| .. |
|
concurrency
|
Pending more CPPLint errors in fluid/operators/math (#10243)
|
7 years ago |
|
detail
|
refine seq_concat
|
7 years ago |
|
detection
|
Cuda speed for generate_proposals_op. (#13596)
|
7 years ago |
|
distributed
|
[Feature] dist op role and lr op role, to support memory optimize with dist training (#13220)
|
7 years ago |
|
math
|
fix bug vtanh
|
7 years ago |
|
nccl
|
cudnn windows support (#12757)
|
7 years ago |
|
reader
|
Tiny refines for lod_tensor_blocking_queue and reshape_op
|
7 years ago |
|
CMakeLists.txt
|
[Accelerate] Refine seq_softmax_op (#13421)
|
7 years ago |
|
accuracy_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
accuracy_op.cu
|
Feature/cuda9 cudnn7 (#10140)
|
7 years ago |
|
accuracy_op.h
|
…
|
|
|
activation_mkldnn_op.cc
|
MKLDNN layout: Support for activation operator
|
7 years ago |
|
activation_op.cc
|
flags (#13542)
|
7 years ago |
|
activation_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
activation_op.h
|
fix windows compile (#13147)
|
7 years ago |
|
adadelta_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
adadelta_op.cu
|
…
|
|
|
adadelta_op.h
|
…
|
|
|
adagrad_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
adagrad_op.cu
|
Feature/cuda9 cudnn7 (#10140)
|
7 years ago |
|
adagrad_op.h
|
…
|
|
|
adam_op.cc
|
fix adam op for selected rows
|
7 years ago |
|
adam_op.cu
|
…
|
|
|
adam_op.h
|
modification
|
7 years ago |
|
adamax_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
adamax_op.cu
|
…
|
|
|
adamax_op.h
|
…
|
|
|
arg_max_op.cc
|
Polish arg_min_max_op
|
7 years ago |
|
arg_max_op.cu
|
Polish arg_min_max_op
|
7 years ago |
|
arg_min_max_op_base.h
|
Polish arg_min_max_op
|
7 years ago |
|
arg_min_op.cc
|
Polish arg_min_max_op
|
7 years ago |
|
arg_min_op.cu
|
Polish arg_min_max_op
|
7 years ago |
|
argsort_op.cc
|
Add python api for argsort_op
|
7 years ago |
|
argsort_op.cu
|
Enhance cuda code & unittest for argsort_op
|
7 years ago |
|
argsort_op.h
|
Simplify the computation in cpu
|
7 years ago |
|
array_operator.h
|
Fix cpplint issues in some operators
|
7 years ago |
|
array_to_lod_tensor_op.cc
|
fix mac compile error
|
7 years ago |
|
assign_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
assign_value_op.cc
|
Register assign_value_op an empty grad_op
|
7 years ago |
|
assign_value_op.cu.cc
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
assign_value_op.h
|
Fix cpplint issues in some operators
|
7 years ago |
|
attention_lstm_op.cc
|
refine infershape hasinput and hasoutput
|
7 years ago |
|
attention_lstm_op.h
|
implement attention lstm cpu forward
|
7 years ago |
|
auc_op.cc
|
Batch AUC (#13567)
|
7 years ago |
|
auc_op.h
|
Batch AUC (#13567)
|
7 years ago |
|
average_accumulates_op.cc
|
small AverageOptimizer enhance. (#11761)
|
7 years ago |
|
average_accumulates_op.cu
|
Fix average_accumulate_op for parallel executor. (#9852)
|
7 years ago |
|
average_accumulates_op.h
|
small AverageOptimizer enhance. (#11761)
|
7 years ago |
|
batch_norm_mkldnn_op.cc
|
Refactor code
|
7 years ago |
|
batch_norm_op.cc
|
Revert "Disable in_place in batch_norm API. (#12736)"
|
7 years ago |
|
batch_norm_op.cu.cc
|
Skip BatchNorm when feature only has 1 element. (#11578)
|
7 years ago |
|
batch_norm_op.h
|
add url of cuda9.0_cudnn7_avx_mkl library
|
7 years ago |
|
batch_size_like.h
|
Simplize API Reference Documentation
|
7 years ago |
|
beam_search_decode_op.cc
|
Feature/template (#13093)
|
7 years ago |
|
beam_search_decode_op.h
|
Fix copying empty tensor in beam_search_decode_op
|
7 years ago |
|
beam_search_decode_op_test.cc
|
Make python unit test of beam_search_op and beam_searc_decode_op run correctly
|
7 years ago |
|
beam_search_op.cc
|
Fix src_idx out of range in beam_search_op
|
7 years ago |
|
beam_search_op.h
|
Make cc_test of beam_search_op and beam_searc_decode_op run correctly
|
7 years ago |
|
beam_search_op_test.cc
|
Make python unit test of beam_search_op and beam_searc_decode_op run correctly
|
7 years ago |
