fd\u6570\u548c\u6790\u6784\u51fd\u6570":74,"\u7c7b\u91cd\u5199":75,"\u7c7b\u9700\u8981\u5b9e\u73b0\u521d\u59cb\u5316":74,"\u7cfb\u6570":75,"\u7cfb\u7edf\u4e2d\u7684\u74f6\u9888\u53ef\u80fd\u548c\u7a0b\u5e8f\u5458\u5f00\u53d1\u8fc7\u7a0b\u4e2d\u60f3\u8c61\u7684\u74f6\u9888\u76f8\u53bb\u751a\u8fdc":107,"\u7cfb\u7edf\u4f1a\u5bf9\u65b0\u589e\u7684op\u81ea\u52a8\u7ed1\u5b9apython":75,"\u7cfb\u7edf\u4f1a\u63d0\u4f9b\u4e00\u4e2a\u5206\u5e03\u5f0f\u5b58\u50a8\u670d\u52a1":91,"\u7cfb\u7edf\u4f1a\u6839\u636e\u6587\u4ef6\u540d\u81ea\u52a8\u6784\u5efaop\u548c\u5176\u5bf9\u5e94\u7684python\u6269\u5c55":75,"\u7ebf\u7a0bid\u53f7":105,"\u7ec6\u8282\u63cf\u8ff0":104,"\u7ecf\u5e38\u4f1a\u6d88\u8017\u657010gb\u7684\u5185\u5b58\u548c\u6570gb\u7684\u663e\u5b58":81,"\u7ed3\u5c3e":75,"\u7ed3\u675f\u6807\u8bb0":114,"\u7ed3\u675f\u9884\u6d4b\u4e4b\u540e":90,"\u7ed3\u6784\u4f53":[89,90],"\u7ed3\u679c\u4f1a\u5199\u5165\u5f53\u524d\u8fd0\u884c\u76ee\u5f55\u4e0b\u7684":90,"\u7ed3\u679c\u5982\u4e0b\u56fe\u6240\u793a":107,"\u7ed3\u679c\u8f93\u51fa\u5230":0,"\u7ed3\u8bba":45,"\u7ed9\u4e2a\u7b80\u5355\u7684":72,"\u7ed9\u5b9aencoder\u8f93\u51fa\u548c\u5f53\u524d\u8bcd":113,"\u7edf\u4e00\u7528":12,"\u7ee7\u627f\u81ea":75,"\u7ee7\u627f\u81eaoperatorbas":75,"\u7ef4\u57fa\u767e\u79d1\u4e2d\u6587\u9875\u9762":111,"\u7ef4\u57fa\u767e\u79d1\u9875\u9762":111,"\u7ef4\u7a7a\u95f4":114,"\u7ef4\u7a7a\u95f4\u5b8c\u6210":114,"\u7f13\u5b58\u6c60\u7684\u51cf\u5c0f":81,"\u7f16\u5199":2,"\u7f16\u5199\u4e86\u4e00\u4e2apaddlepaddle\u7684\u7a0b\u5e8f":2,"\u7f16\u5199\u5b8cyaml\u6587\u4ef6\u540e":97,"\u7f16\u5199\u672c\u6b21\u8bad\u7ec3\u7684yaml\u6587\u4ef6":97,"\u7f16\u5199\u6df1\u5ea6\u5b66\u4e60\u7a0b\u5e8f":107,"\u7f16\u5199\u7684\u90e8\u5206":4,"\u7f16\u5199\u96c6\u7fa4\u4efb\u52a1\u63d0\u4ea4\u548c\u7ba1\u7406\u811a\u672c":94,"\u7f16\u53f7\u4ece0\u5f00\u59cb":81,"\u7f16\u7801\u5411\u91cf":114,"\u7f16\u7801\u5668\u8f93\u51fa":114,"\u7f16\u7801\u6e90\u5e8f\u5217":114,"\u7f16\u8bd1":[0,2,72,116],"\u7f16\u8bd1\u51fa\u7684paddlepaddle\u9884\u6d4b\u5e93\u548c\u5934\u6587\u4ef6":87,"\u7f16\u8bd1\u540e\u7684\u6587\u4ef6\u5c06\u88ab\u5b58\u50a8\u5728\u5de5\u4f5c\u76ee\u5f55":77,"\u7f16\u8bd1\u548c\u5b89\u88c5paddlepaddl":118,"\u7f16\u8bd1\u548c\u5b89\u88c5paddlepaddle\u9884\u6d4b\u5e93":[116,117],"\u7f16\u8bd1\u5668":[116,117,118],"\u7f16\u8bd1\u5668\u6ca1\u6709":45,"\u7f16\u8bd1\u5668\u8981\u6c42\u7cfb\u7edf\u652f\u6301":116,"\u7f16\u8bd1\u578b\u8bed\u8a00":45,"\u7f16\u8bd1\u5b89\u88c5\u4e0e\u5355\u5143\u6d4b\u8bd5":80,"\u7f16\u8bd1\u5b89\u88c5\u7ed3\u675f\u4e4b\u540e":116,"\u7f16\u8bd1\u5b8c\u6210\u4e4b\u540e":77,"\u7f16\u8bd1\u5b8c\u6210\u540e\u4f1a\u5728build":1,"\u7f16\u8bd1\u5de5\u5177\u94fe":116,"\u7f16\u8bd1\u5de5\u5177\u94fe\u6240\u5728\u7684\u7edd\u5bf9\u8def\u5f84":118,"\u7f16\u8bd1\u6027\u80fd\u4f1a\u548c":107,"\u7f16\u8bd1\u6210\u529f\u540e":75,"\u7f16\u8bd1\u6210\u529f\u540e\u5728":87,"\u7f16\u8bd1\u6210\u52a8\u6001\u5e93":103,"\u7f16\u8bd1\u65f6\u4e00\u5b9a\u8981\u5f00\u542f\u4f18\u5316":107,"\u7f16\u8bd1\u65f6\u4f1a\u628a\u5bf9\u5e94\u7684\u5934\u6587\u4ef6\u548c\u5e93\u653e\u5728":42,"\u7f16\u8bd1\u65f6\u53ef\u80fd\u4f1a\u53bb\u6389\u8c03\u8bd5\u4fe1\u606f":107,"\u7f16\u8bd1\u65f6\u6307\u5b9a":107,"\u7f16\u8bd1\u751f\u6210":77,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684docker\u53d1\u884c\u955c\u50cf":62,"\u7f16\u8bd1\u8fd9\u4e2a\u7248\u672c\u7684python":62,"\u7f16\u8bd1c":46,"\u7f16\u8bd1paddlepaddl":1,"\u7f51\u7edc\u5c42\u53ef\u4ee5\u6709\u591a\u4e2a\u8f93\u5165":74,"\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u4e3a":74,"\u7f51\u7edc\u5c42\u7684\u7c7b\u578b":74,"\u7f51\u7edc\u5c42\u7684\u7ec6\u8282\u53ef\u4ee5\u901a\u8fc7\u4e0b\u9762\u8fd9\u4e9b\u4ee3\u7801\u7247\u6bb5\u6765\u6307\u5b9a":74,"\u7f51\u7edc\u5c42\u7684\u8f93\u51fa\u662f\u7ecf\u8fc7\u6fc0\u6d3b\u51fd\u6570\u4e4b\u540e\u7684\u503c":103,"\u7f51\u7edc\u5c42\u914d\u7f6e\u5305\u542b\u4ee5\u4e0b\u51e0\u9879":74,"\u7f51\u7edc\u63a5\u53d7\u4e00\u5e45\u56fe\u7247\u4f5c\u4e3a\u8f93\u5165":90,"\u7f51\u7edc\u7ed3\u6784\u7684\u5e8f\u5217\u5316\u7ed3\u679c\u548c\u6a21\u578b\u53c2\u6570\u5b58\u50a8\u76ee\u5f55":90,"\u7f51\u7edc\u901a\u4fe1":74,"\u7f51\u901f\u6216ssl\u94fe\u63a5\u539f\u56e0":78,"\u800c":[82,107,114],"\u800c\u4e0d\u4f1a\u6539\u53d8\u539f\u6709tensor\u7684shape\u4fe1\u606f":76,"\u800c\u4e0d\u5fc5\u5728\u610fpaddl":46,"\u800c\u4e0d\u652f\u6301pypy\u89e3\u91ca\u5668":45,"\u800c\u4e0d\u662f\u5728layer\u91cc\u5b9e\u73b0":82,"\u800c\u4e0d\u662f\u6e90\u7801\u76ee\u5f55\u91cc":78,"\u800c\u4e0d\u662f\u7279\u5f81\u7684\u96c6\u5408":111,"\u800c\u4e0d\u662f\u76f8\u5bf9":89,"\u800c\u4e0d\u662fc":89,"\u800c\u4e0d\u66b4\u9732\u6982\u5ff5\u7684\u5b9e\u73b0":46,"\u800c\u4e14\u4e2a\u6570\u5e76\u4e0d\u786e\u5b9a":91,"\u800c\u4e14\u5305\u542b\u4e86c":4,"\u800c\u4e14\u5728\u4f20\u8f93\u7684\u8fc7\u7a0b\u4e2d\u4e5f\u53ef\u80fd\u51fa\u73b0\u7f51\u7edc\u4e0d\u7a33\u5b9a\u7684\u60c5\u51b5":28,"\u800c\u4e14cento":4,"\u800c\u4e4b\u524d\u7684\u53c2\u6570\u5c06\u4f1a\u88ab\u5220\u9664":103,"\u800c\u4ece\u5e94\u7528\u7684\u89d2\u5ea6":108,"\u800c\u4f18\u5316\u6027\u80fd\u7684\u9996\u8981\u4efb\u52a1":108,"\u800c\u5176\u4ed6\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":105,"\u800c\u51fa\u73b0\u9636\u6bb5\u6027\u7684\u8fd0\u884c\u505c\u6ede":11,"\u800c\u53cc\u5c42rnn\u662f\u53ef\u4ee5\u5904\u7406\u8fd9\u79cd\u8f93\u5165\u6570\u636e\u7684\u7f51\u7edc\u7ed3\u6784":111,"\u800c\u53cd\u5411\u6d4b\u8bd5\u4e2d":75,"\u800c\u53ea\u9700\u8981\u83b7\u5f97recurr":82,"\u800c\u5728\u8ba1\u7b97\u7ed3\u675f\u4e4b\u540e":76,"\u800c\u5728cpp\u91cc\u9762\u5b9e\u73b0\u8fd9\u4e2ac\u7684\u63a5\u53e3":45,"\u800c\u591a\u8bed\u8a00\u63a5\u53e3\u9700\u8981\u76f4\u63a5\u8bfb\u53d6\u751f\u6210\u7684\u4e8c\u8fdb\u5236":45,"\u800c\u5b89\u88c5\u5305":[4,78],"\u800c\u5b89\u88c5\u5305\u662f":[4,78],"\u800c\u5bf9\u4e8e\u53cc\u5c42\u5e8f\u5217":111,"\u800c\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u5185\u5c42\u7279\u5f81\u6570\u636e\u800c\u8a00":111,"\u800c\u5bf9\u4e8egolang":45,"\u800c\u5bf9\u4e8egolang\u9519\u8bef\u5904\u7406\u5e94\u8be5\u4f7f\u7528\u8fd4\u56de\u503c":45,"\u800c\u5c06\u8fd9\u4e2a\u6bb5\u843d\u7684\u6bcf\u4e00\u53e5\u8bdd\u7528lstm\u7f16\u7801\u6210\u4e00\u4e2a\u5411\u91cf":111,"\u800c\u5f53\u524d\u5df2\u7ecf\u67095":108,"\u800c\u662f\u5c06\u8f93\u5165":[89,90],"\u800c\u662f\u76f4\u63a5\u4ece\u5185\u5b58\u7684\u7f13\u5b58\u91cc\u8bfb\u53d6\u6570\u636e":81,"\u800c\u662f\u76f4\u63a5\u4fee\u6539paddl":46,"\u800c\u662f\u76f4\u63a5\u7528api\u7684\u63a5\u53e3\u8fdc\u7a0b\u8bbf\u95ee":12,"\u800c\u66f4\u6df1\u5165\u7684\u5206\u6790":108,"\u800c\u6709\u4e9b\u53c2\u6570\u9700\u8981\u5728\u96c6\u7fa4\u591a\u673a\u8bad\u7ec3\u4e2d\u4f7f\u7528\u7b49":102,"\u800c\u6e90\u5e8f\u5217\u7684\u7f16\u7801\u5411\u91cf\u53ef\u4ee5\u88ab\u65e0\u8fb9\u754c\u7684memory\u8bbf\u95ee":114,"\u800c\u795e\u7ecf\u7f51\u7edc\u662f\u6211\u4eec\u8981\u642d\u5efa\u7684\u5b9d\u5854":84,"\u800c\u7a00\u758f\u66f4\u65b0\u5728\u53cd\u5411\u4f20\u64ad\u4e4b\u540e\u7684\u6743\u91cd\u66f4\u65b0\u65f6\u8fdb\u884c":105,"\u800c\u8ba1\u7b97\u8fc7\u7a0b\u662f\u7531":76,"\u800c\u8fd9\u4e00\u53e5\u8bdd\u5c31\u53ef\u4ee5\u8868\u793a\u6210\u8fd9\u4e9b\u4f4d\u7f6e\u7684\u6570\u7ec4":111,"\u800c\u8fd9\u6bcf\u4e00\u4e2a\u6570\u7ec4\u5143\u7d20":111,"\u800c\u975e\u76f4\u63a5\u56de\u590d\u7684\u65b9\u5f0f":72,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":1,"\u800ceigenvector":76,"\u800crnn\u662f\u6700\u6d41\u884c\u7684\u9009\u62e9":113,"\u800cswig\u53ea\u80fd\u7b80\u5355\u7684\u66b4\u9732c":45,"\u800ctrainer\u9700\u8981\u8bfb\u53d6\u8bad\u7ec3\u6570\u636e\u8fdb\u884c\u8bad\u7ec3":84,"\u800cy_predict\u662f\u63a5\u6536x\u4f5c\u4e3a\u8f93\u5165":84,"\u8054\u901a":101,"\u80fd\u591f\u5904\u7406\u53cc\u5c42\u5e8f\u5217":113,"\u80fd\u591f\u5bf9\u53cc\u5411\u5e8f\u5217\u8fdb\u884c\u5904\u7406\u7684\u6709":113,"\u80fd\u591f\u8bb0\u5f55\u4e0a\u4e00\u4e2asubseq":113,"\u80fd\u591f\u9488\u5bf9cpu\u548cgpu\u7684\u8ba1\u7b97\u505a\u66f4\u591a\u4f18\u5316":82,"\u80fd\u83b7\u53d6":93,"\u811a\u672c":[0,90,116],"\u811a\u672c\u5f00\u59cb\u65f6":97,"\u811a\u672c\u96c6\u6210\u4e86\u5e8f\u5217\u5316\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u7684\u8fc7\u7a0b":90,"\u81ea\u52a8\u5173\u95ed\u5bf9\u5e94\u7684":72,"\u81ea\u52a8\u5730\u5c06\u8fd9\u4e9b\u9009\u9879\u5e94\u7528\u5230":93,"\u81ea\u52a8\u5b8c\u6210\u8fd9\u4e00\u8fc7\u7a0b":113,"\u81ea\u52a8\u6302\u8f7d\u5206\u5e03\u5f0f\u5b58\u50a8\u76ee\u5f55":11,"\u81ea\u52a8\u6784\u5efa\u72ec\u7acb\u5de5\u5177\u94fe":116,"\u81ea\u52a8\u751f\u6210":77,"\u81ea\u52a8\u83b7\u53d6\u4e0a\u4e00\u4e2a\u751f\u6210\u7684\u8bcd":114,"\u81ea\u52a8\u9009\u62e9":117,"\u81ea\u6b64":[116,117],"\u81ea\u7136\u4e5f\u5c31\u6709\u7ba1\u7406\u5458\u6743\u9650":0,"\u81ea\u7136\u8bed\u8a00\u4e2d\u7684\u53e5\u5b50\u662f\u4e00\u4e2a\u5e8f\u5217":89,"\u81ea\u7136\u8bed\u8a00\u4e2d\u7684\u6bb5\u843d\u662f\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":89,"\u81ea\u7136\u8bed\u8a00\u5904\u7406\u7b49":105,"\u81f3\u4e8e\u4e3a\u4ec0\u4e48\u9700\u8981c":46,"\u81f3\u5c11\u5305\u542bgcc_3":4,"\u81f3\u5c11\u5305\u542bglibcxx_3":4,"\u81f3\u6b64":[72,111],"\u826f\u597d\u7684\u6587\u6863":45,"\u8282\u70b9":[98,101],"\u8282\u7701\u4e86\u4e0d\u5fc5\u8981\u7684\u64cd\u4f5c":42,"\u82e5":74,"\u82e5\u5728paddlepaddle\u7f16\u8bd1\u65f6":83,"\u82e5\u5e0c\u671b\u5f97\u5230\u6700\u5feb\u7684\u6267\u884c\u901f\u5ea6":117,"\u82e5\u5e0c\u671b\u6700\u5feb\u7684\u6267\u884c\u901f\u5ea6":[116,118],"\u82e5\u5e72\u4e2a\u53e5\u5b50\u6784\u6210\u4e00\u4e2a\u6bb5\u843d":110,"\u82e5\u6709\u4e0d\u4e00\u81f4\u4e4b\u5904":108,"\u82e5\u6709\u5fc5\u8981":74,"\u82e5\u672a\u663e\u5f0f\u6307\u5b9a":117,"\u82e5\u6ca1\u6709\u663e\u5f0f\u8bbe\u7f6e":116,"\u82e5\u73af\u5883\u53d8\u91cf":[116,117,118],"\u82e5\u8981\u5bf9\u8fd9\u51e0\u4e2alayer\u4f7f\u7528dropout":82,"\u82e5\u8f93\u51fa\u662f\u5355\u5c42\u5e8f\u5217":110,"\u82e5\u8f93\u51fa\u662f\u53cc\u5c42\u5e8f\u5217":110,"\u82f1\u6587\u6587\u6863":77,"\u82f1\u6587\u6587\u6863\u76ee\u5f55":77,"\u8303\u56f4":105,"\u83b7\u53d6":72,"\u83b7\u53d6\u53ef\u9009\u7684tag":2,"\u83b7\u53d6\u5f53\u524d\u7cfb\u7edf\u652f\u6301\u7684\u5b89\u88c5\u5305\u683c\u5f0f":4,"\u83b7\u53d6\u5f53\u524d\u7cfb\u7edf\u652f\u6301\u7684python\u5305\u7684\u540e\u7f00":78,"\u83b7\u53d6\u6700\u65b0\u7684\u68c0\u67e5\u70b9\u7684\u6587\u4ef6uuid":11,"\u83b7\u53d6\u6e90\u7801":0,"\u83b7\u53d6\u8f93\u51fa\u65f6":90,"\u83b7\u53d6trainer":97,"\u83b7\u5f97\u53c2\u6570\u5c3a\u5bf8":74,"\u83b7\u5f97\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u67d0\u4e00\u5c42\u7684name":81,"\u83b7\u5f97\u57fa\u672c\u7684docker\u5b89\u88c5\u548c\u4f7f\u7528\u65b9\u6cd5":2,"\u83b7\u5f97\u5f53\u524dmini":81,"\u83b7\u5f97\u7684\u503c\u7c7b\u578b\u5747\u4e3a":81,"\u83b7\u5f97\u8ba1\u7b97\u7ed3\u679c":90,"\u83b7\u5f97\u8fd9\u4e9b\u8282\u70b9\u7684ip\u5730\u5740":93,"\u83b7\u5f97head\u548cnode\u8282\u70b9\u7684ip\u5730\u5740":98,"\u865a\u62df\u673a\u4e0a":0,"\u867d\u7136\u4e0d\u9f13\u52b1\u8fd9\u6837":46,"\u867d\u7136\u5f02\u6b65sgd\u65b9\u5f0f\u4f1a\u63d0\u9ad8\u53c2\u6570\u66f4\u65b0\u5e76\u884c\u5ea6":92,"\u867d\u7136paddle\u770b\u8d77\u6765\u5305\u542b\u4e86\u4f17\u591a\u53c2\u6570":102,"\u884c":89,"\u884c\u504f\u79fb":89,"\u884c\u53f7":107,"\u8865\u5145\u4e0a\u6b21\u7684commit":72,"\u8868\u660e\u4e86\u8fd9\u4e9b\u884c\u7684\u6807\u53f7":74,"\u8868\u660e\u8fd9\u4e2a\u5c42\u7684\u4e00\u4e2a\u5b9e\u4f8b\u662f\u5426\u9700\u8981\u504f\u7f6e":74,"\u8868\u793a":75,"\u8868\u793a\u4e3adeviceid":105,"\u8868\u793a\u5bf9\u8f93\u5165\u6570\u636e":42,"\u8868\u793a\u5c06\u5916\u5c42\u7684outer_mem\u4f5c\u4e3a\u5185\u5c42memory\u7684\u521d\u59cb\u72b6\u6001":111,"\u8868\u793a\u5f53\u524d\u96c6\u7fa4\u4f5c\u4e1a\u7684\u8282\u70b9":93,"\u8868\u793a\u6570\u636e\u7c7b\u578b":75,"\u8868\u793a\u7684\u504f\u79fb\u662f\u4ee5":89,"\u8868\u793a\u8bbe\u5907\u7c7b\u578b":75,"\u8868\u793a\u8bcd\u8bed\u5728\u8bcd\u5178\u4e2d\u7684\u5e8f\u53f7":89,"\u8868\u793a\u8bfb\u8005\u6240\u4f7f\u7528\u7684docker\u955c\u50cf\u4ed3\u5e93\u5730\u5740":97,"\u8868\u793a\u8fd9\u4e2ajob\u7684\u540d\u5b57":97,"\u8868\u793a\u9700\u8981\u6784\u5efa\u63a8\u7406\u5e93":118,"\u88ab":72,"\u88ab\u5207\u5206\u6210\u591a\u4e2a\u90e8\u5206":92,"\u88ab\u6269\u5c55\u4e3a\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"\u88ab\u653e\u5728":74,"\u88ab\u79f0\u4e3a":114,"\u8981\u4f7f\u7528\u547d\u4ee4\u884c\u5206\u6790\u5de5\u5177":108,"\u8981\u5728\u5df2\u6709\u7684kubernetes\u96c6\u7fa4\u4e0a\u8fdb\u884cpaddlepaddle\u7684\u5206\u5e03\u5f0f\u8bad\u7ec3":97,"\u8981\u6c42\u5355\u5c42\u5e8f\u5217\u542b\u6709\u5143\u7d20\u7684\u6570\u76ee":110,"\u8981\u751f\u6210\u7684\u76ee\u6807\u5e8f\u5217":113,"\u8981\u8c03\u7528":74,"\u89c6\u9891\u7b49":89,"\u89e3\u51b3\u529e\u6cd5\u662f":78,"\u89e3\u51b3\u65b9\u6848\u662f":83,"\u89e3\u6790\u73af\u5883\u53d8\u91cf\u5f97\u5230":97,"\u89e3\u7801\u5668\u4f7f\u7528":114,"\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":114,"\u89e3\u7801\u5668\u662f\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":114,"\u89e3\u91ca\u578b\u8bed\u8a00\u53ea\u80fd\u8c03\u7528\u52a8\u6001\u5e93":45,"\u89e3\u91ca\u6027\u8bed\u8a00\u5b9e\u9645\u8fd0\u884c\u7684\u4e8c\u8fdb\u5236\u662f\u89e3\u91ca\u5668\u672c\u8eab":45,"\u8ba1\u5212\u5728":[41,42],"\u8ba1\u7b97":[92,114],"\u8ba1\u7b97\u504f\u7f6e\u7684\u68af\u5ea6":74,"\u8ba1\u7b97\u53cd\u5411rnn\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":114,"\u8ba1\u7b97\u53d8\u6362\u77e9\u9635\u7684\u5927\u5c0f\u548c\u683c\u5f0f":74,"\u8ba1\u7b97\u5f53\u524d\u5c42\u6743\u91cd\u7684\u68af\u5ea6":74,"\u8ba1\u7b97\u6548\u7387\u66f4\u9ad8":82,"\u8ba1\u7b97\u6bcf\u4e2a\u8bcd\u7684\u8bcd\u5411\u91cf":114,"\u8ba1\u7b97\u6fc0\u6d3b\u51fd\u6570\u7684\u68af\u5ea6":74,"\u8ba1\u7b97\u7684\u7ec6\u8282\u5c06\u5728\u4e0b\u9762\u7684\u5c0f\u8282\u7ed9\u51fa":74,"\u8ba1\u7b97\u8282\u70b9":92,"\u8ba1\u7b97\u8282\u70b9\u4e4b\u95f4\u4e5f\u4e0d\u4f1a\u76f8\u4e92\u4f9d\u8d56":92,"\u8ba1\u7b97\u8f6c\u6362\u77e9\u9635\u548c\u8f93\u5165\u7684\u68af\u5ea6":74,"\u8ba1\u7b97\u8f93\u5165\u548c\u53c2\u6570\u7684\u68af\u5ea6":74,"\u8ba1\u7b97\u8f93\u5165\u5c42\u7684\u504f\u5dee":74,"\u8ba1\u7b97\u8f93\u51fa":74,"\u8ba1\u7b97\u8fd9\u4e2a\u6587\u4ef6\u7684md5":11,"\u8ba1\u7b97\u96c6\u7fa4\u901a\u5e38\u7531\u4e00\u7ec4":101,"\u8ba1\u7b97\u9700\u8981\u7684\u6570\u636e\u5b58\u653e\u5728":76,"\u8ba9paddle\u6838\u5fc3\u4e2d":46,"\u8bad\u7ec3":102,"\u8bad\u7ec3\u4efb\u52a1\u7684\u8fd0\u884c\u53ef\u80fd\u4f1a\u5360\u6ee1trainer\u548cparamet":11,"\u8bad\u7ec3\u548c\u7eaf\u4f7f\u7528":62,"\u8bad\u7ec3\u5931\u8d25\u65f6\u53ef\u4ee5\u68c0\u67e5\u9519\u8bef\u65e5\u5fd7":93,"\u8bad\u7ec3\u597d\u4e00\u4e2a\u6df1\u5c42\u795e\u7ecf\u7f51\u7edc\u901a\u5e38\u8981\u8017\u8d39\u975e\u5e38\u957f\u7684\u65f6\u95f4":108,"\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u9ed8\u8ba4\u4fdd\u5b58\u5728\u5f53\u524d\u8fd0\u884c\u76ee\u5f55\u4e0b\u7684":90,"\u8bad\u7ec3\u6570\u636e\u6709\u95ee\u9898":81,"\u8bad\u7ec3\u6570\u636e\u683c\u5f0f\u548c\u8bad\u7ec3\u7a0b\u5e8f\u7684":91,"\u8bad\u7ec3\u65f6":97,"\u8bad\u7ec3\u6a21\u578b\u540e":114,"\u8bad\u7ec3\u6a21\u578b\u6b63\u786e\u6027":62,"\u8bad\u7ec3\u7a0b\u5e8f":91,"\u8bad\u7ec3\u7ed3\u675f\u540e\u67e5\u770b\u8f93\u51fa\u7ed3\u679c":97,"\u8bad\u7ec3\u8282\u70b9\u6570\u91cf":97,"\u8bad\u7ec3\u8bed\u8a00\u6a21\u578b\u8ddd\u79bb":81,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u53c2\u6570\u6216\u8005\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u7684\u68af\u5ea6\u5c3a\u5ea6\u8fc7\u5927":81,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6d4b\u8bd5test_period":102,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4e3a\u672c\u5730\u6a21\u5f0f":103,"\u8bad\u7ec3\u8fc7\u7a0b\u662f\u5426\u4f7f\u7528gpu":103,"\u8bad\u7ec3\u914d\u7f6e\u4e2d\u7684\u8bbe\u5907\u5c5e\u6027\u5c06\u4f1a\u65e0\u6548":103,"\u8bad\u7ec3dot_period":102,"\u8bb0\u5f55\u4e0b\u6240\u6709\u5931\u8d25\u7684\u4f8b\u5b50":62,"\u8bb0\u5fc6\u6a21\u5757":114,"\u8bbe\u4e3a\u5df2\u90e8\u7f72\u7684\u5de5\u4f5c\u7a7a\u95f4\u76ee\u5f55":93,"\u8bbe\u4e3a\u672c\u5730":93,"\u8bbe\u5b9a":82,"\u8bbe\u7f6e":[1,46,81,82,116,117],"\u8bbe\u7f6e\u4e3a":74,"\u8bbe\u7f6e\u4e3a\u4e0d\u540c\u7684\u503c":82,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528\u672c\u5730\u8bad\u7ec3\u6216\u8005\u4f7f\u7528\u96c6\u7fa4\u4e0a\u7684\u4e00\u4e2a\u8282\u70b9":103,"\u8bbe\u7f6e\u4e3atrue\u4f7f\u7528gpu\u6a21\u5f0f":103,"\u8bbe\u7f6e\u4e86\u76f8\u540c\u7684\u53d6\u503c":82,"\u8bbe\u7f6e\u5176\u53c2\u6570\u5c5e\u6027":83,"\u8bbe\u7f6e\u53c2\u6570\u7684\u540d\u5b57":83,"\u8bbe\u7f6e\u547d\u4ee4\u884c\u53c2\u6570":81,"\u8bbe\u7f6e\u5b66\u4e60\u7387\u8870\u51cf\u56e0\u5b50\u5206\u6bb5\u51fd\u6570":83,"\u8bbe\u7f6e\u5e8f\u5217\u4fe1\u606f\u7684\u63a5\u53e3":89,"\u8bbe\u7f6e\u6210":83,"\u8bbe\u7f6e\u6210\u4e00\u4e2a\u5c0f\u4e00\u4e9b\u7684\u503c":81,"\u8bbe\u7f6e\u8f93\u51fa\u7684\u5c3a\u5bf8":74,"\u8bbe\u7f6e\u8f93\u51fatensor\u7684\u5f62\u72b6":75,"\u8bbe\u7f6e\u9ed8\u8ba4\u8bbe\u5907\u53f7\u4e3a0":105,"\u8bbe\u7f6egpu":103,"\u8bbf\u95ee\u5bf9\u5e94\u7684\u7f51\u5740":107,"\u8bbf\u95eekubernetes\u7684\u63a5\u53e3\u6765\u67e5\u8be2\u6b64job\u5bf9\u5e94\u7684\u6240\u6709pod\u4fe1\u606f":97,"\u8bc4\u5ba1\u4eba\u4e00\u822c\u4e0d\u505a\u8bc4\u5ba1":72,"\u8bc4\u5ba1\u4eba\u7684\u6bcf\u4e2a\u610f\u89c1\u90fd\u5fc5\u987b\u56de\u590d":72,"\u8bc4\u5ba1\u4eba\u9700\u8981\u9010\u4e00\u67e5\u770b\u6bcf\u4e2acommit\u624d\u80fd\u77e5\u9053\u505a\u4e86\u54ea\u4e9b\u4fee\u6539":72,"\u8bc4\u8bba\u6846\u4e2d\u52a0\u4e0a":72,"\u8bc6\u522b\u6570\u5b57":62,"\u8bcd\u5411\u91cf":62,"\u8bd5\u7740\u8ba9\u8f93\u51fa\u7684\u5206\u6790\u6570\u636e\u548c\u7406\u8bba\u503c\u5bf9\u5e94":108,"\u8be5\u53c2\u6570\u5728\u7f51\u7edc\u914d\u7f6e\u7684output":103,"\u8be5\u53c2\u6570\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u81ea\u52a8\u8bbe\u7f6e":103,"\u8be5\u53c2\u6570\u5df2\u7ecf\u5728\u96c6\u7fa4\u63d0\u4ea4\u73af\u5883\u4e2d\u5b8c\u6210\u8bbe\u7f6e":103,"\u8be5\u53c2\u6570\u5fc5\u987b\u80fd\u88abflag":103,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u65e5\u5fd7\u622a\u65ad\u4fe1\u606f":103,"\u8be5\u53c2\u6570\u6307\u793a\u662f\u5426\u6253\u5370\u9519\u8bef\u622a\u65ad\u65e5\u5fd7":103,"\u8be5\u53c2\u6570\u7528\u4e8e\u6307\u5b9a\u52a8\u6001\u5e93\u8def\u5f84":103,"\u8be5\u53c2\u6570\u7684\u610f\u601d\u662f\u8bad\u7ec3num":103,"\u8be5\u53c2\u6570\u9ed8\u8ba4\u4e3anull":103,"\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":74,"\u8be5\u622a\u65ad\u4f1a\u5f71\u54cd":103,"\u8be5\u6279\u6b21\u7684\u8f93\u5165\u4e2d\u4ec5\u6709\u4e00\u4e2a\u5b50\u96c6\u662f\u975e\u96f6\u7684":74,"\u8be5\u63a5\u53e3\u53ef\u7528\u4e8e\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":1,"\u8be5\u63a5\u53e3\u63a5\u53d7\u4e24\u4e2a\u53c2\u6570":90,"\u8be5\u6570\u76ee\u662f\u63d0\u524d\u5b9a\u4e49\u597d\u7684":103,"\u8be5\u6587\u4ef6\u5bf9\u76f8\u5173gemm":41,"\u8be5\u65f6\u95f4\u53bb\u9664\u6389\u672c\u51fd\u6570\u8c03\u7528\u5176\u4ed6\u51fd\u6570\u7684\u65f6\u95f4":107,"\u8be5\u6a21\u578b\u7684\u8bf4\u660e\u5982\u4e0b\u56fe\u6240\u793a":114,"\u8be5\u7c7b\u7684":75,"\u8be5\u7c7b\u7684\u5b9e\u73b0\u7ec6\u8282\u5728":74,"\u8be5\u7c7b\u7ee7\u627f\u4e8epaddlepaddle\u7684\u57fa\u7c7b":42,"\u8be5\u811a\u672c\u4e2d\u8bb0\u5f55\u4e86\u4ea4\u53c9\u7f16\u8bd1android\u7248paddlepaddle\u5e93\u5e38\u7528\u7684cmake\u914d\u7f6e":116,"\u8be5\u8bed\u53e5\u4f1a\u4e3a\u6bcf\u4e2a\u5c42\u521d\u59cb\u5316\u5176\u6240\u9700\u8981\u7684\u53d8\u91cf\u548c\u8fde\u63a5":74,"\u8be5layer\u662f\u901a\u8fc7\u53c2\u6570":82,"\u8be6\u7ec6\u4ecb\u7ecd\u53ef\u4ee5\u53c2\u8003":111,"\u8be6\u7ec6\u4ecb\u7ecd\u8bf7\u53c2\u8003\u8bbe\u8ba1\u6587\u6863":75,"\u8be6\u7ec6\u4fe1\u606f\u8bf7\u68c0\u67e5":93,"\u8be6\u7ec6\u53c2\u8003":1,"\u8be6\u7ec6\u53ef\u53c2\u8003":72,"\u8be6\u7ec6\u6587\u6863\u53c2\u8003":81,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":1,"\u8be6\u7ec6\u89c1":110,"\u8be6\u7ec6\u8bbe\u8ba1":28,"\u8bed\u610f\u89d2\u8272\u6807\u6ce8":62,"\u8bed\u8a00\u91cd\u6784\u540e\u7684":107,"\u8bf4\u660e":[1,4,11,89],"\u8bf4\u660e\u63d0\u4ea4\u7684\u4ee3\u7801\u5b58\u5728\u95ee\u9898":72,"\u8bf4\u660e\u8fd9\u4e2a\u5c42\u7684\u8f93\u5165":74,"\u8bf7\u4e0d\u8981\u521b\u5efa\u7a7a\u7684":75,"\u8bf7\u4e0d\u8981\u5fd8\u8bb0\u63d0\u524d\u5728\u7269\u7406\u673a\u4e0a\u5b89\u88c5gpu\u6700\u65b0\u9a71\u52a8":2,"\u8bf7\u4fdd\u8bc1travi":72,"\u8bf7\u5148\u4f7f\u7528":[116,117,118],"\u8bf7\u53c2\u7167\u7f51\u7edc\u914d\u7f6e\u7684\u6587\u6863\u4e86\u89e3\u66f4\u8be6\u7ec6\u7684\u4fe1\u606f":105,"\u8bf7\u53c2\u8003":[46,74,75,78,81,84,90,111],"\u8bf7\u53c2\u8003\u6b64":90,"\u8bf7\u53c2\u89c1":72,"\u8bf7\u53c2\u9605":114,"\u8bf7\u5728\u8be5pull":72,"\u8bf7\u5728\u8f93\u5165\u65f6\u8fdb\u884c\u5408\u6cd5\u6027\u68c0\u67e5":89,"\u8bf7\u60a8\u6bcf\u6b21\u63d0\u4ea4\u4ee3\u7801\u65f6":72,"\u8bf7\u60a8\u9075\u5b88\u4ee5\u4e0b\u7ea6\u5b9a":72,"\u8bf7\u6307\u5b9a\u7684paddlepaddle\u5de5\u4f5c\u76ee\u5f55\u7ed9\u73af\u5883\u53d8\u91cf":77,"\u8bf7\u6307\u5b9a\u8be5\u76ee\u5f55":103,"\u8bf7\u663e\u793a\u5730\u8c03\u7528":75,"\u8bf7\u68c0\u67e5python\u7248\u672c\u662f\u5426\u4e3a2":4,"\u8bf7\u6ce8\u610f":[75,96,114],"\u8bf7\u6ce8\u610f\u662f\u5426\u9700\u8981\u4fee\u6539\u7f51\u7edc\u7ed3\u6784":90,"\u8bf7\u6ce8\u610f\u6bcf\u4e2acommit\u7684\u540d\u79f0":72,"\u8bf7\u6ce8\u610f\u8fd9\u4e2a\u547d\u4ee4\u7ed3\u5c3e\u5904\u7684":0,"\u8bf7\u6ce8\u610fcommit\u7684\u6570\u91cf":72,"\u8bf7\u76f4\u63a5\u586b\u51450":83,"\u8bf7\u770b\u4e0b\u9762\u7684\u4f8b\u5b50":105,"\u8bf7\u786e\u4fdd":72,"\u8bf7\u7ed9\u51fa\u603b\u4f53\u7684\u4fee\u6539\u60c5\u51b5":72,"\u8bf7\u7ed9\u51fa\u60a8\u81ea\u5df1\u7684\u53cd\u9a73\u7406\u7531":72,"\u8bf7\u9009\u62e9\u5408\u9002\u7684\u8bcd\u6c47":72,"\u8bf7\u9009\u62e9\u6b63\u786e\u7684\u7248\u672c":78,"\u8bf7\u9075\u5b88":72,"\u8bf7\u91c7\u7528":72,"\u8bf8\u5982\u56fe\u50cf\u5206\u7c7b":105,"\u8bfb\u53d6\u9700\u8981\u7684\u7ed3\u679c\u5373\u53ef":89,"\u8bfb\u53d6volume\u4e2d\u7684\u6570\u636e\u8fdb\u884c\u8fd9\u6b21\u5206\u5e03\u5f0f\u8bad\u7ec3":97,"\u8bfb\u8005\u53ef\u4ee5\u67e5\u770b":97,"\u8bfb\u8005\u9700\u8981\u66ff\u6362\u6210\u81ea\u5df1\u4f7f\u7528\u7684\u4ed3\u5e93\u5730\u5740":97,"\u8c03\u7528":[74,90,117],"\u8c03\u7528\u5bf9\u5e94":76,"\u8c03\u7528\u65b9\u6cd5\u89c1c":[116,117],"\u8c03\u7528\u7528":107,"\u8c03\u7528\u7684\u4e00\u4e9b\u7528\u6237\u5b9a\u4e49\u7684\u5e93\u51fd\u6570":91,"\u8c03\u7528\u7684\u51fd\u6570\u662f\u5426\u652f\u6301\u4e0d\u540c\u8bbe\u5907":75,"\u8c03\u7528\u8be5\u51fd\u6570\u540e":74,"\u8c03\u7528c":[89,90],"\u8d21\u732e\u6587\u6863":77,"\u8d77\u59cb\u5b58\u50a8\u5730\u5740\u4ee5\u6570\u636e\u7684\u5b58\u50a8\u5927\u5c0f\u4e3a\u5355\u4f4d\u7684\u504f\u79fb":89,"\u8df3\u8f6c\u5230":72,"\u8df3\u8fc7":81,"\u8f6c\u5316\u65b9\u6cd5\u5728\u76f8\u5e94\u7684\u9886\u57df\u90fd\u6709\u901a\u7528\u89e3\u51b3\u65b9\u6848":89,"\u8f6c\u6362\u5185\u5b58\u7684\u5de5\u4f5c":42,"\u8f6c\u6362\u5197\u4f59":41,"\u8f6c\u6362\u51fd\u6570":42,"\u8f6c\u6362\u751f\u6210\u7684\u6587\u4ef6\u540d\u4f1a\u662f\u4ee5\u4e0b\u683c\u5f0f":12,"\u8f6c\u6362\u8017\u65f6":41,"\u8f93\u5165":[88,90,110,114],"\u8f93\u5165\u4e86\u6027\u80fd\u5206\u6790\u7ed3\u679c":107,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u5355\u5c42\u5e8f\u5217":113,"\u8f93\u5165\u548c\u8f93\u51fa\u90fd\u662f\u53cc\u5c42\u5e8f\u5217":113,"\u8f93\u5165\u5e8f\u5217\u4e2d\u5143\u7d20\u7684\u603b\u6570":81,"\u8f93\u5165\u6570\u636e\u4e3a\u4e00\u4e2a\u5b8c\u6574\u7684\u65f6\u95f4\u5e8f\u5217":111,"\u8f93\u5165\u6570\u636e\u4e3a\u5728\u5355\u5c42rnn\u6570\u636e\u91cc\u9762":111,"\u8f93\u5165\u6570\u636e\u53ef\u5206\u4e3a":89,"\u8f93\u5165\u6570\u636e\u6574\u4f53\u4e0a\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":111,"\u8f93\u5165\u6570\u636e\u7684\u5b57\u5178\u7ef4\u6570\u662f1\u767e\u4e07":105,"\u8f93\u5165\u6570\u636e\u7c7b\u578b":89,"\u8f93\u5165\u662f\u5426\u662f\u8f6c\u7f6e\u7684":74,"\u8f93\u5165\u662f\u7531\u4e00\u4e2alist\u4e2d\u7684\u7f51\u7edc\u5c42\u5b9e\u4f8b\u7684\u540d\u5b57\u7ec4\u6210\u7684":74,"\u8f93\u5165\u68af\u5ea6":42,"\u8f93\u5165\u7684\u540d\u5b57":74,"\u8f93\u5165\u7684\u5927\u5c0f":74,"\u8f93\u5165\u7684\u7c7b\u578b":74,"\u8f93\u5165\u9700\u8981\u9884\u6d4b\u7684\u5411\u91cf\u7ec4":84,"\u8f93\u51fa":[75,90,110,114],"\u8f93\u51fa\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":113,"\u8f93\u51fa\u4e00\u4e2a\u53cc\u5c42\u5e8f\u5217":113,"\u8f93\u51fa\u4fe1\u606f\u6709\u673a\u5730\u7ec4\u7ec7\u5728\u4e00\u8d77":89,"\u8f93\u51fa\u51fd\u6570":114,"\u8f93\u51fa\u521b\u5efa":[89,90],"\u8f93\u51fa\u5e8f\u5217\u7684\u7c7b\u578b":110,"\u8f93\u51fa\u5e8f\u5217\u7684\u8bcd\u8bed\u6570\u548c\u8f93\u5165\u5e8f\u5217\u4e00\u81f4":113,"\u8f93\u51fa\u6240\u643a\u5e26\u7684\u5e8f\u5217\u4fe1\u606f":89,"\u8f93\u51fa\u6570\u636e\u548c\u8f93\u51fa\u68af\u5ea6":42,"\u8f93\u51fa\u6570\u636e\u548c\u8f93\u51fa\u68af\u5ea6\u7684\u8f6c\u6362":42,"\u8f93\u51fa\u6570\u636e\u662f\u5728\u4e0a\u6587\u4ecb\u7ecd\u7684":89,"\u8f93\u51fa\u6570\u636e\u6709\u673a\u5730\u7ec4\u7ec7\u5728\u4e00\u8d77":90,"\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7":[88,90],"\u8f93\u51fa\u7531":89,"\u8f93\u51fa\u7684\u5e8f\u5217\u4fe1\u606f":89,"\u8f93\u51fa\u7684\u68af\u5ea6":103,"\u8f93\u51fa\u7ed3\u679c\u53ef\u80fd\u4f1a\u968f\u7740\u5bb9\u5668\u7684\u6d88\u8017\u800c\u88ab\u5220\u9664":96,"\u8f93\u51fa\u88ab\u7ec4\u7ec7\u4e3a":89,"\u8f93\u51fa\u88ab\u7ec4\u7ec7\u4e3a\u4e00\u4e2a":89,"\u8f93\u51fa\u90fd\u4f1a\u5bf9\u5e94\u6709\u81ea\u5df1\u7684":[89,90],"\u8fbe\u5230\u5bb9\u707e\u7684\u76ee\u7684":11,"\u8fc7\u4e86\u4e00\u4e2a\u5f88\u7b80\u5355\u7684recurrent