未验证 提交 0c71c839 编写于 作者: Z zhaoyuchen2018 提交者: GitHub

Fix recurrent op not update grade issue (#18581)

* Fix recurrent op fails

For the variable used in outter block,
copy sub block's grad variable to outter block

test=develop
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* Fix unicode error

test=develop
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* Refine test code

test=develop
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* Fix seq2seq case fails

test=develop
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* remove unreasonable code.

test=develop
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* Refine code according to comment

test=develop
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>
上级 d53fa53b
develop 2.0.1-rocm-post Ligoml-patch-1 OliverLPH-patch-1 OliverLPH-patch-2 PaddlePM-patch-1 PaddlePM-patch-2 ZHUI-patch-1 add_default_att add_model_benchmark_ci add_some_yaml_config addfile all_new_design_exec ascendrc ascendrelease cherry_undefined_var compile_windows delete_2.0.1-rocm-post delete_add_default_att delete_all_new_design_exec delete_ascendrc delete_compile_windows delete_delete_addfile delete_disable_iterable_dataset_unittest delete_fix_dataloader_memory_leak delete_fix_imperative_dygraph_error delete_fix_retry_ci delete_fix_undefined_var delete_improve_sccache delete_paddle_tiny_install delete_paralleltest delete_prv-disable-more-cache delete_revert-31068-fix_conv3d_windows delete_revert-31562-mean delete_revert-33630-bug-fix delete_revert-34159-add_npu_bce_logical_dev delete_revert-34910-spinlocks_for_allocator delete_revert-35069-revert-34910-spinlocks_for_allocator delete_revert-36057-dev/read_flags_in_ut dingjiaweiww-patch-1 disable_iterable_dataset_unittest dy2static enable_eager_model_test final_state_gen_python_c final_state_intermediate fix-numpy-issue fix_concat_slice fix_dataloader_memory_leak fix_imperative_dygraph_error fix_npu_ci fix_op_flops fix_retry_ci fix_rnn_docs fix_tensor_type fix_undefined_var fixiscan fixiscan1 fixiscan2 fixiscan3 github/fork/123malin/netifaces github/fork/123malin/tdm_abacus github/fork/AshburnLee/dev_unique github/fork/ForFishes/fix_memory_matmul github/fork/ForFishes/rm_fluid github/fork/LielinJiang/move-2.0-api github/fork/LielinJiang/visual-dl-cb github/fork/LiuChiachi/add-transformer-generate-square-subsequent-mask-api github/fork/LiuChiachi/fix-example-code-for-hapi-Model github/fork/LiuChiachi/remove-input-requirment-in-dygraph-Model github/fork/MrChengmo/fix_ps_profiler github/fork/MrChengmo/update_ps_heter github/fork/PWhiddy/patch-1 github/fork/Shixiaowei02/dev/save_load_upgrade github/fork/TCChenlong/fix_hapi github/fork/TCChenlong/fix_inden github/fork/Thunderbrook/xpu_slice github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var_2 github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var_3 github/fork/XieYunshen/timeout_20S_ut github/fork/ZeyuChen/remove-nltk github/fork/arlesniak/arlesniak/selective__mkldnn_flags github/fork/baiyfbupt/code_doc_mig github/fork/chalsliu/set_timeout github/fork/chen-zhiyu/develop github/fork/chenwhql/ci/try_to_find_test_buffer_shared_memory_reuse_pass_error