|
bilinear_interp_op.cc
|
Add unit tests
|
7 years ago |
|
bilinear_interp_op.cu
|
Make bilinear_interp_op support attrs from input. (#11041)
|
7 years ago |
|
bilinear_interp_op.h
|
Add unit tests
|
7 years ago |
|
bilinear_tensor_product_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
bilinear_tensor_product_op.cu
|
…
|
|
|
bilinear_tensor_product_op.h
|
Clean and extract blas
|
7 years ago |
|
cast_op.cc
|
Add support in cast op kernel
|
7 years ago |
|
cast_op.cu
|
Add support in cast op kernel
|
7 years ago |
|
cast_op.h
|
Feature/template (#13093)
|
7 years ago |
|
channel_close_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
channel_create_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
channel_recv_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
channel_send_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
checkpoint_notify_op.cc
|
enforce rpc client timeout
|
7 years ago |
|
chunk_eval_op.cc
|
update by comment
|
7 years ago |
|
chunk_eval_op.h
|
Fix CPPLint issues with Chunk_eval_op (#9964)
|
7 years ago |
|
clip_by_norm_op.cc
|
fix typo
|
7 years ago |
|
clip_by_norm_op.cu
|
…
|
|
|
clip_by_norm_op.h
|
…
|
|
|
clip_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
clip_op.cu
|
…
|
|
|
clip_op.h
|
fix sparse gradient clip
|
7 years ago |
|
compare_op.cc
|
Polish documentation
|
7 years ago |
|
compare_op.cu
|
…
|
|
|
compare_op.h
|
…
|
|
|
concat_op.cc
|
Fix concat_op InferShape (#13513)
|
7 years ago |
|
concat_op.cu.cc
|
enable more type for splitOp and ConcatOp
|
7 years ago |
|
concat_op.h
|
speed up lod_tensor to array and array to lod_tensor
|
7 years ago |
|
conditional_block_op.cc
|
Fix bug in conditional_block_op. (#12246)
|
7 years ago |
|
conv_cudnn_op.cu.cc
|
Revert "Revert "Add CudnnHolder and use it in Conv and ConvTranspose op""
|
7 years ago |
|
conv_mkldnn_op.cc
|
Fuse Conv+BN+SkipConnectionAdd+ReLU with transpiler temporarily (#13350)
|
7 years ago |
|
conv_op.cc
|
Fuse Conv+BN+SkipConnectionAdd+ReLU with transpiler temporarily (#13350)
|
7 years ago |
|
conv_op.cu.cc
|
…
|
|
|
conv_op.h
|
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into feature/clean_op_maker
|
7 years ago |
|
conv_shift_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
conv_shift_op.cu
|
Feature/cuda9 cudnn7 (#10140)
|
7 years ago |
|
conv_shift_op.h
|
…
|
|
|
conv_transpose_cudnn_op.cu.cc
|
Revert "Revert "Add CudnnHolder and use it in Conv and ConvTranspose op""
|
7 years ago |
|
conv_transpose_op.cc
|
loosen the restriction of output_size in conv2d_transpose (#12292)
|
7 years ago |
|
conv_transpose_op.cu.cc
|
Speed depthwise transposed conv2d. (#11740)
|
7 years ago |
|
conv_transpose_op.h
|
Speed depthwise transposed conv2d. (#11740)
|
7 years ago |
|
cos_sim_op.cc
|
update by comment
|
7 years ago |
|
cos_sim_op.cu
|
…
|
|
|
cos_sim_op.h
|
…
|
|
|
crf_decoding_op.cc
|
update
|
7 years ago |
|
crf_decoding_op.h
|
Optimize CRF Decoding with AVX/AVX2/AVX512F instruction (#12767)
|
7 years ago |
|
crop_op.cc
|
Fix crop op. (#12603)
|
7 years ago |
|
crop_op.cu
|
Fix crop op. (#12603)
|
7 years ago |
|
crop_op.h
|
Fix crop op. (#12603)
|
7 years ago |
|
cross_entropy_op.cc
|
Add ignore_index in cross_entropy op (#13217)
|
7 years ago |
|
cross_entropy_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
cross_entropy_op.h
|
fix warning
|
7 years ago |
|
ctc_align_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
ctc_align_op.cu
|
Fix cpplint errors for a set of operators (#9837)
|
7 years ago |
|
ctc_align_op.h
|
Fix cpplint errors for a set of operators (#9837)
|
7 years ago |
|
cum_op.h
|
…
|
|
|
cumsum_op.cc
|
Polish comsum, DynamicRNN
|
7 years ago |
|
cumsum_op.cu
|
add semicolon to op registry (#10034)
|
7 years ago |
|
decayed_adagrad_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
decayed_adagrad_op.cu
|
…
|
|
|
decayed_adagrad_op.h
|
…
|
|
|
delete_var_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
detection_map_op.cc
|
Refine detection_map doc.
