1177 Commits (c10023006b1001d7ff18c817d4c0b181e7313eda)

Author	SHA1	Message	Date
QI JUN	6b47598103	add data layout (#6832) ... * add data layout * fix ci	7 years ago
QI JUN	ad6d6e9cba	add library type (#6874)	7 years ago
武毅	8d6db25167	Merge pull request #6297 from typhoonzero/simple_dist_train_api ... [Done] API for dist train	7 years ago
Yu Yang	eb612a8261	Merge pull request #6181 from emailweixu/enforce_drop_empty_ig ... Enforce drop_empty_grad=false When the input of an op is duplicable.	7 years ago
dzhwinter	abde3130b7	"remove GPU Sync Interface" (#6793) ... * "remove GPU Sync Interface" * "fix typo" * "fix type cast error" * "fix related Copy with stream" * "fix failed tests with DevicePool" * "fix stupid removed position error"	7 years ago
xuwei06	0bfa1f7c7a	Enforce drop_empty_grad=false When the input of an op is duplicable. ... For input argument with a list of variables, drop_empty_grad is not allowed because it makes the correspondence bewteen a variable and its gradient ambiguous. Use REGISTER_OP_EX to register the op or call InputGrad(?,false) in GradOpDescMaker.	7 years ago
typhoonzero	b848416166	follow comments	7 years ago
typhoonzero	175a4f52c2	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into simple_dist_train_api	7 years ago
typhoonzero	4658f9501e	fix delete ops	7 years ago
fengjiayi	f3cc75d8fb	Fix errors	7 years ago
fengjiayi	0895d1d3cb	Fix a error	7 years ago
fengjiayi	feb05c3a54	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into rename_fill_zero_op_output	7 years ago
fengjiayi	f56f149298	fix_output_name	7 years ago
Yu Yang	091897321f	Rename XXDescBind --> XXDesc (#6797) ... * Rename XXDescBind --> XXDesc * Fix Compile	7 years ago
Yu Yang	0f1c685c60	Add more log of lod_rank_table (#6761)	7 years ago
Yu Yang	59403c3bc3	Add comment of GetSubLoDAndAbsoluteOffset (#6771)	7 years ago
Yu Yang	e445b3ff20	Move framework.proto to proto namespace (#6718) ... * Move framework.proto to proto namespace * Fix compile * Fix compile * Fix Compile	7 years ago
dzhwinter	24fda39220	Feature/global context (#6537) ... * "add DeviceContextPool" * "add devicecontextpool in pybind" * "add comments in python side " * "fix static link error" * "fix CI error" * "add executor.py" * "fix CI error" * "add with gpu macro" * "remove comment out codes" * "add TODO items" * "update init devices"	7 years ago
fengjiayi	dafd449c68	Unify `step_block` and `block` to `sub_block`	7 years ago
typhoonzero	2b47fb3d25	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into simple_dist_train_api	7 years ago
typhoonzero	9508c72685	wip: should fix variable recreate	7 years ago
QI JUN	61ec0b9516	Refine device context (#6433) ... There are mainly following fixes: - take `DeviceContext` as the template parameter of math functors and OpKernel instead of `Place` - remove `eigen_device` interface in base class `DeviceContext` - remove `GetEigenDevice` interface in `ExecutionContext` and base class `DeviceContext` - remove unused `platform::EigenDeviceConverter` - rename `REGISTER_OP_GPU_KERNEL` to `REGISTER_OP_CUDA_KERNEL` - rename `USE_GPU_ONLY_OP` to `USE_CUDA_ONLY_OP`	7 years ago
Yiqun Liu	9592468609	Fix gcc4.9 (#6442) ... * Fix compiling error of gcc4.9. * Refine the check of cxx compiler flags in api/CMakeLists.txt.	7 years ago
Qiao Longfei	36fcc95cab	Nmt decoder train (#6367) ... * init decoder_trainer * can run * fix lod * add sharelod to cross_entropy_grad_op * add avg_cost to fetch list * modify learning rate * can run * optimie code * add early exit * fix print * revert test_understand_sentiment_conv.py * add act to fc	7 years ago
Yu Yang	229c2e7883	Feature/while op sentiment analysis (#6282) ... * Add DataFeeder A v2 API like data feeder for book demos. We can feed data directly from reader. * Fix CI * Add an unittest for while/rnn op forward * Add unittest for raw while op backward * Fix CI * Complete Dynamic RNN	7 years ago