github/fork/chenwhql/dygraph/remove_scale_loss_and_apply_collective_grads github/fork/chenwhql/saveload/add_get_inference_program github/fork/chenwhql/saveload/remove_save_load_config github/fork/cryoco/pass-compatibility-trt github/fork/danleifeng/isempty_api2.0 github/fork/frankwhzhang/api_transfer github/fork/hbwx24/error_msg/cuda_kernel_error_msg github/fork/heavengate/cherry_yolo_box github/fork/heavengate/update_yolo_box github/fork/iclementine/rnn_fix github/fork/iducn/testestse github/fork/jczaja/prv-25537-fix github/fork/jeff41404/release/1.8 github/fork/jiweibo/api_2.0 github/fork/jiweibo/fix_lite_resnet50_test github/fork/juncaipeng/fix_doc_1 github/fork/lfchener/sample_code github/fork/littletomatodonkey/fix_reg_doc github/fork/liym27/dy2stat_update_assign_to_rc20 github/fork/luotao1/profiler_ut github/fork/mapingshuo/add_wait github/fork/mapingshuo/doc_2.0 github/fork/mapingshuo/zero-0.5 github/fork/miraiwk/dev github/fork/pangyoki/add-Categorical-class-branch github/fork/pangyoki/add-multinomial-op-branch github/fork/pangyoki/fix-test_distritbution-CI github/fork/qjing666/doublegrad github/fork/qjing666/fix_hdfs_download github/fork/sandyhouse/add_gather_etc github/fork/sandyhouse/add_send_recv_alltoall_etc github/fork/sandyhouse/pipeline_exe_run github/fork/seiriosPlus/feature/large_scale_kv_save_delta github/fork/seiriosPlus/fix/paddle_errors_fix github/fork/seiriosPlus/fix/paddle_op_errors github/fork/shangzhizhou/fix_test_activation_op_random_bug github/fork/smallv0221/yxp0924 github/fork/smallv0221/yxp0925 github/fork/swtkiwi/del-matplotlib github/fork/tianshuo78520a/kunlun_test github/fork/tianshuo78520a/update_dockerfile github/fork/wanghaoshuang/bert_fuse github/fork/wanghaoshuang/label_smooth github/fork/wanghuancoder/develop_CUDASynchronize github/fork/wanghuancoder/develop_Layer_doc github/fork/wanghuancoder/develop_ParameterList_doc github/fork/wanghuancoder/develop_Sequential_doc github/fork/wanghuancoder/develop_bilinear_tensor_product github/fork/wanghuancoder/develop_coverage_build_sh github/fork/wanghuancoder/develop_in_dynamic_mode_doc github/fork/wanghuancoder/develop_unique_name_doc github/fork/wangxicoding/fleet_meta_combine github/fork/wawltor/error_message_fix_5 github/fork/willthefrog/remove_l2_norm github/fork/windstamp/momentum_op github/fork/windstamp/mv_op_5 github/fork/windstamp/normal_api github/fork/wojtuss/wojtuss/fusion_gru_quantization github/fork/wojtuss/wojtuss/quantization-with-shift github/fork/wzzju/fix_err_info github/fork/wzzju/pure_fp16 github/fork/xiemoyuan/op_error_message github/fork/xiemoyuan/optimize_error_message github/fork/yaoxuefeng6/fix_doc github/fork/yaoxuefeng6/mod_dataset_v2 github/fork/yongqiangma/lod github/fork/ysh329/fix-clip-by-norm-error github/fork/ysh329/fix-error-clip-by-value github/fork/yukavio/error_info github/fork/zhangting2020/conv_filter_grad github/fork/zhangting2020/is_compile_with_cuda github/fork/zhangting2020/place_doc github/fork/zhangting2020/program github/fork/zhhsplendid/fix_any github/fork/zhhsplendid/refine_api2 github/fork/zhhsplendid/refine_api2_test github/fork/zhhsplendid/refine_api_test_ptb_lm