|
7 years ago |
|
detection_map_op.h
|
Revert "Revert "Merge pull request #13431 from chengduoZH/refine_lod""
|
7 years ago |
|
dropout_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
dropout_op.cu
|
Revert "accelerate dropout (#9902)" (#10082)
|
7 years ago |
|
dropout_op.h
|
Revert "accelerate dropout (#9902)" (#10082)
|
7 years ago |
|
dropout_op_test.cc
|
Revert "accelerate dropout (#9902)" (#10082)
|
7 years ago |
|
edit_distance_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
edit_distance_op.cu
|
Feature/cuda9 cudnn7 (#10140)
|
7 years ago |
|
edit_distance_op.h
|
…
|
|
|
elementwise_add_mkldnn_op.cc
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
elementwise_add_op.cc
|
explicit gradient of elementwise_add/elementwise_sub (#11970)
|
7 years ago |
|
elementwise_add_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
elementwise_add_op.h
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
elementwise_div_op.cc
|
explicit gradient of elementwise_add/elementwise_sub (#11970)
|
7 years ago |
|
elementwise_div_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
elementwise_div_op.h
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
elementwise_max_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
elementwise_max_op.cu
|
…
|
|
|
elementwise_max_op.h
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
elementwise_min_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
elementwise_min_op.cu
|
…
|
|
|
elementwise_min_op.h
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
elementwise_mul_op.cc
|
modification
|
7 years ago |
|
elementwise_mul_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
elementwise_mul_op.h
|
modification
|
7 years ago |
|
elementwise_op.h
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
elementwise_op_function.h
|
Feature/op_fuse_pass (#12440)
|
7 years ago |
|
elementwise_pow_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
elementwise_pow_op.cu
|
…
|
|
|
elementwise_pow_op.h
|
…
|
|
|
elementwise_sub_op.cc
|
explicit gradient of elementwise_add/elementwise_sub (#11970)
|
7 years ago |
|
elementwise_sub_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
elementwise_sub_op.h
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
expand_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
expand_op.cu
|
…
|
|
|
expand_op.h
|
Fix CPPLint issues in expand_op, gather_op and get_places_op (#10000)
|
7 years ago |
|
extract_rows_op.cc
|
Revert "Revert "Merge pull request #13431 from chengduoZH/refine_lod""
|
7 years ago |
|
fake_dequantize_op.cc
|
Improve fake_dequantize_op. (#12877)
|
7 years ago |
|
fake_dequantize_op.cu
|
Improve fake_dequantize_op. (#12877)
|
7 years ago |
|
fake_dequantize_op.h
|
Improve fake_dequantize_op. (#12877)
|
7 years ago |
|
fake_quantize_op.cc
|
Improve and fix fake_quantize_op (#13092)
|
7 years ago |
|
fake_quantize_op.cu
|
Fix fake_quantize_op
|
7 years ago |
|
fake_quantize_op.h
|
Improve and fix fake_quantize_op (#13092)
|
7 years ago |
|
fc_mkldnn_op.cc
|
fix bias attri in mkldnn fc
|
7 years ago |
|
fc_op.cc
|
fea/link ir to inference analysis and fc fuse support (#12789)
|
7 years ago |
|
fc_op.h
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
feed_op.cc
|
optimize profiler
|
7 years ago |
|
fetch_barrier_op.cc
|
Refine dist rpc deps (#12899)
|
7 years ago |
|
fetch_op.cc
|
optimize profiler
|
7 years ago |
|
fill_constant_batch_size_like_op.cc
|
Simplize API Reference Documentation
|
7 years ago |
|
fill_constant_batch_size_like_op.cu.cc
|
…
|
|
|
fill_constant_batch_size_like_op.h
|
Fix fill_constant_batch_size_like_op when input is LoDTensor. (#10943)
|
7 years ago |
|
fill_constant_op.cc
|
code optimize
|
7 years ago |
|
fill_op.cc
|
Feature/template (#13093)
|
7 years ago |
|
fill_zeros_like_op.cc
|
Revert "Extend fill_zeros_like_op for zero-filling an LoDTensorArray (#11496)"
|
7 years ago |
|
fill_zeros_like_op.cu.cc
|
…
|
|
|
fill_zeros_like_op.h
|
Revert "Extend fill_zeros_like_op for zero-filling an LoDTensorArray (#11496)"
|
7 years ago |
|
flatten_op.cc
|
Refine reshape_grad and transpose_grad (#13074)
|
7 years ago |
|
ftrl_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
ftrl_op.cu
|
…
|
|
|
ftrl_op.h
|
…
|
|
|
fused_elemwise_activation_op.cc
|
Feature/op_fuse_pass (#12440)
|
7 years ago |
|
fused_elemwise_activation_op.cu
|
Feature/op fusion (#12240)
|
7 years ago |
|
fused_elemwise_activation_op.h
|
Feature/op_fuse_pass (#12440)
|
7 years ago |
|
fusion_gru_op.cc
|
refine infershape hasinput and hasoutput
|
7 years ago |
|
fusion_gru_op.h
|
init fusion gru op
|
7 years ago |
|
fusion_lstm_op.cc
|
refine peephole
|
7 years ago |
|
fusion_lstm_op.h
|
implement attention lstm cpu forward
|
7 years ago |
|
fusion_seqexpand_concat_fc_op.cc
|
fea/fuse attention lstm simplify.with fusion lstm.with sequnce expand (#13006)
|
7 years ago |
|
fusion_seqexpand_concat_fc_op.h
|
rename fusion seq_concat_fc to fusion seqexpand_concat_fc
|
7 years ago |
|
gather.cu.h
|
…
|
|
|
gather.h
|
…
|
|
|
gather_op.cc
|
Faster RCNN Generate Proposal Labels (#12616)
|
7 years ago |
|
gather_op.cu
|
Fix CPPLint issues in expand_op, gather_op and get_places_op (#10000)
|
7 years ago |
|
gather_op.h
|
Fix CPPLint issues in expand_op, gather_op and get_places_op (#10000)
|
7 years ago |
|
gather_test.cc
|
fix some bugs introduced by unfreed memory
|
7 years ago |
|
gaussian_random_batch_size_like_op.cc
|
Update some doc about API reference. (#11495)
|
7 years ago |
|
gaussian_random_mkldnn_op.cc
|
MKLDNN gausian_random tests
|
7 years ago |
|
gaussian_random_op.cc
|
MKLDNN layouts: Gaussian random layout
|
7 years ago |
|
gaussian_random_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
gen_nccl_id_op.cc
|
move dist codes from operaotrs/detail to operators/distributed
|
7 years ago |
|
get_places_op.cc
|
Add API docs.