dangqingqing	4e451a34db	Remove the cuda stream synchronization between each operator.	7 years ago
Yu Yang	d5e3279451	While op forward for sentimental analysis (#6140) ... * Add DataFeeder A v2 API like data feeder for book demos. We can feed data directly from reader. * Fix CI * Add an unittest for while/rnn op forward * Add unittest for raw while op backward * Fix CI	7 years ago
fengjiayi	a38c151243	Add GetInputsElementDim (#6091)	7 years ago
Yu Yang	35453df18f	Fix ShareLoD bug (#6084) ... Fix #6087	7 years ago
武毅	0a8a86e0c9	Send recv op (#5520) ... * WIP send recv op * WIP send recv * put grpc impl in details * put grpc impl in details * update wip * update proto * update proto * update proto * clean cmake * wip on op implementations * wip on op implementations * compile ok adding ut * wip unitest * add extern cares for linking * wip add ut * working version send recv * revert optimizer.py * update test cmake * add libtool to dockerfile * update cmake dependency * update cmake depends * update cmake grpc depends * fix cmake dependency * fix compile error * fix compile * follow comments * update * update copyfrom	7 years ago
QI JUN	b28b2f172b	refine test_recognize_digits_mlp and format codes (#5937)	7 years ago
dzhwinter	45062fe5d7	Feature/copytensor (#5455) ... * "make global tensor function independently" * "replace functor" * "fix inline template error" * "fix tensor array with CopyFrom" * "fix other case use CopyFrom" * "move the op interface hardly" * "fix operators" * "fix typo" * "delete dynamic recurrent rnn and fix gru_unit in debugmode" * "fix unique_ptr copy" * "fix cuda copy" * "fix namespace error" * "removed nccl python test" * "fix include error" * "fix typo" * fix copy util test	7 years ago
QI JUN	3a76062c84	support testing when training and handle dropout and batch_norm operator in testing mode (#5734) ... * is_training to is_test in dropout op * handle dropout and batch_norm operator when prune pdesc in testing mode * handle dropout and batch_norm operator when prune pdesc in testing mode * add get_inference_program method * fix dropout op * fix ci * test data after each batch training * refine code * refine test_book3 * fix ci * follow comments	7 years ago
fengjiayi	50d670ee06	Unify dtype and datatype (#5869) ... * Change all `data_type` in Python to `dtype` * Change `date_type` in C++ to `dtype` * Refine	7 years ago
Qiao Longfei	569f7c4773	enforce shape of backward target to be {1} (#5745) ... * enforce shape of backward target to be {1} * fix test_regularizer.py * rm unused code * fix backward_test * fix a type bug * fix test_program	7 years ago
Abhinav Arora	6cfcf6245a	Adding logical operators for beam search and control flow (#5708)	7 years ago
Yang Yang(Tony)	18f0c40a97	feature/while_grad_op (#5554) ... * first commit * Python API for while op * Python Unittest for simple while_op forward * fix out to be list * Fix UT * VarType * Fix several bugs * Fix bug * Fix bug * Fix Bug * Fix bug * Fix unittest * Remove debug log * Add comments * add PADDLE_ENFORCE * while_grad_op first commit * Add `BlockDescBind::FindRecursiveOrCreateVar()` and fix bugs * not sure how to setdim of while outputs * push for test * add executor vlog * fix bug of while_op cond * Several enhancement for code 1. Backward always infer shape & infer var type. Since there are RENAME variables will be created when creating backward operator, but their shape & var types are not inferenced. 2. Never use SomePtr-> directly, since every pointer could be nullptr if it is a function return value. Add `detail::Ref` to cast pointer to reference safely. 3. Enhance error message for backward. 4. Infer data type of variable in `sum` and `tensor_write` * Fix bugs of while_op gradient * Fix several bugs of while_op grad * fix fill zeros like * fix 3 >= 3 * fix place holder shouldn't be null * fail on sum op * Fix SumOp of TensorList * clean up * pass while test * fix test_array_write_read * pass sum op * Support int/int64 for fill_constant_batch_size_like * Fix compile	7 years ago