_group":111,"\u8fc7\u5b8c\u6240\u6709\u8bad\u7ec3\u6570\u636e\u5373\u4e3a\u4e00\u4e2apass":81,"\u8fc7\u7a0b\u4e2d\u6240\u6709\u65f6\u95f4\u6b65":41,"\u8fd0\u884c":90,"\u8fd0\u884c\u4e00\u4e2a":0,"\u8fd0\u884c\u4e0b\u9762\u547d\u4ee4\u53ef\u4ee5\u8fdb\u884c\u7f16\u8bd1":75,"\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5":0,"\u8fd0\u884c\u5355\u5143\u6d4b\u8bd5\u6d4b\u65f6\u9700\u8981\u7f16\u8bd1\u6574\u4e2a\u5de5\u7a0b":75,"\u8fd0\u884c\u5931\u8d25":105,"\u8fd0\u884c\u5b8c\u6210\u540e":93,"\u8fd0\u884c\u5b8c\u6bd5\u540e\u8f93\u51fa":107,"\u8fd0\u884c\u6027\u80fd\u5206\u6790\u7684\u65f6\u5019":107,"\u8fd0\u884c\u65e5\u5fd7":93,"\u8fd0\u884c\u65f6\u4e5f\u53ef\u80fd\u56e0\u4e3a\u591a\u7ebf\u7a0b\u4ea7\u751f\u6df7\u4e71\u4e0d\u53ef\u8bfb\u7684\u6027\u80fd\u5206\u6790\u7ed3\u679c":107,"\u8fd0\u884c\u65f6\u4f1a\u81ea\u52a8\u627e\u5230\u7cfb\u7edf\u4e2d\u5b89\u88c5\u7684cuda\u548ccudnn\u5e93\u8fdb\u884c\u7f16\u8bd1\u548c\u6267\u884c":1,"\u8fd0\u884c\u65f6c":90,"\u8fd0\u884c\u7684\u4e00\u4e9b\u53c2\u6570\u901a\u8fc7\u8fd9\u79cd\u65b9\u5f0f\u4f20\u9012\u5230\u5bb9\u5668\u5185":97,"\u8fd0\u884c\u8be5\u7f16\u8bd1\u5de5\u5177\u94fe\u9700\u8981\u4e00\u53f0":118,"\u8fd1\u671f\u76ee\u6807":42,"\u8fd4\u56de\u7684\u662f":84,"\u8fd4\u56de\u7b2c\u4e8c\u6b65":62,"\u8fd4\u56de\u7b2ci\u4e2a\u8f93\u5165\u77e9\u9635":74,"\u8fd4\u56depython\u7aef\u7684\u8ba1\u7b97\u7ed3\u679c":75,"\u8fd8\u4f1a\u4e0b\u8f7dmkl":1,"\u8fd8\u4f1a\u5f3a\u5236\u8bbe\u7f6e\u4e00\u4e9bpaddlepaddle\u53c2\u6570\u7684\u503c":116,"\u8fd8\u4f1a\u8f93\u51fa\u4e00\u4e2a":72,"\u8fd8\u53ef\u4ee5\u901a\u8fc7\u51cf\u5c0f\u5b66\u4e60\u7387\u6216\u8005\u5bf9\u6570\u636e\u8fdb\u884c\u5f52\u4e00\u5316\u5904\u7406\u6765\u89e3\u51b3\u8fd9\u7c7b\u95ee\u9898":81,"\u8fd8\u662f\u4ece":12,"\u8fd8\u662f\u865a\u62df\u673a":0,"\u8fd8\u9700\u8981\u5728\u8282\u70b9\u4e0a\u5b89\u88c5\u5bf9\u5e94\u7684gpu\u9a71\u52a8\u4ee5\u53cacuda":101,"\u8fd8\u9700\u8981\u91cd\u5199":75,"\u8fd9":81,"\u8fd98\u79cdlearning_rate_schedule\u53ca\u5176\u5bf9\u5e94\u5b66\u4e60\u7387\u8ba1\u7b97\u65b9\u5f0f\u5982\u4e0b":83,"\u8fd9\u4e00\u4e2a\u5e93":87,"\u8fd9\u4e00\u5757\u7684\u8017\u65f6\u6bd4\u4f8b\u771f\u7684\u592a\u9ad8":108,"\u8fd9\u4e00\u5c42\u8fdb\u884c\u5c01\u88c5":46,"\u8fd9\u4e00\u6570\u636e\u683c\u5f0f\u7684\u8f6c\u6362\u64cd\u4f5c":41,"\u8fd9\u4e00\u6982\u5ff5\u4e0d\u518d\u7410\u788e":46,"\u8fd9\u4e00\u8282\u5bf9\u56fe1\u4e2d\u9884\u6d4b\u4ee3\u7801\u7f16\u5199\u76845\u4e2a\u6b65\u9aa4\u8fdb\u884c\u4ecb\u7ecd\u548c\u8bf4\u660e":90,"\u8fd9\u4e00\u8ba1\u7b97\u5355\u5143":82,"\u8fd9\u4e00\u8fc7\u7a0b\u5bf9\u7528\u6237\u662f\u5b8c\u5168\u900f\u660e\u7684":113,"\u8fd9\u4e09\u4e2a\u5206\u652f":62,"\u8fd9\u4e24\u4e2a\u6307\u6807\u4ee3\u8868\u4e86\u67d0\u4e00\u4e2a\u51fd\u6570\u771f\u5b9e\u7684\u8fd0\u884c\u65f6\u95f4":107,"\u8fd9\u4e2a":[0,4],"\u8fd9\u4e2a\u4efb\u52a1\u7684\u914d\u7f6e\u4e3a":81,"\u8fd9\u4e2a\u4efb\u52a1\u7684dataprovider\u4e3a":81,"\u8fd9\u4e2a\u4f8b\u5b50\u6709\u4e24\u5904\u4e0d\u540c":75,"\u8fd9\u4e2a\u51fd\u6570\u672c\u8eab\u4f1a\u5728\u8ba1\u7b97\u524d\u5c06\u539f\u6570\u636e\u8f6c\u6362\u4e3a\u66f4\u9002\u5408\u82f1\u7279\u5c14\u5e73\u53f0\u7684\u5185\u90e8\u683c\u5f0f":41,"\u8fd9\u4e2a\u51fd\u6570\u7684":114,"\u8fd9\u4e2a\u51fd\u6570\u8fdb\u884c\u53d8\u6362":111,"\u8fd9\u4e2a\u51fd\u6570\u9700\u8981\u8bbe\u7f6e":114,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u7684\u8fde\u63a5\u53c2\u6570\u4e0epaddle\u7684\u5176\u4ed6\u4e8c\u8fdb\u5236":46,"\u8fd9\u4e2a\u53c2\u6570\u4e5f\u4e0d\u4f1a\u4e00\u5e76\u5220\u9664":46,"\u8fd9\u4e2a\u5730\u5740\u6765\u8868\u793a\u6b64\u6b65\u9aa4\u6240\u6784\u5efa\u51fa\u7684\u955c\u50cf":97,"\u8fd9\u4e2a\u57fa\u7c7b":74,"\u8fd9\u4e2a\u5934\u6587\u4ef6\u4e0d\u5047\u8bbe\u5176\u4ed6\u6587\u4ef6\u7684\u5f15\u7528\u987a\u5e8f":46,"\u8fd9\u4e2a\u5e8f\u5217\u7684\u6bcf\u4e2a\u5143\u7d20\u53c8\u662f\u4e00\u4e2a\u5e8f\u5217":113,"\u8fd9\u4e2a\u60c5\u51b5\u4e0b\u6240\u6709\u7684\u6587\u4ef6\u4f1a\u5b58\u5728\u6574\u7406\u8fc7\u7684\u7684\u6587\u4ef6\u76ee\u5f55":77,"\u8fd9\u4e2a\u63a5\u53e3\u9700\u8981\u505a\u5230":45,"\u8fd9\u4e2a\u6570\u636e\u4e5f\u88ab\u5355\u5c42rnn\u7f51\u7edc\u76f4\u63a5\u4f7f\u7528":111,"\u8fd9\u4e2a\u6587\u4ef6\u5177\u6709\u72ec\u7279\u7684\u8bed\u6cd5":45,"\u8fd9\u4e2a\u662f\u76ee\u524d\u63a8\u8350\u7684\u4f7f\u7528\u65b9\u6cd5":77,"\u8fd9\u4e2a\u73af\u5883\u53d8\u91cf\u5173\u95edopenmp\u4f18\u5316":107,"\u8fd9\u4e2a\u76ee\u5f55\u4e2d\u9664\u4e86":46,"\u8fd9\u4e2a\u793a\u4f8b":90,"\u8fd9\u4e2a\u795e\u7ecf\u7f51\u7edc\u5355\u5143\u5c31\u53ebmemori":111,"\u8fd9\u4e2a\u7c7b\u7684\u53c2\u6570\u5305\u62ec":74,"\u8fd9\u4e2a\u7c7b\u9700\u8981\u7ee7\u627f":74,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u4e2d\u7684\u53e6\u4e00\u4e2a\u9879\u76ee\u662f":46,"\u8fd9\u4e2a\u7ed3\u6784\u4f53\u5305\u542b\u4e24\u4e2a\u9879\u76ee":46,"\u8fd9\u4e2a\u811a\u672c\u8c03\u7528":0,"\u8fd9\u4e2a\u8f93\u5165\u4e0d\u53c2\u4e0e":75,"\u8fd9\u4e2a\u8fc7\u7a0b\u5bf9\u7528\u6237\u4e5f\u662f\u900f\u660e\u7684":113,"\u8fd9\u4e2a\u8fc7\u7a0b\u9664\u4e86\u7f16\u8bd1paddlepaddle\u4e3a":72,"\u8fd9\u4e2a\u9009\u62e9":[41,42],"\u8fd9\u4e2a\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u751f\u6210\u4e00\u7cfb\u5217\u6743\u91cd":114,"\u8fd9\u4e2a\u9759\u6001\u5e93\u5305\u542b\u4e86paddle\u7684\u5168\u90e8\u7b26\u53f7":46,"\u8fd9\u4e2ainstance\u53ef\u4ee5\u662f\u5355\u4e2a\u503c":12,"\u8fd9\u4e2aissu":0,"\u8fd9\u4e2ajob\u624d\u7b97\u6210\u529f\u7ed3\u675f":97,"\u8fd9\u4e2alayer\u7684\u8f93\u51fa\u4f1a\u4f5c\u4e3a\u6574\u4e2a":113,"\u8fd9\u4e5f\u4f1a\u6781\u5927\u51cf\u5c11\u6570\u636e\u8bfb\u5165\u7684\u8017\u65f6":81,"\u8fd9\u4e9b\u4f1a\u5728":[41,42],"\u8fd9\u4e9b\u4f8b\u5b50\u90fd\u53ef\u4ee5\u5728":94,"\u8fd9\u4e9b\u51fd\u6570\u4f1a\u5c06\u5bf9\u5e94\u5185\u5bb9\u6dfb\u52a0\u5230":75,"\u8fd9\u4e9b\u51fd\u6570\u4f1a\u6839\u636e\u8f93\u5165\u53c2\u6570\u91cd\u65b0\u8bbe\u7f6e\u5185\u90e8\u548c\u5916\u90e8\u5b58\u50a8":42,"\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":12,"\u8fd9\u4e9b\u53c2\u6570\u53ef\u4ee5\u901a\u8fc7":91,"\u8fd9\u4e9b\u53c2\u6570\u7684\u5177\u4f53\u63cf\u8ff0":97,"\u8fd9\u4e9b\u540d\u5b57\u5fc5\u987b\u8981\u5199\u5bf9":74,"\u8fd9\u4e9b\u6570\u636e\u4f1a\u88ab\u7528\u6765\u66f4\u65b0\u53c2\u6570":81,"\u8fd9\u4e9b\u6570\u636e\u4f7f\u7528\u7684\u5185\u5b58\u4e3b\u8981\u548c\u4e24\u4e2a\u53c2\u6570\u6709\u5173\u7cfb":81,"\u8fd9\u4e9b\u7279\u5f81\u6570\u636e\u4e4b\u95f4\u7684\u987a\u5e8f\u662f\u6709\u610f\u4e49\u7684":111,"\u8fd9\u4e9b\u955c\u50cf\u4e5f\u53ef\u4ee5\u4ece":62,"\u8fd9\u4efd\u6559\u7a0b\u5c55\u793a\u4e86\u5982\u4f55\u5728paddlepaddle\u4e2d\u5b9e\u73b0\u4e00\u4e2a\u81ea\u5b9a\u4e49\u7684\u7f51\u7edc\u5c42":74,"\u8fd9\u4f1a\u63d0\u793a\u5f53\u524d\u76ee\u5f55\u7684\u4e00\u4e9b\u53d8\u5316":72,"\u8fd9\u4f1a\u7ed9\u8bc4\u5ba1\u4eba\u5e26\u6765\u5f88\u5927\u56f0\u6270":72,"\u8fd9\u4f1a\u81ea\u52a8\u8fdb\u884c\u7f51\u7edc\u914d\u7f6e\u4e2d\u58f0\u660e\u7684\u6fc0\u6d3b\u64cd\u4f5c":74,"\u8fd9\u4fbf\u662f\u4e00\u79cd\u53cc\u5c42rnn\u7684\u8f93\u5165\u6570\u636e":111,"\u8fd9\u51e0\u4e2a\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":1,"\u8fd9\u53e5\u8868\u793a\u4f7f\u7528\u57fa\u7c7b":75,"\u8fd9\u53ef\u4ee5\u5e2e\u60a8\u7701\u6389\u82b1\u4e00\u5c0f\u65f6\u5b89\u88c5\u548c\u914d\u7f6e\u5404\u79cd\u5f00\u53d1\u5de5\u5177":0,"\u8fd9\u53ef\u4ee5\u8ba9\u5176\u4ed6\u4eba\u77e5\u9053\u8fd9\u6b21\u63d0\u4ea4\u505a\u4e86\u54ea\u4e9b\u6539\u53d8":72,"\u8fd9\u53ef\u4ee5\u901a\u8fc7":72,"\u8fd9\u548c\u5355\u5c42rnn\u7684\u914d\u7f6e\u662f\u7b49\u4ef7\u7684":111,"\u8fd9\u56db\u4e2a\u5e8f\u5217\u53c8\u5206\u522b\u542b\u67093":89,"\u8fd9\u56db\u6761\u6570\u636e\u540c\u65f6\u5904\u7406\u7684\u53e5\u5b50\u6570\u91cf\u4e3a":111,"\u8fd9\u5728\u6784\u9020\u975e\u5e38\u590d\u6742\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u65f6\u662f\u6709\u7528\u7684":114,"\u8fd9\u5bf9\u4e8e\u901a\u5e38\u7684java\u7684\u5f00\u53d1\u8005\u6765\u8bf4":45,"\u8fd9\u5c06\u4f1a\u5bfc\u81f4\u5355\u5143\u6d4b\u8bd5\u51fa\u9519":75,"\u8fd9\u5c06\u4f1a\u5bfc\u81f4\u7f16\u8bd1\u51fa\u9519":75,"\u8fd9\u610f\u5473\u7740":114,"\u8fd9\u610f\u5473\u7740\u9664\u4e86\u6307\u5b9adevic":105,"\u8fd9\u65f6":[81,90],"\u8fd9\u65f6\u5728\u4f7f\u7528":83,"\u8fd9\u65f6\u7684":89,"\u8fd9\u65f6\u7684\u9700\u8981\u540c\u65f6\u63d0\u4f9b":89,"\u8fd9\u65f6\u884c\u504f\u79fb\u548c\u5217\u53f7\u6307\u5b9a\u7684\u5143\u7d20\u9ed8\u8ba4\u5176\u503c\u4e3a1":89,"\u8fd9\u65f6\u8fdb\u884c\u77e9\u9635\u4e58\u6cd5\u8fd0\u7b97\u5c31\u53ef\u80fd\u5bfc\u81f4\u6d6e\u70b9\u6570\u6ea2\u51fa":81,"\u8fd9\u65f6\u9700\u8981\u8c03\u7528\u521b\u5efa\u5e8f\u5217\u4fe1\u606f\u548c\u4e3a":89,"\u8fd9\u662f\u4e00\u79cd\u6309\u5df2\u8bad\u7ec3\u6837\u672c\u6570\u5206\u6bb5\u53d6\u503c\u7684\u5b66\u4e60\u7387\u9000\u706b\u65b9\u6cd5":83,"\u8fd9\u662f\u4e00\u79cd\u6309\u5df2\u8bad\u7ec3pass\u6570\u5206\u6bb5\u53d6\u503c\u7684\u5b66\u4e60\u7387\u9000\u706b\u65b9\u6cd5":83,"\u8fd9\u662f\u4e00\u79cd\u975e\u5e38\u7075\u6d3b\u7684\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f":110,"\u8fd9\u662f\u56e0\u4e3a":45,"\u8fd9\u662f\u5f00\u6e90\u793e\u533a\u7684\u57fa\u672c\u793c\u8c8c":72,"\u8fd9\u662f\u666e\u901a\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217\u7684dataprovider\u4ee3\u7801":111,"\u8fd9\u662f\u76ee\u524dcmake\u5bfb\u627epython\u7684\u903b\u8f91\u5b58\u5728\u7f3a\u9677":78,"\u8fd9\u6837":[46,92],"\u8fd9\u6837\u4e0b\u4e00\u4e2acpu":42,"\u8fd9\u6837\u4fdd\u5b58\u5728\u5206\u5e03\u5f0f\u5b58\u50a8\u4e2d\u7684\u6570\u636e\u53ef\u4ee5\u88ab\u96c6\u7fa4\u4e2d\u7684\u6bcf\u4e2a\u8282\u70b9\u8bfb\u53d6\u5230":91,"\u8fd9\u6837\u4fdd\u8bc1":62,"\u8fd9\u6837\u4fdd\u8bc1\u8fd0\u884c\u7ed3\u675f\u4e4b\u540e\u7684":0,"\u8fd9\u6837\u505a\u53ef\u4ee5\u6781\u5927\u7684\u51cf\u5c11\u5185\u5b58\u5360\u7528":81,"\u8fd9\u6837\u53ef\u4ee5\u514d\u53bb\u5355\u72ec\u5b89\u88c5\u7f16\u8bd1\u4f9d\u8d56\u7684\u6b65\u9aa4":1,"\u8fd9\u6837\u53ef\u4ee5\u51cf\u5c0fgpu\u5185\u5b58":105,"\u8fd9\u6837\u5982\u679c\u9047\u5230\u95ee\u9898":0,"\u8fd9\u6837\u5bb9\u5668\u7684":97,"\u8fd9\u6837\u5c31\u53ef\u4ee5\u5728\u4e91\u7aef\u6267\u884c\u591a\u79cd\u6570\u636e\u7c7b\u8ba1\u7b97\u4efb\u52a1":12,"\u8fd9\u6837\u5df2\u7ecf\u4f20\u8f93\u6210\u529f\u7684\u90e8\u5206\u5c31\u4e0d\u7528\u91cd\u65b0\u4f20\u8f93\u4e86":28,"\u8fd9\u6837\u5e26\u6765\u7684\u597d\u5904\u5c31\u662f\u4e0d\u9700\u8981\u4e00\u76f4\u6e05\u7a7amemori":42,"\u8fd9\u6837\u5f53\u8be5pull":72,"\u8fd9\u6837\u65e2\u4f7f\u5f97\u6700\u7ec8\u4fdd\u5b58\u7684\u53c2\u6570\u683c\u5f0f\u4e0epaddlepaddle\u4e00\u81f4":42,"\u8fd9\u6837\u6781\u5927\u5730\u63d0\u9ad8\u4e86\u8ba1\u7b97\u7684\u5e76\u884c\u6027":92,"\u8fd9\u6837\u7684\u88c5\u9970\u5668":74,"\u8fd9\u6837\u7684\u8bdd":96,"\u8fd9\u6837\u8bad\u7ec3\u6587\u4ef6\u7684\u4e2a\u6570\u4f1a\u6bd4\u8f83\u591a":91,"\u8fd9\u6b63\u662f\u5b83\u4eec\u901f\u5ea6\u5feb\u7684\u539f\u56e0":108,"\u8fd9\u7528\u4e8e\u5728\u591a\u7ebf\u7a0b\u548c\u591a\u673a\u4e0a\u66f4\u65b0\u53c2\u6570":74,"\u8fd9\u79cd\u521d\u59cb\u5316\u65b9\u5f0f\u5728\u4e00\u822c\u60c5\u51b5\u4e0b\u4e0d\u4f1a\u4ea7\u751f\u5f88\u5dee\u7684\u7ed3\u679c":83,"\u8fd9\u79cd\u60c5\u51b5\u4e0b\u4e0d\u9700\u8981\u91cd\u5199\u8be5\u51fd\u6570":74,"\u8fd9\u79cd\u60c5\u51b5\u591a\u51fa\u73b0\u5728\u4f7f\u7528\u591a\u7ebf\u7a0b\u9884\u6d4b\u65f6":90,"\u8fd9\u79cd\u60c5\u51b5\u5e38\u5e38\u53d1\u751f\u5728":81,"\u8fd9\u79cd\u65b9\u5f0f\u5bf9\u5185\u5b58\u6d88\u8017\u8f83\u5927":82,"\u8fd9\u79cd\u65b9\u5f0f\u5fc5\u987b\u4f7f\u7528paddle\u5b58\u50a8\u7684\u6a21\u578b\u8def\u5f84\u683c\u5f0f":105,"\u8fd9\u79cd\u65b9\u5f0f\u6700\u4e3a\u7b80\u4fbf":87,"\u8fd9\u79cd\u751f\u6210\u6280\u672f\u53ea\u7528\u4e8e\u7c7b\u4f3c\u89e3\u7801\u5668\u7684\u751f\u6210\u8fc7\u7a0b":114,"\u8fd9\u79cd\u7c7b\u578b\u7684\u8f93\u5165\u5fc5\u987b\u901a\u8fc7":113,"\u8fd9\u79cd\u94fe\u63a5\u65b9\u5f0f\u4e3b\u8981\u7528\u4e8e\u79fb\u52a8\u7aef\u9884\u6d4b":87,"\u8fd9\u79cd\u96c6\u7fa4\u8282\u70b9\u7ba1\u7406\u65b9\u5f0f\u4f1a\u5728\u5c06\u6765\u4f7f\u7528":97,"\u8fd9\u7bc7":2,"\u8fd9\u7bc7\u6587\u6863":72,"\u8fd9\u7bc7\u6587\u6863\u4e4b\u540e\u90e8\u5206\u4f1a\u4f7f\u7528":90,"\u8fd9\u7bc7\u6587\u6863\u4e4b\u540e\u90e8\u5206\u4f1a\u7edf\u4e00\u4f7f\u7528":89,"\u8fd9\u7bc7\u6587\u6863\u4e4b\u540e\u90e8\u5206\u5c06\u4f1a\u7edf\u4e00\u4f7f\u7528":89,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd":90,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd\u5728":118,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd\u5728\u4f7f\u7528":89,"\u8fd9\u7bc7\u6587\u6863\u4ecb\u7ecd\u57fa\u4e8e":0,"\u8fd9\u7bc7\u6587\u6863\u7684\u4e4b\u540e\u90e8\u5206\u4f1a\u4f7f\u7528":90,"\u8fd9\u7bc7\u6587\u7ae0":0,"\u8fd9\u7ec4\u8bed\u4e49\u76f8\u540c\u7684\u793a\u4f8b\u914d\u7f6e\u5982\u4e0b":111,"\u8fd9\u884c\u547d\u4ee4\u4e2d":107,"\u8fd9\u901a\u8fc7\u83b7\u5f97\u53cd\u5411\u5faa\u73af\u7f51\u7edc\u7684\u7b2c\u4e00\u4e2a\u5b9e\u4f8b":114,"\u8fd9\u90fd\u9700\u8981\u8fd9\u4e2a\u63a5\u53e3\u6309\u7167\u7ea6\u5b9a\u4fd7\u6210\u7684\u89c4\u5219\u6765\u6ce8\u91ca\u5b8c\u5907":45,"\u8fd9\u91cc":[1,42,72,83,97,114],"\u8fd9\u91cc\u4e0d\u518d\u8d58\u8ff0":75,"\u8fd9\u91cc\u4ecb\u7ecdc":90,"\u8fd9\u91cc\u4f7f\u7528\u4e86\u7528":107,"\u8fd9\u91cc\u4f7f\u7528\u4e86paddlepaddle\u9884\u5b9a\u4e49\u597d\u7684rnn\u5904\u7406\u51fd\u6570":111,"\u8fd9\u91cc\u4f7f\u7528\u7b80\u5355\u7684":81,"\u8fd9\u91cc\u5c06\u4ecb\u7ecdpaddlepaddle\u7684\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":84,"\u8fd9\u91cc\u6211\u4eec\u5c55\u793a\u4e00\u4efd\u7b80\u5316\u8fc7\u7684\u4ee3\u7801":74,"\u8fd9\u91cc\u6211\u4eec\u901a\u8fc7\u5728kubernetes\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2ajob\u6765\u4e0b\u8f7d\u5e76\u5207\u5272\u6570\u636e":97,"\u8fd9\u91cc\u6709\u4e24\u79cd\u6709\u6548\u7684\u89e3\u51b3\u65b9\u6cd5":81,"\u8fd9\u91cc\u68c0\u9a8c\u8fd0\u884c\u65f6\u95f4\u6a21\u578b\u7684\u6536\u655b":93,"\u8fd9\u91cc\u7684dockerimage\u4f5c\u4e3a\u7f16\u8bd1\u73af\u5883\u4ee5\u652f\u6301\u66f4\u591a\u7684linux":62,"\u8fd9\u91cc\u7684eigentensor\u4e4b\u95f4\u7684\u8fd0\u7b97\u53ea\u662f\u6539\u53d8\u4e86\u539f\u6709tensor\u4e2d\u7684\u6570\u636e":76,"\u8fd9\u91cc\u9009\u62e90":62,"\u8fd9\u91cc\u9700\u8981\u7528\u6237\u989d\u5916\u6ce8\u610f":11,"\u8fd9\u9700\u8981\u8054\u5408\u6211\u4eec\u7b2c\u4e8c\u8282":107,"\u8fdb\u4e00\u6b65\u4f18\u5316":42,"\u8fdb\u4e3b\u4ed3\u5e93\u540e":72,"\u8fdb\u5165":62,"\u8fdb\u5165\u5bb9\u5668":96,"\u8fdb\u5165\u5bf9\u5e94\u7684\u76ee\u5f55":78,"\u8fdb\u7a0b\u542f\u52a8\u7684\u5fc5\u8981\u53c2\u6570":97,"\u8fdb\u7a0b\u7684":93,"\u8fdb\u7a0b\u7684\u542f\u52a8\u53c2\u6570":97,"\u8fdb\u7a0b\u7684\u8fd0\u884c\u73af\u5883":97,"\u8fdb\u7a0b\u9700\u8981\u7684":97,"\u8fdb\u800c\u591a\u673a":107,"\u8fdb\u800c\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u547d\u4ee4\u5f00\u542f\u4e00\u4e2ahttp\u670d\u52a1":107,"\u8fdb\u800c\u6307\u5b9a\u4e86python\u53ef\u6267\u884c\u6587\u4ef6\u7684\u8def\u5f84":107,"\u8fdb\u800c\u8fdb\u884c\u4ee3\u7801\u8bc4\u5ba1":62,"\u8fdb\u884c\u4e86":111,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6848":97,"\u8fdb\u884c\u5206\u5e03\u5f0f\u8bad\u7ec3\u7684\u65b9\u6cd5":97,"\u8fdb\u884c\u524d\u5411\u8ba1\u7b97":90,"\u8fdb\u884c\u56de\u590d":72,"\u8fdb\u884c\u5e8f\u5217\u5316":90,"\u8fdb\u884c\u5f00\u53d1":72,"\u8fdb\u884c\u62c6\u89e3":111,"\u8fdb\u884c\u6fc0\u6d3b\u64cd\u4f5c":74,"\u8fdb\u884c\u7f16\u8bd1\u548c\u5b89\u88c5":116,"\u8fdb\u884c\u8bad\u7ec3":90,"\u8fdb\u884c\u8bbe\u7f6e":75,"\u8fdb\u884c\u94fe\u63a5":87,"\u8fdb\u884c\u9884\u6d4b\u4f9d\u8d56\u4e8e\u5c06":87,"\u8fdb\u884c\u9884\u6d4b\u65f6":90,"\u8fdb\u884cpython\u4e0ec":107,"\u8fdb\u9636\u4f7f\u7528":115,"\u8fdb\u9636\u6307\u5357":84,"\u8fde\u63a5":113,"\u8fde\u63a5\u5230pserver\u7684\u7aef\u53e3":91,"\u8fde\u63a5\u5230pserver\u7684\u7aef\u53e3\u4e2a\u6570":91,"\u9000\u51fa\u5bb9\u5668":96,"\u9009\u62e9\u4e0b\u8f7d\u4f7f\u7528\u4e0d\u540c\u7684blas\u5e93\u7684docker\u955c\u50cf":2,"\u9009\u62e9\u662f\u5426\u7f16\u8bd1mkl":42,"\u9009\u62e9\u76ee\u6807\u5206\u652f":72,"\u9009\u62e9\u8def\u5f84\u6765\u52a8\u6001\u52a0\u8f7dnvidia":103,"\u9009\u62e9\u9700\u8981\u53d1\u5e03\u7684\u7248\u672c":62,"\u9009\u9879":[0,1],"\u900f\u4f20\u7528\u6237\u8eab\u4efd\u7684\u529e\u6cd5":28,"\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":102,"\u901a\u5e38":[46,75,107],"\u901a\u5e38\u4f1a\u4f7f\u7528\u73af\u5883\u53d8\u91cf\u914d\u7f6ejob\u7684\u914d\u7f6e\u4fe1\u606f":97,"\u901a\u5e38\u4f1a\u4f7f\u7528mapreduce\u4efb\u52a1\u7684\u8f93\u51fa\u7ed3\u679c\u4f5c\u4e3a\u8bad\u7ec3\u7ed3\u679c":91,"\u901a\u5e38\u4f7f\u7528\u7a00\u758f\u8bad\u7ec3\u6765\u52a0\u901f\u8ba1\u7b97\u8fc7\u7a0b":105,"\u901a\u5e38\u4f7f\u7528cento":4,"\u901a\u5e38\u505a\u6cd5\u662f\u4ece\u4e00\u4e2a\u6bd4\u8f83\u5927\u7684learning_rate\u5f00\u59cb\u8bd5":83,"\u901a\u5e38\u5305\u542b\u4e00\u4e2acpu\u7248\u672c\u548c\u4e00\u4e2agpu\u7248\u672c":62,"\u901a\u5e38\u540d\u5b57\u662f":72,"\u901a\u5e38\u60c5\u51b5\u4e0b":108,"\u901a\u5e38\u6211\u4eec\u4f1a\u5b89\u88c5ceph\u7b49\u5206\u5e03\u5f0f\u6587\u4ef6\u7cfb\u7edf\u6765\u5b58\u50a8\u8bad\u7ec3\u6570\u636e":96,"\u901a\u5e38\u6307\u5c06\u4e00\u4e2a\u6574\u4f53\u62c6\u5206\u6210\u591a\u4efd\u7684\u5176\u4e2d\u7684\u4e00\u4efd":11,"\u901a\u5e38\u6709\u4e24\u4e2a\u65b9\u6cd5\u6765\u6784\u5efa\u57fa\u4e8e":118,"\u901a\u5e38\u7528\u4e8e\u8868\u793a\u79bb\u6563\u7684\u7c7b\u522b\u6807\u7b7e":89,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u4f7f\u7528":114,"\u901a\u5e38\u7684\u505a\u6cd5\u662f\u5c06\u914d\u7f6e\u5b58\u4e8e":74,"\u901a\u5e38\u8981\u6c42\u65f6\u95f4\u6b65\u4e4b\u95f4\u5177\u6709\u4e00\u4e9b\u4f9d\u8d56\u6027":111,"\u901a\u5e38\u89c2\u5bdf\u70ed\u70b9\u51fd\u6570\u95f4\u7684\u8c03\u7528\u5173\u7cfb":107,"\u901a\u5e38\u90fd\u4f1a\u4f7f\u7528\u4e0b\u9762\u8fd9\u4e9b\u547d\u4ee4\u884c\u53c2\u6570":105,"\u901a\u5e38\u9700\u8981\u53bb\u6389\u7f51\u7edc\u4e2d\u7684":90,"\u901a\u7528":102,"\u901a\u8fc7":[72,74,75,81,89,111],"\u901a\u8fc7\u4e24\u4e2a\u5d4c\u5957\u7684":113,"\u901a\u8fc7\u4f7f\u7528":1,"\u901a\u8fc7\u4f7f\u7528\u8fd9\u4e9bapi":41,"\u901a\u8fc7\u51fd\u6570":97,"\u901a\u8fc7\u547d\u4ee4\u884c\u53c2\u6570":81,"\u901a\u8fc7\u591a\u4e2a\u7ebf\u7a0b\u5171\u4eab\u540c\u4e00\u4e2a\u6a21\u578b\u6765\u51cf\u5c11\u5185\u5b58\u5f00\u9500":90,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u7684\u8f93\u51fa":113,"\u901a\u8fc7\u5f15\u7528memory\u5f97\u5230\u8fd9\u4e2alayer\u4e0a\u4e00\u4e2a\u65f6\u523b\u8f93\u51fa":113,"\u901a\u8fc7\u6240\u6709\u5355\u5143\u6d4b\u8bd5":72,"\u901a\u8fc7\u6a21\u578b\u63a8\u65adapi\u7684\u5b9e\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u6837\u4f8b":46,"\u901a\u8fc7\u7075\u6d3b\u4f7f\u7528\u4ee5\u4e0a\u4e24\u4e2a\u63a5\u53e3":90,"\u901a\u8fc7\u7ec4\u5408\u4e0d\u540c\u7684layer":84,"\u901a\u8fc7\u7f51\u7edc\u5c42\u7684\u6807\u8bc6\u7b26\u6765\u6307\u5b9a":74,"\u901a\u8fc7\u8ba1\u7b97\u8282\u70b9\u548c\u53c2\u6570\u670d\u52a1\u5668\u7684\u5206\u5e03\u5f0f\u534f\u4f5c":92,"\u901a\u8fc7\u8be5\u53c2\u6570\u53ef\u83b7\u53d6\u5230\u8f93\u5165\u8f93\u51fa\u4ee5\u53ca\u5c5e\u6027":75,"\u901a\u8fc7\u8c03\u7528":[89,90],"\u901a\u8fc7\u8c03\u7528\u4ee5\u4e0b\u63a5\u53e3\u521b\u5efa\u7a00\u758f\u77e9\u9635":89,"\u901a\u8fc7data":113,"\u901a\u8fc7ssh\u7b49\u65b9\u5f0f\u767b\u5f55\u5230raspberri":118,"\u903b\u8f91\u4e0a\u9ad8\u4e8e\u4e8c\u7ef4\u7684\u6570\u636e":89,"\u903b\u8f91\u5212\u4e0a\u6587\u4ef6\u5206\u5757\u7684\u5355\u4f4d":28,"\u9047\u5230\u8be5\u9519\u8bef\u65f6":82,"\u9075\u5b88\u4ee5\u4e0b\u7ea6\u5b9a":72,"\u9075\u5faa\u4ee5\u4e0b\u6d41\u7a0b":62,"\u90a3\u4e48":[46,74,113],"\u90a3\u4e480\u5c42\u5e8f\u5217\u5373\u4e3a\u4e00\u4e2a\u8bcd\u8bed":113,"\u90a3\u4e48\u4f1a\u643a\u5e26\u6709":89,"\u90a3\u4e48\u53ef\u4ee5\u8ba4\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":83,"\u90a3\u4e48\u5728":75,"\u90a3\u4e48\u5982\u4f55\u5224\u65ad\u8bad\u7ec3\u4e0d\u6536\u655b\u5462":83,"\u90a3\u4e48\u5bf9\u5e94\u7684\u5185\u90e8\u5b58\u50a8\u4e5f\u4f1a\u4e0e\u5b83\u4eec\u5171\u4eab\u5185\u5b58":42,"\u90a3\u4e48\u5c31\u4f1a\u4f7f":42,"\u90a3\u4e48\u5e38\u6570\u8f93\u51fa\u6240\u80fd\u8fbe\u5230\u7684\u6700\u5c0fcost\u662f":83,"\u90a3\u4e48\u6211\u4eec\u4e5f\u5c31\u4e0d\u9700\u8981":0,"\u90a3\u4e48\u6211\u4eec\u53ef\u4ee5\u5224\u65ad\u4e3a\u8bad\u7ec3\u4e0d\u6536\u655b":83,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528":114,"\u90a3\u4e48\u63a8\u8350\u4f7f\u7528\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u65b9\u6cd5":114,"\u90a3\u4e48\u6536\u655b\u53ef\u80fd\u5f88\u6162":83,"\u90a3\u4e48\u6700\u597d\u5c06\u6570\u636e\u6587\u4ef6\u5728\u6bcf\u6b21\u8bfb\u53d6\u4e4b\u524d\u505a\u4e00\u6b21shuffl":81,"\u90a3\u4e48\u7528\u6237\u9700\u8981\u62c9\u53d6\u6240\u6709\u7684\u8fdc\u7a0b\u5206\u652f\u5230\u672c\u673a":78,"\u90a3\u4e48\u7f16\u8bd1\u8fc7\u7a0b\u53ea\u4f1a\u4ea7\u751f":0,"\u90a3\u4e48\u8bad\u7ec3\u6709\u53ef\u80fd\u4e0d\u6536\u655b":83,"\u90a3\u4e48\u8be5\u4f18\u5316\u7b97\u6cd5\u81f3\u5c11\u9700\u8981":81,"\u90a3\u4e48fc1\u548cfc2\u5c42\u5c06\u4f1a\u4f7f\u7528\u7b2c1\u4e2agpu\u6765\u8ba1\u7b97":105,"\u90a3\u4e5f\u5c31\u4e0d\u9700\u8981\u6025\u7740\u4f18\u5316\u6027\u80fd\u5566":108,"\u90a3\u4f30\u8ba1\u8fd9\u91cc\u7684\u6f5c\u529b\u5c31\u6ca1\u5565\u597d\u6316\u7684\u4e86":108,"\u90a3\u51cf\u5c11\u5b66\u4e60\u738710\u500d\u7ee7\u7eed\u8bd5\u9a8c":83,"\u90a3\u6211\u4f1a\u671f\u671b\u5206\u6790\u5de5\u5177\u7edf\u8ba1\u5230\u901f\u5ea6\u662f100gb":108,"\u90a3\u7a0b\u5e8f\u5206\u6790\u5de5\u5177\u662f\u5fc5\u4e0d\u53ef\u5c11\u7684\u5229\u5668":108,"\u90fd\u4e0d\u4f1a\u60f3\u8981\u77e5\u9053next":42,"\u90fd\u4e0d\u9700\u8981":0,"\u90fd\u4f1a\u4ea7\u751f\u5f53\u524d\u5c42\u72b6\u6001\u7684\u6240\u6709\u7ee7\u627f\u7ed3\u679c":103,"\u90fd\u4f1a\u7ba1\u7406\u7ef4\u62a4\u4e00\u4efd\u8bad\u7ec3\u597d\u7684\u6a21\u578b":90,"\u90fd\u4f1a\u9020\u6210\u8bad\u7ec3\u4e2d\u7684\u6570\u636e\u4ecec":81,"\u90fd\u4f7f\u7528":89,"\u90fd\u53ea\u662f\u4ecb\u7ecd\u53cc\u5c42rnn\u7684api\u63a5\u53e3":111,"\u90fd\u53ef\u4ee5\u8fd0\u884c":0,"\u90fd\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528\u4e0b\u9762\u7684\u63a5\u53e3\u4e3a\u539f\u6709\u7684\u6570\u636e\u8f93\u5165\u9644\u52a0\u4e0a\u5e8f\u5217\u4fe1\u606f":89,"\u90fd\u5e94\u4f7f\u7528c":89,"\u90fd\u662f\u4e94\u4f4d\u7684\u6570\u5b57":12,"\u90fd\u662f\u4ee5ext\u5f00\u5934":42,"\u90fd\u662f\u5bf9layer1\u5143\u7d20\u7684\u62f7\u8d1d":110,"\u90fd\u662f\u5c06\u6bcf\u4e00\u53e5\u5206\u597d\u8bcd\u540e\u7684\u53e5\u5b50":111,"\u90fd\u662fabi\u8c03\u7528\u6807\u51c6\u7684":45,"\u90fd\u7528":72,"\u90fd\u7ee7\u627f\u4e8epaddlepaddle\u7684\u57fa\u7c7b":41,"\u90fd\u9700\u8981\u5199\u63d0\u4ea4\u8bf4\u660e":72,"\u90fd\u9700\u8981\u8c03\u7528\u4e00\u6b21":74,"\u914d\u5236\u7f16\u8bd1\u9009\u9879":87,"\u914d\u7f6e\u6253\u5f00":108,"\u914d\u7f6e\u6587\u4ef6\u63a5\u53e3\u662ffc_layer":74,"\u914d\u7f6e\u6587\u4ef6\u91cc\u52a0\u4e24\u884c":0,"\u914d\u7f6e\u7684\u65b9\u6cd5\u53c2\u8003":28,"\u914d\u7f6e\u7b80\u5355\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u7684\u4f8b\u5b50":114,"\u914d\u7f6e\u7f51\u7edc\u5c42\u7684\u8f93\u5165":74,"\u914d\u7f6eapi":110,"\u91c7\u7528\u5747\u5300\u5206\u5e03\u6216\u8005\u9ad8\u65af\u5206\u5e03\u521d\u59cb\u5316":103,"\u91c7\u7528multi":83,"\u91ca\u653e\u5bf9paramters\u5185\u5b58\u7684\u9501\u5b9a":11,"\u91cc":0,"\u91cc\u53ef\u4ee5\u6807\u51c6\u5316\u7f16\u8bd1\u73af\u5883":0,"\u91cc\u5b8c\u6210":75,"\u91cc\u6240\u6709\u7684\u7b26\u53f7\u90fd\u5199\u5165\u81ea\u5df1\u7684\u7a0b\u5e8f\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cc":45,"\u91cc\u7684":0,"\u91cc\u7684\u65e5\u5fd7":93,"\u91cc\u7684\u6e90\u7801":0,"\u91cc\u8fd0\u884c\u7684\u7f16\u8bd1\u5de5\u5177\u5b9e\u9645\u4e0a\u90fd\u662f\u5728\u672c\u673a\u7684":0,"\u91cc\u9009\u62e9\u9700\u8981\u53d1\u5e03\u7684\u5206\u652f":62,"\u91cc\u9762":75,"\u91cc\u9762\u6db5\u76d6\u4e86\u4ea4\u53c9\u7f16\u8bd1android\u7248paddlepaddle\u5e93\u9700\u8981\u7684\u6240\u6709\u7f16\u8bd1\u5de5\u5177":116,"\u91cc\u9762\u6dfb\u52a0":42,"\u91ccstep\u7684\u5185\u5bb9":81,"\u91cd\u5199\u7236\u7c7blayer\u7684":42,"\u91cd\u547d\u540d\u6210":45,"\u91cd\u65b0\u7f16\u8bd1paddlepaddl":108,"\u9488\u5bf9\u4e0d\u540c\u7684":117,"\u9488\u5bf9\u4efb\u52a1\u8fd0\u884c\u5b8c\u6210\u540e\u5bb9\u5668\u81ea\u52a8\u9000\u51fa\u7684\u573a\u666f":96,"\u9488\u5bf9\u5185\u5b58\u548c\u663e\u5b58":81,"\u94fe\u63a5":87,"\u94fe\u63a5\u4e2d\u627e\u5230":4,"\u94fe\u63a5\u4f55\u79cdblas\u5e93\u7b49":1,"\u94fe\u63a5\u5230\u81ea\u5df1\u7684\u7a0b\u5e8f\u91cc":45,"\u94fe\u63a5\u76f8\u5bf9\u5bb9\u6613":87,"\u94fe\u63a5\u9009\u9879":87,"\u94fe\u63a5\u9759\u6001\u5e93":87,"\u9519\u8bef":78,"\u9519\u8bef\u5904\u7406":45,"\u9519\u8bef\u5904\u7406\u65b9\u5f0f\u662f\u8fd4\u56de\u503c":45,"\u9519\u8bef\u5904\u7406\u7684\u65b9\u5f0f\u4e5f\u4e0d\u5c3d\u76f8\u540c":45,"\u9519\u8bef\u7684define_py_data_sources2\u7c7b\u4f3c":83,"\u952e\u6765\u542f\u52a8\u7f16\u8bd1\u4e86":0,"\u955c\u50cf\u91cc\u6709paddlepaddle\u7684\u6e90\u7801\u4e0edemo":96,"\u957f\u5ea6":81,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u5355\u6b65\u51fd\u6570\u548c\u8f93\u51fa\u51fd\u6570":114,"\u95e8\u63a7\u5faa\u73af\u5355\u5143\u7684\u8f93\u51fa\u88ab\u7528\u4f5c\u8f93\u51famemori":114,"\u9644\u52a0\u4e0a\u5e8f\u5217\u4fe1\u606f":89,"\u9650\u5236\u5957\u63a5\u5b57\u53d1\u9001\u7f13\u51b2\u533a\u7684\u5927\u5c0f":103,"\u9650\u5236\u5957\u63a5\u5b57\u63a5\u6536\u7f13\u51b2\u533a\u7684\u5927\u5c0f":103,"\u9664\u4e86\u53ef\u4ee5\u81ea\u52a8\u7f16\u8bd1\u6587\u6863":77,"\u9664\u4e86boot_lay":111,"\u9664\u6784\u9020\u67d0\u79cd\u7c7b\u578b\u7684\u51fd\u6570":46,"\u9664\u6b64\u4e4b\u5916":81,"\u9664\u96f6\u7b49\u95ee\u9898":81,"\u968f\u540e\u53ef\u4ee5\u7528\u8fd9\u