|
7 years ago |
|
go_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
gru_op.cc
|
add some comment
|
7 years ago |
|
gru_op.cu.cc
|
seperate cpu and gpu implementations for gru kernel compute
|
7 years ago |
|
gru_op.h
|
refine packed condition
|
7 years ago |
|
gru_unit_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
gru_unit_op.cu
|
…
|
|
|
gru_unit_op.h
|
fix mac compile error 0903 (#13184)
|
7 years ago |
|
hierarchical_sigmoid_op.cc
|
complete the hsigmoid_op
|
7 years ago |
|
hierarchical_sigmoid_op.h
|
Fix the HierarchicalSigmoidGradOpKernel and refine the codes. Now hsigmoid_op is same with V2 implementation and can pass gradient check.
|
7 years ago |
|
hinge_loss_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
hinge_loss_op.cu
|
…
|
|
|
hinge_loss_op.h
|
…
|
|
|
huber_loss_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
huber_loss_op.cu
|
…
|
|
|
huber_loss_op.h
|
…
|
|
|
im2sequence_op.cc
|
Fix infershape of im2sequence. (#12183)
|
7 years ago |
|
im2sequence_op.cu
|
…
|
|
|
im2sequence_op.h
|
Fix infershape of im2sequence. (#12183)
|
7 years ago |
|
increment_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
increment_op.cu
|
add semicolon to op registry (#10034)
|
7 years ago |
|
increment_op.h
|
kernels of increment op
|
7 years ago |
|
is_empty_op.cc
|
Enhanced is_empty_op for our seq2seq model (#10704)
|
7 years ago |
|
is_empty_op.h
|
Enhanced is_empty_op for our seq2seq model (#10704)
|
7 years ago |
|
l1_norm_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
l1_norm_op.cu
|
…
|
|
|
l1_norm_op.h
|
…
|
|
|
label_smooth_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
label_smooth_op.cu
|
…
|
|
|
label_smooth_op.h
|
fix windows compile (#13147)
|
7 years ago |
|
layer_norm_op.cc
|
Polish documentation
|
7 years ago |
|
layer_norm_op.cu
|
Use double to reduce
|
7 years ago |
|
layer_norm_op.h
|
Clean and extract blas
|
7 years ago |
|
linear_chain_crf_op.cc
|
follow comments
|
7 years ago |
|
linear_chain_crf_op.cu
|
…
|
|
|
linear_chain_crf_op.h
|
Fix cpplint errors for a set of operators (#9837)
|
7 years ago |
|
listen_and_serv_op.cc
|
Polish code
|
7 years ago |
|
listen_and_serv_op.h
|
cache var types
|
7 years ago |
|
load_combine_op.cc
|
Add FP16 option to load_combine op (#10601)
|
7 years ago |
|
load_op.cc
|
add sync in load selectedrows
|
7 years ago |
|
lod_array_length_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
lod_rank_table_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
lod_reset_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
lod_reset_op.cu
|
…
|
|
|
lod_reset_op.h
|
fix errors in lod_reset_op
|
7 years ago |
|
lod_tensor_to_array_op.cc
|
speed up lod_tensor to array and array to lod_tensor
|
7 years ago |
|
log_loss_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
log_loss_op.cu
|
…
|
|
|
log_loss_op.h
|
…
|
|
|
logical_op.cc
|
update
|
7 years ago |
|
logical_op.cu
|
…
|
|
|
logical_op.h
|
…
|
|
|
lookup_sparse_table_op.cc
|
Optimize selected rows for dist lookup table with pthread rwlock (#12635)
|
7 years ago |
|
lookup_table_op.cc
|
make look_up_op supporting tensor ids
|
7 years ago |
|
lookup_table_op.cu
|
Fix MixedVector
|
7 years ago |
|
lookup_table_op.h
|
test/add text-classification test (#13081)
|
7 years ago |
|
lrn_mkldnn_op.cc
|
Function for running MKLDNN primitive added. Unittest added for is_test attribute
|
7 years ago |
|
lrn_op.cc
|
Mkldnn layout (#11040)
|
7 years ago |
|
lrn_op.cu
|
…
|
|
|
lrn_op.h
|
Plain LRN op throws an exception when is_test is set in backward pass
|
7 years ago |
|
lstm_op.cc
|
polish doc
|
7 years ago |
|
lstm_op.cu.cc
|
…
|
|
|
lstm_op.h
|
Clean and extract blas
|
7 years ago |
|
lstm_unit_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
lstm_unit_op.cu
|
Fix cpplint errors for a set of operators (#9837)
|
7 years ago |
|
lstm_unit_op.h
|
…
|
|
|
lstmp_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
lstmp_op.cu
|
…
|
|
|
lstmp_op.h
|
Clean and extract blas
|
7 years ago |
|
margin_rank_loss_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
margin_rank_loss_op.cu
|
…
|
|
|
margin_rank_loss_op.h
|
…
|
|
|
matmul_op.cc
|
modification
|
7 years ago |
|
max_sequence_len_op.cc
|
Simplize API Reference Documentation
|
7 years ago |