QI JUN	5f9f990e62	fix gitignore (#5657) ... * fix gitignore * refine cmake file	7 years ago
Yu Yang	488320a703	Conditional Block Forward (#5530) ... * Conditional Block Forward * Assign Operator. Out=X, when type in [LoDTensor/SelectedRows/LoDTensorArray] * Stash * Add Scope::Rename it is useful in gradient phase of an operator with block * ConditionalBlock Grad Done * Add comments * yapf format code	7 years ago
Yu Yang	7c1755d93f	Assign Operator. (#5531) ... * Assign Operator. Out=X, when type in [LoDTensor/SelectedRows/LoDTensorArray] * Follow comments	7 years ago
Yu Yang	edb22c2f0c	Add Scope::Rename (#5534) ... it is useful in gradient phase of an operator with block	7 years ago
Yang Yang(Tony)	40367d18d4	feature/while_op (#5502) ... * first commit * Python API for while op * Python Unittest for simple while_op forward * fix out to be list * Fix UT * VarType * Fix several bugs * Fix bug * Fix bug * Fix Bug * Fix bug * Fix unittest * Remove debug log * Add comments * add PADDLE_ENFORCE * while_grad_op first commit * Add `BlockDescBind::FindRecursiveOrCreateVar()` and fix bugs * refine code * fix unittest bug	7 years ago
Yu Yang	ef84ff8657	Merge branch 'develop' into feature/increase_cpu	7 years ago
Yang Yu	c9fc7ba9f8	Do not sum output if that output is not a gradient ... * increament is default inplace	7 years ago
Yang Yu	2ac5772970	Merge branch 'develop' of github.com:baidu/Paddle into feature/fill_constant_force_cpu	7 years ago
Yang Yu	0ede2a7311	Fix CI Compile	7 years ago
Yu Yang	91b724821c	Merge pull request #5443 from reyoung/feature/InferKernelKey ... Polish OpWithKernel	7 years ago
Yu Yang	f72729d407	Feature/rnn to array to lod tensor (#5411) ... * Add LoDRankTable LoD Rank Table stores the `level` of `lod` which is ordered by sequence length in descending order. It is useful when implement dynamic RNN and is shared by dynamic RNN memory, dynamic RNN slice input and dynamic RNN slice output operators. * Add skeleton for array_to_lod_tensor and lod_tensor_to_array * Add VarType::LoDTensorArray * Add PyBind of LoDTensorArray * Add InferVarType * Add first unittest * Add ut * Add unittest * Add unittest * Add unittests * update * init * add infershape for lod_tensor_to_array_op * compelete array_to_lod_tensor_op * copy data * clean code * clean code * Fix unittest data * fix bugs * fix compile error * Refine TensorToArrayOp * refactor array_to_lod_tensor * Unittest * fix bugs * Fix unittest * Fix unittest * debug * Debug * Fix unittest * clean code * refactor * use ostream * update test * fix gpu build error * make gpu test pass	7 years ago
Yu Yang	5ee62383bd	Rewrite fill_constant op	7 years ago
Yu Yang	bbdac7f7d8	Polish OpWithKernel ... * Chage `IndicateDataType` to `GetKernelType`. Make it easier to understand. * Change `OpKernelKey` to `OpKernelType` * Make operator developers can customize which kernel the operator will use in runtime.	7 years ago
Yu Yang	f74fb79036	Compare Operator (#5325) ... * Compare Operator * Follow comments	7 years ago
qingqing01	58db07b7bb	Check errors for the cuda kernel calls. (#5436)	7 years ago
武毅	3014645d35	Merge pull request #5403 from typhoonzero/refine_get_cuda_context ... Refine get cuda context	7 years ago
Yu Yang	6cde889b5e	Add unittest, backward of array read/write op (#5409) ... * Use stable_sort in lod_rank_table It is easy to debug and test when use `stable_sort`and the time complexity is not changed. * Add LoDTensorArray * Stash * Better debug message for IsInitialized * Stash * Better debug message for IsInitialized * Complete array read/write op unittests * Add unittest, Gradient of array read/write * Follow comments	7 years ago
Yang Yang(Tony)	701545979b	Update lod_tensor.md (#5383) ... An important change on lod tensor indexing. A higher level offset will be based on its next level rather than an absolute offset.	7 years ago