4e2a\u5f00\u53d1\u955c\u50cf\u5f00\u59cbbuild":72,"\u968f\u673a\u6570\u7684\u79cd\u5b50":103,"\u968f\u673a\u6570seed":102,"\u9694\u5f00":91,"\u96c6\u6210\u5230":41,"\u96c6\u6210\u5230paddlepaddl":42,"\u96c6\u675f\u641c\u7d22\u4f7f\u7528\u5e7f\u5ea6\u4f18\u5148\u641c\u7d22\u7684\u65b9\u5f0f\u6784\u5efa\u67e5\u627e\u6811":103,"\u96c6\u7fa4\u4e0a\u542f\u52a8\u4e00\u4e2a\u5355\u673a\u4f7f\u7528cpu\u7684paddlepaddle\u8bad\u7ec3\u4f5c\u4e1a":96,"\u96c6\u7fa4\u4e2d\u7684\u6bcf\u53f0\u8ba1\u7b97\u673a\u901a\u5e38\u88ab\u6210\u4e3a\u4e00\u4e2a":101,"\u96c6\u7fa4\u4efb\u52a1":93,"\u96c6\u7fa4\u4f5c\u4e1a\u5c06\u4f1a\u5728\u51e0\u79d2\u540e\u542f\u52a8":93,"\u96c6\u7fa4\u6d4b\u8bd5":102,"\u96c6\u7fa4\u7ba1\u7406\u5de5\u5177":94,"\u96c6\u7fa4\u8bad\u7ec3":102,"\u96c6\u7fa4\u8bad\u7ec3\u4e0e\u9884\u6d4b":80,"\u96c6\u7fa4\u8fdb\u7a0b":93,"\u9700\u52a0\u8be5\u6a21\u677f\u53c2\u6570":75,"\u9700\u5728nvvp\u754c\u9762\u4e2d\u9009\u4e0a\u624d\u80fd\u5f00\u542f":108,"\u9700\u6307\u5b9a":87,"\u9700\u63d0\u4f9b\u975e\u96f6\u5143\u7684\u503c":89,"\u9700\u6ce8\u610f":87,"\u9700\u8981":[0,12,75,90],"\u9700\u8981\u4e3a":75,"\u9700\u8981\u4f7f\u7528":81,"\u9700\u8981\u4f7f\u7528\u5176\u5236\u5b9a\u7684\u65b9\u5f0f\u6302\u8f7d\u540e\u5e76\u5bfc\u5165\u6570\u636e":97,"\u9700\u8981\u4f7f\u7528\u6700\u65b0\u7684pip":4,"\u9700\u8981\u4f7f\u7528\u8005\u81ea\u5df1\u4e86\u89e3\u5e76\u5b8c\u6210\u8f6c\u5316":89,"\u9700\u8981\u4fdd\u6301\u5f53\u524d\u5206\u652f\u76ee\u5f55":72,"\u9700\u8981\u4fee\u6539build":62,"\u9700\u8981\u521b\u5efa\u5e76\u586b\u5199":89,"\u9700\u8981\u5347\u7ea7pip\u7248\u672c\u5230\u6700\u65b0":[4,78],"\u9700\u8981\u5355\u72ec":2,"\u9700\u8981\u53ef\u4ee5\u8de8\u5e73\u53f0\u6267\u884c":28,"\u9700\u8981\u540c\u65f6\u63d0\u4f9b\u6bcf\u4e00\u4e2a\u5185\u5c42\u5e8f\u5217\u5728\u6574\u4e2a":89,"\u9700\u8981\u540c\u6b65\u539f\u4ed3\u5e93":72,"\u9700\u8981\u542f\u52a8\u7684\u8282\u70b9\u4e2a\u6570\u4ee5\u53ca":97,"\u9700\u8981\u548c\u8be5op\u7684\u540d\u5b57\u4e00\u6837":75,"\u9700\u8981\u54ea\u4e9b\u5c42\u7684\u8ba1\u7b97\u7ed3\u679c\u4f5c\u4e3a\u8f93\u51fa":90,"\u9700\u8981\u5728":75,"\u9700\u8981\u5728\u521b\u5efa\u5bb9\u5668\u524d\u6302\u8f7d\u5377\u4ee5\u4fbf\u6211\u4eec\u4fdd\u5b58\u8bad\u7ec3\u7ed3\u679c":96,"\u9700\u8981\u5728\u7cfb\u7edf\u91cc\u5148\u5b89\u88c5\u597ddocker\u5de5\u5177\u5305":77,"\u9700\u8981\u5728cmake\u7684\u65f6\u5019":46,"\u9700\u8981\u5728macos\u7cfb\u7edf\u4e0a\u8fdb\u884c":117,"\u9700\u8981\u5c06\u5176parameter\u8bbe\u7f6e\u6210":81,"\u9700\u8981\u5c06\u7f51\u7edc\u7ed3\u6784\u4f7f\u7528":90,"\u9700\u8981\u5c06bugfix\u7684\u5206\u652f\u540c\u65f6merge\u5230":62,"\u9700\u8981\u5c06cuda\u76f8\u5173\u7684\u5e93\u8bbe\u7f6e\u5230":87,"\u9700\u8981\u5c06paddl":87,"\u9700\u8981\u5f15\u7528":46,"\u9700\u8981\u5f3a\u8c03\u7684\u662f":0,"\u9700\u8981\u601d\u8003\u5b8c\u6210\u4ee5\u4e0b\u5de5\u4f5c":[89,90],"\u9700\u8981\u624b\u52a8\u8fdb\u884c\u89e3\u538b":90,"\u9700\u8981\u6267\u884c":[1,4,86],"\u9700\u8981\u6307\u5b9a":87,"\u9700\u8981\u6307\u5b9a\u4e0e\u67d0\u4e00\u4e2a\u8f93\u5165\u7684\u5e8f\u5217\u4fe1\u606f\u662f\u4e00\u81f4\u7684":111,"\u9700\u8981\u6307\u5b9alayer\u7684\u8f93\u5165\u6765\u6e90":84,"\u9700\u8981\u660e\u786e\u6307\u5b9a":103,"\u9700\u8981\u663e\u5f0f\u5730\u94fe\u63a5":87,"\u9700\u8981\u663e\u793a\u5730\u94fe\u63a5":87,"\u9700\u8981\u663e\u793a\u5730\u94fe\u63a5mkl\u7684\u52a8\u6001\u5e93":87,"\u9700\u8981\u6709\u7a33\u5b9a\u7684\u5bfc\u51fa\u7b26\u53f7":45,"\u9700\u8981\u6839\u636e\u4e0d\u540c\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u6765\u7ed1\u5b9a\u4e00\u4e2a":97,"\u9700\u8981\u6ce8\u610f":75,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[42,62,81,103],"\u9700\u8981\u6ce8\u610f\u7684\u662f\u68af\u5ea6\u68c0\u67e5\u4ec5\u4ec5\u9a8c\u8bc1\u4e86\u68af\u5ea6\u7684\u8ba1\u7b97":74,"\u9700\u8981\u6ce8\u610f\u7684\u662fpaddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5b50\u5e8f\u5217\u6570\u76ee\u4e00\u6837\u7684\u591a\u8f93\u5165\u53cc\u5c42rnn":111,"\u9700\u8981\u7528\u5230\u7684\u7f16\u8bd1\u5de5\u5177\u548c\u7cfb\u7edf\u5e93":116,"\u9700\u8981\u7528\u6237\u663e\u5f0f\u8bbe\u5b9a":82,"\u9700\u8981\u7d2f\u52a0\u4e0d\u540clayer\u4f20\u8fc7\u6765\u7684\u68af\u5ea6":42,"\u9700\u8981\u81ea\u5df1\u94fe\u63a5mkl\u94fe\u63a5\u5e93":87,"\u9700\u8981\u88ab\u66b4\u9732\u5230\u5176\u4ed6\u8bed\u8a00":46,"\u9700\u8981\u8bf7\u7ba1\u7406\u5458\u5b89\u88c5\u548c\u914d\u7f6e\u597d":0,"\u9700\u8981\u9075\u5faa\u4ee5\u4e0b\u7ea6\u5b9a":113,"\u9700\u8981\u91cd\u547d\u540dwheel\u5305\u4e2dplatform\u76f8\u5173\u7684\u540e\u7f00":62,"\u9700\u8981\u989d\u5916\u6ce8\u610f\u7684\u662f":76,"\u9700\u9644\u52a0\u53cc\u5c42\u5e8f\u5217\u4fe1\u606f":89,"\u9700\u9644\u52a0\u5e8f\u5217\u4fe1\u606f":89,"\u975e\u5e38\u6570":74,"\u975e\u5e8f\u5217\u8f93\u5165\u4e0d\u643a\u5e26":89,"\u975e\u5e8f\u5217\u8f93\u5165\u65e0\u9700\u6784\u9020":89,"\u975e\u96f6\u5143\u4e2a\u6570":89,"\u975e\u96f6\u5143\u7d20\u7684\u503c":89,"\u975e\u96f6\u5143\u7d20\u7684\u5217\u53f7":89,"\u975e\u96f6\u6570\u5b57\u7684\u4e2a\u6570":74,"\u9762\u5411\u67b6\u6784\u4e3a32\u4f4darm\u67b6\u6784":116,"\u9762\u5411\u67b6\u6784\u4e3a64\u4f4darm64\u67b6\u6784":116,"\u9879\u76ee\u5728\u52aa\u529b\u5f00\u59cb\u652f\u6301\u5176\u4ed6\u4e0d\u9700\u8981":0,"\u987a\u5e8f":111,"\u9884\u63d0\u4ea4\u94a9\u5b50":72,"\u9884\u6d4b\u4ee3\u7801\u66f4\u591a\u8be6\u7ec6\u793a\u4f8b\u4ee3\u7801\u8bf7\u53c2\u8003":90,"\u9884\u6d4b\u4f7f\u7528\u7684\u7f51\u7edc\u7ed3\u6784\u5f80\u5f80\u4e0d\u540c\u4e8e\u8bad\u7ec3":90,"\u9884\u6d4b\u65f6":90,"\u9884\u6d4b\u65f6\u53ea\u9700\u52a0\u8f7d\u4e00\u4e2a\u6587\u4ef6\u4fbf\u4e8e\u53d1\u5e03":90,"\u9884\u6d4b\u7a0b\u5e8f\u5f00\u53d1\u4e24\u5927\u90e8\u5206":90,"\u9996\u5148":[74,111,114],"\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":1,"\u9996\u5148\u5b89\u88c5\u5e76\u5728\u5f53\u524d\u76ee\u5f55\u8fd0\u884c\u5b83":72,"\u9996\u5148\u5b9a\u4e49":75,"\u9996\u5148\u5bf9\u8f93\u5165\u505a\u4e00\u4e2a\u5c0f\u7684\u6270\u52a8":74,"\u9996\u5148\u6211\u4eec\u9700\u8981\u63a8\u5bfc\u8be5\u7f51\u7edc\u5c42\u7684":74,"\u9996\u5148\u6784\u9020\u5934\u4fe1\u606f":83,"\u9996\u5148\u901a\u8fc7":72,"\u9996\u5148\u9700\u8981\u52a0\u8f7d\u76f8\u5e94\u7684python\u5e93":84,"\u9a71\u52a8":77,"\u9ad8\u4eae\u90e8\u5206":111,"\u9ad8\u5ea6":89,"\u9ad8\u5ea6\u652f\u6301\u7075\u6d3b\u548c\u9ad8\u6548\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u914d\u7f6e":114,"\u9ad8\u65af\u5206\u5e03":83,"\u9ed8\u8ba4":103,"\u9ed8\u8ba40":91,"\u9ed8\u8ba41":91,"\u9ed8\u8ba4127":91,"\u9ed8\u8ba4256k":28,"\u9ed8\u8ba47164":91,"\u9ed8\u8ba4\u4e0d\u663e\u793a":103,"\u9ed8\u8ba4\u4e0d\u8bbe\u7f6e":113,"\u9ed8\u8ba4\u4e3a0":[103,105],"\u9ed8\u8ba4\u4e3a1":[89,105],"\u9ed8\u8ba4\u4e3a100":105,"\u9ed8\u8ba4\u4e3a4096mb":103,"\u9ed8\u8ba4\u4e3a\u7b2c\u4e00\u4e2a\u8f93\u5165":113,"\u9ed8\u8ba4\u4e3anull":103,"\u9ed8\u8ba4\u4f1a\u5c06a\u548cb":81,"\u9ed8\u8ba4\u4f7f\u7528concurrentremoteparameterupdat":103,"\u9ed8\u8ba4\u4f7f\u7528mkl":1,"\u9ed8\u8ba4\u503c":[1,105,110,117],"\u9ed8\u8ba4\u503c\u4e3a":[116,117,118],"\u9ed8\u8ba4\u503c\u4e3a\u73af\u5883\u53d8\u91cf":117,"\u9ed8\u8ba4\u521d\u59cb\u72b6\u4e3a0":113,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b":[83,93],"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u6309\u7167float\u7cbe\u5ea6\u8ba1\u7b97":83,"\u9ed8\u8ba4\u6307\u5b9a\u7b2c\u4e00\u4e2a\u8f93\u5165":111,"\u9ed8\u8ba4\u662f\u4f7f\u7528mkl\u7684\u955c\u50cf":2,"\u9ed8\u8ba4\u6ca1\u6709\u5b89\u88c5vim":2,"\u9ed8\u8ba4\u7684paddlepaddle\u751f\u4ea7\u73af\u5883\u955c\u50cf":96,"\u9ed8\u8ba4\u7f16\u8bd1\u6240\u6709\u67b6\u6784":117,"\u9ed8\u8ba4\u8bbe\u7f6e\u4e3a":41,"\u9ed8\u8ba4\u8bbe\u7f6e\u4e3a\u771f":105,"\u9ed8\u8ba4\u8bbe\u7f6e\u6210\u73af\u5883\u53d8\u91cf":[116,118],"\u9ed8\u8ba4\u8c03\u7528":0,"\u9ed8\u8ba4fals":91,"abi\u7684paddlepaddle\u5e93":116,"abstract":[19,20,27,31,52,61,63],"android\u5b98\u65b9\u63d0\u4f9b\u7684":116,"android\u5e73\u53f0\u4e0a\u4f7f\u7528\u7684c":116,"android\u5e73\u53f0\u53ef\u9009\u914d\u7f6e\u53c2\u6570":116,"android\u7684docker\u5f00\u53d1\u955c\u50cf\u5411\u7528\u6237\u63d0\u4f9b\u4e24\u4e2a\u53ef\u914d\u7f6e\u7684\u53c2\u6570":116,"api\u4e0d\u4f1a\u76f4\u63a5\u52a0\u8f7d":90,"api\u4e0d\u5c0f\u4e8e21":116,"api\u4e2d":89,"api\u4e2d\u4f7f\u7528":45,"api\u4e2d\u7684\u4e00\u7ef4\u6570\u7ec4":89,"api\u4e2d\u7684\u77e9\u9635\u6765\u8868\u793a":89,"api\u4e2d\u795e\u7ecf\u7f51\u7edc\u7684\u4e00\u4e2a\u8f93\u5165":90,"api\u4f7f\u7528\u4e2d\u7684\u4e00\u4e2a\u91cd\u8981\u6982\u5ff5":90,"api\u4f7f\u7528\u6d41\u7a0b":88,"api\u4f7f\u7528\u6d41\u7a0b\u793a\u610f\u56fe":90,"api\u4f7f\u7528\u793a\u4f8b":90,"api\u521b\u5efa\u7684gradientmachine\u7c7b\u7684\u5bf9\u8c61":90,"api\u53ef\u4ee5\u901a\u8fc7\u5206\u522b\u6307\u5b9a\u5e8f\u5217\u5316\u540e\u7684\u7f51\u7edc\u7ed3\u6784\u6587\u4ef6\u548c\u53c2\u6570\u76ee\u5f55\u6765\u52a0\u8f7d\u8bad\u7ec3\u597d\u7684\u6a21\u578b":90,"api\u53ef\u4ee5\u901a\u8fc7\u6307\u5b9a":90,"api\u5b8c\u6210\u5206\u5e03\u5f0f\u8bad\u7ec3":91,"api\u5bf9\u6bd4\u4ecb\u7ecd":112,"api\u5bfc\u51fa\u7684\u52a8\u6001\u5e93":46,"api\u5bfc\u51fa\u7684\u9759\u6001\u5e93":46,"api\u5e93\u5c06\u88ab\u5b89\u88c5\u5230":116,"api\u5f00\u53d1\u5305\u5e76\u5b89\u88c5":4,"api\u5f00\u53d1\u9884\u6d4b\u7a0b\u5e8f\u65f6":87,"api\u5f00\u53d1\u9884\u6d4b\u7a0b\u5e8f\u9700\u8981\u4e00\u4e2a\u8bad\u7ec3\u597d\u7684\u6a21\u578b":90,"api\u6240\u9700\u7684\u4f9d\u8d56":87,"api\u63a5\u53d7\u7684\u7c7b\u578b\u5168\u662f":46,"api\u63a5\u53e3":28,"api\u63a5\u53e3\u751f\u6210":75,"api\u63a5\u53e3\u7684\u53c2\u6570\u8f6c\u53d1\u7ed9":46,"api\u63a5\u53e3\u7684\u751f\u6210":75,"api\u63d0\u4f9b\u7684":90,"api\u652f\u6301\u7684\u6240\u6709\u8f93\u5165\u6570\u636e\u7c7b\u578b\u548c\u4ed6\u4eec\u7684\u7ec4\u7ec7\u65b9\u5f0f":90,"api\u6587\u6863":[116,117],"api\u65f6":46,"api\u65f6\u4e3a\u8f93\u51fa":90,"api\u65f6\u6240\u552f\u4e00\u9700\u8981\u5f15\u5165\u7684\u5934\u6587\u4ef6":46,"api\u662f\u591a\u8bed\u8a00api\u7684\u57fa\u7840\u90e8\u5206":46,"api\u66b4\u9732\u7684\u7c7b\u578b":46,"api\u6765\u9884\u6d4b":[116,117],"api\u751f\u6210\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u4f1a\u88ab\u5b89\u88c5\u5230":46,"api\u7684\u5934\u6587\u4ef6":[116,117,118],"api\u7684\u5b9e\u4f8b":46,"api\u7684\u5b9e\u73b0\u7ec6\u8282":46,"api\u7684\u63a5\u53e3":46,"api\u7684\u65f6\u5019\u63a8\u8350paddle\u4e0d\u5d4c\u5165python\u89e3\u91ca\u5668":46,"api\u7684\u7f16\u8bd1\u9009\u9879\u9ed8\u8ba4\u5173\u95ed":46,"api\u76ee\u5f55\u7ed3\u6784\u5982\u4e0a\u56fe\u8868\u6240\u793a":46,"api\u76f8\u5173\u63a5\u53e3":89,"api\u7ea7\u522b":116,"api\u7ea7\u522b\u4e3a21":116,"api\u83b7\u5f97\u4e86\u795e\u7ecf\u7f51\u7edc\u7684\u53c2\u6570\u5b9e\u4f8b":46,"api\u8bad\u7ec3":90,"api\u9700\u8981\u521b\u5efa\u7684\u6570\u636e\u7c7b\u578b":89,"api\u9759\u6001\u5e93":117,"api\u9884\u6d4b\u5e93":[106,117],"api\u9884\u6d4b\u65f6":90,"apis\u505a\u4e86\u5c01\u88c5":41,"app\u4e2d":[116,117],"apple\u5b98\u65b9\u4e3aios\u5f00\u53d1\u63d0\u4f9b\u4e86\u5b8c\u6574\u7684\u4ea4\u53c9\u7f16\u8bd1\u5de5\u5177\u548c\u96c6\u6210\u5f00\u53d1\u73af\u5883":117,"async_sgd\u8fdb\u884c\u8bad\u7ec3\u65f6":83,"avx\u662f\u4e00\u79cdcpu\u6307\u4ee4\u96c6":2,"avx\u7248\u672c":2,"avx\u7684\u955c\u50cf":2,"batch\u4e2d\u5305\u542b":81,"batch\u7684\u6743\u91cd":81,"batches\u4e2a\u6279\u6b21\u4fdd\u5b58\u4e00\u6b21\u53c2\u6570":103,"batches\u6b21":103,"block\u6784\u6210\u4e00\u4e2amodel":11,"book\u4e00\u5b9a\u662f\u60a8\u6700\u597d\u7684\u9009\u62e9":2,"book\u4e2d\u6240\u6709\u7ae0\u8282\u529f\u80fd\u7684\u6b63\u786e\u6027":62,"book\u662f\u4e3a\u7528\u6237\u548c\u5f00\u53d1\u8005\u5236\u4f5c\u7684\u4e00\u4e2a\u4ea4\u4e92\u5f0f\u7684jupyt":2,"book\u7684":84,"book\u7684docker\u955c\u50cf":2,"boolean":[27,29,37,45,68],"break":[9,66,68,70,71],"bugfix\u5206\u652f\u4e5f\u662f\u5728\u5f00\u53d1\u8005\u81ea\u5df1\u7684fork\u7248\u672c\u5e93\u7ef4\u62a4":62,"bugfix\u5206\u652f\u9700\u8981\u5206\u522b\u7ed9\u4e3b\u7248\u672c\u5e93\u7684":62,"byte":[28,44,83],"c99\u662f\u76ee\u524dc\u6700\u5e7f\u6cdb\u7684\u4f7f\u7528\u6807\u51c6":45,"c\u6709\u6807\u51c6\u7684abi":45,"c\u8bed\u8a00\u662f\u6709\u5bfc\u51fa\u7b26\u53f7\u7684\u6807\u51c6\u7684":45,"case":[13,19,21,22,27,31,40,46,52,56,58,59,68,95,108,119],"cc\u4e2d\u7684":76,"cells\u7b49":82,"char":15,"ci\u73af\u5883\u4f7f\u7528":62,"ci\u7f16\u8bd1wheel\u5b8c\u6210\u540e\u4f1a\u81ea\u52a8\u5c06docker\u955c\u50cfpush\u5230dockerhub":62,"class":[5,8,19,20,21,22,25,26,30,31,32,33,35,36,38,40,45,49,50,54,55,59,60,61,63,64,65,67,70,74,75,76,83,109],"cmake\u4e2d\u5c06":108,"cmake\u5b98\u65b9\u5bf9android\u5e73\u53f0\u7684\u4ea4\u53c9\u7f16\u8bd1\u63d0\u4f9b\u4e86\u901a\u7528\u7684\u652f\u6301":116,"cmake\u627e\u5230\u7684python\u5e93\u548cpython\u89e3\u91ca\u5668\u7248\u672c\u53ef\u80fd\u6709\u4e0d\u4e00\u81f4\u73b0\u8c61":78,"cmake\u7cfb\u7edf\u5bf9\u4ea4\u53c9\u7f16\u8bd1\u63d0\u4f9b\u4e86\u652f\u6301":116,"cmake\u7f16\u8bd1\u65f6":1,"cmake\u7f16\u8bd1\u7684\u76ee\u6807\u5e73\u53f0":[116,117,118],"cmake\u914d\u7f6e\u4e2d\u5c06":108,"cmake\u914d\u7f6e\u5b8c\u6210\u540e":[116,117,118],"compute\u51fd\u6570":41,"const":[8,13,15,20,30,32,38,39,53,54,56,60,63,65,67,69,70,74,75,76],"container\u4e2d":96,"core\u4e2d\u7684\u6a21\u578b\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u53c2\u6570":46,"core\u4e2d\u8fd9\u4e00\u7c7b\u578b\u63a5\u53e3\u7684\u667a\u80fd\u6307\u9488":46,"core\u662f\u5426\u8fd8\u5728\u4f7f\u7528\u8fd9\u4e2a\u5b9e\u4f8b":46,"core\u6982\u5ff5":46,"cost\u63a5\u6536y_predict\u4e0ey\u4f5c\u4e3a\u8f93\u5165":84,"cost\u8fd8\u5927\u4e8e\u8fd9\u4e2a\u6570":83,"count\u4e2agpu\u4e0a\u4f7f\u7528\u6570\u636e\u5e76\u884c\u6765\u8ba1\u7b97\u67d0\u4e00\u5c42":105,"count\u548cgpu":105,"csr\u5b58\u50a8\u683c\u5f0f\u901a\u8fc7":89,"cuda\u5171\u4eabkernel\u5b9e\u73b0\u5728":75,"cuda\u5b9e\u73b0\u5171\u4eab\u540c\u4e00\u4e2a":75,"cuda\u5b9e\u73b0\u5728":75,"cuda\u5e93":103,"cuda\u7684\u4ee3\u7801\u53ef\u4ee5\u590d\u7528":75,"cuda\u76f8\u5173\u5e93\u4f1a\u5728\u9884\u6d4b\u7a0b\u5e8f\u8fd0\u884c\u65f6\u52a8\u6001\u88c5\u8f7d":87,"cudnn\u5e93":[1,103],"cumtime\u7684\u6bcf\u6b21\u8c03\u7528\u5e73\u5747\u65f6\u95f4":107,"data\u5230\u5206\u5e03\u5f0f\u5b58\u50a8\u8865\u5145\u8bad\u7ec3\u6570\u636e":12,"data\u63a5\u53e3\u5206\u914d\u5b9e\u9645\u7684\u5185\u5b58":76,"data\u76ee\u5f55\u4e2d\u5b58\u653e\u5207\u5206\u597d\u7684\u6570\u636e":97,"dataprovider\u5171\u8fd4\u56de\u4e24\u4e2a\u6570\u636e":111,"dataprovider\u5171\u8fd4\u56de\u4e24\u7ec4\u6570\u636e":111,"dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":81,"decoder\u5faa\u73af\u5c55\u5f00\u7684\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u4f1a\u5f15\u7528\u5168\u90e8\u7ed3\u679c":113,"decoder\u63a5\u53d7\u4e24\u4e2a\u8f93\u5165":113,"decoder\u6bcf\u6b21\u9884\u6d4b\u4ea7\u751f\u4e0b\u4e00\u4e2a\u6700\u53ef\u80fd\u7684\u8bcd\u8bed":113,"decoer\u67b6\u6784":113,"default":[5,8,9,19,21,25,34,38,44,47,48,56,57,63,64,65,68,71,95,96,97,105,116,119],"device\u5c31\u80fd\u62ff\u5230\u6b63\u786e\u7684\u6570\u636e":42,"dist\u76ee\u5f55\u4e0b\u751f\u6210\u8f93\u51fa\u7684whl\u5305":1,"dnn\u4e09\u8005\u5173\u7cfb\u5982\u4e0b\u8868":42,"dnn\u4e2d\u7684":42,"dnn\u4e2d\u7684\u6392\u5217\u65b9\u5f0f\u4e0d\u6b62\u8fd9\u4e00\u79cd":42,"dnn\u4f1a\u4f5c\u4e3a\u7b2c\u4e09\u65b9\u5e93\u96c6\u6210\u8fdbpaddlepaddl":42,"dnn\u4f1a\u7528\u5230":42,"dnn\u5171\u540c\u4f7f\u7528":42,"dnn\u524d\u540e\u7684cnn\u7f51\u7edc\u6027\u80fd":42,"dnn\u5728\u53d1\u5e03":42,"dnn\u5b9e\u73b0":42,"dnn\u5e0c\u671b\u7684\u683c\u5f0f":42,"dnn\u6570\u5b66\u5e93":1,"dnn\u6570\u636e\u7684\u4e0d\u540c\u683c\u5f0f\u4ee5\u53ca\u76f8\u4e92\u4e4b\u95f4\u7684\u8f6c\u6362":42,"dnn\u7684":42,"dnn\u7684\u5e93\u76ee\u524d\u53ea\u6709\u52a8\u6001\u5e93":42,"dnn\u7684\u6027\u80fd":42,"dnn\u7684\u60c5\u51b5\u4e0b":42,"dnn\u7684\u64cd\u4f5c\u90fd\u662f\u76f4\u63a5\u8986\u76d6\u7684\u5f62\u5f0f":42,"dnn\u7684\u6d4b\u8bd5":42,"dnn\u7684\u73af\u5883\u4e0b":42,"dnn\u7684\u76f8\u5173\u529f\u80fd":42,"dnn\u7684\u7ed3\u679c":42,"dnn\u7684\u9ad8\u6027\u80fd\u683c\u5f0f\u4e0epaddlepaddle\u539f\u6709\u7684":42,"dnn\u7684layer":42,"dnn\u7684layers\u90fd\u4f1a\u7ee7\u627f\u4e8e":42,"docker\u5b89\u88c5\u65b9\u5f0f\u53ef\u4ee5\u8fdb\u5165docker\u5bb9\u5668\u6267\u884c":101,"docker\u5b89\u88c5\u8bf7\u53c2\u8003":77,"docker\u5b89\u88c5\u8bf7\u53c2\u8003docker\u7684\u5b98\u7f51":77,"docker\u5b98\u7f51":2,"docker\u5bb9\u5668\u4e2d\u5c06\u9ed8\u8ba4\u4f7f\u7528":116,"docker\u7684\u5b98\u7f51":77,"docker\u7f16\u8bd1\u73af\u5883\u955c\u50cf\u5b8c\u6210\u7f16\u8bd1":1,"docker\u80fd\u5728\u6240\u6709\u4e3b\u8981\u64cd\u4f5c\u7cfb\u7edf":116,"docker\u955c\u50cf":2,"docker\u955c\u50cf\u4e3a\u4e86\u51cf\u5c0f\u4f53\u79ef":2,"docker\u955c\u50cf\u9ed8\u8ba4":2,"dockerhub\u7f51\u7ad9":2,"double\u7c7b\u578b\u65f6\u4e3a8":83,"dropout\u7684\u6bd4\u4f8b":74,"eigenscalar\u7684\u8f6c\u6362":76,"encode\u6210\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":111,"encoder\u548cdecoder\u53ef\u4ee5\u662f\u80fd\u591f\u5904\u7406\u5e8f\u5217\u7684\u4efb\u610f\u795e\u7ecf\u7f51\u7edc\u5355\u5143":113,"encoder\u8f93\u51fa":113,"entropy\u4f5c\u4e3acost":83,"enum":[13,15,21,47,54,55,64,65,71],"export":[2,31,36,77,78,91],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":84,"fc1\u548cfc2\u5c42\u5728gpu\u4e0a\u8ba1\u7b97":105,"fc3\u5c42\u4f7f\u7528cpu\u8ba1\u7b97":105,"final":[6,7,22,36,48,49,66,70],"flatten\u65b9\u6cd5\u662f\u628apaddle\u4e2d\u7684\u4e00\u4e2atensor\u8fdb\u884creshape\u64cd\u4f5c":76,"float":[25,30,38,65,67,69,75,76,89,108],"float\u7b49":105,"forward\u7684output\u7684\u503c":81,"from\u65b9\u6cd5\u662f\u628apaddle\u4e2d\u7684\u4e00\u7ef4tensor\u8f6c\u4e3aeigen\u7684\u4e00\u7ef4tensor":76,"from\u662feigentensor\u6a21\u677f\u63d0\u4f9b\u7684\u4e00\u4e2a\u63a5\u53e3":76,"full\u53c2\u6570\u63d0\u4ea4":79,"function":[5,7,8,10,14,15,16,18,19,20,21,22,25,26,30,32,35,38,43,48,49,52,53,54,55,56,58,59,60,61,63,65,70,107,114,119],"function\u4f7f\u7528":82,"git\u6d41\u5206\u652f\u6a21\u578b":72,"github\u9996\u9875":72,"glibc\u81f3\u5c11\u5305\u542bglibc_2":4,"golang\u53ef\u4ee5\u4f7f\u7528":45,"golang\u7684":45,"google\u5f00\u6e90\u7684\u5bb9\u5668\u96c6\u7fa4\u7684\u8c03\u5ea6\u6846\u67b6":94,"gpu\u4e8c\u8fdb\u5236\u6587\u4ef6":1,"gpu\u5219\u8fd8\u9700\u8981\u9ad8\u5e76\u884c\u6027":108,"gpu\u6027\u80fd\u8c03\u4f18":106,"gpu\u6267\u884c":76,"gpu\u6838\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u6307\u5b9a":103,"gpu\u7684docker\u955c\u50cf\u7684\u65f6\u5019":78,"gpu\u7b49":62,"group\u6559\u7a0b":112,"group\u7684\u5b9e\u73b0\u65b9\u5f0f":82,"gru\u6216lstm":114,"h\u5e76\u4e0d\u56f0\u96be":45,"html\u5373\u53ef\u8bbf\u95ee\u672c\u5730\u6587\u6863":77,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":84,"id\u6307\u5b9a\u4f7f\u7528\u54ea\u4e2agpu\u6838":103,"id\u6307\u5b9a\u7684gpu":105,"id\u65e0\u6548":103,"image\u91cc":96,"images\u6570\u636e\u96c6\u4e0a\u4f20\u5230\u4e91\u7aef\u7684":12,"imikolov\u6570\u636e\u96c6":91,"import":[4,5,8,9,19,21,24,32,34,36,37,43,48,49,55,63,66,75,84,86,90,91,95],"infer\u63a5\u53e3\u7684\u8fd4\u56de\u503c\u662f\u4e00\u4e2apython":81,"ingress\u9700\u8981\u628apfsclient\u7684\u8eab\u4efd\u4fe1\u606f\u4f20\u7ed9pfsserv":28,"instance\u4e0e\u751f\u6210\u6570\u636e\u96c6\u65f6":12,"instance\u5305\u6db5\u4e24\u4e2a\u503c":12,"instance\u662f\u4e00\u6a21\u4e00\u6837\u7684":12,"int":[7,8,13,14,15,18,19,21,23,37,38,41,42,43,45,46,54,55,57,58,64,65,67,70,71,74,76,89,91,105],"interface\u6587\u4ef6\u7684\u5199\u6cd5\u975e\u5e38":45,"ios\u5e73\u53f0\u53ef\u9009\u914d\u7f6e\u53c2\u6570":117,"ip\u548c\u4efb\u52a1\u8282\u70b9\u4e2a\u6570\u7b49":94,"issue\u7f16\u53f7":72,"job\u662f\u672c\u6b21\u8bad\u7ec3\u5bf9\u5e94\u7684job":97,"job\u7684\u540d\u5b57":97,"kernel\u5b9e\u73b0":75,"kernel\u6ce8\u518ccpu\u5b9e\u73b0\u5728":75,"kernel\u7684\u5b9e\u73b0\u57fa\u4e8eeigen":75,"kubernetes\u4e3a\u8fd9\u6b21\u8bad\u7ec3\u521b\u5efa\u4e863\u4e2apod\u5e76\u4e14\u8c03\u5ea6\u5230\u4e863\u4e2anode\u4e0a\u8fd0\u884c":97,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":94,"kubernetes\u5355\u673a\u8bad\u7ec3":94,"kubernetes\u53ef\u4ee5\u901a\u8fc7yaml\u6587\u4ef6\u6765\u521b\u5efa\u76f8\u5173\u5bf9\u8c61":97,"kubernetes\u5c31\u4f1a\u521b\u5efa3\u4e2apod\u4f5c\u4e3apaddlepaddle\u8282\u70b9\u7136\u540e\u62c9\u53d6\u955c\u50cf":97,"kubernetes\u6709job\u7c7b\u578b\u7684\u8d44\u6e90\u6765\u652f\u6301":96,"label\u662f\u539f\u59cb\u6570\u636e\u4e2d\u5bf9\u4e8e\u6bcf\u4e00\u53e5\u8bdd\u7684\u5206\u7c7b\u6807\u7b7e":111,"labels\u662f\u6bcf\u7ec4\u5185\u6bcf\u4e2a\u53e5\u5b50\u7684\u6807\u7b7e":111,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"layer1\u5fc5\u987b\u662f\u4e00\u4e2a\u5355\u5c42\u5e8f\u5217":110,"layer\u4f5c\u4e3a\u4e00\u4e2a\u6574\u4f53\u6765\u5b9e\u73b0":82,"layer\u62ff\u5230\u7684\u7528\u6237\u8f93\u5165":113,"layer\u65f6":[42,82],"layer\u662f\u6211\u4eec\u7684\u79ef\u6728":84,"layer\u7684\u540e\u9762\u63a5\u6709cpu":42,"layer\u7c7b\u53ef\u4ee5\u81ea\u52a8\u8ba1\u7b97\u4e0a\u9762\u7684\u5bfc\u6570":74,"layer\u8ba1\u7b97\u7684\u8f93\u51fa":82,"linux\u4e2d":2,"list\u4e2d":90,"list\u4f5c\u4e3a\u68c0\u67e5\u5217\u8868":62,"list\u5982\u4e0b\u6240\u793a":105,"list\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b\u5217\u8868":105,"long":21,"memory\u4e0d\u80fd\u72ec\u7acb\u5b58\u5728":113,"memory\u4e5f\u53ef\u4ee5\u5177\u6709":114,"memory\u4e5f\u53ef\u4ee5\u662f\u5e8f\u5217":114,"memory\u53ea\u80fd\u5728":113,"memory\u53ef\u4ee5\u7f13\u5b58\u4e0a\u4e00\u4e2a\u65f6\u523b\u67d0\u4e00\u4e2a\u795e\u7ecf\u5143\u7684\u8f93\u51fa":111,"memory\u6307\u5411\u4e00\u4e2alay":113,"memory\u662f\u5728\u5355\u6b65\u51fd\u6570\u4e2d\u5faa\u73af\u4f7f\u7528\u7684\u72b6\u6001":114,"memory\u662fpaddlepaddle\u5b9e\u73b0rnn\u65f6\u5019\u4f7f\u7528\u7684\u4e00\u4e2a\u6982\u5ff5":111,"memory\u7684":114,"memory\u7684\u521d\u59cb\u72b6\u6001":113,"memory\u7684\u65f6\u95f4\u5e8f\u5217\u957f\u5ea6\u4e00\u81f4\u7684\u60c5\u51b5":111,"memory\u7684\u66f4\u591a\u8ba8\u8bba\u8bf7\u53c2\u8003\u8bba\u6587":113,"memory\u7684\u8f93\u51fa\u5b9a\u4e49\u5728":114,"memory\u7684i":113,"memory\u9ed8\u8ba4\u521d\u59cb\u5316\u4e3a0":113,"mkl\u5e93\u7684":41,"mklml\u4ee5\u53camkl":42,"mklml\u53ef\u4ee5\u4e0emkl":42,"mklml\u7684\u5e93\u76ee\u524d\u90fd\u662f\u52a8\u6001\u5e93":42,"mnist\u624b\u5199\u6570\u5b57\u8bc6\u522b\u76ee\u5f55":90,"mode\u4e0b\u7684\u7ed3\u679c":41,"model\u505a\u5206\u652f\u7ba1\u7406":62,"name\u7ec4\u5408\u53ef\u4ee5\u627e\u5230\u672c\u6b21\u8bad\u7ec3\u9700\u8981\u7684\u6587\u4ef6\u8def\u5f84":97,"ndarray\u7c7b\u578b\u7684\u503c\u548c\u6574\u578b\u7684\u503c":12,"ndk\u4e2d\u5305\u542b\u4e86\u6240\u6709android":116,"new":[6,7,8,9,10,13,14,15,16,17,20,21,22,25,30,31,40,41,43,47,49,52,57,58,59,61,65,66,70,72,74,95,119],"note\u7684\u4e66\u5199":62,"null":[36,74,89,103],"num\u51b3\u5b9a\u603b\u7aef\u53e3\u4e2a\u6570":91,"num_gradient_servers\u53c2\u6570":97,"num_samples_processed\u4e3a\u5df2\u8bad\u7ec3\u6837\u672c\u6570":83,"only\u7684\u4e8c\u8fdb\u5236":1,"op\u4e0d\u9700\u8981\u5b9a\u4e49opprotomak":75,"op\u5355\u5143\u6d4b\u8bd5\u7ee7\u627f\u81ea":75,"op\u5b9a\u4e49":75,"op\u6709\u8ba1\u7b97\u51fd\u6570":75,"op\u6ce8\u518c\u5b9e\u73b0\u5728":75,"op\u7684\u4fe1\u606f":42,"op\u8ba1\u7b97\u51fd\u6570\u7684\u57fa\u7c7b":75,"openmp\u7528\u4e8e\u63d0\u9ad8mklml\u7684\u6027\u80fd":42,"opprotomake\u5b9a\u4e49":75,"org\u5de5\u5177\u7684\u8be6\u7ec6\u4fe1\u606f":77,"org\u76ee\u524d\u9075\u5faa":62,"outer_mem\u662f\u4e00\u4e2a\u5b50\u53e5\u7684\u6700\u540e\u4e00\u4e2a\u5411\u91cf":111,"output\u53ef\u4ee5\u662f\u4efb\u610f\u7ef4\u5ea6\u7684tensor":76,"output\u6587\u4ef6\u5939\u5b58\u653e\u8bad\u7ec3\u7ed3\u679c\u4e0e\u65e5\u5fd7":97,"output\u7684\u539f\u6709shape\u4fe1\u606f\u4e0d\u53d8":76,"packages\u91cc\u9762":78,"packages\u91cc\u9762\u7684python\u5305":78,"packed\u4f18\u5316\u540elayer\u7684\u6d4b\u8bd5":41,"packed\u76f8\u5173\u529f\u80fd":41,"paddepaddle\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":1,"paddle\u4e00\u4e2a\u52a8\u6001\u5e93\u53ef\u4ee5\u5728\u4efb\u4f55linux\u7cfb\u7edf\u4e0a\u8fd0\u884c":45,"paddle\u4e2d\u7ecf\u5e38\u4f1a\u5c06\u65f6\u95f4\u5e8f\u5217\u6210\u4e3a":111,"paddle\u4e8c\u8fdb\u5236\u5728\u8fd0\u884c\u65f6\u6355\u83b7\u4e86\u6d6e\u70b9\u6570\u5f02\u5e38":81,"paddle\u5185\u5d4c\u7684python\u89e3\u91ca\u5668\u548c\u5916\u90e8\u4f7f\u7528\u7684python\u5982\u679c\u7248\u672c\u4e0d\u540c":45,"paddle\u5185\u90e8\u7684\u7c7b\u4e3ac":45,"paddle\u7684\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0\u5305\u62ec\u4e00\u4e0b\u51e0\u4e2a\u65b9\u9762":45,"paddle\u7684\u7c7b\u578b\u5168\u90e8\u9000\u5316\u6210":46,"paddle\u7684\u94fe\u63a5\u65b9\u5f0f\u6bd4\u8f83\u590d\u6742":45,"paddle\u7684c":46,"paddle\u8bad\u7ec3\u4efb\u52a1":12,"paddle\u8def\u5f84\u4e0b":46,"paddle\u9700\u8981\u4e00\u4e2a\u591a\u8bed\u8a00\u63a5\u53e3":45,"paddle\u9700\u8981\u66b4\u9732\u7684api\u5f88\u591a":46,"paddle\u9759\u6001\u5e93\u94fe\u63a5\u590d\u6742":45,"paddle_\u7c7b\u578b\u540d":46,"paddle_\u7c7b\u578b\u540d_\u51fd\u6570\u540d":46,"paddlepaddle\u4e2d":[110,113],"paddlepaddle\u4e2d\u4e00\u4e2a\u8ba1\u7b97\u5c42\u7684\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f\u548c\u8f93\u5165\u6570\u636e\u7ec4\u7ec7\u65b9\u5f0f\u5b8c\u5168\u76f8\u540c":89,"paddlepaddle\u4e2d\u7684\u8bb8\u591alayer\u5e76\u4e0d\u5728\u610f\u8f93\u5165\u662f\u5426\u662f\u65f6\u95f4\u5e8f\u5217":111,"paddlepaddle\u4e2d\u7684cudnn\u90e8\u5206\u4f7f\u7528\u7684\u4e5f\u662f":42,"paddlepaddle\u4e2d\u795e\u7ecf\u7f51\u7edc\u8ba1\u7b97\u5c42\u8f93\u5165":89,"paddlepaddle\u4e2d\u8fd8\u5305\u542b":82,"paddlepaddle\u4e2d\u901a\u8fc7reader\u6765\u52a0\u