|
maxout_op.cc
|
modified error info for maxout op
|
7 years ago |
|
maxout_op.cu.cc
|
…
|
|
|
maxout_op.h
|
…
|
|
|
mean_iou_op.cc
|
Add mean IOU op. (#10519)
|
7 years ago |
|
mean_iou_op.cu
|
Add mean IOU op. (#10519)
|
7 years ago |
|
mean_iou_op.h
|
Add mean IOU op. (#10519)
|
7 years ago |
|
mean_op.cc
|
polish doc: mean
|
7 years ago |
|
mean_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
mean_op.h
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
merge_ids_op.cc
|
update comment
|
7 years ago |
|
merge_ids_op.h
|
optimize code and comment
|
7 years ago |
|
merge_lod_tensor_op.cc
|
Make IfElse operator works and fix unit testing. (#11972)
|
7 years ago |
|
minus_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
minus_op.cu
|
…
|
|
|
minus_op.h
|
…
|
|
|
mkldnn_activation_op.h
|
MKL-DNN activations improvements
|
7 years ago |
|
modified_huber_loss_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
modified_huber_loss_op.cu
|
…
|
|
|
modified_huber_loss_op.h
|
…
|
|
|
momentum_op.cc
|
Fix serious bug in nesterov momentum optimizer. (#12231)
|
7 years ago |
|
momentum_op.cu
|
Fix serious bug in nesterov momentum optimizer. (#12231)
|
7 years ago |
|
momentum_op.h
|
Fix serious bug in nesterov momentum optimizer. (#12231)
|
7 years ago |
|
mul_op.cc
|
Merge develop
|
7 years ago |
|
mul_op.cu.cc
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
mul_op.h
|
Process elemwise grad op's lod. mul_op's lod
|
7 years ago |
|
multiplex_op.cc
|
Polish documentation
|
7 years ago |
|
multiplex_op.cu
|
fix errors in multiplex_op
|
7 years ago |
|
multiplex_op.h
|
…
|
|
|
nccl_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
nccl_op.cu.cc
|
Fix CPPLint errors in multiclass_nms, nccl, nce, reduce and save_load_combine (#10032)
|
7 years ago |
|
nccl_op_test.cu.cc
|
Move fluid::framework::InitDevices into fluid::platform (#11757)
|
7 years ago |
|
nce_op.cc
|
Polish the doc of nce layer
|
7 years ago |
|
nce_op.h
|
Fix CPPLint errors in multiclass_nms, nccl, nce, reduce and save_load_combine (#10032)
|
7 years ago |
|
norm_op.cc
|
Make the normalization operator more general and fix bug in l2_normalize. (#11348)
|
7 years ago |
|
norm_op.cu
|
Implement norm_op by CUDA instead of Eigen. (#13273)
|
7 years ago |
|
norm_op.h
|
Implement norm_op by CUDA instead of Eigen. (#13273)
|
7 years ago |
|
one_hot_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
one_hot_op.cu
|
Feature/template (#13093)
|
7 years ago |
|
one_hot_op.h
|
Feature/template (#13093)
|
7 years ago |
|
pad2d_op.cc
|
Add pad2d op. (#12950)
|
7 years ago |
|
pad2d_op.cu
|
Add pad2d op. (#12950)
|
7 years ago |
|
pad_constant_like_op.cc
|
Enhance pad_constant_like_op (#12999)
|
7 years ago |
|
pad_constant_like_op.cu
|
Add pad_constant_like_op (#12943)
|
7 years ago |
|
pad_constant_like_op.h
|
Add pad_constant_like_op (#12943)
|
7 years ago |
|
pad_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
pad_op.cu
|
…
|
|
|
pad_op.h
|
Add pad_constant_like_op (#12943)
|
7 years ago |
|
parallel_do_op.cc
|
graph to program pass
|
7 years ago |
|
pool_cudnn_op.cu.cc
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
pool_mkldnn_op.cc
|
MKLDNN Pooling: inline functions handling ceiled mode
|
7 years ago |
|
pool_op.cc
|
fix errors
|
7 years ago |
|
pool_op.cu.cc
|
…
|
|
|
pool_op.h
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
pool_with_index_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
pool_with_index_op.cu.cc
|
add semicolon to op registry (#10034)
|
7 years ago |
|
pool_with_index_op.h
|
Fix CPPLint errors in operators (#9828)
|
7 years ago |
|
positive_negative_pair_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
positive_negative_pair_op.h
|
hide utils to legacy
|
7 years ago |
|
precision_recall_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
precision_recall_op.h
|
…
|
|
|
prefetch_op.cc
|
Fix rpcclient's wait action in aync env. (#13307)
|
7 years ago |
|
prelu_op.cc
|
add error_info prelu_op
|
7 years ago |
|
prelu_op.h
|
refine prelu
|
7 years ago |
|
print_op.cc
|
Polish PrintOp (#12895)
|
7 years ago |
|
proximal_adagrad_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
proximal_adagrad_op.cu
|
…
|
|
|
proximal_adagrad_op.h
|
…
|
|
|