Yu Yang	c9b57dcc83	ReadFromArray/WriteToArray op (#5407) ... * Use stable_sort in lod_rank_table It is easy to debug and test when use `stable_sort`and the time complexity is not changed. * Add LoDTensorArray * Stash * Better debug message for IsInitialized * Stash * Better debug message for IsInitialized * Complete array read/write op unittests	7 years ago
typhoonzero	272f3e6d43	refine get cuda context	7 years ago
fengjiayi	d05c182e93	Add LoD's slice and append function (#5368) ... * Add GetFineGrainedLoDLength and AppendLoD * Follow comments and fix bugs * fix a compile error * fix a compile bug	7 years ago
Yu Yang	2be4c3cb62	Feature/lod tensor array (#5379) ... * Use stable_sort in lod_rank_table It is easy to debug and test when use `stable_sort`and the time complexity is not changed. * Add LoDTensorArray	7 years ago
Yu Yang	ea2fc4cc51	Use stable_sort in lod_rank_table (#5378) ... It is easy to debug and test when use `stable_sort`and the time complexity is not changed.	7 years ago
Qiao Longfei	906e2565a7	Add acc test to image classification (#5336) ... * add acc layer * memory log level change from 3 to 10 * use gaussian random to init conv parameters * use initializer * fix import * batch_norm use helper to create persistable var * refine code * train only 2 batches for test * use g_program and g_init_program * use XavierInitializer to init fc parameter	7 years ago
Yu Yang	74849158e3	Add LoDRankTable (#5349) ... * Add LoDRankTable LoD Rank Table stores the `level` of `lod` which is ordered by sequence length in descending order. It is useful when implement dynamic RNN and is shared by dynamic RNN memory, dynamic RNN slice input and dynamic RNN slice output operators. * Add InferVarType	7 years ago
Tao Luo	a8dee3b315	Merge pull request #5251 from qingqing01/lstm_fix ... Refine activation function pointer for LSTM operator.	7 years ago
Yi Wang	1ed5ae7a14	Fix comparing between signed and unsigned values (#5328)	7 years ago
Yu Yang	0a32e74d13	Rewrite StaticRNN with Executor (#5224) ... * Init commit * Make executor use ProgramDescBind * Change Attribute from BlockDesc to BlockDescBind * Since we will get the program desc in RNN, just BlockDesc is not enough. * Add DeviceContext to Executor API * Rewrite RNN * Pass Python * AddBiasOp does not care num_flatten_dims * Stash * Fix MacOS Compile * Pass RNN forward * add python test * refactor test * Make compile pass * add gradopmaker * First draft done * Polish code * add grad op maker and grad infershape * Polish code * Fix backward.cc bug * Fix infershape * Rename function * add backward test * simplify recurrent test * Update * Pass unittest * Add comments & refine test * Add comments * refactor test * Complete Unittest * fix StepScopes enforce * Remove unused unittest * no type error * Update * Make RNN Pass unittest	7 years ago
dangqingqing	03bfd7611a	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_fix	7 years ago
Cao Ying	8401039feb	Merge pull request #5084 from lcy-seso/crf ... Add the LinearChainCrf operator.	7 years ago
qingqing01	36d20609a2	Merge pull request #5115 from qingqing01/lstm_bp ... Add backward implementation for LSTM operator.	7 years ago
Yu Yang	0318f47e1c	Enhance in backward (#5262) ... Set gradient's data type based on its forward variable	7 years ago
Yu Yang	1363ddb6d7	Feature/executor use program bind (#5196) ... * Init commit * Make executor use ProgramDescBind * Change Attribute from BlockDesc to BlockDescBind * Since we will get the program desc in RNN, just BlockDesc is not enough.	7 years ago
Qiao Longfei	ee11f00642	add shareLod (#5259) ... * add shareLod * fix sequence_conv grad infershape	7 years ago
caoying03	2ac9a3d8dc	follow comments.	7 years ago
dangqingqing	1c8a0c4bd4	Refine activation function pointer for LSTM operator.	7 years ago