8f7d\u6570\u636e":84,"paddlepaddle\u4e3a\u4ea4\u53c9\u7f16\u8bd1\u63d0\u4f9b\u4e86\u5de5\u5177\u94fe\u914d\u7f6e\u6587\u6863":[116,117],"paddlepaddle\u4e3a\u6df1\u5ea6\u5b66\u4e60\u7814\u7a76\u4eba\u5458\u63d0\u4f9b\u4e86\u4e30\u5bcc\u7684api":84,"paddlepaddle\u4e3ano":2,"paddlepaddle\u4e3b\u8981\u4f7f\u7528":1,"paddlepaddle\u4f1a\u81ea\u52a8\u8bbe\u5b9a":82,"paddlepaddle\u4f7f\u7528\u540c\u6b65\u5c4f\u969c":92,"paddlepaddle\u4f7f\u7528\u5747\u503c0":83,"paddlepaddle\u4f7f\u7528avx":78,"paddlepaddle\u4f7f\u7528git":62,"paddlepaddle\u4fdd\u5b58\u7684\u6a21\u578b\u53c2\u6570\u6587\u4ef6\u5185\u5bb9\u753116\u5b57\u8282\u5934\u4fe1\u606f\u548c\u7f51\u7edc\u53c2\u6570\u4e24\u90e8\u5206\u7ec4\u6210":83,"paddlepaddle\u4fdd\u5b58\u7684\u6a21\u578b\u53c2\u6570\u6587\u4ef6\u524d16\u5b57\u8282\u4e3a\u5934\u4fe1\u606f":83,"paddlepaddle\u53d1\u5e03\u7684\u5b89\u88c5\u5305\u4f1a\u5c3d\u91cf\u5bf9\u9f50":4,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528\u591a\u79cd\u5206\u5e03\u5f0f\u8ba1\u7b97\u5e73\u53f0\u6784\u5efa\u5206\u5e03\u5f0f\u8ba1\u7b97\u4efb\u52a1":94,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528\u5e38\u7528\u7684python\u5305\u7ba1\u7406\u5de5\u5177":4,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528cudnn":1,"paddlepaddle\u53ef\u4ee5\u540c\u65f6\u652f\u6301\u540c\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":92,"paddlepaddle\u53ef\u4ee5\u6bd4\u8f83\u7b80\u5355\u7684\u5224\u65ad\u54ea\u4e9b\u8f93\u51fa\u662f\u5e94\u8be5\u8de8\u8d8a\u65f6\u95f4\u6b65\u7684":111,"paddlepaddle\u53ef\u4ee5\u901a\u8fc7\u8be5\u673a\u5236\u5224\u65ad\u662f\u5426\u5df2\u7ecf\u6536\u96c6\u9f50\u6240\u6709\u7684\u68af\u5ea6":74,"paddlepaddle\u5728\u5b9e\u73b0rnn\u7684\u65f6\u5019":111,"paddlepaddle\u5728\u6fc0\u6d3b\u51fd\u6570\u91cc\u5b9e\u73b0dropout":82,"paddlepaddle\u5728\u7f16\u8bd1\u65f6":1,"paddlepaddle\u5b58\u7684\u662f\u6709\u503c\u4f4d\u7f6e\u7684\u7d22\u5f15":84,"paddlepaddle\u5b89\u88c5\u5305\u7531\u4e8e\u4e0d\u4ec5\u4ec5\u5305\u542b":4,"paddlepaddle\u5c06\u4f1a\u6839\u636e":117,"paddlepaddle\u5c06\u4f1a\u81ea\u52a8\u9009\u62e9":117,"paddlepaddle\u5c06\u6839\u636e":116,"paddlepaddle\u5c06\u81ea\u52a8\u4e0b\u8f7d\u548c\u7f16\u8bd1\u6240\u6709\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93":[116,117,118],"paddlepaddle\u5e93\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6210":117,"paddlepaddle\u5f00\u53d1\u8fc7\u7a0b\u4f7f\u7528":62,"paddlepaddle\u63d0\u4f9b":3,"paddlepaddle\u63d0\u4f9b\u4e13\u7528\u7684":12,"paddlepaddle\u63d0\u4f9b\u7684":82,"paddlepaddle\u652f\u6301":1,"paddlepaddle\u652f\u6301\u4e0d\u540c\u7c7b\u578b\u7684\u8f93\u5165\u6570\u636e":84,"paddlepaddle\u652f\u6301\u4f7f\u7528pip\u5feb\u901f\u5b89\u88c5":86,"paddlepaddle\u652f\u6301\u975e\u5e38\u591a\u7684\u4f18\u5316\u7b97\u6cd5":81,"paddlepaddle\u652f\u6301sparse\u7684\u8bad\u7ec3":81,"paddlepaddle\u6587\u6863\u4f7f\u7528":77,"paddlepaddle\u662f\u6e90\u4e8e\u767e\u5ea6\u7684\u4e00\u4e2a\u6df1\u5ea6\u5b66\u4e60\u5e73\u53f0":84,"paddlepaddle\u6bcf\u6b21\u53d1\u65b0\u7684\u7248\u672c":62,"paddlepaddle\u6bcf\u6b21\u53d1\u7248\u672c\u9996\u5148\u8981\u4fdd\u8bc1paddlepaddl":62,"paddlepaddle\u7684":96,"paddlepaddle\u7684\u4e3b\u7248\u672c\u5e93\u9075\u5faa":62,"paddlepaddle\u7684\u5185\u5b58\u5360\u7528\u4e3b\u8981\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u65b9\u9762":81,"paddlepaddle\u7684\u53c2\u6570\u4f7f\u7528\u540d\u5b57":83,"paddlepaddle\u7684\u5404\u7248\u672c\u955c\u50cf\u53ef\u4ee5\u53c2\u8003":96,"paddlepaddle\u7684\u5b89\u88c5\u53ef\u4ee5\u53c2\u8003":101,"paddlepaddle\u7684\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6210":116,"paddlepaddle\u7684\u6240\u6709layer\u90fd\u6709\u552f\u4e00\u7684nam":82,"paddlepaddle\u7684\u6587\u6863\u5305\u62ec\u82f1\u6587\u6587\u6863":77,"paddlepaddle\u7684\u6587\u6863\u6784\u5efa\u6709\u4e09\u79cd\u65b9\u5f0f":77,"paddlepaddle\u7684\u6e90\u7801":72,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":1,"paddlepaddle\u7684activation\u4f1a\u76f4\u63a5\u4f7f\u7528":42,"paddlepaddle\u7684bas":74,"paddlepaddle\u7684c":116,"paddlepaddle\u7684cmake\u7cfb\u7edf\u4f1a\u81ea\u52a8\u7f16\u8bd1\u6240\u6709\u7684\u7b2c\u4e09\u65b9\u4f9d\u8d56\u5e93":117,"paddlepaddle\u7684cmake\u7cfb\u7edf\u5c06\u6839\u636e\u8be5\u503c\u81ea\u52a8\u63a8\u5bfc\u548c\u8bbe\u7f6e\u9700\u8981\u4f7f\u7528\u7684\u4ea4\u53c9\u7f16\u8bd1\u5668":116,"paddlepaddle\u7684cmake\u7cfb\u7edf\u5c06\u6839\u636e\u8be5\u503c\u81ea\u52a8\u8bbe\u7f6e\u9700\u8981\u4f7f\u7528\u7684\u4ea4\u53c9\u7f16\u8bd1\u5668":118,"paddlepaddle\u7684cmake\u7cfb\u7edf\u624d\u8ba4\u4e3a\u5728\u662f\u5728\u4ea4\u53c9\u7f16\u8bd1raspberri":118,"paddlepaddle\u7684cmake\u7cfb\u7edf\u624d\u8ba4\u4e3a\u662f\u5728\u4ea4\u53c9\u7f16\u8bd1android\u7cfb\u7edf\u7684\u7248\u672c":116,"paddlepaddle\u7684dock":96,"paddlepaddle\u7684softmax\u4e0d\u80fd\u6307\u5b9a\u8ba1\u7b97\u7ef4\u5ea6":82,"paddlepaddle\u76ee\u524d\u53ea\u652f\u6301\u5728\u6bcf\u4e2a\u65f6\u95f4\u6b65\u4e2d":111,"paddlepaddle\u76ee\u524d\u63d0\u4f9b\u4e24\u79cd\u53c2\u6570\u521d\u59cb\u5316\u7684\u65b9\u5f0f":83,"paddlepaddle\u76ee\u524d\u652f\u63018\u79cdlearning_rate_schedul":83,"paddlepaddle\u7f16\u8bd1\u9700\u8981\u4f7f\u7528\u5230\u4e0b\u9762\u7684\u4f9d\u8d56":1,"paddlepaddle\u82e5\u68c0\u6d4b\u5230\u7528\u6237\u4f7f\u7528\u7684cmake\u7248\u672c\u4e0d\u4f4e\u4e8e3":116,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u68af\u5ea6\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":113,"paddlepaddle\u8d1f\u8d23\u5b8c\u6210\u4fe1\u606f\u548c\u8bef\u5dee\u5728\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u4f20\u64ad":113,"paddlepaddle\u955c\u50cf\u9700\u8981\u63d0\u4f9b":97,"pass\u4e2a\u6a21\u578b\u5230\u7b2c":103,"pass\u5c06\u4e0d\u8d77\u4f5c\u7528":103,"pass\u8f6e\u5f00\u59cb\u8bad\u7ec3":103,"pass\u8f6e\u7684\u6a21\u578b\u7528\u4e8e\u6d4b\u8bd5":103,"passes\u8f6e":103,"patch\u53f7":62,"patch\u53f7\u52a0\u4e00":62,"path\u6307\u5b9a\u6d4b\u8bd5\u7684\u6a21\u578b":105,"perftools\u6765\u8fdb\u884c\u6027\u80fd\u5206\u6790":107,"period\u4e2a\u6279\u6b21\u5bf9\u6240\u6709\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6d4b\u8bd5":103,"period\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":103,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u53c2\u6570\u7edf\u8ba1":103,"period\u4e2a\u6279\u6b21\u8f93\u51fa\u7b26\u53f7":103,"period\u6574\u9664":103,"period\u8f6e\u4fdd\u5b58\u8bad\u7ec3\u53c2\u6570":103,"pfsclient\u9700\u8981\u548cingress\u4e4b\u95f4\u505a\u53cc\u5411\u9a8c\u8bc1":28,"pfsclient\u9700\u8981\u5728\u4f20\u8f93\u5b8c\u6bd5\u6700\u540e\u4e00\u4e2achunk\u7684\u65f6\u5019\u68c0\u67e5destination\u6587\u4ef6\u7684md5\u503c\u662f\u5426\u548csource\u6587\u4ef6\u4e00\u81f4":28,"pfsserver\u63d0\u4f9brest":28,"pi\u5e73\u53f0\u4e0a\u9002\u7528\u7684paddlepaddle\u7684\u65b9\u6cd5\u548c\u6b65\u9aa4":118,"pi\u7248\u672c\u7684\u5e93":118,"pi\u7248\u672cpaddlepaddle\u5e93\u65f6":118,"pi\u7684\u914d\u7f6e\u4fe1\u606f\u5728":118,"pi\u7cfb\u7edf\u4e0a\u6765\u6784\u5efa":118,"pi\u7cfb\u7edf\u7684\u7248\u672c":118,"pip\u548cdocker\u7684\u5b89\u88c5\u65b9\u5f0f":3,"pserver\u5730\u5740\u7b49\u53c2\u6570\u4f7ftrainer\u53ef\u4ee5\u6b63\u786e\u8fde\u63a5\u5230pserv":91,"pserver\u76d1\u542c\u7684\u8d77\u59cb\u7aef\u53e3":91,"public":[8,20,30,33,38,54,60,63,65,66,67,70,74,75,76,95,96],"pwd\u53d8\u91cf\u4f1a\u5c55\u5f00\u4e3a\u5f53\u524d\u8def\u5f84\u7684\u7edd\u5bf9\u8def\u5f84":2,"py\u4e2d":62,"py\u7a0b\u5e8f":4,"pydataprovider\u4f7f\u7528\u7684\u662f\u5f02\u6b65\u52a0\u8f7d":81,"pypi\u4e0a\u7684package\u540d\u79f0\u4e3apaddlepaddle\u548cpaddlepaddl":62,"pypi\u4e0d\u652f\u6301\u8986\u76d6\u4e0a\u4f20":62,"pypi\u5b89\u88c5\u5305\u53ef\u4ee5\u5728":4,"python\u5b89\u88c5\u5305\u652f\u6301linux":78,"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":74,"python\u5e93yep":107,"python\u6807\u51c6\u5e93\u4e2d\u63d0\u4f9b\u4e86\u6027\u80fd\u5206\u6790\u7684\u5de5\u5177\u5305":107,"reader\u7684\u4f7f\u7528\u65b9\u5f0f\u90fd\u662f\u4e00\u81f4\u7684":12,"reader\u8f93\u51fa\u7684data":12,"recommendation\u6587\u4ef6\u5939\u5185\u5b58\u653e\u8bad\u7ec3\u6587\u4ef6":97,"request\u524d":72,"request\u7684":72,"request\u88ab\u5408\u5e76\u540e":72,"resnet\u7684mkl":42,"return":[5,6,7,8,12,13,15,18,19,20,21,26,30,32,33,34,36,38,39,40,43,48,49,50,54,55,56,60,63,65,67,70,74,76,81,83,84,95,97,114],"rnn\u5373\u65f6\u95f4\u9012\u5f52\u795e\u7ecf\u7f51\u7edc":111,"rnn\u5bf9\u4e8e\u6bcf\u4e00\u4e2a\u65f6\u95f4\u6b65\u901a\u8fc7\u4e86\u4e00\u4e2alstm\u7f51\u7edc":111,"rnn\u603b\u662f\u5f15\u7528\u4e0a\u4e00\u65f6\u523b\u9884\u6d4b\u51fa\u7684\u8bcd\u7684\u8bcd\u5411\u91cf":113,"rnn\u6a21\u578b":106,"rnn\u90e8\u5206\u4e2d":41,"rnn\u914d\u7f6e":112,"root\u66ff\u6362\u4e3apaddlepaddle\u9884\u6d4b\u5e93\u7684\u5b89\u88c5\u8def\u5f84":87,"s3\u4e4b\u7c7b\u7684\u5206\u5e03\u5f0f\u5b58\u50a8\u4e4b\u4e0a":12,"search\u7684\u65b9\u6cd5":103,"sentences\u662f\u53cc\u5c42\u65f6\u95f4\u5e8f\u5217\u7684\u6570\u636e":111,"seq\u53c2\u6570\u5fc5\u987b\u4e3afals":113,"server\u4e2a\u6279\u6b21\u6253\u5370\u65e5\u5fd7\u8fdb\u5ea6":103,"server\u4e4b\u4e0a":11,"server\u4e4b\u95f4\u7684\u7f51\u7edc\u5e26\u5bbd":11,"server\u4f1a\u6682\u505c\u53c2\u6570\u66f4\u65b0\u5e76\u7b49\u5f85":11,"server\u4f1a\u83b7\u53d6parameters\u5185\u5b58\u7684":11,"server\u5185\u5b58\u4e2d\u7684\u6a21\u578b\u6570\u636e\u7684\u5b8c\u6574\u955c\u50cf":11,"server\u540c\u6b65\u7684\u4fdd\u5b58\u4e00\u4e2a\u7279\u5b9a\u65f6\u95f4\u70b9\u7684\u5168\u5c40\u68c0\u67e5\u70b9":11,"server\u5728\u96c6\u7fa4\u4e2d\u542f\u52a8\u540e":11,"server\u6545\u969c\u540e\u88abkubernetes\u91cd\u65b0\u542f\u52a8":11,"server\u6b64\u65f6\u8fd8\u9700\u8981\u901a\u8fc7\u7f51\u7edc\u8bbf\u95ee\u5206\u5e03\u5f0f\u5b58\u50a8\u4ee5\u4fdd\u5b58\u5feb\u7167":11,"server\u751f\u6210\u4e00\u4e2auuid":11,"server\u7684\u5355\u70b9\u6216\u591a\u70b9\u540c\u65f6\u6545\u969c":11,"server\u7684\u6570\u636e\u5feb\u7167":11,"server\u7684\u68c0\u67e5\u70b9\u5404\u81ea\u72ec\u7acb\u4fdd\u5b58":11,"server\u7b2c\u4e00\u6b21\u542f\u52a8\u6216\u4efb\u610f\u65f6\u95f4paramet":11,"short":[30,34,57,63,66,70],"simd\u6307\u4ee4\u63d0\u9ad8cpu\u6267\u884c\u6548\u7387":78,"size\u4e3a512":103,"size\u53ef\u80fd\u4f1a\u5bf9\u8bad\u7ec3\u7ed3\u679c\u4ea7\u751f\u5f71\u54cd":81,"size\u672c\u8eab\u662f\u795e\u7ecf\u7f51\u7edc\u7684\u8d85\u53c2\u6570":81,"softmax\u6fc0\u6d3b\u7684\u8f93\u51fa\u7684\u548c\u603b\u662f1":74,"sparse\u8bad\u7ec3\u9700\u8981\u8bad\u7ec3\u7279\u5f81\u662f":81,"static":[15,46,63,65,95,119],"step\u51fd\u6570\u4e2d\u7684memori":113,"step\u51fd\u6570\u5185\u90e8\u53ef\u4ee5\u81ea\u7531\u7ec4\u5408paddlepaddle\u652f\u6301\u7684\u5404\u79cdlay":113,"store\u4e0b\u8f7d\u5b89\u88c5xcode\u5373\u53ef":117,"subseq\u7684\u6bcf\u4e2a\u5143\u7d20\u662f\u4e00\u4e2a0\u5c42\u5e8f\u5217":110,"super":[57,74],"swig\u652f\u6301\u7684\u8bed\u8a00\u6216\u8005\u89e3\u91ca\u5668\u6709\u5c40\u9650":45,"swig\u66b4\u9732\u7684\u63a5\u53e3\u4fdd\u7559\u4e86c":45,"swig\u751f\u6210\u7684\u4ee3\u7801\u4e0d\u80fd\u4fdd\u8bc1\u591a\u8bed\u8a00\u4ee3\u7801\u98ce\u683c\u7684\u4e00\u81f4\u6027":45,"swig\u76f4\u63a5\u8bfb\u53d6c":45,"swig\u9700\u8981\u5199\u4e00\u4e2ainterface\u6587\u4ef6":45,"switch":[8,46,95],"tag\u4e3a":62,"tag\u53ef\u4ee5\u662flatest\u6216latest":62,"tag\u7684\u66f4\u65b0\u65f6\u95f4\u662f\u5426\u5728\u4e0a\u8ff0\u7f16\u8bd1wheel\u5305\u5b8c\u6210\u540e\u662f\u5426\u6700\u65b0":62,"tensor\u5230\u5bf9eigentensor\u7684\u8f6c\u6362":76,"tensor\u5230eigentensor":76,"tensor\u5b9a\u4e49\u5728framework\u76ee\u5f55\u4e0b":76,"tensor\u662f\u4e00\u4e2a\u6b63\u5728\u5f00\u53d1\u4e2d\u7684\u6a21\u5757":76,"tensor\u6a21\u5757\u5bf9el":76,"tensor\u6a21\u5757\u6765\u5b9e\u73b0":75,"tensor\u6a21\u5757\u7684\u6587\u6863\u8f83\u5c11":76,"tensor\u6a21\u5757\u7684\u8be6\u7ec6\u4ecb\u7ecd\u8bf7\u53c2\u8003":76,"tests\u7684paddlepaddl":72,"tflops\u4e86":108,"throw":95,"tottime\u7684\u6bcf\u6b21\u8c03\u7528\u5e73\u5747\u65f6\u95f4":107,"trainer\u542f\u52a8\u9700\u8981\u4f20\u5165\u7aef\u53e3":91,"trainer\u63a5\u6536\u4e09\u4e2a\u53c2\u6570":84,"trainer\u8282\u70b9\u4e2a\u6570":91,"trainer\u9700\u8981\u548cpserver\u4fdd\u6301\u7f51\u7edc\u8054\u901a\u4ee5\u5b8c\u6210\u8bad\u7ec3":91,"true":[5,7,8,13,31,37,41,50,55,56,57,58,62,65,70,74,81,83,89,90,95,97,105,114],"true\u8868\u793a\u53cd\u5411\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":114,"try":[9,10,13,14,15,31,36,40,43,58,63,66,78],"type\u5b57\u6bb5\u5747\u4e0d\u5c3d\u76f8\u540c":46,"type\u6307\u5b9a\u4e3a":107,"ubuntu\u4e0b\u5b89\u88c5\u547d\u4ee4\u4e3a":107,"unit\u5728\u4e00\u4e2a\u65f6\u95f4\u6b65\u5185\u8ba1\u7b97\u5f97\u5230\u7684\u4e2d\u95f4\u503c":82,"unsupported\u6a21\u5757":75,"update\u53c2\u6570\u65f6\u624d\u6709\u6548":103,"v1\u7248\u672c":78,"var":[7,8,19,20,21,25,32,33,35,37,50,55,56,57,59,63,68,69,70,77],"vector\u662frank\u4e3a1\u7684tensor":76,"void":[8,13,15,20,27,30,32,33,38,43,44,45,46,55,56,64,65,67,69,74,75,76],"wheel\u5305":62,"while":[8,17,20,21,31,36,39,40,49,52,53,58,61,63,67,97],"wise\u8ba1\u7b97\u63d0\u4f9b\u4e86\u5f3a\u5927\u7684\u652f\u6301":76,"wmt14\u6570\u636e\u7684\u63d0\u4f9b\u6587\u4ef6\u5728":114,"words\u5373\u4e3a\u8fd9\u4e2a\u6570\u636e\u4e2d\u7684\u5355\u5c42\u65f6\u95f4\u5e8f\u5217":111,"words\u662f\u539f\u59cb\u6570\u636e\u4e2d\u7684\u6bcf\u4e00\u53e5\u8bdd":111,"x86_64\u548cmaco":78,"x\u4e0ey\u4e3a\u4e4b\u524d\u63cf\u8ff0\u7684\u8f93\u5165\u5c42":84,"x\u548cwindow":116,"y\u8868\u793a\u8f93\u5165\u6570\u636e\u662f\u4e00\u4e2a\u7ef4\u5ea6\u4e3a1\u7684\u7a20\u5bc6\u5411\u91cf":84,"yaml\u6587\u4ef6\u4e2d\u5404\u4e2a\u5b57\u6bb5\u7684\u5177\u4f53\u542b\u4e49":97,"yaml\u6587\u4ef6\u63cf\u8ff0\u4e86\u8fd9\u6b21\u8bad\u7ec3\u4f7f\u7528\u7684docker\u955c\u50cf":97,"zero\u4e09\u79cd\u64cd\u4f5c":103,AGE:[95,96],AWS:[12,94,99,100],Added:66,And:[13,17,18,27,34,36,47,51,54,58,63,67,95],But:[6,33,39,47,54,63,78,119],For:[5,7,8,14,15,16,18,19,20,22,25,26,31,32,33,35,38,40,44,47,48,49,52,53,54,55,56,57,58,59,60,61,64,65,66,67,69,71,108,119],IDE:0,IDs:[17,21,49],IRs:22,Into:95,Its:[32,64,95],K8s:119,NOT:57,Not:[5,10,40,66,119],OPs:[22,24],One:[6,17,44,47,63,66,69],Ops:[59,61,65],PFS:28,QoS:96,Such:[38,57,66,70],TLS:[5,28,95],That:[20,51],The:[5,6,7,9,10,14,16,17,18,20,21,22,24,25,26,29,30,32,36,39,40,43,44,46,48,49,51,52,54,55,56,57,58,61,63,64,65,66,67,69,70,71,74,75,76,89,95,97,109],Their:10,Then:[19,22,33,38,40,51,54,56,95],There:[5,8,9,10,15,17,18,21,22,29,30,31,36,40,47,48,49,52,53,54,57,61,63,64,67,95],These:[7,8,20,25,30,35,50,61,64,65,66],Use:[5,23,29,58,59,66,95],Used:[59,67],Uses:40,Using:[10,31,52,58,59,61,63],VPS:95,Will:84,With:[19,20,25,31,51,55,66,70],YES:18,Yes:[2,42],___embedding_0__:97,___embedding_1__:97,___fc_layer_0__:95,__align__:30,__cuda_align__:30,__device__:30,__doc__:65,__file__:18,__forceinline__:30,__fp16:30,__global__:30,__hadd:30,__half:30,__half_raw:30,__impl__:65,__init__:[25,26,34,40,50,57,70,74,107],__main__:34,__name__:34,__rnn_step__:114,__square_error_cost_0__:97,__va_args__:60,__x:30,_addup_repetitive_outputs_:7,_append_backward_ops_:[7,25],_append_backward_vars_:7,_binari:9,_create_global_var:57,_def:40,_dtype:36,_filer:43,_filter:43,_fwd:43,_fwd_pd:43,_input:43,_librari:9,_live_in:40,_live_out:40,_loss:34,_op:[36,75],_output:43,_presucessor:40,_program:40,_recurrent_group:114,_remove_no_grad_branch_:7,_reorder_input:43,_source_language_embed:114,_src_input:43,_sucessor:40,_target_language_embed:114,_test:9,_update_op:26,_use:40,_value_index:36,a75:30,a_op:75,a_prev:66,aaaaa:12,aaaaaaaaaaaaa:95,abil:34,abl:[5,7,20,21,22,38,50,54,57,119],about:[6,8,9,18,24,29,32,40,48,58,63,65,66,67,95],abov:[5,6,7,8,9,10,14,19,21,22,30,31,32,33,35,43,48,49,50,51,52,54,55,57,65,66,68,69,70,95,108,119],abs:[6,34,81],abs_numerical_grad:6,absolut:[6,81],acc:22,acceler:[11,42,51,52],accept:[5,59],access:[5,9,14,17,18,19,21,22,57],accessmod:95,accessor:57,accord:[6,7,15,22,24,35,49,59,70],account:[59,119],accoust:66,accrodingli:13,accumul:[10,15,26,51,52,66],accur:[6,17,54],accuraci:[26,66],achiev:[20,24,51,52,66,67],acquir:31,across:[7,22,48,66],act1:36,act2:36,act:[8,20,22,36,49,57,70,81,84,86,109,114],act_output:65,act_typ:36,action:[21,95],activ:[9,36,40,49,54,57,61,65,81,84,86,114],actual:[13,25,31,34,36,43,47,52,65,67,69],actual_layout:43,adadelta:81,adagrad:[52,64],adagradoptim:50,adam:[5,15,22,34,83],adapt:[54,69],add:[6,7,8,9,13,17,20,21,22,24,26,30,33,37,39,50,52,56,57,59,61,63,67,68,72,76,78,109],add_activ:57,add_bia:57,add_depend:9,add_execut:9,add_input:[48,74],add_memori:48,add_output:48,add_scalar:[8,49,55],add_sum:57,add_test:[9,74],add_to:82,add_two:[8,48],add_unittest_without_exec:74,addattr:[65,75],addbia:74,addcom:[65,75],added:[8,24,25,30,47,51,52,61,72],adding:61,addinput:[65,75],addit:[7,21,51,54,59,61,70],addition:48,addmemori:43,addop:33,addoutput:75,addprimit:43,addprimitivedesc:43,addr:10,address:[10,15,19,22,89,119],addrow:74,addtyp:65,adjust:[7,25],admin:119,administr:[17,119],adopt:[30,34],advanc:6,advantag:[6,30,31,52,58],adversari:[34,58],affect:8,affili:49,afford:14,aforement:9,after:[7,8,9,14,15,17,22,23,24,25,27,29,30,40,43,51,53,54,57,66,72,95],aftern:54,again:[5,6,10,52],against:95,age:97,agg_level:[110,111],aggreg:[26,51,95],aggregatelevel:[110,111],ago:9,agre:[81,84],ahead:66,alexnet_pass1:105,alexnet_pass2:105,algo:43,algo_hrnn_demo:111,algorithm:[7,14,25,40,43,49,52,61,66,69],all:[5,7,8,9,10,13,15,17,18,19,20,21,22,23,25,27,29,31,34,35,36,39,40,43,44,46,47,48,49,50,51,52,54,55,57,59,65,66,67,78,81,84,95,97,109,113,119],all_output_nam:7,alloc:[15,18,40,43,67,76,109],allow:[5,15,19,20,22,25,31,52,61,95],allow_only_one_model_on_one_gpu:[102,103,105],allreduc:51,alpha:[9,61],alreadi:[9,10,31,43,57,63,78,95],also:[5,7,8,9,13,16,20,21,22,30,31,33,34,35,36,39,40,47,48,49,52,53,54,55,56,57,58,61,63,65,66,67,70,71,108,119],although:[7,51],altogeth:119,alwai:[9,21,44,64,95,97],amazon:[95,96],amazonaw:95,amazonec2fullaccess:95,amazonelasticfilesystemfullaccess:95,amazonroute53domainsfullaccess:95,amazonroute53fullaccess:95,amazons3fullaccess:95,amazonvpcfullaccess:95,ambigu:[58,66],amd64:95,amd:47,amend:72,amodei:66,among:95,amort:51,analys:54,analysi:54,analyz:40,ancestor:[55,57],andd:95,andrew:40,android:116,android_abi:116,android_api:116,android_arm_neon:116,android_native_api_level:116,android_standalone_toolchain:116,android_toolchain:116,androideabi:116,ani:[5,9,10,15,17,18,19,21,22,27,30,31,38,40,44,49,51,52,57,58,60,61,66,69,81,84,95],announc:30,anoth:[5,7,8,18,20,21,31,32,43,49,57,63,65,67,69,95],anroid_arm_mod:116,ans:95,answer:[19,31,95],anymor:51,anyth:[49,58,95],anytim:34,apach:[42,81,84],api:[4,5,7,9,15,16,18,19,20,26,28,33,34,36,48,53,54,59,62,70,71,87,89,90,91,95,97,107,108,116,119],api_pydataprovider2:81,api_shar:9,api_test:9,api_trainer_config_helpers_lay:114,api_v2:110,apiserv:95,apivers:[95,96,97],appar:7,appear:[19,31,35,67],appel:40,append:[7,25,26,49,57,58,66,91,97,114],append_backward:[7,50,107],append_clip_op:25,append_op:[25,39,57],append_oper:57,appleyard:108,appli:[34,35,51,54,63],applic:[19,21,30,31,32,35,57,59,81,84,95,96,108,119],applyl1:13,appreci:66,approach:[22,23,24,51,52,53,61,66,119],approxim:52,apt:[2,107],arbitrari:[22,44],arch:116,archetectur:66,architectur:[30,66],archiv:[45,46,87],area:34,arg:[7,36,50,65,75,83,97],argpars:97,args_ext:97,argu:56,argument:[7,8,13,14,22,50,53,56,57,87,89,90,97,116],argumentpars:97,arithmet:30,arm64:[116,117],arm64_standalone_toolchain:116,arm:[30,116,117,118],arm_standalone_toolchain:116,armeabi:116,armv7:[30,117],armv8:30,arn:95,around:[17,40,57,95,119],arrai:[15,19,21,35,49,55,57,58,59,70,75,81,83,84,89],arrang:70,arrari:89,array_to_lod_tensor:40,arrow:34,articl:[32,35,72],artifact:[62,95],artifici:40,arxiv:[34,66],as_step_input:8,asgd:52,ask:[7,10,17],asr:66,assgin:40,assign:[7,14,19,23,25,30,32,51,66,89,95,119],assigne:66,assignmemt:40,associ:[53,60],assum:[8,22,43],assumpt:22,ast:19,astyp:[58,75],asyc:10,async:[10,24,102],async_count:[102,103],async_lagged_grad_discard_ratio:103,async_lagged_ratio_default:[102,103],async_lagged_ratio_min:[102,103],asynchron:[10,21,51,54],atom:23,attr:[8,19,36,39,43,55,56,57,65,75,81,82,83,114],attr_map:65,attrdesc:55,attribu:43,attribut:[7,8,24,25,39,55,57,59,63,65,70],attributemap:75,attrproto:65,attrtyp:[55,65,75],attrvalu:65,auc:[26,102],audio:66,augment:66,authent:95,author:[28,66,81,84,95],auto:[1,8,13,23,32,43,45,56,59,63,69,70,74,75,76,108],autom:95,automat:[5,7,15,22,24,25,33,50,59,65,66,95],avail:[10,15,24,30,31,40,95,119],averag:[14,81],average_test_period:[102,103],avg:[108,110],avg_cost:[22,109],avoid:[6,8,10,21,22,39,43,51,52,53,108],avx:2,awai:31,await:96,awar:[5,19,21,26,32,48,57,95],awk:98,awni:66,aws:28,aws_account_id:95,awsaccountid:95,awskeymanagementservicepowerus:95,axi:81,axiom:21,b363:96,b8561f5c79193550d64fa47418a9e67ebdd71546186e840f88de5026b8097465:96,ba5f:95,back:[7,10,22,30,34,52],background:[61,66],backpropag:[6,7],backward:[6,8,13,15,25,34,41,42,50,52,53,56,60,61,68,74,75,109],backward_first:114,backward_op:6,backwardactiv:74,bacth:20,baidu:[31,66,96],bake:22,balanc:[24,51,95],bandwidth:[30,51],bare:[96,119],barrier:92,barrierstatset:108,basci:36,base:[0,5,14,20,25,26,30,31,38,43,47,50,51,52,54,59,60,61,67,70,95,109],baseerrorclipattr:25,baselin:66,basematrix:74,bash:[0,1,2,72,95,96,97,101],basi:[81,84],basic:[22,36,43,54,55,59,60,66,70],batch:[5,8,10,12,13,20,21,22,26,27,31,34,37,47,48,49,51,52,66,70,81,84,89,95,96,97],batch_id:[34,81,84],batch_im:34,batch_label:34,batch_norm:[34,66],batch_read:[12,20,58],batch_siz:[22,34,41,49,81,84],batch_szi:34,batch_z:34,batchnorm:[34,66],batchread:20,batchsiz:74,bazel:9,bbbbb:12,bcm2708:118,bdist_wheel:62,beacus:36,beam:114,beam_gen:114,beam_search:[49,113,114],beam_siz:[49,102,103,105,114],becaus:[5,6,8,9,10,15,30,49,53,57,58,61,63,64,68,69,70,71,95],becom:[23,24,63,67],been:[7,9,14,21,31],befor:[7,10,17,21,25,29,32,35,47,52,53,54,58,61,75,78,81,95,119],begin:[7,13,15,20,26,29,35,49,51],beginiter:5,beginn:114,beginpass:5,begintrain:5,behavior:21,behind:[31,70],being:[7,17,25,31,56,58],belong:[20,22,63],below:[8,10,15,22,24,30,31,44,53,58,61,70,71,95],benchmark:[44,66],benefit:[17,18,49],besid:[22,40,47,51],best:[9,43],besteffort:96,beta:34,better:[9,31,40,43,49,69,95,119],between:[7,9,10,15,22,24,30,31,43,46,51,53,60,63,95],bia:[49,57,74],bias_attr:[57,81,83,114],biases_:74,biasparameter_:74,biassiz:74,bidi:96,bidirect:66,bidirectional_lstm:82,big:[19,24,40,119],bigger:10,bilinearfwdbwd:108,bin:[2,90,91,95,96,97,101],binari:[9,18,22,30,32,34,44,90,95],bind:[19,21,30,33,63,67],bit:30,bitcod:117,black:34,blank:95,block0:40,block1:40,block2:40,block:[7,11,13,15,19,21,22,23,24,25,26,27,31,38,40,47,48,50,67,71,76],block_expand:66,block_id:[19,27],blockdesc:[8,35,57,59],blockdescbind:38,blockingcount:23,blueprint:49,book:[2,59,66,77,109,114],bool:[8,20,30,37,39,41,42,43,56,57,64,65,70,71,74,89,103,105],boost:[47,66,67],boot:[113,114,119],boot_lay:114,boot_stat:70,bootstrapp:119,borrow:[34,70],bos_id:114,both:[5,8,9,10,17,20,21,22,24,30,31,34,38,40,47,49,51,54,56,64,66,67,95],bottl:51,bottleneck:54,bottom:66,bound:40,boundari:22,box:34,brace:[8,35],brain:17,branch:[5,8,9,22,31,37,55,62,68,72],break_if:70,brief:[9,15,30,67,76],bring:[31,40],broadcast:[10,51,59,119],broken:72,browser:95,bsd:[0,21,51],bsp:21,bucket_nam:95,buf:13,buf_siz:22,buffer:[13,21,43,44,52,58,63,109],buffer_s:21,buffered_read:58,bug:[72,95],build:[0,1,9,18,22,35,36,40,42,52,61,62,65,66,72,77,78,87,95,97,99,100,101,107,116,117,118],build_doc:77,build_model:34,buildtool:62,built:[9,19,22,30,32,40,47,51,65,66,70,107,108,119],bulk:21,bunch:44,button:95,c11:45,c703c041:72,c99:46,c99e:95,cach:[30,81],cache_pass_in_mem:81,cachetyp:81,cacul:26,caff:[8,31],caffe2:[8,19,21,31],caffe_poli:83,calcul:[6,7,10,15,23,26,30,40],calcut:40,calendar:54,call:[5,6,7,8,13,14,15,16,18,19,21,22,25,32,34,35,40,48,49,50,54,57,59,60,63,65,67,70,95,97,107,108],callback:[25,74],caller:[6,95],can:[5,6,7,8,9,10,13,14,17,18,19,20,21,22,24,25,27,30,31,32,33,34,35,36,38,39,40,43,47,48,49,50,51,52,54,55,56,57,58,59,60,61,65,67,69,70,71,95,108,119],cancel:17,candid:[49,66],cannot:[20,59,63,69,70,78],cantain:36,capabl:[30,53,59],capac:[61,95],capi:[45,87],capi_priv:87,capi_prvi:46,caption:49,card:51,care:[18,40,58,66,67,119],carefulli:66,carpedm20:34,cast:[30,69],cast_to_op_attr:65,cat:[2,97,98],categori:10,categoryfil:96,caus:[10,29],caution:95,cbla:[41,87],cc_:9,cc_binari:9,cc_test:9,cclient:16,cduadevicecontext:[47,67],cento:[4,119],central:61,ceph:12,cephf:[12,18,28],cer:66,certain:[20,39,47,50,54,63,67],certif:[5,28,78,95],cffi:45,cfg:[40,96],cgo:45,ch1:21,ch2:21,chain:[7,35],challeng:[6,10,31,37,67],chan:21,chanc:[5,30],chang:[9,14,18,22,31,43,53,55,58,60,62,63,66,72,95],changes:43,channel:[19,108],chapter:[48,49,66],chapter_data:48,chapter_out:48,charact:[21,66],check:[7,8,9,25,43,56,59,68,72,74,78,83,89,95,103],check_attr:65,check_eq:74,check_grad:[6,75],check_l:74,check_output:75,check_sparse_distribution_batch:[102,103],check_sparse_distribution_in_pserv:[102,103],check_sparse_distribution_ratio:[102,103],check_sparse_distribution_unbalance_degre:[102,103],checker:[6,59],checkgrad:103,checkgrad_ep:103,checkmark:119,checkout:72,checkpoint:[24,56],checksum:28,child:8,chip:31,chmod:95,choic:[9,31],choos:[21,39],chosen:[34,47],chunk:[14,28],circl:35,circular:21,circumst:67,claim:95,claimnam:[95,97],clang:[30,45,72,116],clariti:49,classic:[40,66],classif:35,classifi:34,classification_cost:81,claster:95,clean:[8,9,27,53,59,72,78],clear:[9,49,53,63,69],clearer:[53,57],clearli:63,cli:95,click:95,client:[13,16,59],clip:103,clip_op:25,clip_op_desc:25,clone:[0,1,77,87,107,116,118],close:[21,58,72],close_channel:21,cloud:[9,10,18,28,29,59,119],cludform:95,cluster:[5,8,10,15,22,66,91,93,97],cluster_test_fil:91,cluster_train:[81,93],cluster_train_fil:91,cluster_train_v2:[93,94,98],cm469:95,cmake:[0,1,46,72,75,77,78,87,108,116,117,118],cmake_build_typ:[116,117,118],cmake_c:[116,117],cmake_system_nam:[116,117,118],cmakefil:78,cmakelist:[9,41,42,74],cmatrix:[45,46],cname:95,cnn:96,coars:33,code:[5,7,9,17,20,21,22,24,27,30,33,34,35,39,44,47,50,52,53,54,56,58,59,60,61,65,70,74,95,96],codebas:59,coded_stream:83,codedinputstream:83,colindic:89,collabor:10,collect:54,collectbia:74,colum:89,column:[35,58,89],com:[0,1,2,9,34,62,72,77,78,87,95,96,107,109,116,118,119],combin:[40,50,59,63],come:[22,26,40,55,66,70],comma:15,command:[0,9,13,18,29,74,95,96,97,99,100,105],commandlin:[97,108],comment:[9,36,65,66,97],commit:[9,96],common:[12,61,67],commonli:[29,61],commun:[10,15,16,21,22,24,51,95],compani:31,compar:[6,9,19,59],comparison:[9,31],compat:[30,33,51],compil:[0,9,22,31,36,38,40,47,51,60,64,65,71,101,116,117,118],complaint:9,complet:[7,8,10,14,15,25,28,35,44,47,59,95,96,97,119],complex:[17,21,40,49,59],complianc:[81,84],complic:[22,33,58,69,70],compon:[21,22,36,66,70,71],compos:[5,21,33,36,48,57,59],composit:33,compress:[14,89],compris:7,comput:[5,6,10,21,22,24,27,30,31,32,36,40,44,47,50,51,52,54,60,63,66,67,69,71,75,76,95,109],computationgraph:36,con:51,concat:[34,114],concaten:[34,48,70,81],concentr:59,concept:[5,19,20,21,31,33,34,36,43,48,49,52,53,55,63,70,71],conceptu:[21,27,31,34,36],concern:[5,21,26],concis:[34,70],concret:[59,67],concurr:[10,17,24,54],cond:[8,31,37,55],condit:[14,22,31,37,43,66,68,81,84,96],condtion:34,conf:[83,93,97,111],conf_paddle_gradient_num:[95,97],conf_paddle_n:[95,97],conf_paddle_port:[95,97],conf_paddle_ports_num:[95,97],conf_paddle_ports_num_spars:[95,97],config:[12,29,49,74,84,87,95,96,97,102,103,105,119],config_:[13,103],config_arg:[102,103,105],config_lay:74,config_len:15,config_pars:[41,42,74],config_proto:15,configmap:22,configprotostr:83,configur:[7,13,15,17,18,22,24,31,36,39,57,66,67,69,74,86,119],confirm:29,conflict:[63,72],confus:[34,39],connect:[18,19,22,24,66,79,95,96,119],consid:[7,56,67,119],consider:[47,66],consist:[14,20,21,32,44,55,58,59,60,65,66,71],consol:95,consolid:8,constant:[36,38,39,47,83],constraint:63,construct:[5,27,36,40,48,57,59,63,65],constructbackwardgraph:35,constructoptimizationgraph:35,constructor:[25,30,54,57,59,63,65],consum:10,consumpt:40,contact:17,contain:[0,5,7,8,14,27,34,36,43,44,47,53,54,57,59,60,63,64,65,66,70,71,95,96,97],containerport:95,content:[15,29,44,49,77],content_dir:77,content_len:15,context:[25,43,63,64,67,75,76,81,109,114],contin:95,continu:[7,10,44,66],contrib:61,contribut:[61,66],contributor:59,control:[8,19,21,68,95,96,119],controlflowgraph:40,conv2d:34,conv:[34,43,69],conv_fwd:43,conv_pool_2:22,conveni:[5,7,36,50,65,66],convent:[7,15],convers:[30,31],convert:[12,22,23,24,30,31,32,43,58,60,66],convlut:66,convolut:[34,47,57,67],convolution_algorithm_opt:43,cool:72,cooper:66,coordin:[10,15],copi:[5,14,17,29,35,48,49,51,52,70,81,84,95],copy_from:25,copyright:[81,84],copyvariablewithtensor:69,core:[7,36,39,46,52,53,70,109],coreo:[95,119],corner:59,corpu:66,correct:[6,7,30,51,95],correctli:[7,30,34],corresond:30,correspoind:5,correspond:[5,7,8,9,25,30,36,37,43,47,48,49,57,59,60,61,65,67,68,83],corss_entropi:5,cortex:30,cos:65,cosin:65,cosineop:65,cosineopproto:65,cosineopprotomak:65,cost:[5,7,22,35,50,51,55,56,69,81,84,109],cost_np:56,could:[5,6,14,19,21,22,23,24,30,31,32,48,50,52,53,55,57,58,60,68,95],count:[10,18,26,56,58,66,91,96,103,105,108],counter:[10,14,23,35],cours:[18,47],cover:[31,66],cp27:4,cp27m:[4,62],cp27mu:[4,62],cpp:[6,13,33,41,42,45,46,53,59,71,74,83,97,108,111],cprofil:107,cprofilev:107,cpu:[0,6,18,30,39,47,52,53,54,59,61,62,67,69,75,76,96,105,108,109],cpu_avx_mkl:4,cpu_avx_openbla:4,cpu_kernel:39,cpu_noavx_openbla:4,cpu_ns_:54,cpu_per_pserv:22,cpu_per_train:22,cpudevicecontext:[47,67,75],cpuelapsedu:54,cpuengin:42,cpuinfo:2,cpuplac:[22,39,43,47,67,69,75,76,109],cpusparsematrix:46,crash:[10,108],creat:[5,6,8,10,15,19,20,23,25,26,27,28,29,30,31,33,34,35,43,47,48,50,51,52,53,57,60,61,66,72,74,77,81,83,84,89,96,97,98,119],create_backward_pass:50,create_bias_paramet:74,create_block:57,create_doc_str:65,create_input_paramet:74,create_local_scop:27,create_oper:33,create_optimization_pass:50,create_paramet:57,create_python_ops_creatation_funct:65,create_rnn:8,create_rnn_op:48,create_tmp_var:57,create_tmp_vari:57,create_var:57,create_whileloop:70,creategradientoper:60,creatememori:43,createop:65,createoper:8,createprimitivedesc:43,createstack:95,createvari:8,creation:[33,95],creationd:95,creator:[12,59,60],creator_:60,credenti:29,crf:[67,69],critic:34,crlf:72,crop:67,crop_grad:67,cropgradkernel:67,cropkernel:67,cross:[57,83,116,117,118],cross_entropi:[5,22,34,40,69],crt:28,csc:74,csr:[74,89],csv:83,ctc_error_evalu:66,ctest:[0,1,72,75],ctor:57,ctrl:[0,93],ctx:[43,69,75,76],cubla:47,cublas_handle_:67,cublashandle_t:67,cuda7:[4,86],cuda8:[1,2,4],cuda:[0,9,32,47,54,59,67,75,103,108],cuda_context:32,cuda_dir:[102,103],cuda_fp16:30,cuda_so:[2,78],cuda_visible_devic:81,cudaconfigurecal:108,cudadevicecontext:[32,47,67,75],cudadevicegetattribut:108,cudaelapsedu:54,cudaevent_t:54,cudaeventcr:108,cudaeventcreatewithflag:108,cudafre:108,cudagetdevic:108,cudagetdevicecount:108,cudagetdeviceproperti:108,cudagetlasterror:108,cudahostalloc:108,cudalaunch:108,cudamalloc:108,cudamemcpi:108,cudaplac:[47,67,69],cudaprofilerstart:108,cudaprofilerstop:108,cudaprofilestop:108,cudaruntimegetvers:108,cudasetdevic:108,cudasetupargu:108,cudastream_t:67,cudastreamcr:108,cudastreamcreatewithflag:108,cudastreamsynchron:108,cudeviceget:108,cudevicegetattribut:108,cudevicegetcount:108,cudevicegetnam:108,cudevicetotalmem:108,cudnn:[9,39,43,47,67,69],cudnn_conv_workspace_limit_in_mb:[102,103],cudnn_dir:[102,103],cudnn_kernel:39,cudnnv5:1,cudrivergetvers:108,cuinit:108,cumtim:107,cur_mem:49,curl:95,curli:[8,35],current:[7,8,9,10,13,15,19,24,26,31,39,47,48,49,52,53,54,57,63,69,70,77,95],current_block:[55,57],current_oper:55,current_word:[81,114],curv:5,custom:[5,18,30,34,49,52,59,66,95],custom_batch_read:58,cut:70,cxx:[116,117],cxx_compil:[116,117,118],cxx_flag:[116,117],cxxabi_1:4,cycl:10,cython:45,d3e0:95,d_b0:34,d_b1:34,d_b2:34,d_block:34,d_f:34,d_g:34,d_h0:34,d_h0_bn:34,d_h0_relu:34,d_h1:34,d_h1_bn:34,d_h1_relu:34,d_h2:34,d_loss:34,d_loss_fak:34,d_loss_real:34,d_optim:34,d_step:34,d_t:34,d_w0:34,d_w1:34,d_w2:34,dandroid_abi:116,dandroid_arm_mod:116,dandroid_arm_neon:116,dandroid_standalone_toolchain:116,dario:66,darwin:95,dash:34,dat:12,data:[5,6,8,12,13,14,21,24,26,28,30,31,34,35,36,38,39,40,43,44,47,48,49,50,51,52,53,55,57,59,61,63,64,65,66,67,70,71,76,81,84,86,89,92,96,97,99,102,109,114],data_batch:81,data_i:34,data_lay:[13,81],data_layout_:69,data_read:58,data_reader_creator_random_imag:58,data_shar:70,data_typ:[44,64,66,69,71,82,84,86,89,114],data_type_:[39,47,69],data_x:34,datacent:[12,29],datacenter1:12,datacenter2:12,datacenter_1:12,datacenter_2:12,datacenter_nam:12,datafeed:109,dataflow:36,dataflow_analysi:40,datalayout:69,dataparallel:22,dataprovid:[81,83,97],dataprovider_convert:66,dataset:[12,18,22,52,58,66,84,86,91,107,114],datatransform:69,datatyp:[39,43,64,66,69,71],dcgan:34,dcmake_build_typ:[77,87],dcmake_install_prefix:[87,116,117,118],dcmake_system_nam:[116,117,118],dcuda_arch_nam:1,dcudnn_root:1,ddim:[20,47,67,76],dead:10,deal:[7,119],deb:72,debug:[6,7,22,29,31,57,77,107],debug_str:36,decayr:13,decent:14,decid:[5,17,34,44,52,60,61,64],declar:[8,34,48],decod:[66,113,114],decoder_boot:114,decoder_dim:49,decoder_group_nam:114,decoder_input:[49,81,114],decoder_mem:[49,114],decoder_s:[81,114],decoder_st:114,deconv:34,decor:20,decrement:23,decrementcount:23,decrypt:95,deduc:59,deep:[7,17,21,27,34,35,40,42,54,59,61,66,67,108],deeper:32,deepspeech2:41,def:[5,6,7,12,18,25,26,33,34,36,39,40,48,49,50,57,58,65,70,74,75,81,83,84,97,114],def_block:34,default_block:34,default_decor:97,default_devic:105,default_main_program:109,default_param_attr:57,default_st:70,default_startup_program:109,default_valu:105,defaultdict:40,defaultinfervartyp:38,defect:53,defer:17,defin:[5,7,8,9,10,17,19,20,23,24,25,30,31,32,33,34,36,39,40,47,48,51,55,57,58,59,63,65,67,70,75,81,84,109],define_py_data_sources2:83,definit:[7,8,10,14,22,27,32,39,55,60,65,70,109],definiton:33,delai:[52,67],delet:[18,28,72],deletestack:95,delimit:83,deliv:119,delta:6,demand:[10,67],demo:[59,96,99],dens:[15,16,64,66,95],dense_arrai:82,dense_vector:[84,86,89],dense_vector_sequ:89,dense_vector_sub_sequ:89,densescann:66,dep:9,depart:66,depend:[8,9,10,18,20,21,22,24,36,51,56,64,119],dependent_var:56,deploi:119,deploy:[36,44,59,95,119],deprec:66,depth:[8,31,66],dequeu:24,deriv:[5,20,22,25,37,50],desc:[8,25,43,44,57,65,70],desc_:8,descend:70,descent:[10,52],descproto:44,describ:[5,7,8,9,14,19,22,27,32,39,43,44,48,49,53,55,57,59,64,65,69,71,95,96],describestack:95,describestackev:95,describestackresourc:95,descripotor:43,descript:[8,9,38,42,44,47,60,64,66,69,71,95,97],descriptor:[21,43,69],deseri:[44,53],deserializ:59,desgin:35,design:[7,13,20,39,40,45,52,54,61,119],desir:[10,22,52,95,96],destin:[15,29],destroi:[8,27],destruct:63,destructor:54,detail:[6,7,14,18,22,24,29,31,34,36,40,43,44,47,48,54,57,61,63,67,69,70,71,95,119],detect:[38,72],determin:[8,22,40,47,59],dev:[0,2,59,72,78,107,116,119],dev_ctx:[8,43,54],devel:62,develop:[7,9,31,38,53,54,57,60,62,66,72,109,117],devic:[2,6,19,22,26,30,36,42,43,47,51,53,54,59,69,76,78,105,109],device_:54,device_context:[43,75],devicecontext:[8,47,54,75],deviceid:[42,105],deviceid_:42,deviceplac:67,devot:66,dhcp:119,diagram:48,diamond:34,dict:[7,57,83,97],dict_dim:81,dict_siz:[13,49],dictionari:[5,6,57,81],did:53,diff:[72,81],diff_mat:6,differ:[6,7,8,9,10,15,17,20,22,23,24,25,26,27,30,31,34,36,37,40,43,47,49,51,52,54,56,60,63,66,68,69,70,71,95],differenti:33,difficult:[6,31],dig:95,digraph:36,dilat:43,dim0:75,dim1:75,dim:[13,43,44,48,59,64,67,71,74,75,76],dim_:[67,76],dimens:[34,59,64,66,67,70,76,81],dimension:89,dios_arch:117,dios_enable_bitcod:117,dios_platform:117,dios_use_veclib_for_bla:117,dir:[78,97,116],direcit:66,direct:[31,40,52,66],directli:[9,16,18,20,22,30,39,53,65,69,70],directori:[9,12,17,28,29,67,72,77,96,108],disabl:[54,83],disadvantag:[52,57],discard:[10,14,49,72,103],discexp:83,discov:10,discoveri:95,discrim:34,discuss:[5,8,14,15,16,22,43,66],disk:44,dispatch:[22,53],displai:[18,29],dist:[1,62,78],dist_train:[5,18],distinguish:9,distribut:[8,14,15,16,17,19,21,26,32,51,59,66,71,81,84,94,99,100,103,119],distribute_test:[102,103],distributedli:22,disucss:5,divid:[7,26,65,71],diy_beam_search_prob_so:[102,103],dnn:[43,66,78],dns:95,do_forward_backward:58,doc:[36,48,70,75,77,91,93,97],doc_cn:77,docker:[1,2,62,72,77,78,95,96,97,99,100,101,116,119],docker_build:5,docker_clust:[93,98],docker_push:5,dockerfil:[0,72,97,116,118],document:[6,20,22,28,35,48,49,54,59,66],doe:[10,14,15,17,18,19,20,22,24,27,30,36,40,48,53,57,59,60,61,109],doesn:[5,8,19,21,58],doing:[13,17,22,35],domain:95,don:[5,9,33,35,40,58,66,95],done:[7,9,10,14,15,22,23,38,40,44,52,60,61,66,72,95,97,108],dot:75,dot_period:[97,103,105],doubl:[22,30,35,54,69,75,103],down:[66,108],download:[9,10,13,17,28,78,96,119],doxygen:72,dozen:9,dpython_execut:78,dpython_include_dir:78,dpython_librari:78,draw:49,drive:63,drop:[20,49],drop_fc:82,drop_rat:82,dropout:82,dropout_r:82,drpi_arm_neon:118,drpi_toolchain:118,drwxr:96,ds2:66,dst:[15,43],dst_primitive_desc:43,dtoh:108,dtype:[21,22,36,57,83,109],due:[14,17,34,40,49,57,107],dummi:14,dump:[44,90],dump_config:90,dump_v2_config:90,duplic:24,durat:14,dure:[7,8,10,14,17,18,26,31,40,51,52,54,57,59,66,71,95,119],duse_eigen_for_bla:116,dwith_c_api:[46,87,116,117,118],dwith_doc:77,dwith_golang:87,dwith_gpu:[1,77,87,118],dwith_mkl:[77,87],dwith_profil:108,dwith_python:[46,87,118],dwith_swig_pi:[46,87,116,117,118],dwith_test:[1,75,117],dwith_tim:108,dynam:[15,46,48,57,58],dynamic_cast:74,dynamic_recurrent_op:70,e2e:119,each:[6,7,9,10,13,14,15,17,18,19,20,21,22,25,26,27,32,35,38,40,43,47,48,49,51,53,54,56,57,58,59,60,63,64,65,66,67,68,69,70,71,95,119],eager:31,earli:[30,32],eas:38,easi:[6,7,49,52,58,59,61],easier:[5,24,30,31,58,68,70],easili:[5,34,51,54,58,60,63,67],echo:[2,78],edg:40,edit:[21,95],editor:57,edu:[95,96],eeoi3ezpr86c:95,effect:95,effici:[22,44,58,66,67],effort:[22,66],efs:95,efs_dns_nam:95,efsvol:95,egd:40,eigen:[30,47,52,59,61,67,75],eigen_device_:67,eigen_use_gpu:75,eigenmatrix:76,eigentensor:76,eigenvector:76,either:[5,22,34,37,38,48,52,61,81,84],elabor:66,elb:95,elbapis:95,electr:40,electron:96,element:[6,14,21,24,36,49,59,75],element_typ:15,elementari:59,elif:[5,65,68],els:[2,5,13,18,21,22,24,25,31,34,37,38,39,40,63,65,68,74],elsewher:54,emac:0,emailweixu:9,emb1:13,emb2:[13,111],emb:[81,83,96],emb_para:83,emb_param_fil:83,emb_sum:81,embed:[5,8,13,24,38,49,64,70,83,114],embedding_lay:[13,81,111],embedding_nam:114,embedding_s:114,emplace_back:74,emploi:[7,25,65],empti:[7,10,20,21,49],emul:30,enabl:[0,8,9,14,19,24,25,36,54,95,108],enable_grad_shar:[102,103],enable_parallel_vector:103,enc_proj:114,enc_vec:114,encapsul:15,encod:[14,49],encoded_proj:114,encoded_sequ:114,encoded_vector:114,encoder_ctx:49,encoder_ctx_expand:49,encoder_ctx_proj:49,encoder_dim:49,encoder_out_seq:49,encoder_s:114,encount:13,encourag:[22,27],encrypt:95,encrypt_decrypt:95,end2end:119,end:[7,8,22,25,32,36,40,49,53,54,58,63,66,68,72,114],end_pass:5,endforwardbackward:81,endian:44,endif:[47,54],enditer:[5,81,84],endpass:[5,84],endpoint:[12,95],endtrain:5,engin:[18,42,43,66],english:66,enough:[7,8,39,40,47],enqueu:24,ensur:[10,43,51,63],enter:[8,27],enterpris:59,entir:[15,17],entiti:[8,63],entranc:27,entri:[14,18,38,95],entropi:57,entry_point:18,enumer:[47,83],env:[77,81,95,97,107],environ:[5,20,22,78,95,96,108],environmenterror:91,eol:72,eos_id:114,epoch:34,epol:21,equal:[10,70,75],equat:[40,75],equival:[5,8,19,21,25,31,37,65,119],eras:20,erlang:21,error:[5,6,14,29,30,31,43,63,66,78,83,87,95,103],error_clip:25,error_clip_callback:25,error_clipping_threshold:81,errorclipbyvalu:25,especi:42,essenc:[5,7],essenti:[5,27,30],establish:19,estim:[5,24,52],eta:96,etc:[8,21,22,26,43,51,52,58,63,66,95,119],etcd:[10,14,15,17],etcd_addr:15,eth0:[95,97],etyp:21,eval:[8,26,34,59],eval_program:26,eval_result:26,evalu:[17,36,56,66,90,108,109],even:[5,30,51,57,58],evenli:[15,95],event:[81,84,96],event_:54,event_block:54,event_handl:[5,81,84],eventkind:54,eventlist:54,eventu:[22,70],everi:[5,10,14,15,17,25,26,35,36,38,40,43,47,48,51,57,63,65,69,84,109],everyth:[22,24,34],evid:53,evolv:31,exactli:[20,81,95],exampl:[8,18,20,22,24,26,29,31,32,33,34,35,36,38,40,43,47,48,49,53,54,55,57,58,59,60,61,64,67,68,69,70,95,109],exc_path:78,except:[17,19,31,35,54,66,70,84],excess:40,exchang:53,exe:[22,109],execut:[9,10,14,18,19,20,21,22,26,27,32,34,36,40,43,51,54,60,71,95],executioncontext:[43,69,75,76],executor:[19,22,26,30,31,32,34,50,55,57,69,71,107,109],exist:[5,8,10,20,29,31,49,57,58,60,65,67,70,76,95],exit:[15,29,96],exp:83,expand:49,expand_a:110,expand_level:110,expandlevel:110,expect:69,expected_desc:43,expected_kernel_kei:69,experi:[44,66],expert:9,expir:10,explain:[10,19,31,33,35],explan:[18,19,22,63,69],explicit:[20,54,70,74],explicitli:[5,22,27],explod:25,explor:[49,61],expos:[7,16,21,43,44,67,70,95],express:[5,24,26,36,40,81,84,95],extend:[52,70],extens:[17,24,49],extent:46,extern:[9,42,45,46,59,66],extern_mklml:78,external_librari:9,extingrad_:42,extinval_:42,extoutgrad_:42,extoutval_:42,extra:[22,61,67,119],extraattr:105,extract:[31,53,66,95],extralayerattribut:[81,82],extrem:[19,31],f1205:83,f120da72:96,f7e3:95,fa0wx:96,fabric:[93,94],face:[9,61],fact:[19,31,51,55,57],factori:45,fail:[10,14,49,78,83,96,103],failur:[10,15],fake:34,fake_imag:58,faked_imag:34,fall:[30,56],falloc:28,fals:[6,7,8,31,37,39,41,48,55,56,58,64,71,74,75,81,84,86,89,91,96,105,114],false_block:[8,37,55],false_label:58,false_neg:26,false_posit:26,false_read:58,familiar:21,faq:115,far:[25,70],fashion:22,fast:[14,31,108],faster:[10,31],fastest:31,father:7,fault:[14,59],fbd1f2bb71f4:96,fc1:[36,74,105],fc1_bia:36,fc1_weight:36,fc2:[36,105],fc3:[36,105],fc4:105,fc8a365:95,fc8a:95,fc_layer:[57,65,81,83,105],fc_op:65,fc_out:8,fc_output:65,fc_without_b:8,fclayer:74,fcop:33,feasibl:52,featur:[7,22,30,36,51,54,66,72],feed:[5,22,35,48,61,84,109],feed_dict:34,feed_list:109,feeder:[22,109],fetch:[10,13,20,22,56,72,78,109],fetch_list:[22,57,109],fetch_op:56,few:[9,10,21,22,40,52,58,64,66],fewer:[21,57],fft:66,field:[8,36,38,44,56,57,60,64,65,81,95],fifth:35,figur:[5,9,22,24,34,42,48,54,57,66],file:[5,7,9,10,12,14,15,17,18,20,21,28,29,31,32,36,44,46,58,59,66,67,71,72,75,81,84,109,119],file_nam:83,filelist:66,filenam:[12,57,81,107],fileoffset:28,filesystem:[17,18,22,28,95],fill:[10,14,47,57,89,95],fill_zero_grad:59,fill_zeros_like_op:7,filter:[25,43],find:[8,10,17,21,30,36,43,49,63,69],find_var:6,findmemori:43,findop:8,findprimit:43,findprimitivedesc:43,findvar:[8,63],fine:[14,33],fingerprint:95,finish:[10,14,17,18,27,40,51,65,95,96],first:[5,7,8,10,14,17,18,19,22,27,29,31,34,35,36,43,48,49,55,56,57,59,64,65,66,67,68,70,75,76,89,95,119],first_seq:114,firstseen:96,fit:[30,39,40,44,49,59],five:55,fix:[22,40,45,57,66,72,89],flag:[41,42,54],flatten0:36,flatten:[36,55,57,76],flatten_result:81,flexibl:[5,15,22,31,35,39,48,49,52,58,67,70],flist:91,fliud:19,float16:21,float16_t:30,float32:[22,30,33,34,57,58,75,83,109],float_to_half_rn:30,floor:83,flow:[8,19,21,48,54,62,68],fluid:[7,20,22,24,27,47,54,57,67,68,69,107],fluid_cuda_create_tensor:32,fluid_cuda_mult:32,fluid_cuda_read:32,fly:7,fmt:[21,83],fname:83,fnt03:95,focu:[21,36],folder:[9,12,18,29,95],follow:[5,6,7,8,9,10,14,18,19,20,21,22,24,27,30,31,32,33,34,35,36,37,38,40,43,47,48,49,51,52,54,55,56,57,58,59,60,61,63,64,65,66,67,68,69,70,95,99,100,109,119],footprint:32,forbid:5,forc:[51,57,69],force_cpu:39,force_cudnn:39,force_load:45,forest:8,forev:21,forget:5,form:[20,21,26],formal:69,format:[6,14,20,22,30,31,47,49,66,70,72,74,86,89,95],former:[5,9,31,40,52],formula:[6,40],forth:34,forward:[6,7,8,13,15,25,31,34,41,42,43,44,50,53,55,58,59,60,61,64,74],forward_infer:43,forward_list:54,forward_op:6,forward_train:43,forwardactiv:74,found:[30,55,61,63],four:[21,26,31,35,43,47],foward:56,fp16:[30,59,71],fp32:[47,59,69,71],fp64:[47,71],fparam:83,fpga:[47,109],fpgadevicecontext:67,fpgaengin:42,fpgaplac:[47,67],frame:[27,59,66,70],framework:[5,7,8,21,25,26,30,31,36,47,51,52,54,55,59,61,63,65,67,75,107,109],free:[32,67,119],freememoryop:32,frequenc:66,frequent:[14,58,59,61,67],fresh:17,friend:63,friendli:34,from:[6,7,8,9,10,12,13,14,15,19,20,21,22,24,25,26,29,30,31,33,34,35,36,37,39,40,43,48,49,50,51,53,55,57,58,59,60,63,66,67,68,69,70,75,76,78,90,95,96,107,108,113,119],from_no_sequ:110,from_sequ:110,from_tar:84,fromfil:[58,83],fromstr:83,front:[36,40],fulfil:108,full:[10,17,21,48,51,52,119],full_matrix_project:114,fulli:[22,24,66,119],fullsiz:13,fullyconnect:[36,57],fullyconnectedlay:74,func:[14,19,32,60],functor:[33,36],fundament:[20,21,24,30,59],further:[20,65,119],futur:[17,22,30,40,48,59],fvs:65,fwd_desc:43,fwd_op:60,fwd_primit:43,fwd_primitive_desc:43,fwd_var:25,g_b0:34,g_b1:34,g_b2:34,g_block:34,g_command_config_arg:[41,42],g_h0:34,g_h0_bn:34,g_h0_relu:34,g_h1:34,g_h1_bn:34,g_h1_relu:34,g_h2:34,g_im:34,g_loss:34,g_optim:34,g_program:57,g_state:54,g_step:34,g_w0:34,g_w1:34,g_w2:34,gan:5,gangliao:9,gatedrecurrentlay:41,gather:[7,40,51,53,75],gaussian_normal_random:34,gcc:[0,1,30,32,45,59,116,118],gcc_3:4,gcreators_:65,gemm:41,gen_proto_pi:77,gen_rand_param:83,gender:97,gendrated_id:49,gener:[5,6,7,8,9,10,12,14,15,17,19,22,31,33,38,40,43,47,51,52,55,56,57,58,59,60,64,65,66,67,68,70,84,89,95,97,105,108],generated_id:49,generated_scor:49,generatedinput:[113,114],genr:97,get:[2,6,7,8,9,10,14,15,17,18,20,28,31,34,36,39,40,41,42,43,47,48,49,54,57,59,60,63,65,69,70,74,81,89,93,95,96,98,107],get_all_op_proto:65,get_block:57,get_config_arg:105,get_data:96,get_dim:6,get_float_el:6,get_grad:81,get_grad_op_desc:7,get_input_lay:74,get_numeric_gradi:6,get_numerical_gradi:6,get_output:6,get_program:40,get_sample_from_lin:81,get_support:[4,78],get_symbol:36,get_tensor:6,get_vari:8,get_worker_addr:19,getactualkerneltyp:39,getattr:25,getbatchs:74,geteigendevic:76,getengin:43,getenv:[5,18,91,97],getexpectedkerneltyp:[39,43,69],gethostbynam:97,gethostnam:97,getidmap:97,getinfervartyp:38,getinput:74,getinputgrad:74,getinputvalu:74,getkerneltyp:30,getkerneltypeforvar:69,getlayeroutput:81,getlibrari:43,getmat:13,getoptconfig:13,getoutputgrad:74,getoutputvalu:74,getparam:13,getparameterconfig:13,getparameterptr:74,getparameterspars:13,getparametersremot:13,getplac:[43,67,75,76],getpodlist:97,getsiz:74,gettask:14,gettempl:95,gettensor:69,gettranspos:74,getw:74,getweight:74,getwgrad:74,gflag:87,gflags_complet:87,gflags_declar:87,git:[0,1,62,72,77,78,87,116,118],github:[0,1,9,34,47,62,72,77,78,87,107,109,116,118],give:[10,48,57,59,69,95],given:[7,15,17,21,24,25,31,33,34,49,58,61,70],glibc:[116,118],glibc_2:4,glibcxx_3:4,glide:9,global:[0,5,8,9,10,32,36,39,53,54,59,63,65,67,69,95,108],global_block:57,globalstat:108,globalstatinfo:108,glog:87,gnueabihf:118,go_librari:9,go_test:9,goal:[20,21,24,30,35,51,59,66],gob:14,godep:9,godoc:45,goe:[10,31,37,63,109],going:[7,33,52,119],golang:9,good:[21,34,52,57,58,61,119],googl:[5,54,59,83,87,107,116],googleapi:95,googlenet:42,goroutin:[19,21],got:[39,63],govern:[81,84],gpg2:95,gpg:95,gprof:107,gprotos_:65,gpu:[0,2,4,6,18,21,26,30,40,47,51,52,53,54,59,61,62,67,69,78,86,89,91,105,108,109,119],gpu_id:[81,103,105],gpu_per_train:22,gpudevic:67,gpugpu_id:102,gpukernel:59,grab:10,grad:[6,7,15,25,42,57,64,81,103],grad_info_map:7,grad_n:25,grad_nam:25,grad_op:25,grad_op_class:59,grad_op_desc:25,grad_op_maker_:60,grad_op_typ:[59,60],grad_op_type_:60,grad_s_block:7,grad_share_block_num:[102,103],grad_to_var:[7,25],grad_var_nam:6,gradient:[10,14,21,23,25,35,38,50,51,52,53,57,59,64,90,91,92,97,103],gradient_clipping_threshold:81,gradient_flat:6,gradient_machin:[46,87],gradientmachin:[46,53,97],gradientmachine_:13,gradopdescmak:[38,60],gradopdescmakerbas:60,gradopmak:60,grain:33,gram:66,grant:95,graph:[7,8,9,10,19,21,22,23,24,26,31,34,48,51,52,55],great:[24,66,119],greater:[25,52,89],greaterthan:65,greedi:66,green:[19,34],grep:[2,98],groudtruth:114,group:[14,36,43,67,119],group_input1:114,group_input2:114,group_input:114,grpc:119,gru:[49,66],gru_decod:114,gru_decoder_with_attent:114,gru_out:49,gru_step:[49,114],grumemori:[82,114],gserver:[41,42,74],gsizex:108,gtx:40,guarante:[43,57],guard:13,guest:4,guid:[28,40,59,95,96],gzip:[14,96],h0_bn:34,h_prev:8,hadoop:5,half:[30,95],half_to_float:30,hand:[40,59,66,67],handi:9,handl:[5,7,18,19,22,36,40,43,47,53,58,63,67,70,109],handler:8,hannun:66,happen:[14,65],hard:[22,31,49,66,70,95],hardwar:[31,32,67],has:[5,6,7,8,9,10,14,15,17,20,21,22,24,25,26,30,31,34,36,40,44,47,49,51,54,55,59,64,65,67,69,89,95,108,109,119],has_kei:[7,25],has_var_recurs:7,hasdependentvar:56,hash:[47,51],hasn:31,hasnext:20,have:[5,6,7,8,9,10,14,15,17,18,20,21,22,24,25,27,30,31,32,33,34,35,39,40,43,44,47,48,49,51,52,53,54,55,57,58,59,60,63,64,66,67,69,71,95,119],haven:31,hdf:12,head:[75,98],header:[15,44,46,59,67,83],headip:98,height:[8,45,58,74,75,83,89],height_:64,held:10,hello:5,help:[8,29,31,36,43,49,58,59,70,72],helper:[22,43,60,70],henc:[22,52,57,60,61,63],here:[5,9,10,16,21,24,25,27,29,31,35,36,43,47,48,58,61,65,89,95,119],heterogen:[22,24,54],heurist:[24,49],hidden:[50,57,82,83,95],hidden_a:83,hidden_b:83,hidden_out:8,hierach:113,hierarch:[55,57,59],hierarchi:59,high:[30,51,66,67,119],higher:[33,48,70],highest:8,highli:[66,70],him:5,hint:39,histor:33,hl_get_sync_flag:74,hold:[5,7,10,14,16,20,21,30,34,36,38,40,63,65,67,69,76,95],holder_:[67,76],home:[2,12,29,95,96,97,98,107],honor:14,host:[9,18,54,95,96],host_c:[116,117,118],hostfil:98,hostnam:95,hostnetwork:97,hostpath:[96,97],hostport:95,hous:86,how:[5,8,10,14,19,20,21,22,27,29,31,33,36,39,43,48,49,53,54,61,65,69,95],howev:[6,7,17,21,22,27,31,40,47,52,53,57,58,60,61,64,65,66,67,69,95],howto:[91,93,97],hpp:[30,45],htod:108,http:[0,1,2,9,18,34,62,72,77,78,81,84,87,95,96,107,109,116,118,119],hub:62,huge:52,human:[54,66],hyper:34,hyperparamet:61,i1116:97,i1117:108,i386:117,iOS:117,iamfullaccess:95,iamusersshkei:95,icc:32,icml:66,id_rsa:98,idea:[9,21,31,32,52,58,61],ideal:[22,69],ident:[60,95],identifi:[37,47],idmap:97,ids:[49,81,89],ids_arrai:89,idx:[14,20,34,40,74],ies:29,if_els:68,if_else_op:7,ifdef:[47,54],ifels:[8,55],ifelseop:55,iii:66,iil:83,illustr:[10,15,21,22,33,48],im_siz:34,imag:[0,5,22,31,34,35,49,50,55,58,66,72,95,96,97,99,100,119],image_a:58,image_b:58,image_conv_lay:66,image_fil:58,image_lay:58,image_nam:5,image_path:58,image_reader_cr:58,imagenet:12,imagepullpolici:[95,97],images_reader_cr:58,imagin:35,imgsiz:108,imgsizei:108,imgsizex:108,immedi:[40,43,52,61,95],immutable_paramet:5,imper:19,imperfect:59,implement:[8,14,15,16,17,18,19,21,22,24,31,33,36,37,38,40,43,45,46,47,49,53,56,63,65,66,67,68,69,70],implemet:13,impli:[9,81,84],implicitli:19,imposs:[20,49,119],impractic:69,improv:[23,24,40,59,66,95],in_arg:89,inarg:13,inbound:95,inc_path:78,includ:[5,8,9,15,18,20,21,30,31,34,36,40,45,46,48,49,54,55,57,59,65,72,75,87,95,108,116,117,118],inclus:49,incom:[19,39],increas:[10,14,30,83],increment:[21,26,35,40],incupd:74,inde:21,independ:[6,7,15,23,63,67,119],index:[6,7,8,10,14,19,55,57,70,95],indic:[7,8,15,27,34,48,55,60,64,67,70,89,95],indice_map:70,indices_map:70,individu:[10,51,95],industri:[10,44,119],ineffici:[53,69],infer:[5,7,8,10,26,31,37,38,39,40,41,45,47,56,57,59,64,66,69,84,86],infer_shap:57,infer_var_type_:38,inferer:66,inferfac:38,inferior:17,infershap:[8,57,59,75,76],infershapecontext:[75,76],infervartypefn:38,info:[30,48,74,81,84,93,97,119],inform:[8,18,29,36,39,40,43,44,47,48,51,57,61,63,64,95],infrastructur:[31,95],ingrad_:42,ingredi:[21,66],inherit:[8,20,50,59,67],ininst:5,init:[8,23,34,42,48,49,74,84,86,91,95,97],init_attr:57,init_model_path:[102,103,105],initi:[7,9,14,19,22,23,24,26,35,48,51,52,57,61,65,70,86,103,109],initial_max:83,initial_mean:83,initial_min:83,initial_std:83,initialize_op_attr:57,initrd:119,inlin:[67,76,95],inner:[81,89],inner_pos_arrai:89,inner_seq_pos_arrai:89,input0:76,input1:76,input:[6,7,8,13,17,19,20,22,23,24,25,26,30,31,32,33,34,35,36,38,39,40,42,43,47,48,49,52,53,56,57,58,59,60,63,65,66,67,69,70,74,75,76,81,82,83,84,86,89,97,105,109,110,113,114],input_data:74,input_data_target:74,input_hassub_sequence_data:74,input_index:74,input_label:74,input_lay:74,input_nam:5,input_seg:70,input_sequence_data:74,input_sequence_label:74,input_sparse_float_value_data:74,input_sparse_non_value_data:74,input_t:74,input_to_check:6,input_typ:81,input_valu:6,input_var:[6,57],inputbuff:13,inputdef:74,inputgradi:60,inputlayers_:74,inputs_to_check:6,inputsizechang:43,insert:[7,25,32,51,56,59,60,72],insid:[7,10,22,24,25,26,39,43,53,54,58,59,60,95],inspir:54,instal:[1,2,4,18,42,62,72,77,78,86,87,96,101,107,116,117,118],install_android:116,instanc:[6,8,10,12,16,19,20,22,23,25,27,32,37,43,48,49,52,57,59,60],instance_ip:95,instanti:[10,27,109],instead:[6,7,9,13,18,19,21,22,30,31,35,36,66],instrins:30,instruct:[8,35],int16:71,int32:[47,55,70,71,103],int64:[22,28,47,64,69,71],int64_t:54,int8:47,integ:[14,18,19,21,30,45,49,84],integer_sequ:81,integer_valu:[81,84,89],integer_value_sequ:[49,66,89,114],integer_value_sub_sequ:89,integr:119,intel:[31,47,67],intellig:40,inteloptimizedpaddl:42,intens:66,inter:22,interact:[22,95],interchang:[35,59],interconnect:51,interest:[19,30,51],interfac:[8,14,18,20,29,36,51,53,59,60,66,67,95,119],intermedi:[22,29,32,34,40,50,66],intern:[30,66,95,107],internel:42,internet:[9,10,119],interpret:[27,31,32,71],intrins:[19,27,30],introduc:[8,10,20,34,41,44,61,63,65],intuit:[17,59],inval_:42,invalid:[58,63],invent:31,invoc:[9,33,59],invok:[7,20,22,25,53,57,59,60,65,69,95,108],involv:49,ios:117,ios_arch:117,ios_deployment_target:117,ios_development_root:117,ios_enable_bitcod:117,ios_platform:117,ios_sdk_root:117,ios_use_veclib_for_bla:117,ip_str:97,ips:[95,97],ipt:[57,65,83,114],ipx:119,ipython:5,is_cpu_plac:43,is_inf:90,is_mkldnn_librari:43,is_seq:114,is_stat:83,is_target:56,is_tensor:65,is_test:43,is_traget:56,isbinari:89,isinst:[25,81,84],ismkldnnkernel:43,ispodallrun:97,isspars:74,issu:[9,34,66],issue_numb:72,istag:62,item:[17,30,58,86,97,119],iter:[5,10,22,31,32,40,43,52,54,58,66,70],iter_multiple_input_and_param:57,its:[5,7,8,10,14,19,20,21,24,25,26,31,32,34,35,36,38,40,44,48,49,51,52,53,56,57,59,60,63,64,65,67,69,95,108],itself:[7,10,17,20,32,43,52,63],ivector:[89,90],ivs:65,java:[8,45,55,59],jeremi:108,job:[7,17,19,22,25,59,97,102,103,105],job_desc:22,job_dispatch_packag:93,job_nam:[18,95,97],job_namespac:[95,97],job_path:[95,97],job_path_output:97,job_workspac:93,jobdesc:22,jobnam:[22,97],jobpath:[95,97],jobport0:95,jobport1:95,jobport2:95,jobport3:95,jobselector:97,jobserv:18,join:10,jpg:20,json:[36,66,95,96],juditski:52,jupyt:18,just:[9,14,15,19,22,31,32,34,38,43,52,53,57,58,59,60,61,63,64,95],jx4xr:95,jypyt:5,k8s:[19,97,119],k8s_data:[95,97],k8s_job:5,k8s_token:5,k8s_train:[95,97],k8s_user:5,kafka:12,kcpu:54,kcuda:54,kdisabl:54,kebilinearinterpbw:108,kebilinearinterpfw:108,keep:[10,21,31,32,35,49,52,57,63,65,119],kei:[0,6,7,8,10,12,14,28,30,39,43,59,60,65,66,70,72,81,97,108],kenlm:66,kept:[40,57],kera:61,kernel:[6,21,30,32,39,42,52,54,61,64,66,67,69,75,76],kernel_hint:39,kernel_type_for_var:69,kerneltyp:[39,43],key1:103,key2:103,key_pair_nam:95,keyid:95,keymetadata:95,keypair:95,keyserv:95,keystat:95,keyusag:95,keyword:[57,68,97],kforcecpu:39,kill:[10,95],kind:[5,6,10,16,20,22,25,32,35,39,43,50,51,54,67,69,71,81,84,95,96,97],kind_:54,kmark:54,kms:95,knchw8c:47,knchw:47,knhwc:47,know:[5,14,19,20,40,44,95],knowledg:66,known:[7,8,21,31,33,48],kpoprang:54,kpushrang:54,kqueue:21,kselectedrow:64,kstate:54,kube_cluster_tl:5,kube_ctrl_start_job:5,kube_get_workers_addr:19,kube_list_containers_in_job_and_return_current_containers_rank:5,kubeconfig:95,kubectl:[93,96,97,98],kuberent:[10,95],kubernet:[5,10,19,22,59,94,97,99,100,119],kubernetes_service_host:5,kusecudnn:39,kusemkldnn:39,kwarg:[26,36,57,65],l1_regularization_op:61,l2_regularization_op:61,l2regular:81,l93:13,label:[22,26,31,34,35,36,50,55,58,66,69,81,84,96,109],label_fil:58,label_lay:58,label_path:58,labelselector:97,lag:103,lambda:[19,25],lan:101,languag:[19,21,31,35,40,54,59,63,66,68,81,84],larg:[22,24,25,40,44,52,66,72],larger:[40,68],larger_than:[8,37,55],last:[7,25,40,48,54,55,110],last_seq:[49,111],lastseen:96,latenc:[30,66,95],later:[9,59,61,66,67,76,95],latest:[2,8,9,10,17,62,72,77,78,96,97,116],latter:[52,70],launch:[43,95],launcher:5,law:[81,84],layer1:[81,110],layer2:[81,110],layer:[7,8,13,19,21,22,24,31,34,35,37,50,52,55,58,59,61,65,66,67,70,74,81,83,84,86,89,90,109,110,113,114],layer_0:74,layer_att:82,layer_attr:[81,82,105,114],layer_expand:110,layer_first_seq:110,layer_help:39,layer_last_seq:110,layer_nam:81,layer_num:105,layer_pool:110,layer_s:89,layer_typ:[41,42],layerbas:74,layerconfig:74,layergradutil:74,layerhelp:[39,57],layermap:74,layers_test:78,layout:[43,69],layout_:[39,47],layouttyp:39,lazi:[52,61],ld_library_path:87,lead:[20,40,47],leaki:34,learing_r:50,learn:[2,5,7,15,17,21,22,24,27,34,35,40,42,49,51,52,54,58,59,61,67,108],learning_r:[15,22,81,83,109],learning_rate_arg:83,learning_rate_decay_a:83,learning_rate_decay_b:83,learning_rate_schedul:83,leas:10,least:10,leav:[8,95],lectur:40,left:8,left_scor:81,legal:65,len:[15,19,28,31,57,74,86,97],length:[15,30,41,44,48,49,59,66,70,96],leran:40,less:[5,25,119],less_equ:68,less_than:[5,40],let02:96,let:[5,8,17,19,21,32,33,35,39,43,47,48,49,50,60,67,69,95],level:[30,33,36,44,48,49,54,67,70,71,89,113],lgtest:9,lgtest_main:9,lib64:[2,78,103],lib:[1,46,87,107,116,117,118],lib_path:78,libapi:9,libari:46,libc:4,libcuda:[2,78],libgcc_:4,libgflag:87,libglog:87,libgoogl:107,libiomp5:42,libmkldnn:42,libmklml_intel:42,libnvidia:[2,78],libopenbla:87,libpaddl:[45,46,59,72,107],libpaddle_capi:46,libpaddle_capi_engin:87,libpaddle_capi_lay:87,libpaddle_capi_shar:87,libpaddle_capi_whol:87,libpaddle_gserv:46,libpaddle_math:46,libprotobuf:[83,87],librari:[9,16,21,22,42,43,46,51,66,69,103],library_:47,library_type_:69,librarydevicecontext:47,librarytyp:69,libstdc:4,libz:87,licens:[42,51,81,84],life:10,lifecycl:[54,119],lifetim:63,lightweight:33,like:[7,8,9,10,13,18,19,21,27,31,32,33,34,35,36,38,43,47,51,52,57,58,59,60,61,63,64,66,68,69,70,95,109,119],limit:[31,40,44,49,59,61,81,83,84,108],linaro:118,line:[9,13,18,21,29,35,52,55,57,59,61,72,81,83,95,105],line_count:83,linear:[49,81,83,84,86],lineno:107,link1:30,link2:30,link:[9,28,29,63,95,113,119],linux:[0,4,21,28,95,116,118],linux_x86_64:[4,62,78],lipo:117,list:[5,7,8,9,14,18,19,27,29,31,34,47,50,53,54,57,60,63,70,81,95,105,107],listdir:91,listen:[10,19,22],listen_and_do:19,listenanddo:19,lite:87,littl:[15,39,44],live:109,live_in:40,live_out:40,load:[5,10,22,34,51,57,84,95,97],load_missing_parameter_strategi:[102,103,105],load_mnist:34,load_paramet:83,loadsave_parameters_in_pserv:[13,102,103],local:[1,6,8,10,16,17,21,35,40,48,55,57,59,97,102,103],local_scop:6,localhost:[2,77],localip:97,localpath:29,locat:[9,31,47,54,67,70],lock:[9,10,14,15],lod:[21,44,48,64,70,71],lod_desc:[64,71],lod_expand:49,lod_level:[57,64,71],lod_tensor:[48,64,71],lodtenosr:20,lodtensor:[20,21,38,44,59,71],lodtensordesc:[44,64],log:[4,14,22,29,34,74,79,83,91,93,95,96,97,98,103],log_barrier_abstract:[102,103],log_barrier_lowest_nod:[102,103],log_barrier_show_log:[102,103],log_clip:[102,103],log_error_clip:[102,103],log_period:[96,97,103,105],log_period_serv:[102,103],logger:81,logic:[17,22,24,25,34,38,50,51,53,63,70],logit:[34,69],longer:[10,22,40],look:[8,18,19,21,31,32,35,52,57,60,61,66,95,109],lookahead:66,lookup:[38,49,109],lookup_t:40,loop:[6,8,20,31,40,54,58,63],loop_var:70,loss:[7,22,34,36,50,52,61,66],lot:[22,47,49,52,57,61,67,119],low:[50,51,66,67,70],low_rnn:48,lower:[30,48,49],lower_level_rnn:48,lpaddle_capi_engin:87,lpaddle_capi_lay:87,lpaddle_capi_shar:46,lpaddle_capi_whol:46,lrelu:34,lstm:[96,114],lstm_last:111,lstmemori:[82,114],lstmemory_group:82,lstmemory_unit:82,lstmlayer:41,luckili:40,mac:[46,116],machin:[22,24,31,34,40,42,51,52,61,81,90,95,98,113,119],machine_transl:114,maco:[0,4],macro:[33,47,60],made:[10,15,31],mai:[6,8,22,26,30,32,39,40,43,51,54,58,59,63,66,69,81,84,95],main:[19,21,25,31,32,36,51,55,59,87,95,107],main_program:[7,26],mainli:[16,40,47,67],maintain:[8,14,52,57,59,95],majel:9,major:[22,30,69],make:[0,1,5,7,8,9,10,14,15,17,21,22,23,30,31,35,48,49,52,53,57,58,59,61,66,70,72,74,75,77,78,87,95,108,116,117,118,119],make_chan:21,make_channel:21,make_ddim:76,make_function_oper:33,make_vari:65,maker:[59,60],malloc:67,man:28,manag:[10,15,16,19,21,22,29,54,63,67,77],mandarin:66,mani:[7,9,14,19,21,31,34,39,40,49,53,54,57,59,60,63,64,65,68,69,70],manili:36,manipul:[31,57,60],manner:[52,61,66,67],mantain:40,manual:[22,50,52,60,83,119],manufactur:31,manylinux1:4,manylinux1_x86_64:[4,62,78],manylinux:62,map:[5,8,14,25,43,47,57,60,63,65,67,70,84,87,119],map_fn:70,mapreduc:5,mark:[7,24,34,35,48,49,54,63,119],marker:54,market:30,master:[5,17,59,62,118],mastermind:9,mat:[45,46,89],mat_cache_row:13,mat_norm:13,mat_normal_shar:13,mat_sparse_row:13,mat_sparse_row_auto_grow:13,mat_sparse_row_id:13,mat_sparse_row_prefetch:13,mat_sparse_row_prefetch_full_s:13,mat_value_shar:13,match:[9,30,68,81],matchbox:119,math:[42,45,59,74,75,108],mathemat:61,matmul:[8,36,48,70,75],matrix:[13,45,46,74,75,87,89,90],matrixptr:74,matrixtyp:46,mattyp:13,max:[6,23,25,40,57,83,105,108,110],max_diff:6,max_length:[49,114],max_relative_error:[6,75],maxim:25,maximum:[8,15],maxoutfunctor:67,mayb:[8,43],md5:11,mean:[7,9,22,23,25,36,49,56,58,63,66,69,81,95,103,109,119],meant:70,measur:26,mechan:[7,16,20,26,43,57,60,95],mem:[8,18,49],mem_per_pserv:22,mem_per_train:22,member:[5,25,35,36,47,53,57,63],memcpi:[53,108],memori:[7,8,13,14,18,30,32,42,43,44,47,49,52,54,59,69,76,96,108,109,114],memory_nam:82,memory_optim:40,memory_test:0,memory_threshold_on_load_data:[102,103],memoryalloc:67,memorydesc:43,mention:[7,9,14,22,24,31,48,51,52,54],merg:[15,17,23,26,42,48,51,53,72,90],merge_model:90,merge_v2_model:90,merge_v2_modelss:90,messag:[8,19,21,27,31,32,35,44,54,55,56,57,59,60,64,71,72,78,96],metadata:[28,95,96,97],metal:119,metaphor:35,metaplotlib:5,method:[6,8,17,19,20,22,23,25,30,34,35,36,39,50,51,57,58,59,63,64,69,70,107],methodolog:52,metric:[26,54],mfs:97,microarchitectur:30,might:[8,9,19,21,31,40,55,66,95],min:[23,25,57,95,105,108],min_block:8,min_count:24,min_desc:8,min_pool_s:81,mini:[8,10,21,26,27,31,37,48,81],mini_batch:58,minibatch:[8,26,35,37,55],minim:[8,22,24,25,31,34,50,59,109],minimum:66,minsizerel:[116,117,118],minu:60,minus_grad:60,minusgradop:60,minusop:60,minusopgradmak:60,minusopprotoandcheckermak:60,minut:[10,17,95],mip:116,mirror:9,mislead:15,miss:34,mistak:31,mit:95,mix:[54,70,114],mkdir:[1,29,77,87,95,98],mkl:[1,43,59,67,69,78,87],mkl_packed_:41,mkldnn:[42,47,69],mkldnn_:42,mkldnnactiv:42,mkldnnbase:42,mkldnnlayer:42,mkldnnmatrix:42,mkldnnstream:42,mkldnntester:42,mklml:[42,78],mklml_lnx_2018:78,mklpack:41,mklpackedgatedrecurrentlay:41,mklpackedgemm:41,mklpackedlstmlay:41,mklpackedrecurrentlay:41,mlp:36,mnist:[12,22,34,35,55,58,59,90,107],mnist_random_image_batch_read:58,mnist_train:58,mnist_train_batch_read:58,mnist_v2:90,mnt:97,mobil:[30,31,40,59,77],mode:[30,41,51,72,97],model:[7,8,10,11,19,22,24,25,26,35,40,41,50,51,52,59,61,66,69,70,77,84,86,90,95,105],model_list:[103,105],model_path:105,modelparallel:22,modern:40,modif:66,modifi:[22,30,36,61,72,95],modul:[22,33,34,49,66,70,75,83,107],modular:49,momentum:[63,81,84],momentumop:107,mon:96,monitor:[21,54],month:9,more:[5,6,7,9,10,14,17,18,20,21,22,24,29,30,31,32,33,35,39,40,43,47,48,49,50,54,57,58,59,61,66,67,68,70,83,108,109,119],most:[5,7,9,17,21,22,32,35,36,47,49,52,54,58,61,66,67,68,109,119],mostli:[30,119],motiv:59,mount:[18,95],mountpath:[95,96,97],move:[10,14,20,29,31,52,95,119],movidiu:31,movie_id:97,mpi:[21,51,98],mpirun:98,mse:[31,35,50,55],much:[10,31,43,50,58,61,70],mul:[33,40,57,74,75],mul_grad:75,mul_op:75,mul_result:57,mulgradkernel:75,mulkernel:75,mulop:[33,75],mulopgrad:75,mulopmak:75,mult:[19,32],multi:[26,51,53,69,119],multigradientmachin:53,multipl:[5,6,14,15,17,19,21,22,24,26,31,32,33,39,51,54,59,66,69,71,84,95],multiple_input:57,multiple_param_attr:57,multipli:19,must:[7,15,20,25,40,43,44,47,54,56,57,58,59,65,71,74,75,76,81,91,95],mutabl:[67,76],mutable_data:[43,67,75,76],mutex:21,mutuable_data:67,mxnet:[8,19,21,31],my_cluster_nam:95,my_cost:83,my_external_dns_nam:95,my_lib:91,myerrorclip:25,mypaddl:[96,97],naiv:19,name:[5,6,7,8,10,12,13,15,18,19,20,22,26,30,33,36,39,42,43,44,46,47,49,54,55,57,59,62,64,65,70,71,74,83,84,86,96,97,99,100,105,108,109,114,119],name_:54,name_prefix:12,namespac:[8,37,45,57,74,75,96,97],nativ:30,natur:[14,17,24,49,70],ncall:107,nccl1:51,nccl2:51,ncclinit:51,nchw8:69,nchw8c:69,nchw:[42,47],ndarrai:[12,81],ndk:116,nearest:30,nearli:[6,20],necess:70,necessari:[7,8,15,17,25,26,40,44,49,53,57,65,70],necessarili:19,neck:51,need:[5,6,7,9,13,14,15,17,18,20,21,22,24,25,26,29,31,32,33,34,39,40,43,47,49,50,51,52,53,54,56,57,59,60,61,63,64,65,66,67,70,78,95,97,108,119],need_tran:83,neighberhood:51,neon:30,nervana:31,nessesari:66,nest:[7,8,54,55,71,89],net:[8,34,48,63,90],netop:[8,59],network:[5,6,7,8,10,13,22,24,26,34,36,40,41,42,48,50,52,54,57,58,61,63,65,66,67,69,71,82,84,86,89,90,97,105,111,119],network_config:105,networkadministr:95,neural:[5,7,8,10,22,36,40,41,42,48,52,61,63,67,69,71,86,111,113],neuralnetwork:53,never:[40,58,63,95,96,97],new_block_idx:57,new_op_desc:25,new_scop:69,new_stat:48,newblock:57,newbuff:43,newest:15,newli:[30,119],newop:8,newopdesc:57,newprogram:57,newscop:69,newvardesc:57,next:[7,10,16,20,21,25,49,51,70,95],nextlay:42,nfs4:95,nfs:[95,97],nfsdir:97,nfsver:95,nic:[97,102,103],nil:[14,21],nmt_without_attent:81,nnz:[74,89],no_grad_dict:7,no_grad_set:[6,7,75],no_gradi:7,no_sequ:84,node0:97,node1ip:98,node2ip:98,node3ip:98,node:[9,17,19,22,24,36,40,49,51,59,95,96,97,98,119],node_0:[95,97],node_1:[95,97],node_2:[95,97],node_id:91,nodeattr:36,nodeentri:36,nodefil:93,nodesep:36,nohup:91,nois:[10,34],noisi:34,non:[10,21,30,31,64,95],none:[5,6,7,8,21,25,26,34,36,37,48,49,50,55,57,65,70,109,114],noneedtran:43,nor:19,norm:[34,47],normal:[52,66,96,97],notat:40,note:[5,7,8,13,14,18,40,44,47,51,58,59,67,69,76,95],notebook:[2,18],noteworthi:31,noth:[39,57,63,72],notic:[25,31,51,60],notimplementederror:25,notin:69,notingradi:75,notion:70,notori:6,now:[7,9,10,20,21,24,34,44,47,52,59,60,61,63,95,113],nproc:0,nullptr:[43,54,60,63,74],num:[91,97,103],num_class:36,num_gradient_serv:[91,102,103],num_hidden:36,num_parameter_serv:5,num_pass:[84,96,97,102,103,105],num_pserv:22,num_row:64,num_samples_process:83,num_shard:12,num_step:70,num_train:22,number:[8,10,12,24,26,40,52,54,58,59,65,70,95],numdevices_:105,numeric_grad:6,numerical_grad:6,numlogicaldevices_:105,numpi:[1,12,30,34,57,58,75,81,83,84],numreal:13,numsampl:108,numtimeout:14,nv_:9,nv_gpu:0,nv_librari:9,nv_test:9,nvcc:[9,30,32],nvidia:[0,2,30,47,51,67,78],nvlink:51,nvprof:54,obj:83,object:[5,13,20,22,25,26,34,36,40,45,50,54,57,59,61,63,108],obtain:[17,20,52,67,81,84],obvious:[9,47],occup:[40,97],occupi:[30,54],occur:40,occurr:8,oct:96,off:[0,1,46,72,77,87,101,116,117,118,119],offer:[8,59,65],offici:[9,95],offlin:[10,12,119],offset:[13,89],often:[13,36,40,47],ograd:74,old:[6,15,17,49,59],older:31,omega:61,omit:81,omp_num_thread:107,ompi_comm_world_rank:91,onc:[10,14,19,22,24,26,31,35,52,68,95],one:[5,6,7,8,10,13,14,15,17,18,19,20,21,22,25,26,27,30,31,32,33,34,36,38,39,43,44,47,48,49,50,51,52,53,55,56,57,58,59,60,63,64,66,67,68,69,70,89,95,109,119],onehotcrossentropyopkernel:75,ones:[33,34,59],onli:[5,6,7,9,13,14,15,16,17,18,19,20,22,24,25,26,27,29,30,31,34,35,40,43,48,49,50,51,53,54,57,59,64,65,66,67,68,69,70,89,95,113,119],onlin:[10,12,40,58],only_cpu:6,onnx:31,onto:[22,24,95],op1:[40,69],op1_2_op2:69,op1_to_op2:69,op2:[40,69],op3:40,op_:75,op_check:75,op_class:[59,65],op_desc:[25,38,56],op_info:109,op_kei:43,op_maker_class:[59,65],op_proto:65,op_registri:109,op_siz:25,op_test:75,op_typ:[59,75],op_unique_kei:43,opattrcheck:75,opcreat:65,opdesc:[8,25,35,55,56,57,59,60,65,71],opdescbind:[38,60],opdescbuild:8,open:[5,12,31,34,42,58,81,83,84,95],openbla:[1,2,87],openmpi:[94,98],opensourc:51,oper:[6,8,19,21,22,23,24,26,27,30,31,32,34,35,36,38,39,48,49,50,51,54,56,61,63,66,67,69,71,75,76,95,109],operand:30,operartor:76,operator_grad:6,operator_list:54,operatorbas:[8,33,59,60,65,75],operatorwithkernel:[69,75],opinfo:[38,59,60],opinfomak:38,opinfomap:60,opkernel:[75,76],opkernelkei:59,opkerneltyp:[47,69],opmak:65,opproto:75,opprotoandcheckermak:[60,75],opprotomak:[65,75],opregist:65,opregistri:65,ops:[6,7,8,9,19,20,32,35,36,52,55,56,57,59,67,75,119],ops_:8,ops_test:9,opt:[1,5,50,56,65,97],opt_op_list:50,optest:75,optim:[6,7,22,23,24,32,34,51,52,53,55,59,61,64,66,81,83,84,109],optimis:50,optimize_op_attr:57,optimzi:81,option:[5,9,22,34,39,44,55,56,57,59,64,65,66,71,116,119],optmization_op_list:50,opts_np:56,optyp:[38,65],opwithkernel:64,order:[7,35,44,54,58,61,70,95,97,107,119],oregon:95,org:[12,28,34,81,84],orient:65,origin:[6,30,34,63,70,72],other:[8,10,15,19,29,30,31,32,38,40,43,47,48,52,56,61,63,65,66,67,69,95,109,119],otherwis:[5,7,10,15,17,34,38,43,58,66],our:[5,7,9,20,21,22,24,34,38,40,47,51,52,63,70,95],out:[5,8,9,14,17,20,22,25,31,36,40,43,48,49,57,69,75,76,81,95,107,113,114],out_dir:[95,97],out_mem:114,outer_mem:111,outgrad_:42,output:[5,6,7,8,12,17,19,20,24,25,29,32,33,34,35,36,37,38,40,43,44,48,49,52,55,56,57,58,59,60,63,64,65,67,69,70,75,76,81,84,90,93,96,97,105,114],output_:42,output_all_step:48,output_arg_nam:25,output_fil:90,output_lay:[81,84,86],output_mem:114,output_nam:6,output_num:48,output_path:12,output_seg:70,outputbuff:13,outputgradi:60,outsid:[22,63],outter:89,outter_pos_arrai:89,outter_seq_pos_arrai:89,outupt:70,outv:74,outval_:42,over:[5,31,40,51,52,70],overal:[34,52,54,119],overfit:61,overload:[30,39],overrid:[8,10,29,43,67,75,76],overview:[14,15,16,67],overwrit:29,own:[7,15,17,25,27,36,38,50,51,52,61,65,95],pack:[70,83],packag:[14,18,19,33,42,62,78,95,107],pad:[43,66],paddl:[0,1,2,4,5,8,9,10,12,18,20,22,29,32,33,34,37,41,42,43,44,45,46,48,49,53,55,59,61,62,65,66,67,70,72,74,75,77,81,83,84,86,87,89,90,91,93,95,96,97,98,101,105,107,108,109,114,116,119],paddle_arguments_get_sequence_start_po:89,paddle_arguments_set_id:89,paddle_arguments_set_sequence_start_po:89,paddle_arguments_set_valu:89,paddle_begin_init_param:15,paddle_capi:87,paddle_dir:75,paddle_doc:77,paddle_docs_cn:77,paddle_element_typ:15,paddle_element_type_float32:15,paddle_element_type_float64:15,paddle_element_type_int32:15,paddle_element_type_int64:15,paddle_element_type_uint32:15,paddle_element_type_uint64:15,paddle_enforc:[8,20,43],paddle_enforce_eq:[75,76],paddle_error:[45,46],paddle_exampl:18,paddle_finish_init_param:15,paddle_get_param:15,paddle_gradi:15,paddle_gradient_machine_create_shared_param:90,paddle_gradient_machine_forward:90,paddle_gradient_machine_load_parameter_from_disk:90,paddle_init:90,paddle_init_num_gradient_serv:91,paddle_init_param:15,paddle_init_port:91,paddle_init_ports_num:91,paddle_init_ports_num_for_spars:91,paddle_init_pserv:91,paddle_init_trainer_count:91,paddle_init_trainer_id:91,paddle_init_use_gpu:91,paddle_ivector:89,paddle_ivector_cr:89,paddle_job:18,paddle_manylinux_devel:1,paddle_matrix:[45,46,89,90],paddle_matrix_cr:[46,89],paddle_matrix_create_spars:89,paddle_matrix_get_row:89,paddle_matrix_get_shap:45,paddle_matrix_shap:45,paddle_matrix_sparse_copy_from:89,paddle_n:97,paddle_new_etcd_pserver_cli:15,paddle_new_pserver_cli:15,paddle_on_cloud:18,paddle_output:96,paddle_paramet:15,paddle_port:97,paddle_ports_num:97,paddle_ports_num_spars:97,paddle_process_by_paddl:97,paddle_pserver2:93,paddle_pserver_cli:15,paddle_pserver_client_releas:15,paddle_r:89,paddle_root:87,paddle_save_model:15,paddle_send_grad:15,paddle_server_num:97,paddle_train:[46,62,93,97],paddle_with_cuda:54,paddle_with_mkldnn:47,paddlepaddl:[0,1,2,4,9,10,12,15,16,17,18,19,22,28,29,33,34,35,37,39,44,48,49,50,53,54,57,58,59,63,70,71,72,81,84,86,87,89,90,93,94,96,97,99,100,101,107,108,114,116,118,119],paddlepaddle_gpu:4,paddlepaddlebook:2,paddlepaddlehub:2,page:95,pair:[7,8,22,35,50,54,59],pakcag:9,panic:21,paper:[34,66],para:13,paradigm:[19,27,59],paragraph:48,paragraph_data:48,paragraph_out:48,parallel:[19,21,22,24,51,54,59,69,95,96,97,105,108],parallel_for:19,parallel_nn:[102,103],paralleldo:23,parallelfor:19,paralleliz:66,param:[6,8,15,53,57,67,76,81,83],param_attr:[13,57,81,83,114],param_config_proto:15,param_fil:[83,90],paramattr:[81,83,114],paramet:[6,7,8,9,11,13,17,19,22,23,25,27,29,31,32,34,35,36,38,44,48,50,51,55,58,63,65,66,70,81,83,84,86,89,92,93,97,103,109],parameter_block_s:[102,103],parameter_block_size_for_spars:[102,103],parameter_list:[7,50],parameter_nam:5,parameter_serv:5,parameter_valu:13,parameterattribut:13,parameterclient2:97,parameterclient_:13,parametermap:74,parametermutex_:13,parameters_:74,parameters_and_grad:50,parameterserver2:13,parameterset:5,parameterupdat:53,parameterupdater_:13,params_grad:50,params_pass_4:90,params_pass_90:84,params_pass_:84,paramt:95,paraspars:74,parent:[8,19,55,57,59],parent_:[8,63],parent_idx:57,parenthes:59,pars:[0,9,22,36,95],parse_known_arg:97,parsefromstr:83,parser:97,part:[7,8,17,22,31,43,44,55,57,66,67,119],particular:[35,44,59,66,69],partit:[10,12,22,24,59,95],paserv:97,pass:[7,8,10,21,25,26,31,34,40,44,50,52,53,56,57,58,59,61,63,66,68,70,72,81,84,95,96,97,103,105,108],pass_id:[22,81,84],pass_idx:58,pass_manu:83,passtyp:74,past:[5,95],patch:28,path:[10,14,15,18,40,49,58,66,87,95,96,97,103,116,117,118],path_to_paddlepaddle_working_directori:77,pattern:[10,45,52,61,95],paus:[10,17],pcie:51,pd_api:89,peer:[51,79],pem:[5,12,95],pend:[10,14],pep425tag:[4,78],per:[10,15,51,52,58,61],percal:107,perf_test:107,perfectli:66,perform:[6,15,21,22,26,30,31,34,40,51,53,54,58,59,61,66,67,69,102,107,108],perftool:[54,107],period:[10,17,103],permiss:[81,84,95],persist:[27,64,66,71,95],persistentvolum:95,persistentvolumeclaim:[95,97],person:[5,39],perspect:59,perturb:6,pex:119,pfs:[12,29],pfsclient:12,pfspath:29,pgp:95,phase:[43,49,51,52,58,60,66,119],philosophi:[52,61],photo:34,physic:119,pick:95,pickl:[91,98],pictur:51,piec:[19,54],pillow:18,pip:[1,4,62,72,77,78,86,107],pipelin:[26,66],pivot:43,pixel:22,place:[7,8,10,17,22,24,27,39,43,51,59,69,76,109],place_:[39,47,67,69],placehold:[34,67,76],placement:24,plain:[18,44,46,47],plan:[10,19,43,59,66],platform:[4,8,32,43,47,54,67,69,75,76,95,109,116,117],pleas:[5,10,14,15,16,19,21,32,36,47,48,57,58,59,66,67,71,77,78,95,97],plot:5,plug:[51,52],pnpairvalid:102,pod:[12,18,19,95,96,97],pod_nam:95,podip:97,podlist:97,point:[8,10,18,21,30,40,43,51,67,76,89,108,119],pointer:[8,15,36,40,47,57,59,63,67,76,89],poli:83,polici:95,poll:21,pollut:17,polyak:52,ponit:36,pool3:74,pool:[23,40,66],pooling_lay:81,pooling_typ:[81,110],pop:[8,27],popul:15,popular:[9,34,36,54],port:[9,19,91,95,96,97,102,103],port_num:102,portabl:36,portal:77,ports_num:[91,97,103],ports_num_for_spars:[13,91,97,102,103,105],pose:10,posit:89,possibl:[5,8,14,21,24,40,57,61,71],post:[18,28],postpon:61,potenti:[30,108],pow:83,power:[30,40,51,66,119],ppo_workspac:77,pprof:107,pre:[5,15,39,40,95],pre_activ:57,pre_bia:57,pre_stat:[48,70],preambl:57,precis:[26,30,52],precompil:27,pred:[36,40],predecessor:40,predict:[22,61,81,86,90],predict_fil:[102,103],predict_output_dir:[102,103],prefer:[31,39],prefetch:[13,74],prefix:[10,12,49,66,95],pregel:21,pregrad:74,prepand:57,prepar:[6,18,53,66,91,98,99],prepend:57,prepend_oper:57,preprocess:[66,70],present:[5,7,8,54,70,72],preserv:29,prev_batch_st:[102,103],prevent:[5,10,14,17,25,61],preview:59,previou:[7,10,24,29,48,49,95],previous_memori:8,price:[59,86],prim:43,primari:[31,35],primarili:[52,61],primit:[30,42,43,51,70],primitive_desc:43,primitivedesc:43,principl:[5,9,47],print:[4,5,21,22,31,36,57,68,78,84,86,98],print_graphviz:36,printallstatu:108,println:21,printstatu:108,prioriti:59,privat:[8,46,54,57,63,64,65,67,70,72,76],privileg:95,pro:51,prob:86,probabl:[49,66],problem:[5,6,9,17,20,31,34,35,52,59,61],proc:2,proce:[10,58,95],procedur:[8,44],process:[5,7,8,12,13,14,17,19,20,21,22,26,27,31,32,36,40,42,44,51,61,65,69,81,83,95,97],processor:[30,108],prod:72,produc:[10,31,36,58],product:[18,31,84,95],productgraph:96,prof:107,profil:[29,54,66,107,108],profilerst:54,profl:107,proflier:[54,108],prog:97,program:[5,7,12,15,17,22,24,27,35,37,40,50,51,54,58,59,63,68,71,97,108],programdesc:[19,22,27,31,40,44,56,57,60,71],programm:[22,31,57],progress:[10,14],project:[18,46,66],promis:49,prompt:[29,31],prone:5,pronunc:66,prop_kind:43,propag:[7,31,52],proper:39,properli:39,properti:[36,61],propos:[8,23,24,49,50,51,52,70],proprietari:42,protect:[20,30,65,74,75],proto:[21,39,44,47,55,59,65,71,75],proto_:65,protobuf:[8,18,19,22,27,31,32,35,36,40,44,55,57,59,60,65,83,87,90],protocol:[109,119],protomak:75,provi:91,provid:[5,8,15,18,19,26,27,30,31,34,36,38,39,47,51,52,54,57,61,65,66,67,68,70,81,86,95,102,119],provis:[95,119],prune:8,ps_desir:10,pserver:[13,15,16,18,59,91,93,95,97,102,103],pserver_addr:15,pserver_cpu:18,pserver_id:11,pserver_mem:18,pserver_num_thread:[13,102,103],pseudo:[5,7,18,60,70],pseudocod:70,psize:74,ptr:[46,67],pub:98,pull:[2,9,59,62,116],purpos:[10,22,24,39,108],push:[8,27,31,54,62,97],push_back:74,put:[9,10,13,24,40,43,57,67],pvc:95,pwd:[0,1,2,72,77,116],pxe:119,py_paddl:78,pybind:[8,21,30],pydataprovid:81,pydataprovider2:97,python2:107,python:[0,1,4,5,8,16,20,21,26,27,31,33,34,35,36,39,45,49,53,54,59,62,67,70,72,74,75,77,78,81,86,90,91,98,107,109,114],pythonpath:78,pytorch:[31,54],qualcomm:30,queri:95,question:[5,19,24,65,95],queue:[21,24],quick:36,quick_start:[18,95,96,97,99],quick_start_data:96,quickli:[49,57,59],quickstart:96,quit:49,r14b:116,rais:[25,36,91],rajathkmp:34,ran:[24,108],rand:[34,83,89,103,105,108],rand_max:89,random:[12,21,34,47,53,57,58,75,83],random_imag:12,randomli:[17,89],rang:[12,19,22,30,34,40,54,58,65,97],rank0:51,rank1:51,rank:[5,70,95],rankdir:36,rapid:60,raspberrypi:118,raspbian:118,rasspberri:118,rate:[15,66,81,97],rather:[7,18,34,70,95],ratio:103,raw:44,rdma_tcp:[102,103],reach:[10,40,51],read:[5,7,10,12,19,20,21,22,24,31,32,58,59,66,70,83,95,119],read_from_arrai:40,read_from_realistic_imag:5,read_from_rng:5,read_lock:11,read_minibatch:31,read_mnist_imag:5,read_next_from_fil:81,read_paramet:83,read_ranking_model_data:5,readabl:[54,59],reader:[12,22,30,34,35,55,66,84,91,107],reader_cr:12,reader_creator_bool:58,reader_creator_random_imag:58,reader_creator_random_image_and_label:58,readi:[10,21,95,96,119],readlockguard:13,readm:[46,72],readnext:20,readwritebuffer_:13,readwritemani:95,real:[13,34,58],realist:5,realiti:66,realiz:[8,48],realli:[31,61],reason:[5,6,10,21,31,96],receiv:[10,18,21,22,24,48],recent:[40,52],recognit:66,recommend:[5,97],record:[14,43,54,65,95],recordev:54,recordio:[5,12,14,20],recov:[10,70],recover:59,recoveri:14,recurr:[41,48,63,66,111,112],recurrent_group:[66,81,82,111,113,114],recurrent_op:70,recurrentgradientmachin:[46,49,70],recurrentlay:41,recurs:[7,8,9,29,40,59],recv:[19,22,24,51,95],recvparametertyp:13,red:34,reduc:[24,30,51,59,107],reduce_by_kei:59,reduce_mean:34,refactor:[22,24,35,49,52,53,57,61,70],refer:[8,10,14,15,16,19,30,36,43,47,48,51,55,57,59,61,63,67,70,71],referenc:14,refine_unknown_arg:97,reflect:14,reg:65,regard:[20,119],region:63,regist:[21,40,47,60,67,69],register_gpu_profil:108,register_lay:74,register_op:[33,59,60,65,75],register_op_cpu_kernel:[67,75],register_op_cuda_kernel:[67,75],register_op_without_gradi:[59,75],register_oper:[38,60],register_tim:13,register_timer_info:108,registerop:65,registr:109,registri:[18,38,67,96,119],regular:[7,81,95],reinit:20,reiniti:[20,43],rel:[6,17,61],relat:[10,17,18,30,47,54,63,69,119],relationship:[60,67],releas:[62,66,78,87,95,116,117,118],reli:[6,19,49,50,52,61],reliabl:[10,61],relu1:36,relu2:36,relu:[34,36,40],relwithdebinfo:107,remain:70,remot:[9,13,22,59,72,95,103,105],remoteparameterupdat:[13,16],remov:[7,22,29,31,49,72],renam:[7,29,30],reorder:43,reorder_primit:43,repeat:[8,35,55,56,64,65,71],repeatedli:[35,40],replac:[9,14,38,52,60,66],replic:22,replicaset:18,repo:[9,118],report:[14,30,31,54],reportdataset:14,repositori:[77,116],repres:[7,8,14,19,22,24,25,31,36,44,47,49,52,57,59,61,64,67,70,71,95],represent:[15,22,32,34,35,40,47,49,64],request:[9,10,13,17,19,59,62,95,96,119],requir:[5,7,10,15,17,18,20,22,24,25,29,30,36,40,42,48,52,54,55,56,59,61,64,65,66,71,77,81,84,95,119],requisit:40,research:[22,31],reserv:[29,81,84],reserveoutput:74,reset:[10,26,79],reset_program:26,resetingrad:42,resetinvalu:42,resetoutgrad:42,resetoutvalu:42,resetxxx:42,reshap:[6,58,83],residu:21,resiz:[13,67,75,76],resolv:[9,72,96],resourc:[22,27,51,54,67,95],respect:[6,20,25,30,34,48],respons:[13,21,22,26,34,51,52,53,61,95,96],rest:[8,18,28,32,69,119],restart:[10,15,95,96,119],restartpolici:[95,96,97],restor:[6,52],restrict:[61,63,107],result:[6,7,14,21,26,34,35,36,40,44,49,50,51,53,84,95,108,109],resum:17,retran:95,retriev:[8,49,63],reuqest:62,reus:[8,17,49,58,59],rev:0,revamp:22,reveal:5,revers:[7,113,114],review:[19,72,96],reviews_electronics_5:96,rewrit:[9,21],rid:[20,31],right:[6,7,8,9,18,26,40,59,61,81,84],right_scor:81,ring:51,risk:7,rkt:[0,18],rmsprop:[52,81],rmspropoptim:52,rnn:[8,31,34,49,57,59,63,66,102,113,114],rnn_bias_attr:114,rnn_layer_attr:114,rnn_out:114,rnn_output:70,rnn_use_batch:[41,102,103],rnnalgorithm:49,rnnstep:70,roadmap:[66,70],rocmplac:47,role:[5,14,15,22,51,95],rollback:57,root:[7,51,95,96,97],roughli:66,round:[30,51],routin:[30,42,51],row:[13,21,89],row_offset:89,rowoffset:89,rows_:64,rpc:14,rpcserver:14,rpi:118,rpi_arm_neon:118,rpi_toolchain:118,rsize:95,rstrip:97,rtk:119,rule:[7,22,25,31,35,95],run:[0,1,2,5,6,7,8,9,10,18,19,20,21,22,23,24,26,30,31,32,33,34,35,36,40,43,47,48,50,51,52,54,55,56,57,59,62,63,64,66,67,68,69,72,76,77,78,93,95,96,97,99,100,101,107,108,116,119],run_test:1,runinitfunct:[97,108],runnabl:24,running_on_cloud:18,runserv:77,runtim:[8,19,21,22,38,48,59,71,78],runtime_table_:8,s_block:7,safe:18,sai:[20,32,35,37,40,58],said:31,same:[5,6,14,15,17,19,20,21,22,33,34,36,39,40,48,49,51,57,59,60,63,66,69,70,76,95],sampl:[26,34,57,65,89],sampler:34,satifi:40,satisfi:[9,43,64,95],save:[10,12,14,15,18,19,22,35,36,40,44,52,64,71,95,96],save_dir:[96,97,103,105],save_only_on:[102,103],save_parameter_to_tar:84,savetxt:83,saving_period:[97,102,103],saving_period_by_batch:[102,103,105],scalabl:59,scalar:[7,8,37,68,70],scale:[22,24,52,60,65,66,75],scaleop:75,scaleopmak:[59,75],scan:[7,14,40,59],scatter:[7,51],scenario:49,schdule:95,schedul:[14,18,24,95],scheme:[13,61],scienc:40,scope:[6,19,23,27,32,69,109],score:[49,81],score_diff:81,scorer:66,scp:98,script:[1,51,77,93,95,98,116],sdk:117,search:[10,63,114],second:[5,19,29,31,34,36,48,49,55,56,58,63,65,75],secret:95,section:[7,24,31,57,95],see:[5,7,10,19,21,24,30,31,57,66,81,83,84,95],seed:[83,103,108],seem:[9,21,30,31,66],seen:61,segment:[48,70],sel:21,select:[49,95],selected_generation_scor:49,selected_id:49,selected_row:[64,71],selected_rows_desc:[64,71],selected_scor:49,selectedrow:[38,71],selector:96,self:[6,25,26,34,36,40,41,42,44,50,57,70,74,75],self_addr:19,semant:[5,49,62],semaphor:21,semat:5,send:[10,15,19,22,24,39,51,59,65,95],send_back_parameter_typ:13,sendbackparameterspars:13,sendbackparametertyp:13,sendparameterrequest:13,sendparameterrespons:13,sens:[52,61],sent:[5,15,19,22,59,65,71,96],sentanc:81,sentenc:[31,48,49,70,114],sentence_input:70,separ:[15,22,33,52,60,61],seper:70,seq_len:70,seq_po:89,seq_pool:110,seq_pos_arrai:89,seqtoseq:83,sequenc:[7,8,19,27,31,35,41,50,55,66,70,81,84,89,111,113],sequence_layer_group:111,sequence_nest_layer_group:111,sequence_start_posit:89,sequencegen:111,sequencetyp:84,sequenti:[8,19,21],seri:[20,111],serial:[8,14,44,53,59],serializ:[59,71],serv:[22,30,59,70,95],server:[1,5,9,13,16,17,22,32,51,59,79,91,92,93,97,103,119],serverless:10,servic:119,sess:[34,36,50],session:[36,50,56],set:[0,5,7,10,18,20,34,38,40,43,47,48,49,54,56,57,59,60,63,66,67,70,75,76,77,81,83,84,89,95,96,108],set_active_typ:74,set_attr:25,set_drop_r:74,set_float_el:6,set_input:25,set_output:25,set_shap:20,set_siz:74,set_typ:[25,74],setdatatyp:64,setdefault:75,setp:95,setq:0,settotalbyteslimit:83,setup:[22,52,62,74,75,119],seven:66,sever:[6,13,22,24,34,48,49,51,53,54,57,64,67,70,95],sexstant:119,sgd:[5,10,18,24,52,53,64,84,92,102,109],sgd_optim:109,shall:[7,9],shape:[6,7,8,20,22,34,37,47,48,55,57,59,64,66,67,84,109],shapes_:20,shard:[10,11,12,13,14,15,17,22,24,92,95],share:[9,20,34,46,53,57,59,61,66,67,70],shared_librari:9,shared_ptr:[43,45,46,63,67,76],shell:95,shoul:6,should:[5,6,7,8,15,18,21,22,25,26,30,32,33,34,38,39,43,47,48,49,50,52,53,54,55,58,59,60,61,64,65,66,68,69,70,71,75,77,95,113],should_be_fals:5,should_be_tru:5,show:[0,7,8,10,20,21,29,31,37,40,44,48,51,52,55,68,70,95],show_check_sparse_distribution_log:[102,103],show_layer_stat:[102,103],show_parameter_stats_period:[96,102,103,105],shown:[5,22,26,51,54,66,95],shrunk:25,shuf:81,shuffl:[20,22,81],shuffleread:20,sid:95,side:[22,26,40,53],sig:95,sigint:93,sigmod:65,sigmod_op:65,sigmod_output:65,sigmoid:[8,65,70],sign:[28,44,95],signatur:95,signific:66,similar:[8,19,21,22,24,27,31,49,52,54,58,59,61,66,67,69,70,95,119],similarli:[31,40],simpl:[19,24,30,32,35,36,40,48,52,55,61,63,65,66,70,97],simple_attent:114,simple_gru:114,simple_lstm:82,simple_rnn:114,simpler:53,simplest:95,simpli:[5,15,22],simplifi:[5,49,57,65,66],simul:[31,117],simultan:95,sinc:[10,14,16,17,21,22,23,24,31,38,40,43,47,52,57,58,60,61,70,95,119],singl:[7,10,20,22,24,26,30,39,51,59,63,66],singleton:[19,23],sit:22,site:[9,95,107],situat:[7,56,69],size:[10,12,13,15,21,22,30,34,40,44,49,52,57,58,64,65,66,67,70,74,75,76,81,83,84,86,89,109,114],size_in_byt:43,size_t:[13,20,67,70,74],sizeof:[8,89],skip:[7,58,72,83,95],sleep:97,slice:19,slide:10,slight:31,slightli:34,small:[6,19,32,34,42,49],small_messag:[102,103],smaller:[6,10,30,49],smart:63,snap:96,snapdragon:30,snapshot:[11,17,95],snippet:[21,33,50,95],sock:18,sock_recv_buf_s:[102,103],sock_send_buf_s:[102,103],socket:97,softmax:[5,8,22,24,31,36,37,49,55,74,81,114],softmax_param:83,softmaxoutput:36,softwar:[30,54,81,84,119],solid:34,solut:[51,119],solv:[5,7,20,40,59],some:[5,7,8,9,13,14,15,17,18,20,21,22,24,25,30,32,33,34,35,39,40,43,47,48,49,50,55,56,57,58,59,60,63,67,69,70,95,119],some_c_api_funct:46,some_inst:46,some_op:[38,48,70],some_python_class:45,somecppclass:45,somegotyp:45,someth:[7,13,57],sometim:[54,58],somewhat:15,somewher:63,soon:10,sort:[70,95,97],sort_by_length:70,sortagrad:66,sourc:[6,9,29,31,34,42,44,46,49,58,59,95],source_dict_dim:[49,114],source_dict_s:49,source_language_word:[49,114],space:[24,30,57,61,66],span:54,spars:[13,21,74,81,89,91,95,97,103,105],sparse_binary_vector:[81,84,89],sparse_binary_vector_sequ:89,sparse_binary_vector_sub_sequ:89,sparse_float_vector:84,sparse_remot:13,sparse_upd:[13,81],sparse_vector:[81,89],sparse_vector_sequ:89,sparse_vector_sub_sequ:89,sparseparam:74,sparseprefetchrowcpumatrix:74,spec:[95,96,97],special:[7,15,22,30,32,38,47,49,50],specialvartypeinfer:38,specif:[7,9,10,20,22,25,29,32,49,59,63,67,81,84],specifi:[5,6,13,14,15,18,19,21,22,23,25,26,27,29,34,44,54,57,63,65,70,77,95],spectrogram:66,speech:66,speed:[30,44,51,52,119],speedup:54,sphinx:[45,77],split:[17,19,20,23,31,37,49,59,70,91,95],split_count:[91,95,97],spread:7,squar:36,square_error_cost:[84,109],squash:72,srand:[89,103],src:[9,43,78,91,93,97],src_backward:114,src_dict:83,src_dict_path:83,src_embed:[49,114],src_forward:114,src_primitive_desc:43,src_word_id:[49,114],src_word_vec:49,sreializ:71,ssh:[95,98],ssh_server:93,sstabl:5,stabil:[6,40],stabl:[62,95],stack:[27,59,70,95],stage:[9,16,23,34,40,43,66,71,72],stale:10,stamp:78,standalon:116,standard:[21,31,59,61,66],stanford:[6,96],star:9,start:[7,9,10,13,14,15,17,18,21,22,23,49,51,53,54,72,78,89,96,97,103],start_mpi_train:98,start_op_idx:7,start_paddl:97,start_pass:[102,103],start_pserv:[102,103],startpaddl:97,startup:[10,18,31,95],stat:[103,108],state:[8,10,26,27,48,49,54,63,66,70,82,96,113],statem:40,statement:[21,31,35,40,95],statfulset:97,static_cast:[43,76],staticinput:[113,114],statist:[26,54],statset:108,statu:[18,49,72,95,96,97,108],status:96,std:[9,13,20,36,38,39,43,45,46,54,56,59,60,63,65,67,74,75,76,103],stdbuf:91,stderr:93,stdout:93,step:[6,8,10,15,21,22,24,26,31,34,35,41,49,52,53,57,59,65,66,70,95,111,113,114,119],step_gradi:7,step_id:70,step_input:70,step_net:8,step_output:70,step_scop:59,stepnet:[8,48,59,63],still:[7,14,17,22,31,40,60],stirng:57,stmt1482205552000:95,stmt1482205746000:95,stochast:[10,14,17,52],stop:57,stop_gradi:57,storag:[28,30,95],store:[6,8,9,13,27,36,38,44,47,49,53,55,57,59,60,61,63,70,95],str:[7,18,70,97,105],straight:[55,58,64],straightforward:43,strategi:[10,57,103],stream:[22,43,54,67],stream_:67,strict:58,stride:[43,47,66],string:[7,8,14,29,36,39,44,54,55,56,57,59,60,63,64,65,71,74,75,95,103],strip:83,struct:[14,15,28,30,38,39,46,47,54,60,65,69,83],structur:[7,8,14,31,34,44,49,55,57,59,64,95],sts:95,stuff:72,style:[59,65],sub:[5,7,17,19,24,34,40,48,51,53,57],sub_block:7,sub_sequ:84,subclass:57,subcommand:29,subgraph:[24,34],submiss:22,submit:[43,59,95],subnet0:95,subnet:[5,95],subobjectpath:96,subseq:[110,113],subsequ:51,subsequenceinput:111,subtract:6,succ:40,succeed:[14,96],success:[15,95,96],successfulcr:96,sucess:40,sucessor:40,sudo:[0,95],suffer:6,suffix:[18,91],suggest:9,suit:119,suitabl:[64,67],sum:[7,8,11,23,38,57],sum_op:7,summar:[34,54],summari:54,sumopgradmak:60,sumpool:81,supercomput:40,suppli:64,support:[4,6,8,10,17,18,19,21,22,24,31,33,34,40,43,44,47,49,52,53,54,56,58,59,60,61,64,66,69,95,119],support_inplac:40,suppos:[9,19,33,64,89],suppress:29,sure:95,svs:65,swagger:28,swig:[1,16,45,46],switch_ord:82,switchop:8,sychron:51,symbol:[8,36,46],symbols_ready_:8,symbolt:[8,59],symlink:72,sync:[10,52,61],sync_with_cpp:107,syncflag:74,synchron:[10,14,21,43,51,54,95],syntax:[19,27,31,49,58],sysroot:116,system:[8,9,10,15,17,21,22,24,28,33,34,40,42,66,81],tabl:[8,19,31,38,44,64,71],tablelookup:64,tablelookupgrad:64,tablelookupop:64,tag:[2,62,72,78,101],tail:49,tainer_id:97,take:[5,7,8,9,10,17,19,20,21,22,25,27,30,32,34,35,37,38,40,43,47,52,55,56,57,58,59,60,67,69,70,95],taken:[25,36,40,47,70],talk:[15,32],tanh:[34,49,74,114],tar:[78,84,90,95],tarbal:95,target:[7,8,9,25,27,34,36,50,56,59],target_block:[7,25],target_dict_dim:114,target_dict_s:49,target_language_word:114,target_link_librari:9,target_word:49,targetinlink:111,task13:66,task14:66,task:[22,44,49,54,65,90],task_queu:14,taskentri:14,taskqueu:14,tbd:[16,43,66,111],tcp:[95,103],tear:108,technic:[7,10],techniqu:40,technolog:31,tee:96,tell:[10,14,15,49,65],templat:[33,43,65,67,75,76,96,97,119],tempor:66,temporari:[7,18,27,40,52,57],tempori:40,tensor:[6,9,19,21,23,24,30,31,32,34,36,38,39,43,44,47,48,49,64,70,71,75,109],tensor_arrai:19,tensor_array_read:70,tensor_array_s:70,tensor_array_stack:70,tensor_array_unstack:70,tensor_array_writ:70,tensor_data:44,tensor_in:69,tensor_s:6,tensor_test:9,tensor_to_check:6,tensorarrai:23,tensorarraydesc:70,tensordesc:[44,64],tensorflow:[8,19,21,22,24,31,34,37,61,70],term:[10,60,61,66],termin:96,terminolog:40,tessorarrai:70,test1:12,test:[5,6,9,36,46,52,58,62,72,74,75,76,86,89,91,98,103,105,108,109],test_:75,test_all_data_in_one_period:96,test_check_grad_ingore_i:75,test_check_grad_ingore_x:75,test_check_grad_norm:75,test_check_output:75,test_compar:78,test_comparespars:78,test_comparetwonet:78,test_comparetwoopt:78,test_config_pars:78,test_data_dir:91,test_fcgrad:74,test_gpuprofil:108,test_layergrad:74,test_list:83,test_mkldnn:42,test_mklpack:41,test_mul_op:[1,75],test_networkcompar:78,test_pass:[102,103,105],test_period:[102,103,105],test_predict:78,test_pydataprovid:78,test_pydataprovider2:78,test_pydataproviderwrapp:78,test_recurrent_machine_gener:78,test_recurrentgradientmachin:[78,111],test_swig_api:78,test_train:78,test_traineronepass:78,test_wait:[102,103],testa:5,testb:5,testbilinearfwdbwd:108,testconfig:74,testfcgrad:74,testfclay:74,testlayergrad:74,testmulop:75,testq:5,testutil:74,text1:29,text:[5,44,48,54,66,95],tflop:108,tftp:119,tgz:[4,78],than:[7,10,18,19,20,25,31,32,33,34,57,59,61,68,70,83,89,95,119],the_step:31,theano:31,thei:[5,7,9,10,15,17,19,20,21,24,25,29,31,34,35,39,40,49,50,54,57,59,65,70,71,95,108],them:[5,6,7,9,10,13,18,20,21,24,25,31,32,33,38,39,40,49,57,58,59,60,63,64,65,69,70,71,95,108],themselv:[7,9],theori:31,therefor:[7,40,52],therein:8,theta:34,theta_d:34,theta_g:34,thi:[4,5,6,7,8,9,10,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,30,31,32,33,34,35,36,39,40,43,47,48,49,50,51,52,53,54,55,57,58,59,60,61,64,65,66,67,69,70,81,84,89,95,108,109,119],thin:38,thing:[22,34,59,67],think:[5,9],third:[10,36,89],third_parti:[42,78,87,116,117,118],those:[8,9,10,33,35,36,37,55],though:[70,119],thought:9,thread:[19,21,23,54,108],thread_count:23,thread_id:54,thread_id_:54,thread_local_rand_use_global_se:[102,103],thread_pool:23,threadid:105,threadloc:108,threadpool:19,three:[6,7,10,21,26,30,31,32,35,43,49,50,53,54,55,58,66,67],threshold:[10,14,25,103],through:[7,9,10,14,16,26,40,50,52,77,109],throughout:27,throughput:108,thrust:59,thu:[17,26,36,40,66,95],tier:96,time:[5,6,9,10,14,17,20,21,22,24,25,31,33,38,40,41,47,48,49,51,54,57,58,59,60,64,65,66,70,71,89,96,97,103,107,108,111,119],timelin:[54,59],timeo:95,timeout:[10,14],timer:108,timestamp:11,timestep:63,titan:40,titl:97,tls:28,tmp:57,to_no_sequ:110,to_sequ:[110,111],to_your_paddle_clone_path:77,todo:[8,10,14,17,49,65,66],togeth:[7,70],token:[5,66,114],toler:6,too:[6,19,21,25,43,69,70],took:119,tool:[54,77,95,97,116,118],toolchain:116,toolkit:66,top:[48,49,66],top_k:49,top_level_rnn:48,topic:43,topk_generated_scor:49,topk_id:49,topk_scor:49,toplevel:0,topolog:[5,10,22,36,40,44,53],topoloi:36,torch:[8,31],tostr:83,total:[10,24,26,51,54,58,96,108,119],total_pass:58,tottim:107,touch:78,toward:31,trace:[8,32,34],track:[10,14,36,57,72],tradit:[8,30,66],traffic:22,train:[2,7,8,12,14,15,17,19,20,25,26,27,31,32,34,35,40,41,44,51,52,53,54,55,56,57,59,61,64,66,67,71,79,81,84,91,93,94,96,97,98,99,100,103,105,114],train_arg:97,train_args_dict:97,train_args_list:97,train_config_dir:[95,97],train_data:91,train_data_dir:91,train_i:84,train_id:95,train_list:[83,91],train_loop:31,train_read:[22,84],train_x:84,trainabl:[44,57],trainer:[5,11,12,13,14,16,22,24,32,41,42,52,53,59,74,84,91,92,97,103,105],trainer_config:[90,95,96,97],trainer_config_help:74,trainer_count:[81,86,91,95,96,97,102,103,105],trainer_cpu:18,trainer_cr:18,trainer_gpu:18,trainer_id:[91,95,97,103],trainer_intern:13,trainer_mem:18,trainer_packag:18,trainer_prog:22,trainerconfighelp:83,trainerid:[17,97],trainingjob:22,trainonebatch:13,tran:[43,74],trans_var:69,transact:[10,14],transcript:66,transfer:[40,54],transform:[59,66],translat:[40,81],translation_id:49,translation_scor:49,transpar:49,transpil:19,travers:[7,35,40],travi:72,treat:[8,15,40],treatment:[15,30],tree:[8,19,27,31,57,97,109,118],trg_dic_siz:49,trg_embed:[49,114],trick:49,tricki:45,trigger:[17,20,53],trivial:[49,70],true_block:[8,37,55],true_imag:58,true_label:58,true_neg:26,true_posit:26,true_read:58,tune:[66,102,107],tupl:[7,57,58],turn:[57,58,113],tutori:[95,97,98,99,100],twice:[24,34],twine:62,two:[5,7,15,16,17,18,19,20,21,22,26,29,30,31,32,34,35,38,40,44,47,49,52,54,55,58,59,60,61,63,64,65,66,69,70,71,75,76,95,108],txt:[9,18,29,41,42,74,77,91,95,98],type:[5,7,8,10,13,14,17,18,20,22,28,29,30,32,38,39,43,44,45,46,48,49,55,56,57,58,59,60,61,64,65,66,67,69,71,74,75,76,84,86,89,95,96,105,114],type_nam:65,typedef:[15,30,45,46,47,67],typeerror:25,typeid:65,typenam:[33,65,67,75,76],typic:22,ubuntu:[4,62,86],ubyt:58,uci_h:86,uid:96,uint16_t:30,uint32:[28,44],uint32_t:54,uint64:[44,45],uint64_t:[45,89],unawar:15,unbias:6,unblock:21,unbound:40,unbuff:21,unclear:17,uncreat:7,under:[9,14,24,51,69,81,84,95],underli:[20,49],understand:[31,57,66,119],understand_senti:114,undeterminist:108,uni:66,unidirect:66,unifi:[20,27,36,64],uniform:[12,34,57,58],uniform_random:57,uniniti:7,uninstal:[1,78],uniqu:[5,8,10,17,18,43,47,57,63,95],unique_nam:57,unique_name_gener:57,unique_ptr:[60,63,67,74],unit:[9,52,54,61,67],unittest:[46,75,78],unix:21,unk:[64,71],unless:[81,84],unlik:49,unnecessari:[7,66],unordered_map:63,unpack:70,unrol:48,unseen:61,unsign:[15,30],unstack:70,unstack_from:70,unsupervis:34,unsupport:75,until:[10,15,21,23,24,31,40,63,95,97],untrack:72,unzip:116,updat:[7,10,14,15,22,28,30,34,48,49,50,51,52,53,63,66,70,72,105,107],update_equ:84,update_memori:8,update_op:50,updatecallback:74,updatestack:95,upgrad:[4,51,78],upload:[10,18,21,28,62],upon:10,upstream:[72,78],uri:95,usag:[30,37,40,53,57,68,93,97],use:[5,6,8,9,10,16,20,21,22,23,24,27,30,34,36,38,39,40,43,47,49,50,51,53,54,57,63,64,65,66,69,70,71,72,75,81,84,89,91,95,97,108],use_cpu:39,use_cudnn:39,use_eigen_bla:116,use_eigen_for_bla:[116,117],use_gpu:[81,84,86,91,96,97,102,103,105],use_mkl_pack:41,use_mkldnn:[39,42],use_old_updat:[13,102,103],use_sparse_remote_updat:13,used:[5,6,8,9,10,16,17,20,21,22,25,27,30,31,34,36,40,48,49,52,53,54,57,58,59,61,63,65,67,69,70,95,108],useful:[6,30,40,57,63,69],usegpu:[74,89],user:[5,6,7,8,9,12,14,17,18,19,22,23,24,25,26,27,29,33,34,35,36,38,39,43,47,49,50,51,52,54,57,58,59,60,61,63,65,67,69,70,95,119],user_id:97,user_nam:12,usercert:12,userkei:12,usernam:[12,72,116],uses:[6,10,17,19,21,22,30,40,47,48,49,53,54,67,69,95],using:[5,6,7,8,9,10,14,15,17,18,21,22,27,29,30,31,33,34,36,38,40,48,50,52,55,57,58,60,61,63,65,66,67,75,86,95],usr:[1,2,78,91,95,97,103],usual:[7,18,40,47,54,55,61,67,95,108],util:[22,41,42,51,90,97,108,119],uuid:[11,17],v7a:116,v8a:116,val:7,valgrind:107,valid:[58,59,63,95],valu:[6,7,8,10,19,21,25,26,36,37,40,42,44,48,49,50,52,53,55,59,63,64,65,68,70,71,74,81,89,95,97,105,116],value1:103,value2:103,value_:64,valueerror:[36,81],values_:70,vanilla:114,var_nam:[7,69],var_recurs:25,vardesc:[8,35,55,57,59,64],vardescbuild:8,vari:95,variabl:[5,6,8,19,20,21,22,24,25,26,27,32,34,35,36,37,38,47,48,49,50,52,55,56,60,61,64,65,66,69,70,95,96,109],variablenamemap:75,varialbl:34,variant:[38,47,67,70],varibal:7,varibl:36,varienc:70,varient:70,variou:[8,21,30,40,61],varproto:65,vars_:[8,63],vartyp:[64,71],vartypeinfer:38,vec2seq:66,vec:83,veclib:117,vector:[5,8,13,15,20,36,37,43,48,49,54,57,59,60,64,66,68,70,87,89],vendor:9,verbos:29,veri:[9,14,19,24,27,31,33,34,40,43,49,53,58,61,63,66,67,70],verifi:8,version:[7,9,18,22,25,29,32,34,36,37,44,49,62,66,72,81,84,95,101,102,103,108,117],versu:5,via:[7,10,47,72,95,119],view:[44,47],vim:2,virtual:[20,25,38,39,60,67],virtualenv:0,visibl:[17,63],visit:7,visual:49,vlog:13,vocabulari:66,volum:[77,96,97],volumemount:[95,96,97],volumn:95,wai:[5,7,15,17,21,27,31,39,40,49,52,57,58,61,70],wait:[10,15,21,23,97],wangkuiyi:9,want:[5,18,19,21,26,34,39,47,52,54,56,58,61,63,67,68,69,70],warn:[29,78,83,97],warp_ctc:66,warranti:[81,84],wast:51,watch:10,wbia:95,weight:[41,44,61,74],weightlist:74,weights_:74,weights_primitive_desc:43,weights_t:74,welcom:[9,66],well:[7,18,21,22,24,31,33,34,61,64,66,95],wer:66,were:[9,21,31],west:95,wget:[78,116],wgt:43,what:[9,20,31,34,49,57,65,69,72,119],wheel:4,when:[7,8,9,10,13,14,15,18,19,22,24,25,26,27,29,30,31,32,36,49,51,52,53,54,55,57,59,67,70,95,108,119],whenev:[57,66],where:[5,7,8,10,17,19,22,31,32,35,47,48,49,52,55,59,61,67,70,109],wherea:[8,14,33,37,67,68,71],whether:[6,7,8,20,27,54,58,64,70,89],which:[5,6,7,8,9,10,12,14,15,17,18,19,20,21,22,23,25,27,30,31,32,33,34,36,38,40,43,44,47,48,49,50,51,53,55,56,57,58,59,60,63,64,65,68,69,70,71,84,95,107,119],while_grad:40,while_loop:[49,70],while_op:7,whileloop:70,whileop:8,white:66,whl:[1,4],who:[7,33,35,51,57],whoever:15,whole:[7,20,34,37,40,45,46,48,51,56,65,66,87,95,119],wholli:20,whose:[6,7,10,17,25,48,59,60,65,70],why:[6,46,108],wide:[9,25,34,93,98],width:[13,45,58,74,75,83,89],wiki:9,window:[0,52,66],wirt:36,wise:[24,59,66],with_avx:[1,72,101,116,117],with_bia:65,with_c_api:[1,87,116,117,118],with_doc:1,with_doubl:[1,74,101],with_dso:1,with_golang:[1,87,116],with_gpu:[0,1,72,87,101,116,117],with_mkl:[1,41,42,87,116],with_mkldnn:42,with_mklml:42,with_profil:108,with_python:[1,87,101,116,117],with_rdma:[101,116,117],with_style_check:[1,72],with_swig_pi:[1,87,116,117],with_test:[1,72,75],with_tim:[101,108],within:[14,22,31,66],without:[7,10,15,20,21,54,57,58,59,66,81,84],wloop:70,wmt14:114,word2vec:[18,81,91,93],word:[7,24,35,38,40,48,49,59,65,66,70,81,113],word_dict:[91,98],word_dim:83,word_id:81,word_vector_dim:[49,114],wordcount:66,work:[2,5,8,9,10,22,27,30,31,39,50,52,54,57,72,77,95,96,97,119],worker:[24,71,95],workercount:95,workflow:[59,95],workspac:[91,93],worri:6,worth:109,would:[8,9,10,17,21,22,23,24,31,33,34,35,43,50,52,53,57,58,64,66,70,95,119],wouldn:[31,35],wrap:[20,31,33,34,51,119],wrapper:[9,20,21,33,51,52,60,70,108],write:[5,10,17,19,20,21,22,24,30,31,32,33,36,38,43,50,52,57,58,59,60,67,70,81,83,84,95],write_lock:11,write_to_arrai:40,writer:[5,57],written:[7,8,19,24,27,34,44,52,59,60,64],wrong:58,wrote:36,wsize:95,www:[81,84],x64:118,x86:[116,117],x86_64:[116,117],x_neg:6,x_po:6,xarg:[2,74,78,98],xcode:117,xcodebuild:117,xgbe0:103,xgbe1:103,xpu:31,xrang:[6,31,34,54,58,74,84,86],xx_layer:39,xxx:[5,70],xxxx:11,xxxxxxxxx:95,xxxxxxxxxx:95,xxxxxxxxxxxxx:95,xxxxxxxxxxxxxxxxxxx:95,y_dim:34,y_neg:6,y_po:6,y_predict:[84,86,109],yaml:[9,93,95,96,97,98,119],yancey1989:18,yapf:72,year:31,yep:[54,107],yet:[31,66,119],yield:[5,12,20,58,81,84],you:[6,18,22,30,63,81,84,95,119],your:[5,9,13,18,29,59,78,95,116,117,118,119],your_access_key_id:95,your_param_nam:83,your_repo:97,your_secrete_access_kei:95,your_source_root:46,yuang:31,yuyang:107,z_dim:34,z_size:34,zero:[6,7,10,21,34,49,53,57,64,95,103],zip:[57,97,116],zlib:87,zone:95,zxf:78,zxvf:95},titles:["\u7528Docker\u7f16\u8bd1\u548c\u6d4b\u8bd5PaddlePaddle","\u4ece\u6e90\u7801\u7f16\u8bd1","\u4f7f\u7528Docker\u5b89\u88c5\u8fd0\u884c","\u5b89\u88c5\u4e0e\u7f16\u8bd1","\u4f7f\u7528pip\u5b89\u88c5","PaddlePaddle Design Doc","Auto Gradient Check Design","Backward Building","Design Doc: Block and Scope","Required CMake Function","Design Doc: Distributed Training","\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9\uff08Checkpointing\uff09","\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1","Alalysis of large model