github/fork/zhhsplendid/refine_api_test_resnet github/fork/zhhsplendid/refine_api_test_simnet github/fork/zhiqiu/dev/refine_initializer github/fork/zhiqiu/dev/remove_inplace_argument github/fork/zlsh80826/nvinfer_plugin_var_len_cuda11 improve_sccache incubate/infrt inplace_addto make_flag_adding_easier move_embedding_to_phi move_histogram_to_pten move_sgd_to_phi move_slice_to_pten move_temporal_shift_to_phi move_yolo_box_to_phi npu_fix_alloc numel paddle_tiny_install paralleltest preln_ernie prv-disable-more-cache prv-md-even-more prv-onednn-2.5 pten_tensor_refactor release/1.6 release/1.7 release/1.8 release/2.0 release/2.0-alpha release/2.0-beta release/2.0-rc release/2.0-rc1 release/2.1 release/2.2 release/2.3 release/2.3-fc-ernie-fix release/2.4 revert-24981-add_device_attr_for_regulization revert-26856-strategy_example2 revert-27520-disable_pr revert-31068-fix_conv3d_windows revert-31562-mean revert-32290-develop-hardlabel revert-33037-forci revert-33475-fix_cifar_label_dimension revert-33630-bug-fix revert-34159-add_npu_bce_logical_dev revert-34406-add_copy_from_tensor revert-34910-spinlocks_for_allocator revert-35069-revert-34910-spinlocks_for_allocator revert-36057-dev/read_flags_in_ut revert-36201-refine_fast_threaded_ssa_graph_executor revert-36985-add_license revert-37318-refactor_dygraph_to_eager revert-37926-eager_coreops_500 revert-37956-revert-37727-pylayer_support_tuple revert-38100-mingdong revert-38301-allocation_rearrange_pr revert-38703-numpy_bf16_package_reupload revert-38732-remove_useless_header_in_elementwise_mul_grad revert-38959-Reduce_Grad revert-39143-adjust_empty revert-39227-move_trace_op_to_pten revert-39268-dev/remove_concat_fluid_kernel revert-40170-support_partial_grad revert-41056-revert-40727-move_some_activaion_to_phi revert-41065-revert-40993-mv_ele_floordiv_pow revert-41068-revert-40790-phi_new revert-41944-smaller_inference_api_test revert-42149-do-not-reset-default-stream-for-stream-safe-cuda-allocator revert-43155-fix_ut_tempfile revert-43882-revert-41944-smaller_inference_api_test revert-45808-phi/simplify_size_op revert-46827-deform_comment rocm_dev_0217 support_weight_transpose test_benchmark_ci test_feature_precision_test_c test_model_benchmark test_model_benchmark_ci zhiqiu-patch-1 v2.4.0-rc0 v2.3.2 v2.3.1 v2.3.0 v2.3.0-rc0 v2.2.2 v2.2.1 v2.2.0 v2.2.0-rc0 v2.2.0-bak0 v2.1.3 v2.1.2 v2.1.1 v2.1.0 v2.1.0-rc0 v2.0.2 v2.0.1 v2.0.0 v2.0.0-rc1 v2.0.0-rc0 v2.0.0-beta0 v2.0.0-alpha0 v1.8.5 v1.8.4 v1.8.3 v1.8.2 v1.8.1 v1.8.0 v1.7.2 v1.7.1 v1.7.0 v1.6.3 v1.6.2 v1.6.1 v1.6.0 v1.6.0-rc0
无相关合并请求
......@@ -576,7 +576,7 @@ class StaticRNN(object):
if in_var_name not in local_inputs:
params.append(in_var_name)
parameters = [parent_block.var(name) for name in params]
parameters = [parent_block.var(name) for name in set(params)]
step_scope = parent_block.create_var(
type=core.VarDesc.VarType.STEP_SCOPES)
......
......@@ -464,5 +464,143 @@ class RecurrentOpNoMemBootTest(RecurrentOpTest1):
return rnn()
class RecurrentOpSubBlockTest(RecurrentOpTest1):
'''
Test RNNOp with subblock variable
equation:
y_ = emb * w1
h_t = \concat([x, h_{t-1}])
h_t = h_t * w2
h_t = \\unsqueeze(h_t, 1)