caoying03	dd2be3daba	Merge branch 'develop' into crf	7 years ago
caoying03	86fd6b6373	add gpu kernel by copying inputs/outputs between cpu and gpu.	7 years ago
dangqingqing	1d7c03e789	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp	7 years ago
Qiao Longfei	0049ce0479	03 image classification (#5192) ... * add batch_norm_layer * add img_conv_group layer and test * add check to Tensor.type() * forward can run * with backward * change label data time from int32 to int64 * refine code * follow comment	7 years ago
dzhwinter	833d0ad0fe	Merge pull request #4838 from dzhwinter/feature/multigpu ... Feature/multigpu	7 years ago
dzhwinter	71305e5f90	"polish code based on comment"	7 years ago
Yu Yang	b84e822651	Cast Operator (#5149) ... * Cast Operator Cast input variable to other data type * Fix compile error * Add cast op * Follow comments	7 years ago
Yu Yang	46a13e37d7	Polish Accuracy Op (#5191) ... * Accuracy does not support float/double, only support integers * Polish error message when an operator does not support some device.	7 years ago
Yu Yang	8f6c0a0fad	Extract InferShape to many cc files (#5174) ... * Shrink Operator.h * Fix CI compile	7 years ago
fengjiayi	6783dcee9e	Python API for inference model saving/load (#5020) ... * Add `dump_to_file()` for ProgrameDescBind in pybind * Update * Add utility.py * typo * Fix bugs * Move add_feed/fetch_components to untility.py * Compelete dump * Follow comments * Change output of Prune() from inference to pointer * Expose Prune() to Python * Compelete save/load API of inference model * Fix errors * Debuging * Compelete unit tests * follow comments	7 years ago
Yu Yang	2a5edec03e	Add debug logs in scope, meta_cache and memory (#5170) ... * Add debug logs in scope, meta_cache and memory * Add missing deps	7 years ago
chengduo	5a95792925	Merge pull request #4636 from chengduoZH/Add_pool_cudnn_op ... Add pool2d cudnn op	7 years ago
chengduoZH	cadee843b8	follow comments	7 years ago
caoying03	bce4f7d6eb	follow comments.	7 years ago
Yu Yang	b44f4ccbeb	Make InferShape as a field in OpInfo (#5139) ... * Op developer can add `InferShape` to any operator	7 years ago
QI JUN	7f8574c0f5	add sparse support for sum op (#5093) ... * add sparse support for sum op * typo fix * fix gpu build error * fix unittest error * typo fix * infer var type and shape in op_test * follow comments * fix build error * bypass some unittests depend on NetOp	7 years ago
Yu Yang	be00b0c4d6	Gradient check use graph (#5027) ... * Simplize Gradient Check * Stash * Extract apply_backward_pass to backward.py Rename apply_backward_pass to append_backward_ops * Use graph API to check gradient * Fix ci * Fix CI * Fix backward for double precision * Stash * Fix CI * Fix ci * Ignore GRU test * Ignore xe op * Fix CI * Fix softmax with xe gradient The correct equation should be IG = OG * (d_softmax_with_xe()) * Fix typo * Fix merge error * Disable LRN	7 years ago
Dong Zhihong	6cce5268ed	"fixed based on comment"	7 years ago
fengjiayi	aa379ccb5e	Add functions of restoring ProgramDescBind from ProgramDesc (#5109) ... * compelete restoring program_bind from program_desc * Fix bugs * fix compile errors * fix errors and add unit tests * rename some vars * Follow comments	7 years ago
Dong Zhihong	16a39d24f3	fix conflict	7 years ago
dangqingqing	ac3370a467	Add unit testing for gemv and fix the gradien check for bais.	7 years ago
Yu Yang	efc2464f6c	Feature/save op (#5090) ... * Init * Stash * Polish SaveLoadOp * Fix CI * Polish code * Save GPU Tensor * Stash * Fix CI	7 years ago
Yan Chunwei	39a6f43bcc	fix bug (#5106)	7 years ago
caoying03	4c63086921	Merge branch 'develop' into crf	7 years ago
Yan Chunwei	9a0233de9f	Feature/tensor array lod pack (#5007)	7 years ago
Dong Zhihong	bab5e53c03	Merge remote-tracking branch 'origin/develop' into feature/multigpu	7 years ago
Yu Yang	6c0b383672	Add VarType::STEP_SCOPES for RNN (#5056)	7 years ago