distributed training in Paddle","Design Doc: Master Server","Design Doc: The Client Library of Parameter Server","Design Doc: Remote Parameter Updater for Cluster Train","Design Doc: Save Model","Submit a Distributed Training Job","Design Doc: Concurrent Programming with Fluid","C++ Data Feeding","Design Doc: CSP in PaddlePaddle Fluid","Design Doc: Distributed Training Architecture","Design Doc: Execute the Program with Multi CPU","Design Doc: Parameter Server","Error Clip","Evaluator Design","Executor Design Doc","FileManager\u8bbe\u8ba1\u6587\u6863","PFSClient","Design Doc: float16","Design Doc: PaddlePaddle Fluid","PaddlePaddle Fluid: Towards a Compiled Programming Language","Design Doc: Functions, Operators, and Layers","Design for GAN","Design Doc: Computations as a Graph","Survey on Graph","The IfElse Operator","Design Doc: InferVarType","Problem","Memory Optimization","Intel\u00ae MKL Packed on PaddlePaddle: Design Doc","Intel\u00ae MKL-DNN on PaddlePaddle: Design Doc","Design Doc: Add MKLDNN Kernel in Fluid Operator","Design Doc: Model Format","Paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0","C-API \u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863","Design Doc: The Keys of Operator Kernel Type","RNNOp design","Design: Sequence Decoder Generating LoDTensors","Optimizer Design","Design Doc: NCCL support in Paddle Fluid","Averaging Parameter in PaddlePaddle","Design Doc: The C++ Class Parameters","Introduction","Design Doc: PaddlePaddle Programs","Prune","Design Doc: Python API","Python Data Reader Design Doc","Design Doc: Refactorization Overview","Design Doc: Gradient Operators Registration","Regularization in PaddlePaddle","PaddlePaddle\u53d1\u884c\u89c4\u8303","Design of Scope in Paddle","Design Doc: Selected Rows","Interaction between C++ and Python","DeepSpeech2 on PaddlePaddle: Design Doc","Design Doc: Supporting new Device/Library","Design Doc: Switch","Background","Design for TensorArray","Background","\u5982\u4f55\u8d21\u732e\u4ee3\u7801","\u5f00\u53d1\u6807\u51c6","\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42","\u5982\u4f55\u5199\u65b0\u7684Operator","\u5728Paddle\u4e2d\u5982\u4f55\u4f7f\u7528Eigen","\u5982\u4f55\u8d21\u732e\u6587\u6863","\u7f16\u8bd1\u5b89\u88c5\u4e0e\u5355\u5143\u6d4b\u8bd5","\u96c6\u7fa4\u8bad\u7ec3\u4e0e\u9884\u6d4b","FAQ","\u672c\u5730\u8bad\u7ec3\u4e0e\u9884\u6d4b","\u6a21\u578b\u914d\u7f6e","\u53c2\u6570\u8bbe\u7f6e","\u57fa\u672c\u4f7f\u7528\u6982\u5ff5","\u65b0\u624b\u5165\u95e8","\u5feb\u901f\u5f00\u59cb","\u5b89\u88c5\u4e0e\u7f16\u8bd1C-API\u9884\u6d4b\u5e93","C-API\u9884\u6d4b\u5e93","\u8f93\u5165/\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7","C-API\u4f7f\u7528\u6d41\u7a0b","\u542f\u52a8\u53c2\u6570\u8bf4\u660e","\u5206\u5e03\u5f0f\u8bad\u7ec3","\u4f7f\u7528fabric\u542f\u52a8\u96c6\u7fa4\u8bad\u7ec3","\u5728\u4e0d\u540c\u96c6\u7fa4\u4e2d\u8fd0\u884c","Distributed PaddlePaddle Training on AWS with Kubernetes","Kubernetes\u5355\u673a\u8bad\u7ec3","Kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3","\u5728OpenMPI\u96c6\u7fa4\u4e2d\u63d0\u4ea4\u8bad\u7ec3\u4f5c\u4e1a","<no title>","<no title>","\u73af\u5883\u51c6\u5907","\u53c2\u6570\u6982\u8ff0","\u7ec6\u8282\u63cf\u8ff0","\u547d\u4ee4\u884c\u53c2\u6570\u8bbe\u7f6e","\u4f7f\u7528\u6848\u4f8b","\u8fdb\u9636\u4f7f\u7528","Python\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790","GPU\u6027\u80fd\u8c03\u4f18","PaddlePaddle Fluid Source Code Overview","\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684Layer","\u5355\u53cc\u5c42RNN API\u5bf9\u6bd4\u4ecb\u7ecd","RNN\u6a21\u578b","Recurrent Group\u6559\u7a0b","RNN\u914d\u7f6e","PaddlePaddle \u6587\u6863","Android\u5e73\u53f0\u7f16\u8bd1\u6307\u5357","iOS\u5e73\u53f0\u7f16\u8bd1\u6307\u5357","Raspberry Pi\u5e73\u53f0\u7f16\u8bd1\u6307\u5357","Cluster bootstrapping tool survey"],titleterms:{"\u4e00\u4e9b\u7ec6\u8282\u7684\u8865\u5145":97,"\u4e0a\u4f20\u8bad\u7ec3\u6587\u4ef6":12,"\u4e0b\u8f7d\u6570\u636e":96,"\u4e0b\u8f7dmklml\u5e93\u5931\u8d25":78,"\u4e0d\u4f7f\u7528":45,"\u4e0d\u4f7f\u7528swig\u8fd9\u79cd\u4ee3\u7801\u751f\u6210\u5668":45,"\u4e0d\u540c\u7684":82,"\u4e0d\u5bfc\u51fapaddle\u5185\u90e8\u7684\u7ed3\u6784\u4f53":45,"\u4e0d\u5f15\u7528\u5176\u4ed6\u52a8\u6001\u5e93":45,"\u4e24\u79cd\u4f7f\u7528":82,"\u4e3a\u4ec0\u4e48\u8981":0,"\u4e3a\u4ec0\u4e48\u9700\u8981\u6027\u80fd\u5206\u6790":108,"\u4ec0\u4e48\u662f\u6027\u80fd\u5206\u6790":108,"\u4ec5\u4ec5\u4f7f\u7528void":45,"\u4ece\u5feb\u7167\u6062\u590d":11,"\u4ece\u6e90\u7801\u7f16\u8bd1":1,"\u4ee3\u7801\u8981\u6c42":72,"\u4f7f\u7528":[72,96],"\u4f7f\u7528\u52a8\u6001\u5e93\u6765\u5206\u53d1paddl":45,"\u4f7f\u7528\u6848\u4f8b":105,"\u4f7f\u7528\u6a21\u578b\u521d\u59cb\u5316\u7f51\u7edc":105,"\u4f7f\u7528\u6d41\u7a0b":90,"\u4f7f\u7528\u73af\u5883\u53d8\u91cf":97,"\u4f7f\u7528\u8f6c\u6362\u5e93":12,"\u4f7f\u7528docker\u542f\u52a8paddlepaddl":2,"\u4f7f\u7528docker\u5b89\u88c5\u8fd0\u884c":2,"\u4f7f\u7528docker\u6267\u884cgpu\u8bad\u7ec3":2,"\u4f7f\u7528docker\u6784\u5efa":77,"\u4f7f\u7528fabric\u542f\u52a8\u96c6\u7fa4\u8bad\u7ec3":93,"\u4f7f\u7528paddlepaddl":77,"\u4f7f\u7528pip\u5b89\u88c5":4,"\u4fdd\u6301\u672c\u5730\u4ed3\u5e93\u6700\u65b0":72,"\u4fee\u6539\u542f\u52a8\u811a\u672c":96,"\u514b\u9686":72,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5934\u6587\u4ef6":46,"\u5177\u4f53\u67d0\u79cd\u7c7b\u578b\u7684\u5b9e\u73b0\u6587\u4ef6":46,"\u5185\u7f6e\u5b9a\u65f6\u5668":108,"\u5199\u68af\u5ea6\u68c0\u67e5\u5355\u5143\u6d4b\u8bd5":74,"\u51c6\u5907\u4e00\u4e2alinux\u96c6\u7fa4":93,"\u51c6\u5907\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883":[116,117],"\u51c6\u5907\u6570\u636e\u96c6":91,"\u51c6\u5907\u8bad\u7ec3\u6570\u636e":97,"\u51c6\u5907\u8bad\u7ec3\u7a0b\u5e8f":91,"\u51c6\u5907\u9884\u6d4b\u6a21\u578b":90,"\u51c6\u5907openmpi\u96c6\u7fa4":98,"\u51cf\u5c11\u6570\u636e\u8f7d\u5165\u7684\u8017\u65f6":81,"\u51cf\u5c11dataprovider\u7f13\u51b2\u6c60\u5185\u5b58":81,"\u51fa\u73b0":82,"\u5206\u5757\u6587\u4ef6\u4f20\u8f93":28,"\u5206\u5e03\u5f0f\u8bad\u7ec3":92,"\u5206\u652f\u89c4\u8303":62,"\u521b\u5efa\u672c\u5730\u5206\u652f":72,"\u521b\u5efa\u795e\u7ecf\u7f51\u7edc\u8f93\u5165":90,"\u521b\u5efajob":97,"\u521b\u5efapaddlepaddl":96,"\u521d\u59cb\u5316paddlepaddle\u8fd0\u884c\u73af\u5883":90,"\u5220\u9664\u672c\u5730\u5206\u652f":72,"\u5220\u9664\u8fdc\u7a0b\u5206\u652f":72,"\u5229\u7528\u66f4\u591a\u7684\u8ba1\u7b97\u8d44\u6e90":81,"\u5230\u8fdc\u7a0b\u4ed3\u5e93":72,"\u5236\u4f5c\u955c\u50cf":97,"\u5236\u4f5cdocker\u955c\u50cf":96,"\u524d\u5411\u8ba1\u7b97":90,"\u524d\u5411operator\u5355\u6d4b":75,"\u52a0\u8f7d\u6a21\u578b":90,"\u52a0\u8f7dpaddlepaddl":84,"\u52a0\u901f\u6267\u884c":11,"\u52a0\u901f\u8bad\u7ec3\u901f\u5ea6":81,"\u52a8\u6001\u5e93\u4e2d\u4e0d\u5d4c\u5165\u4efb\u4f55\u5176\u4ed6\u8bed\u8a00\u7684\u89e3\u91ca\u5668":45,"\u52a8\u6001\u6269\u5bb9":11,"\u5355\u5143\u6d4b\u8bd5":103,"\u5355\u53cc\u5c42rnn":111,"\u539f\u56e0":45,"\u539f\u56e0\u5217\u8868":45,"\u53c2\u6570\u5185\u5b58":81,"\u53c2\u6570\u670d\u52a1\u5668\u548c\u5206\u5e03\u5f0f\u901a\u4fe1":103,"\u53c2\u6570\u6982\u8ff0":102,"\u53c2\u6570\u8bbe\u7f6e":83,"\u53c2\u8003\u6587\u6863":28,"\u53c2\u8003\u8d44\u6599":108,"\u53cc\u5c42rnn":111,"\u53cc\u5c42rnn\u4ecb\u7ecd":113,"\u53cc\u5c42rnn\u7684\u4f7f\u7528":113,"\u53cd\u5411operator\u5355\u6d4b":75,"\u53d1\u5e03docker\u955c\u50cf":62,"\u53d1\u5e03wheel\u5305\u5230pypi":62,"\u53ef\u80fd\u78b0\u5230\u7684\u95ee\u9898":0,"\u5404\u4e2a\u7248\u672c\u6700\u65b0\u7684whl\u5305":4,"\u540d\u8bcd\u89e3\u91ca":28,"\u5411\u91cf":103,"\u542f\u52a8\u4efb\u52a1":97,"\u542f\u52a8\u53c2\u6570\u670d\u52a1\u5668":91,"\u542f\u52a8\u53c2\u6570\u8bf4\u660e":91,"\u542f\u52a8\u8ba1\u7b97\u8282\u70b9":91,"\u542f\u52a8\u96c6\u7fa4\u4f5c\u4e1a":[93,98],"\u5440":0,"\u547d\u4ee4\u884c\u53c2\u6570\u8bbe\u7f6e":104,"\u548c":110,"\u5728\u4e0d\u540c\u8bbe\u5907\u4e0a\u6307\u5b9a\u5c42":105,"\u5728\u4e0d\u540c\u96c6\u7fa4\u4e2d\u8fd0\u884c":94,"\u5728docker\u4e2d\u6267\u884cpaddlepaddle\u8bad\u7ec3\u7a0b\u5e8f":2,"\u5728openmpi\u96c6\u7fa4\u4e2d\u63d0\u4ea4\u8bad\u7ec3\u4f5c\u4e1a":98,"\u5728paddle\u4e2d\u5982\u4f55\u4f7f\u7528eigen":76,"\u57fa\u4e8edocker\u5bb9\u5668\u7684\u7f16\u8bd1\u65b9\u5f0f":116,"\u57fa\u4e8elinux\u4ea4\u53c9\u7f16\u8bd1\u73af\u5883\u7684\u7f16\u8bd1\u65b9\u5f0f":116,"\u57fa\u672c\u4f7f\u7528\u6982\u5ff5":[84,89],"\u57fa\u672c\u539f\u7406":113,"\u57fa\u672c\u8981\u6c42":45,"\u5982\u4f55\u4e66\u5199\u6587\u6863":77,"\u5982\u4f55\u4f7f\u7528":82,"\u5982\u4f55\u5171\u4eab\u53c2\u6570":83,"\u5982\u4f55\u5199\u65b0\u7684oper":75,"\u5982\u4f55\u51cf\u5c11\u5185\u5b58\u5360\u7528":81,"\u5982\u4f55\u521d\u59cb\u5316\u53c2\u6570":83,"\u5982\u4f55\u52a0\u8f7d\u9884\u8bad\u7ec3\u53c2\u6570":83,"\u5982\u4f55\u52a0\u901f\u8bad\u7ec3\u901f\u5ea6":81,"\u5982\u4f55\u548c\u660e\u6587\u8fdb\u884c\u76f8\u4e92\u8f6c\u5316":83,"\u5982\u4f55\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u53c2\u6570\u7684\u6743\u91cd\u548c\u68af\u5ea6":81,"\u5982\u4f55\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u83b7\u5f97\u67d0\u4e00\u4e2alayer\u7684output":81,"\u5982\u4f55\u6307\u5b9agpu\u8bbe\u5907":81,"\u5982\u4f55\u66f4\u65b0www":77,"\u5982\u4f55\u6784\u5efa\u6587\u6863":77,"\u5982\u4f55\u8bbe\u7f6e\u5b66\u4e60\u7387\u9000\u706b":83,"\u5982\u4f55\u8c03\u7528":81,"\u5982\u4f55\u8d21\u732e\u4ee3\u7801":72,"\u5982\u4f55\u8d21\u732e\u6587\u6863":77,"\u5982\u4f55\u8fdb\u884c\u6027\u80fd\u5206\u6790":108,"\u5982\u4f55\u9009\u62e9sgd\u7b97\u6cd5\u7684\u5b66\u4e60\u7387":83,"\u5b50\u5e8f\u5217\u95f4\u65e0memori":111,"\u5b50\u5e8f\u5217\u95f4\u6709memori":111,"\u5b58\u50a8\u7684\u53c2\u6570\u683c\u5f0f\u662f\u4ec0\u4e48":83,"\u5b89\u88c5":4,"\u5b89\u88c5\u4e0e\u7f16\u8bd1":3,"\u5b89\u88c5\u4e0e\u7f16\u8bd1c":87,"\u5b89\u88c5\u4ea4\u53c9\u7f16\u8bd1\u5668":118,"\u5b89\u88c5\u6d41\u7a0b":3,"\u5b9a\u4e49operator\u7c7b":75,"\u5b9a\u4e49opkernel\u7c7b":75,"\u5b9a\u4e49protomaker\u7c7b":75,"\u5b9e\u73b0":45,"\u5b9e\u73b0\u5355\u5143\u6d4b\u8bd5":75,"\u5b9e\u73b0\u65b0\u7684\u7f51\u7edc\u5c42":74,"\u5b9e\u73b0\u65b9\u5f0f":46,"\u5b9e\u73b0\u8ba1\u7b97":76,"\u5b9e\u73b0c":[74,75],"\u5b9e\u73b0python\u5c01\u88c5":74,"\u5bfb\u627e\u6027\u80fd\u74f6\u9888":107,"\u5bfc\u51fac":45,"\u5c06\u547d\u4ee4\u53c2\u6570\u4f20\u7ed9\u7f51\u7edc\u914d\u7f6e":105,"\u5de5\u5177":108,"\u5e38\u89c1\u95ee\u9898\u548c\u89e3\u51b3\u65b9\u6cd5":4,"\u5e38\u89c1\u95ee\u9898\u89e3\u7b54":3,"\u5e76\u5b8c\u6210":72,"\u5efa\u7acb":72,"\u5f00\u53d1\u6807\u51c6":73,"\u5f00\u59cb\u5f00\u53d1":72,"\u5f02\u6b65\u968f\u673a\u68af\u5ea6\u4e0b\u964d":103,"\u5feb\u7167\u4fdd\u5b58\u7684\u8bbe\u8ba1\u5982\u4e0b":11,"\u5feb\u901f\u4f7f\u7528":86,"\u5feb\u901f\u5b89\u88c5":86,"\u5feb\u901f\u5f00\u59cb":86,"\u6027\u80fd\u5206\u6790\u5c0f\u6280\u5de7":108,"\u6027\u80fd\u5206\u6790\u5de5\u5177\u4ecb\u7ecd":108,"\u6027\u80fd\u8c03\u4f18":103,"\u603b\u4f53\u6d41\u7a0b":0,"\u603b\u7ed3":89,"\u6216\u8005\u662f":78,"\u6267\u884c\u5355\u5143\u6d4b\u8bd5":1,"\u627e\u5230\u7684pythonlibs\u548cpythoninterp\u7248\u672c\u4e0d\u4e00\u81f4":78,"\u62a5importerror":78,"\u6307\u9488\u4f5c\u4e3a\u7c7b\u578b\u7684\u53e5\u67c4":45,"\u63a5\u53e3\u8f93\u51fa\u591a\u4e2alayer\u7684\u9884\u6d4b\u7ed3\u679c":81,"\u63a8\u5bfc\u65b9\u7a0b":74,"\u63a8\u6d4b\u6267\u884c":11,"\u63d0\u4ea4":72,"\u63d0\u4ea4\u4ee3\u7801\u7684\u4e00\u4e9b\u7ea6\u5b9a":72,"\u63d0\u4ea4\u955c\u50cf":96,"\u642d\u5efa\u795e\u7ecf\u7f51\u7edc":84,"\u652f\u6301\u53cc\u5c42\u5e8f\u5217\u4f5c\u4e3a\u8f93\u5165\u7684layer":110,"\u652f\u6301\u7528\u6237\u81ea\u5b9a\u4e49\u7684\u6570\u636e\u9884\u5904\u7406job":12,"\u6570\u636e\u652f\u6301":103,"\u6574\u4f53\u65b9\u6848":97,"\u6587\u4ef6\u4f20\u8f93\u4f18\u5316":28,"\u6587\u4ef6\u8bbf\u95ee\u65b9\u5f0f":12,"\u6587\u4ef6\u8bbf\u95ee\u7684\u6743\u9650":12,"\u6587\u4ef6\u9884\u5904\u7406":12,"\u6587\u6863":115,"\u65b0\u624b\u5165\u95e8":85,"\u65e5\u5fd7\u4e2d\u4fdd\u5b58\u5747\u4e3a\u7f51\u7edc\u901a\u4fe1\u7c7b\u9519\u8bef":79,"\u65f6\u95f4\u5e8f\u5217":111,"\u65f6\u95f4\u6b65":111,"\u66b4\u9732\u63a5\u53e3\u539f\u5219":46,"\u672c\u5730\u6d4b\u8bd5":105,"\u672c\u5730\u8bad\u7ec3":105,"\u672c\u5730\u8bad\u7ec3\u4e0e\u9884\u6d4b":81,"\u672f\u8bed":11,"\u6784\u5efa\u548c\u6d4b\u8bd5":72,"\u6784\u5efapaddlepaddle\u7684android\u5f00\u53d1\u955c\u50cf":116,"\u67b6\u6784\u56fe":28,"\u67e5\u770b\u6027\u80fd\u5206\u6790\u6587\u4ef6":107,"\u67e5\u770b\u8bad\u7ec3\u7ed3\u679c":96,"\u67e5\u770b\u8f93\u51fa":97,"\u6846\u67b6\u751f\u6210":28,"\u6848\u4f8b\u4e00":105,"\u6848\u4f8b\u4e8c":105,"\u68c0\u67e5\u6a21\u578b\u8f93\u51fa":93,"\u68c0\u67e5\u96c6\u7fa4\u8bad\u7ec3\u7ed3\u679c":93,"\u6982\u5ff5\u7b80\u4ecb":75,"\u6982\u5ff5\u89e3\u91ca":12,"\u6982\u8ff0":[87,110,113],"\u6a21\u5757":28,"\u6a21\u578b\u53c2\u6570\u68c0\u67e5\u70b9":11,"\u6a21\u578b\u63a8\u65ad\u5b9e\u73b0\u6587\u6863":46,"\u6a21\u578b\u914d\u7f6e":[82,111],"\u6a21\u578b\u914d\u7f6e\u7684\u6a21\u578b\u914d\u7f6e":111,"\u6ce8\u518coper":75,"\u6ce8\u610f\u4e8b\u9879":[75,90],"\u6d41\u7a0b\u4ecb\u7ecd":12,"\u6d4b\u8bd5":103,"\u6df7\u5408\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790":107,"\u6e05\u7406":90,"\u73af\u5883\u51c6\u5907":101,"\u751f\u6210\u5e8f\u5217":114,"\u751f\u6210\u6027\u80fd\u5206\u6790\u6587\u4ef6":107,"\u751f\u6210\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":113,"\u751f\u6210sparse\u6587\u4ef6":28,"\u7528\u6237\u4f7f\u7528\u6d41\u7a0b":28,"\u7528docker\u7f16\u8bd1\u548c\u6d4b\u8bd5paddlepaddl":0,"\u7684\u533a\u522b":82,"\u7684\u53c2\u6570":82,"\u7684\u65b9\u6cd5\u6709\u4f55\u533a\u522b":82,"\u76ee\u5f55\u7ed3\u6784":46,"\u76ee\u6807":28,"\u76f4\u63a5\u6784\u5efa":77,"\u76f8\u5173\u6982\u5ff5":113,"\u77e9\u9635":103,"\u793a\u4f8b1":111,"\u793a\u4f8b2":111,"\u793a\u4f8b3":111,"\u793a\u4f8b4":111,"\u793a\u4f8b\u7a0b\u5e8f":12,"\u795e\u7ecf\u5143\u6fc0\u6d3b\u5185\u5b58":81,"\u7a00\u758f\u8bad\u7ec3":105,"\u7aef\u6570\u636e\u7c7b\u578b\u8bf4\u660e":89,"\u7b26\u53f7":45,"\u7b80\u5355\u95e8\u63a7\u5faa\u73af\u795e\u7ecf\u7f51\u7edc":114,"\u7c7b":[45,74,75],"\u7ebf\u6027\u56de\u5f52\u5b8c\u6574\u793a\u4f8b":84,"\u7ec4\u7ec7\u5e8f\u5217\u4fe1\u606f":89,"\u7ec4\u7ec7\u8f93\u5165\u6570\u636e":[89,90],"\u7ec6\u8282\u63cf\u8ff0":103,"\u7ec8\u6b62\u96c6\u7fa4\u4f5c\u4e1a":93,"\u7ed1\u5b9apython":75,"\u7f16\u5199\u9884\u6d4b\u4ee3\u7801":90,"\u7f16\u5199yaml\u6587\u4ef6":96,"\u7f16\u8bd1":75,"\u7f16\u8bd1\u4f9d\u8d56":1,"\u7f16\u8bd1\u548c\u5b89\u88c5":[116,117,118],"\u7f16\u8bd1\u548c\u6267\u884c":75,"\u7f16\u8bd1\u5b89\u88c5\u4e0e\u5355\u5143\u6d4b\u8bd5":78,"\u7f16\u8bd1\u5b89\u88c5\u540e\u6267\u884c":78,"\u7f16\u8bd1\u65b9\u6cd5":1,"\u7f16\u8bd1\u6d41\u7a0b":3,"\u7f16\u8bd1\u9009\u9879":[1,46],"\u7f16\u8bd1\u9009\u9879\u7684\u8bbe\u7f6e":1,"\u7f16\u8bd1\u9009\u9879\u8bf4\u660e":1,"\u7f16\u8bd1paddlepaddl":116,"\u7f29\u5bb9":11,"\u800c\u662f\u624b\u5199\u591a\u8bed\u8a00\u7ed1\u5b9a":45,"\u80cc\u666f":45,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":103,"\u83b7\u53d6paddlepaddle\u7684docker\u955c\u50cf":2,"\u8986\u76d6\u4e0d\u4e00\u81f4\u7684\u90e8\u5206":28,"\u8bad\u7ec3":103,"\u8bad\u7ec3\u56e0\u6b64\u9000\u51fa\u600e\u4e48\u529e":81,"\u8bad\u7ec3\u6570\u636e\u5b58\u50a8":12,"\u8bad\u7ec3\u6570\u636e\u7684\u5b58\u50a8\u548c\u5206\u53d1":12,"\u8bad\u7ec3\u6a21\u578b":84,"\u8bad\u7ec3\u6d41\u7a0b\u7684\u4f7f\u7528\u65b9\u6cd5":113,"\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0":81,"\u8bcd\u6c47\u8868":111,"\u8be6\u7ec6\u6559\u7a0b":108,"\u8bfb\u53d6\u53cc\u5c42\u5e8f\u5217\u6570\u636e":111,"\u8f6c\u6362\u5e93":12,"\u8f93\u5165":[89,113],"\u8f93\u5165\u4e0d\u7b49\u957f":111,"\u8f93\u5165\u793a\u4f8b":113,"\u8f93\u51fa":113,"\u8f93\u51fa\u6570\u636e":89,"\u8f93\u51fa\u6570\u636e\u7c7b\u578b":89,"\u8f93\u51fa\u6570\u636e\u7ec4\u7ec7":89,"\u8fd0\u884c\u5bb9\u5668":96,"\u8fd0\u884c\u73af\u5883\u4f9d\u8d56":4,"\u8fd0\u884cdocker":78,"\u8fd9\u4e2a\u52a8\u6001\u5e93\u4f7f\u7528c99\u6807\u51c6\u7684\u5934\u6587\u4ef6\u5bfc\u51fa\u4e00\u4e9b\u51fd\u6570":45,"\u8fdb\u884c\u8bad\u7ec3":[12,96],"\u8fdb\u9636\u4f7f\u7528":106,"\u901a\u7528":103,"\u9047\u5230":78,"\u914d\u7f6e\u4ea4\u53c9\u7f16\u8bd1\u53c2\u6570":[116,117,118],"\u914d\u7f6e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u67b6\u6784":114,"\u914d\u7f6e\u7f51\u7edc":84,"\u94a9\u5b50":72,"\u94fe\u63a5\u8bf4\u660e":87,"\u9519\u8bef\u600e\u4e48\u529e":82,"\u968f\u673a\u6570":103,"\u96c6\u7fa4\u591a\u8282\u70b9\u8bad\u7ec3":79,"\u96c6\u7fa4\u8bad\u7ec3":105,"\u96c6\u7fa4\u8bad\u7ec3\u4e0e\u9884\u6d4b":79,"\u9700\u8981\u7684\u8f6f\u786c\u4ef6":0,"\u975e\u6cd5\u6307\u4ee4":78,"abstract":[22,23,24,51,119],"android\u5e73\u53f0\u7f16\u8bd1\u6307\u5357":116,"api\u4f7f\u7528\u6d41\u7a0b":90,"api\u5bf9\u6bd4\u4ecb\u7ecd":111,"api\u5e93":116,"api\u9884\u6d4b\u5e93":[87,88],"beam_search\u7684\u751f\u6210":111,"book\u4e2d\u6240\u6709\u7ae0\u8282":62,"book\u6559\u7a0b":2,"case":7,"class":[34,53,57],"cmake\u6e90\u7801\u7f16\u8bd1":78,"filemanager\u8bbe\u8ba1\u6587\u6863":28,"final":39,"float":81,"function":[9,33,34,57],"gpu\u548ccpu\u6df7\u5408\u4f7f\u7528":105,"gpu\u6027\u80fd\u8c03\u4f18":108,"gpu\u955c\u50cf\u51fa\u73b0":78,"group\u6559\u7a0b":113,"import":78,"ios\u5e73\u53f0\u7f16\u8bd1\u6307\u5357":117,"kubernetes\u5206\u5e03\u5f0f\u8bad\u7ec3":97,"kubernetes\u5355\u673a\u8bad\u7ec3":96,"new":67,"org\u5de5\u5177":77,"paddle\u52a8\u6001\u5e93\u4e2d":45,"paddle\u591a\u8bed\u8a00\u63a5\u53e3\u5b9e\u73b0":45,"paddle\u7248\u672c\u53f7\u4e3a0":78,"paddlepaddle\u53d1\u884c\u89c4\u8303":62,"paddlepaddle\u56de\u5f52\u6d4b\u8bd5\u5217\u8868":62,"paddlepaddle\u662f\u5426\u652f\u6301\u7ef4\u6570\u53ef\u53d8\u7684\u6570\u636e\u8f93\u5165":82,"paddlepaddle\u73af\u5883\u4f9d\u8d56":4,"paddlepaddle\u7684softmax\u80fd\u5426\u6307\u5b9a\u8ba1\u7b97\u7684\u7ef4\u5ea6":82,"paddlepaddle\u7f16\u8bd1\u4f9d\u8d56":1,"pi\u5e73\u53f0\u7f16\u8bd1\u6307\u5357":118,"pod\u95f4\u901a\u4fe1":97,"python\u4e0ec":107,"python\u4ee3\u7801\u7684\u6027\u80fd\u5206\u6790":107,"python\u76f8\u5173\u7684\u5355\u5143\u6d4b\u8bd5\u90fd\u8fc7\u4e0d\u4e86":78,"return":[57,58],"rnn\u6a21\u578b":112,"rnn\u914d\u7f6e":114,"switch":[43,67,68],"tensor\u4f7f\u7528\u6837\u4f8b":76,"tensor\u5230eigentensor\u7684\u8f6c\u6362":76,"tensor\u6a21\u5757":76,AWS:95,DNS:95,EFS:95,For:9,KMS:95,The:[8,15,19,27,31,34,35,37,38,47,50,53,59,60,68],Use:[8,55],Using:[9,15],With:18,about:34,absolut:49,access:95,account:95,action:[41,42],activ:42,actor:21,add:[40,43,95],address:95,advanc:67,alalysi:13,algorithm:[6,10,22,48,56],all:[63,70],analog:19,analysi:[22,40],anneal:83,api:[22,41,42,46,50,52,57,61,65],appendix:119,arbitrari:31,architectur:[22,54],argument:[29,58],arrai:6,asset:95,associ:[63,95],assumpt:119,async:103,attent:114,attribut:[40,61],auto:6,averag:52,aws:95,background:[6,24,41,67,68,69,70,71],backward:[7,31,35,59],base:[18,49],basic:[40,67,119],batch:58,batch_siz:58,beam:[49,66],becaus:83,benchmark:[41,42],benefit:[24,59],between:[5,21,57,59,65,67],big:83,binari:8,bla:1,block:[8,32,34,35,55,57,59],blockdesc:55,bootstrap:119,bring:119,bucket:95,build:[7,34,59],can:63,capi:46,capi_priv:46,challeng:[7,24,56],chang:49,channel:21,check:6,checkpoint:[10,11,17],choic:39,choos:[9,95],client:15,clip:25,clone:72,close:6,cloudform:95,cluster:[16,95,119],cmake:[9,41,42],code:[18,32,57,109],commit:72,compar:119,comparis:57,compat:31,compil:[8,30,32,55,59,109],complet:31,compos:58,comput:[8,35,43,59,61],con:119,concept:[57,59,95],concern:42,conclus:[17,36,119],concurr:[19,21],condit:34,configur:95,construct:35,content:[41,42,46,66,78,79,81,82,83,95,108,110],control:[40,59],contruct:40,convert:17,convolut:66,core:[6,57,95],corner:7,cpu:23,creat:[7,21,58,59,63,95],createreaderop:20,creation:[14,52,61],creator:58,credenti:95,csp:21,ctc:66,cuda:[1,30,78],cudnn:1,current:[30,60],custom:58,data:[10,20,22,58,69,95],dataflow:40,dataprovid:103,dataset:[10,14],datatyp:47,decod:49,decor:58,decoratedread:20,deep:[8,31],deepspeech2:66,defin:95,definit:71,delet:95,demo:[34,95],dens:17,depend:[34,66],deploi:18,describ:[31,50],descript:[29,59],design:[5,6,8,10,14,15,16,17,19,21,22,23,24,26,27,30,31,33,34,35,38,41,42,43,44,47,48,49,50,51,53,55,57,58,59,60,63,64,66,67,68,70],destroi:[63,95],detail:[13,66],develop:59,devic:67,devicecontext:67,dictionari:58,differ:[59,67],directori:95,discrimin:34,discuss:[24,34],dispatch:[10,14],distribut:[5,10,13,18,22,24,95],dnn:42,doc:[5,8,10,14,15,16,17,19,21,22,23,24,27,30,31,33,35,38,41,42,43,44,47,51,53,55,57,58,59,60,64,66,67,68],docker:[0,18],doe:58,down:95,download:95,driver:78,drop_out:82,duplic:82,dure:[49,58],dylib:46,dynam:[10,70],dynet:36,ec2:95,eigen:76,elast:95,elect:17,els:8,engin:34,enough:6,entri:58,environ:18,error:25,evalu:26,event:[5,54],evolut:31,exampl:[5,9,19,21,37,46],except:81,execut:[8,23,31,55,59],executor:27,expand:110,explan:6,extern:95,faq:80,fault:10,feed:20,file:[8,95],fileread:20,find:95,first_seq:110,float16:30,flow:40,fluid:[19,21,31,32,43,51,109],fork:72,format:[8,10,44],forward:35,frame:8,framework:[6,76],from:[5,17,65],functor:67,futur:[31,66],gan:34,gate:114,gener:[32,34,49,119],give:58,global:[55,57],gpu:103,grad_op:7,gradient:[6,7,15,42,60],graph:[35,36,40,59,61],group:95,gru:103,handler:[5,45],happen:17,hardwar:30,helper:57,hierarchi:8,high:[50,52,61,65],how:[6,13,52,58,59,67],iam:95,ifels:37,ifelseop:8,illeg:78,imag:18,implement:[6,7,9,13,23,25,26,30,44,48,51,52,57,58,59,60,61],infer:81,infershap:[55,64],infervartyp:38,ingredi:5,ingress:28,initi:[15,34,95],insid:63,inspect:95,instal:[95,119],instanc:95,instead:58,instruct:78,insuffici:78,integr:[67,95],intel:[41,42],interact:65,interfac:[6,10,15,16,27,50,58,63],intermedi:59,introduc:[49,70],introduct:[54,61],isn:58,issu:[30,72],job:[10,18,95,96],kei:[41,47,95],kernel:[43,47,59],kube:95,kubectl:95,kubernet:[18,95,96],languag:[8,32],larg:13,last_seq:110,layer:[5,33,41,42,57,82],layout:47,learn:[8,31,83],leval:65,level:[50,52,61,65],libpaddle_capi_shar:46,libpaddle_capi_whol:46,librari:[15,30,47,59,67],limit:22,list:[11,58],live:40,load:21,local:[22,63,95],lod:49,lodtensor:[48,49,70],lodtensordesc:71,logic:14,low:[52,61,65],lstm:103,machin:49,macro:59,main:34,make:40,manag:9,map:[58,59],master:[10,14,18,19],math:67,mathemat:6,matrix:42,member:34,memori:[40,48,67,82,111,113],messag:[65,83],method:49,might:34,migrat:59,mileston:59,mini:58,minibatch:21,mkl:[41,42],mkldnn:43,mkldnn_helper:43,mkldnndevicecontext:43,model:[5,13,15,17,21,31,34,44,49,114],modul:[59,67,78],more:34,motiv:[7,21,27,44,51,56],multi:[23,32],multipl:58,mxnet:36,name:[63,78,82,95],nativ:32,nccl:51,necess:57,necessari:59,need:58,nest:48,network:[59,114],neural:114,nlp:103,norm:61,note:6,numer:6,numpi:6,nvprof:108,nvvp:108,object:10,offset:49,onli:[58,63],op_mak:59,oper:[20,33,37,40,43,47,52,55,57,59,60,64,70],opinfomap:59,opkernel:[59,67],opproto:65,ops:61,optim:[10,15,35,40,50,57],option:29,opwithkernel:59,order:29,org:77,origin:59,orthogon:63,other:42,output:95,overview:[17,25,27,41,42,59,63,66,109],pack:[41,49],packag:9,paddl:[13,51,58,63,76,78,82],paddlejob:18,paddlepaddl:[5,8,21,31,32,41,42,52,55,61,62,66,77,78,95,109,115],pair:95,paradigm:31,parallel_nn:105,paramet:[5,10,15,16,18,21,24,42,52,53,57,61,95],parameteraverageoptim:52,parent:63,part:35,partit:15,path:[17,29],penalti:61,perform:[52,103],persist:14,pfsclient:[28,29],pfsserver:28,place:[40,47,67],placement:22,platform:78,point:[41,81,95],polici:40,pool:110,pose:[38,60],potenti:39,pre:72,prefetch:58,prepar:95,principl:43,privat:95,pro:119,problem:[26,38,39,40,47,50,60,69],procedur:119,process:[10,15,18,50,59],program:[8,19,21,23,31,32,55,57],programdesc:[32,55],project:9,propos:[38,60,61],protobuf:64,protocol:83,provid:58,prune:56,pserver:17,pull:72,push:72,python:[6,18,22,41,42,48,50,52,57,58,61,65,71,89],qualiti:59,queue:[10,14],raspberri:118,rate:83,reader:[5,20,58],readerbas:20,readerhold:20,readop:20,realiz:59,recoveri:10,recurr:[82,113,114],recv:21,refactor:59,refer:[6,22,24,40,41,42,66],region:95,regist:[38,59,65],registr:[59,60],registri:59,regular:[15,61],reject:83,rel:49,relat:[20,59,70],remot:16,remoteexecutor:22,render:95,represent:[8,59],request:72,requir:[9,34],retri:14,reus:57,rnn:[48,70,103,111],rnnop:[8,48,59],route53:95,row:[64,66],rpc:21,run:[27,109],runtim:18,save:17,scale:10,scope:[8,48,59,63],search:[49,66],secur:95,select:[15,21,64],selectedrow:64,semant:68,send:21,separ:59,sequenc:[49,114],server:[10,14,15,18,21,24,95],servic:95,setup:95,sextant:119,sgd:103,shape:49,share:[5,7,40,63],should:63,shuffl:58,simpl:49,singl:58,solut:[38,39,40,41,47,56,60,69],sourc:109,spars:[15,16,17,64],stack:8,start:[5,95],statement:26,step2:90,step:[48,90],storag:61,store:10,strategi:40,subcommond:29,submit:18,suffici:58,suitabl:9,sulut:43,summar:[5,19],summari:44,support:[30,51,67,70,78],survei:[30,36,61,119],synopsi:29,syntax:21,system:[31,95],tabl:[46,66],task:[10,14,66],tear:95,tecton:119,templat:95,tensor:[59,67,76],tensorarrai:[49,70],tensordesc:71,tensorflow:36,test:[41,42,43],theori:6,thi:[63,78],think:34,three:70,time:109,timelin:17,todo:[11,12,23],togeth:63,toler:10,too:83,tool:[9,119],topic:67,toward:32,train:[5,10,13,16,18,22,50,58,95],trainer:[10,15,17,18,21,95],transform:69,translat:49,transpil:[22,23,24,32,40,51],tune:103,ture:31,two:6,type:[21,47],uniform:70,unit:[41,42,43],unpack:49,updat:[5,16,17,95],usag:[7,25,48,49,58],use:[13,58],user:10,valu:57,vardesc:71,variabl:[7,40,57,59,63,71],verifi:95,version:[19,30,78],volum:95,vpc:95,what:[13,17],wheel:78,when:[17,63],whl:78,why:[30,31,52,58,59,70],work:66,worker:19}})
\ No newline at end of file