h_t = \dot_attention(h_t, y_)
h_t = \squeeze(h_t, 1)
y = h_t
vars:
- x
- w1
- w2
memories:
- h
outputs:
- y
'''
class PySimpleRNN5(PyRNNBase):
def __init__(self, input_shape, output_shape):
super(RecurrentOpSubBlockTest.PySimpleRNN5, self).__init__(
input_shape, output_shape)
seq_len, batch_size, input_dim = input_shape
self.w1 = np.random.uniform(
-0.1, 0.1, size=(input_dim, input_dim)).astype("float32")
self.w2 = np.random.uniform(
-0.1, 0.1, size=(input_dim * 2, input_dim)).astype("float32")
self.emb = np.random.uniform(
-0.1, 0.1, size=(seq_len, batch_size,
input_dim)).astype("float32")
men_dim = (seq_len, batch_size, input_dim)
self.mems = np.zeros(shape=men_dim).astype("float32")
self.oy = np.matmul(self.emb, self.w1)
def step(self, step_id, x):
def dot_attention(query, memory):
attn = np.matmul(query, memory.transpose((0, 2, 1)))
weight = softmax(attn)
weight_memory = np.matmul(weight, memory)
return weight_memory, weight
def softmax(x):
return np.exp(x) / sum(np.exp(x))
if step_id == 0:
pre_mem = np.zeros_like(x)
else:
pre_mem = self.mems[step_id - 1]
concat_in = np.concatenate([x, pre_mem], 1)
new_mem = np.matmul(concat_in, self.w2)
new_mem = np.expand_dims(new_mem, 1)
new_mem, _ = dot_attention(new_mem, self.oy)
new_mem = np.squeeze(new_mem, 1)
self.mems[step_id] = new_mem
self.y[step_id] = self.mems[step_id]
input_dim = 2
batch_size = 3
sent_len = 3
def setUp(self):
self.setup_program()
self.data_field = {"x", "emb", "w1", "w2"}
self.input_shape = (self.sent_len, self.batch_size, self.input_dim)
self.output_shape = (self.sent_len, self.batch_size, self.input_dim)
self.py_rnn = RecurrentOpSubBlockTest.PySimpleRNN5(self.input_shape,
self.output_shape)
with fluid.program_guard(self.main_program, self.startup_program):
rnn_out = self.create_rnn_op()
self.output = layers.mean(rnn_out)
def create_rnn_op(self):
x = layers.data(
shape=[self.sent_len, self.batch_size, self.input_dim],
dtype='float32',
name='x',
append_batch_size=False)
x.stop_gradient = False
emb = layers.data(
name='emb',
shape=[self.sent_len, self.batch_size, self.input_dim],
dtype='float32',
append_batch_size=False)
emb.stop_gradient = False
w1 = layers.data(
shape=[self.input_dim, self.input_dim],
dtype='float32',
name='w1',
append_batch_size=False)
w1.stop_gradient = False
w2 = layers.data(
shape=[self.input_dim * 2, self.input_dim],
dtype='float32',
name='w2',
append_batch_size=False)
w2.stop_gradient = False
rnn = layers.StaticRNN()
def dot_attention(query, memory):
attn = layers.matmul(query, memory, transpose_y=True)
weight = layers.softmax(attn)
weight_memory = layers.matmul(weight, memory)
return weight_memory, weight
y = layers.matmul(emb, w1)
with rnn.step():
pre_h = rnn.memory(
shape=(self.sent_len, self.input_dim),
batch_ref=x,
init_value=0.0)
step_in = rnn.step_input(x)
concat_in = layers.concat([step_in, pre_h], 1)
new_h = layers.matmul(concat_in, w2)
new_h = layers.unsqueeze(new_h, [1])
new_h, _ = dot_attention(new_h, y)
new_h = layers.squeeze(new_h, [1])
rnn.update_memory(pre_h, new_h)
rnn.step_output(new_h)
return rnn()
if __name__ == '__main__':
unittest.main()
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册
反馈
建议
客服 返回
顶部