proximal_gd_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
proximal_gd_op.cu
|
…
|
|
|
proximal_gd_op.h
|
…
|
|
|
random_crop_op.cc
|
Refine random crop
|
7 years ago |
|
random_crop_op.cu
|
Fix GPU compile
|
7 years ago |
|
random_crop_op.h
|
Refine random crop
|
7 years ago |
|
rank_loss_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
rank_loss_op.cu
|
…
|
|
|
rank_loss_op.h
|
…
|
|
|
read_op.cc
|
Revert "Some trivial optimization (#13530)"
|
7 years ago |
|
recurrent_op.cc
|
Revert "Merge pull request #11628 from PaddlePaddle/revert-11102-mozga-intel/Sum_mkldnn_layout"
|
7 years ago |
|
recv_op.cc
|
Fix rpcclient's wait action in aync env. (#13307)
|
7 years ago |
|
reduce_max_op.cc
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_max_op.cu
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_mean_op.cc
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_mean_op.cu
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_mean_op.h
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_min_max_op.h
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_min_op.cc
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_min_op.cu
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_op.h
|
optimize code
|
7 years ago |
|
reduce_op_function.h
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_prod_op.cc
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_prod_op.cu
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_prod_op.h
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_sum_op.cc
|
optimize code
|
7 years ago |
|
reduce_sum_op.cu
|
split reduce op into multiple libraries, accelerate the compiling (#11029)
|
7 years ago |
|
reduce_sum_op.h
|
optimize code
|
7 years ago |
|
reorder_lod_tensor_by_rank_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
reshape_op.cc
|
Refine reshape_grad and transpose_grad (#13074)
|
7 years ago |
|
reverse_op.cc
|
Refine code
|
7 years ago |
|
reverse_op.cu
|
Refine code
|
7 years ago |
|
reverse_op.h
|
fix a bug
|
7 years ago |
|
rmsprop_op.cc
|
Add centered mode rmsprop (#13161)
|
7 years ago |
|
rmsprop_op.cu
|
…
|
|
|
rmsprop_op.h
|
Add centered mode rmsprop (#13161)
|
7 years ago |
|
rnn_memory_helper_op.cc
|
Add error info & remove data sharing between input and output in rnn_memory_helper_op
|
7 years ago |
|
roi_pool_op.cc
|
update by comment
|
7 years ago |
|
roi_pool_op.cu
|
Refine and fix some code for faster-rcnn. (#13135)
|
7 years ago |
|
roi_pool_op.h
|
Refine and fix some code for faster-rcnn. (#13135)
|
7 years ago |
|
row_conv_op.cc
|
Polish code
|
7 years ago |
|
row_conv_op.cu
|
wrap_shfl_x_sync
|
7 years ago |
|
row_conv_op.h
|
…
|
|
|
sampling_id_op.cc
|
fix
|
7 years ago |
|
sampling_id_op.cu
|
merge cpu and gpu
|
7 years ago |
|
sampling_id_op.h
|
refine
|
7 years ago |
|
save_combine_op.cc
|
memory (#13143)
|
7 years ago |
|
save_load_combine_op_test.cc
|
Add FP16 option to load_combine op (#10601)
|
7 years ago |
|
save_load_op_test.cc
|
unittest case fix
|
7 years ago |
|
save_op.cc
|
memory (#13143)
|
7 years ago |
|
scale_op.cc
|
Add distributed unit tests about text_classification/simnet-bow/ctr (#12812)
|
7 years ago |
|
scale_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
scale_op.h
|
modification
|
7 years ago |
|
scatter.cu.h
|
…
|
|
|
scatter.h
|
…
|
|
|
scatter_op.cc
|
Fix scatter_op python API (#12742)
|
7 years ago |
|
scatter_op.cu
|
Fix CPPLint errors in operators (#9826)
|
7 years ago |
|
scatter_op.h
|
Fix scatter_op python API (#12742)
|
7 years ago |
|
scatter_test.cc
|
Fix CPPLint errors in operators (#9826)
|
7 years ago |
|
select_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
send_barrier_op.cc
|
Refine dist rpc deps (#12899)
|
7 years ago |
|
send_op.cc
|
Fix rpcclient's wait action in aync env. (#13307)
|
7 years ago |
|
send_recv_op_test.cc
|
fix compile error
|
7 years ago |
|
send_recv_util.h
|
polish and test
|
7 years ago |
|
sequence_concat_op.cc
|
remove test_seq_concat_op
|
7 years ago |
|
sequence_concat_op.cu.cc
|
refine seq_concat
|
7 years ago |
|
sequence_concat_op.h
|
refine seq_concat
|
7 years ago |
|
sequence_conv_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sequence_conv_op.cu.cc
|
…
|
|
|
sequence_conv_op.h
|
Clean MatMul
|
7 years ago |
|
sequence_enumerate_op.cc
|
doc: refine API and doc
|
7 years ago |
|
sequence_enumerate_op.cu
|
fix: change to enumerate by sentence
|
7 years ago |
|
sequence_enumerate_op.h
|
fix: change to enumerate by sentence
|
7 years ago |
|
sequence_erase_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sequence_erase_op.cu
|
Feature/cuda9 cudnn7 (#10140)
|
7 years ago |
|
sequence_erase_op.h
|
Fix CPPLint errors in operators (#9826)
|
7 years ago |
|
sequence_expand_as_op.cc
|
Add sequence_expand_as_op (#13420)
|
7 years ago |
|
sequence_expand_as_op.cu
|
Add sequence_expand_as_op (#13420)
|
7 years ago |
|
sequence_expand_as_op.h
|
Add sequence_expand_as_op (#13420)
|
7 years ago |
|
sequence_expand_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sequence_expand_op.cu
|
Feature/cuda9 cudnn7 (#10140)
|
7 years ago |
|
sequence_expand_op.h
|
remove broadcast in sequence_expand
|
7 years ago |
|
sequence_mask_op.cc
|
Get sequence length in sequence_pad op & fix sequence_mask op
|
7 years ago |
|
sequence_mask_op.cu
|
Get sequence length in sequence_pad op & fix sequence_mask op
|
7 years ago |
|
sequence_mask_op.h
|
Feature/template (#13093)
|
7 years ago |
|
sequence_pad_op.cc
|
Get sequence length in sequence_pad op & fix sequence_mask op
|
7 years ago |
|
sequence_pad_op.cu
|
Complete sequence_pad_op and its CPU kernel. Add unittests
|
7 years ago |
|
sequence_pad_op.h
|
Get sequence length in sequence_pad op & fix sequence_mask op
|
7 years ago |
|
sequence_pool_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sequence_pool_op.cu
|
…
|
|
|
sequence_pool_op.h
|
Fix CPPLint errors in operators (#9826)
|
7 years ago |
|
sequence_reshape_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sequence_reshape_op.cu
|
…
|
|
|
sequence_reshape_op.h
|
…
|
|
|
sequence_scatter_op.cc
|
[WIP]Sequence Scatter Op (#12625)
|
7 years ago |
|
sequence_scatter_op.h
|
[WIP]Sequence Scatter Op (#12625)
|
7 years ago |
|
sequence_slice_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sequence_slice_op.cu
|
add semicolon to op registry (#10034)
|
7 years ago |
|
sequence_slice_op.h
|
Fix bug in sequence_slice_op
|
7 years ago |
|
sequence_softmax_cudnn_op.cu.cc
|
[Accelerate] Refine seq_softmax_op (#13421)
|
7 years ago |
|
sequence_softmax_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sequence_softmax_op.cu
|
[Accelerate] Refine seq_softmax_op (#13421)
|
7 years ago |
|
sequence_softmax_op.h
|
[Accelerate] Refine seq_softmax_op (#13421)
|
7 years ago |
|
sgd_op.cc
|
add inplace attribute to op_proto_maker (#10665)
|
7 years ago |
|
sgd_op.cu
|
refine sgd_op (#13626)
|
7 years ago |
|
sgd_op.h
|
Optimize selected rows for dist lookup table with pthread rwlock (#12635)
|
7 years ago |
|
shape_op.cc
|
Add flatten op interface and enhance APIs about detection to support variable-length image. (#12422)
|
7 years ago |
|
shape_op.cu
|
Add flatten op interface and enhance APIs about detection to support variable-length image. (#12422)
|
7 years ago |
|
shape_op.h
|
Add flatten op interface and enhance APIs about detection to support variable-length image. (#12422)
|
7 years ago |
|
shrink_rnn_memory_op.cc
|
Revert "Some trivial optimization (#13530)"
|
7 years ago |
|
sigmoid_cross_entropy_with_logits_op.cc
|
update doc for sigmoid_cross_entropy_with_logits
|
7 years ago |
|
sigmoid_cross_entropy_with_logits_op.cu
|
…
|
|
|
sigmoid_cross_entropy_with_logits_op.h
|
…
|
|
|
sign_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
sign_op.cu
|
…
|
|
|
sign_op.h
|
…
|
|
|
slice_op.cc
|
Implement slice grad operator. #8130 (#12330)
|
7 years ago |
|
slice_op.cu
|
Implement slice grad operator. #8130 (#12330)
|
7 years ago |
|
slice_op.h
|
Implement slice grad operator. #8130 (#12330)
|
7 years ago |
|
smooth_l1_loss_op.cc
|
Refine smooth L1 loss. (#10713)
|
7 years ago |
|
smooth_l1_loss_op.cu
|
…
|
|
|
smooth_l1_loss_op.h
|
…
|
|
|
softmax_cudnn_op.cu.cc
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
softmax_mkldnn_op.cc
|
update softmax_mkldnn_op
|
7 years ago |
|
softmax_op.cc
|
Merge pull request #12485 from JiayiFeng/dev_ops_tensor_support
|
7 years ago |
|
softmax_op.cu.cc
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
softmax_op.h
|
code clean
|
7 years ago |
|
softmax_with_cross_entropy_op.cc
|
Add ignore_index in cross_entropy op (#13217)
|
7 years ago |
|
softmax_with_cross_entropy_op.cu
|
code fix (#13365)
|
7 years ago |
|
softmax_with_cross_entropy_op.h
|
Add ignore_index in cross_entropy op (#13217)
|
7 years ago |
|
split_byref_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
split_byref_op.cu.cc
|
update
|
7 years ago |
|
split_byref_op.h
|
VariableResponse support deserialize var into local scope
|
7 years ago |
|
split_ids_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
split_ids_op.h
|
clear mutable rows for the output of split_ids_op
|
7 years ago |
|
split_lod_tensor_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
split_op.cc
|
enable more type for splitOp and ConcatOp
|
7 years ago |
|
split_op.cu.cc
|
enable more type for splitOp and ConcatOp
|
7 years ago |
|
split_op.h
|
wip split byref op
|
7 years ago |
|
split_selected_rows_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
split_selected_rows_op.cu
|
…
|
|
|
split_selected_rows_op.h
|
…
|
|
|
spp_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
spp_op.cu.cc
|
…
|
|
|
spp_op.h
|
Fix CPPLint issues in spp_op, sum_op, topk_op, transpose_op, unpool_op and warpctc_op
|
7 years ago |
|
squared_l2_distance_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
squared_l2_distance_op.cu
|
…
|
|
|
squared_l2_distance_op.h
|
…
|
|
|
squared_l2_norm_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
squared_l2_norm_op.cu
|
…
|
|
|
squared_l2_norm_op.h
|
…
|
|
|
squeeze_op.cc
|
Refine reshape_grad and transpose_grad (#13074)
|
7 years ago |
|
stack_op.cc
|
use for_range
|
7 years ago |
|
stack_op.cu
|
use for_range
|
7 years ago |
|
stack_op.h
|
speed up stack_op
|
7 years ago |
|
strided_memcpy.h
|
Fix CPPLint errors in operators (#9828)
|
7 years ago |
|
strided_memcpy_test.cc
|
…
|
|
|
sum_mkldnn_op.cc
|
bug fix when all inputs are empty
|
7 years ago |
|
sum_op.cc
|
Revert "Merge pull request #11628 from PaddlePaddle/revert-11102-mozga-intel/Sum_mkldnn_layout"
|
7 years ago |
|
sum_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
sum_op.h
|
Merge pull request #13525 from reyoung/fix_mixed_vector
|
7 years ago |
|
tensor_array_read_write_op.cc
|
Fix and enhance beam_search_op and beam_searc_decode_op to be comparable with python beam search
|
7 years ago |
|
tensorrt_engine_op.cc
|
fix comments
|
7 years ago |
|
tensorrt_engine_op.cu.cc
|
1. change tensorrt op from cpu to gpu
|
7 years ago |
|
tensorrt_engine_op.h
|
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into add_ut_for_trt
|
7 years ago |
|
tensorrt_engine_op_test.cc
|
fix ut error
|
7 years ago |
|
test_send_nccl_id.cc
|
move dist codes from operaotrs/detail to operators/distributed
|
7 years ago |
|
top_k_op.cc
|
Revert changes for debug
|
7 years ago |
|
top_k_op.cu
|
Revert ""cherry picked operators changes" (#12184)" (#12747)
|
7 years ago |
|
top_k_op.h
|
reserve vector space before loop in top-k
|
7 years ago |
|
transpose_op.cc
|
Refine reshape_grad and transpose_grad (#13074)
|
7 years ago |
|
transpose_op.cu.cc
|
Refine reshape_grad and transpose_grad (#13074)
|
7 years ago |
|
transpose_op.h
|
Fix CPPLint issues in spp_op, sum_op, topk_op, transpose_op, unpool_op and warpctc_op
|
7 years ago |
|
truncated_gaussian_random_op.cc
|
Add truncated gaussian initializer. (#13000)
|
7 years ago |
|
truncated_gaussian_random_op.cu
|
Add truncated gaussian initializer. (#13000)
|
7 years ago |
|
uniform_random_batch_size_like_op.cc
|
"fix based comments"
|
7 years ago |
|
uniform_random_op.cc
|
code optimize
|
7 years ago |
|
uniform_random_op.cu
|
code optimize
|
7 years ago |
|
unpool_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
unpool_op.cu.cc
|
…
|
|
|
unpool_op.h
|
Fix CPPLint issues in spp_op, sum_op, topk_op, transpose_op, unpool_op and warpctc_op
|
7 years ago |
|
unsqueeze_op.cc
|
Refine reshape_grad and transpose_grad (#13074)
|
7 years ago |
|
unstack_op.cc
|
add unstack_op
|
7 years ago |
|
unstack_op.h
|
add unstack_op
|
7 years ago |
|
warpctc_op.cc
|
Clean OpProtoAndCheckerMaker
|
7 years ago |
|
warpctc_op.cu.cc
|
…
|
|
|
warpctc_op.h
|
Complete sequence_padding GPU kernel
|
7 years ago |
|
while_op.cc
|
Fix the nested dyn_rnn (#13417)
|
7 years ago |
tensor-tang