6cce5268ed
"fixed based on comment"
8 years ago
16a39d24f3
fix conflict
8 years ago
6bc261b933
fix ci
8 years ago
25df892932
remove conflict
8 years ago
66476fc7b7
Add proximal adagrad optimizer ( #5128 )
8 years ago
65dbbd57af
Add and pass unittests.
8 years ago
db1bb8224a
follow comments
8 years ago
8e3ecf5d11
Merge pull request #4814 from chengduoZH/Add_sequence_project_op
...
Add sequence_conv_op and sequence_projection functor
8 years ago
9d142d5060
Local response normalize. ( #4426 )
...
Add local response normalize
8 years ago
bd680f157f
fix compiling warning.
8 years ago
99c6f44a5a
follow comments
8 years ago
746f2a2e36
only compute the first max value in backward
8 years ago
bcc0dad7df
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
8 years ago
ac3370a467
Add unit testing for gemv and fix the gradien check for bais.
8 years ago
dcb3da591e
refine code
8 years ago
56bbfd1af2
Add deconv3d op
8 years ago
aa3de3571d
Polish unit test for xe, generate probablities ( #5096 )
...
* Cross Entropy Wrong
* Fix XE
* Polish gradient check for xe
* Fix compile
8 years ago
06c7c8c80e
Add CPU kernel.
8 years ago
1bb0e2943b
Add pool2d cudnn
8 years ago
eafbbc11a0
write conv2d and conv3d together
8 years ago
2e0298744e
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
8 years ago
cd38286684
Add gradient check unit testing and fix bug.
8 years ago
efc2464f6c
Feature/save op ( #5090 )
...
* Init
* Stash
* Polish SaveLoadOp
* Fix CI
* Polish code
* Save GPU Tensor
* Stash
* Fix CI
8 years ago
4b9cf0e8b1
"add disable"
8 years ago
9b0f092853
change input data type
8 years ago
6cc2ce010a
add fill constant batch size like op ( #5057 )
8 years ago
626ff3b79e
"polish cmake file"
8 years ago
368737e41b
Merge branch 'develop' into maxseq
8 years ago
b0a267c0b8
Adding the squared L2 norm operator for L2 regularization ( #5030 )
...
* Adding the L2 loss operator for L2 regularization
* Renaming l2_loss op to squared_l2_norm_op
* Addressing code review feedback
8 years ago
4c63086921
Merge branch 'develop' into crf
8 years ago
2573ac1448
"remove python side test case to another PR."
8 years ago
4e165f4ea3
"fix create output variable bug"
8 years ago
56b723c40d
Cudnn batch norm op ( #5067 )
...
* init cudnn batch norm op
* rename batch_norm_cudnn_op.cc batch_norm_op.cu
* correct name style
* add ExtractNCWHD, simplify code
* fix ExtractNCWHD
* use CUDNN_ENFORCE instead of PADDLE_ENFORCE
8 years ago
61c1b0469a
"fix multigpu testcase"
8 years ago
38d3adfeb6
"add multioperator testcase"
8 years ago
c22f7fcd17
add positive_negative_pair_op evaluator
8 years ago
94992a990b
"add multiop testcase"
8 years ago
bab5e53c03
Merge remote-tracking branch 'origin/develop' into feature/multigpu
8 years ago
11cf3e3a43
"refactorization of nccl test case"
8 years ago
6d1493a460
"add bcast c++ test case"
8 years ago
5200c657a7
"move Tensor to LoDTensor"
8 years ago
c74107bfdc
fix backward computation.
8 years ago
d2bd7357d2
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
8 years ago
3d8b6ebcf8
Add LSTM backward implenmentation.
8 years ago
9f7c9875a9
fix doc
8 years ago
3f8a7b55b8
remove conflict
8 years ago
ee998a9c44
CPU Batch Norm Op ( #4964 )
...
* init batch norm op
* prepare input output
* compute mean_out var_out save_mean save_var on CPU
* active is test
* use eigen to do computation
* complete batch norm forward
* set default momentum to 0.9
* add batch norm grad op in CPU
* add tensor_format and NHWC support, add python test
* add test training
* add batch norm gradient test
* improve comment, fix foward Python UnitTest
* add gradient test
* fix eigen warning
* follow name style
* fix a bug
* change float to T
* add simple forward test
* test with different place
* add backward test
* refine python test
* remove old python test code
* code clean
* follow code style
* update comment
8 years ago
63fb41b399
"redefine the initop from kernel to OpBase"
8 years ago
1e8474b9f1
"delete python ops testcase"
8 years ago
fd2eb55071
"Serialize LoDTensor, Save/Restore model" ( #4602 )
...
* "add model format design doc"
* "add restore function"
* "add parse protobuf"
* "move necessary information to saver.proto"
* "format code"
* "add gpu option"
* "add lod info"
* "add saveop python test wrapper"
* "checkpoint reuse save operator"
* "rewrite model format design doc"
* "async support needed"
* "fix run once"
* "fix doc based on comments"
* "refine based on comments"
* "fix based comments"
* "remove persistable flag from framework.proto"
* "add IndicateDataType to restore op"
* "add save test"
* "modify save restore code"
* "modified the restore logic"
* rm checkpoint_op.cc
* rm test_checkpoint_op.py
* "get inputs outputs name from execution context"
* Saving each variable to a independent file
* Fix bugs
* Rewrite save_restore_op_test with new Python framework
* Move `SaveOp` and `RestoreOp` from OpWithKernel to OpBase
* Refine unit test of SaveOp and RestoreOp
* fix compile errorwq
8 years ago
0990c87bf6
checkin nccl operator
8 years ago
da1181bfc6
Merge remote-tracking branch 'origin/develop' into feature/multigpu
8 years ago
ef257e6d96
write nccl c++ test case
8 years ago
05239b6ff5
fix functor
8 years ago
5939a17c47
Follow comments and adapt to new interface.
8 years ago
6a630f2798
Merge branch 'develop' into crf
8 years ago
0190d5d644
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into check_out_lod
8 years ago
2e783663fa
Enable to output LoD in fetch_op and check output LoD in the op unit test.
8 years ago
b15c69f59d
remove conflict
8 years ago
427644b2fa
fix the computation kernels.
8 years ago
4c6bccbe20
fix doc and remove useless code
8 years ago
0521161070
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into fix-3923-r
8 years ago
9a3f50d806
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into fix-3923-r
8 years ago
97f1b98759
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_expand_op
8 years ago
296167446c
Rewrite sequence expand op
8 years ago
61dbf4b926
fix im2col interface
8 years ago
35434c33d0
fix compiler error in im2col.h
8 years ago
154dbb4697
Add unit test
8 years ago
23bf6b2c15
Merge pull request #4887 from chengduoZH/fix_im2col_kocf_for_sequence
...
Add up, down, left and right padding for im2col.
8 years ago
8fdc315acc
Merge pull request #4739 from zchen0211/develop
...
deconv op implementing ...
8 years ago
50f04dcae3
"add init allreduce test"
8 years ago
f6106ffa7e
merge develop into multigpu
8 years ago
ec47565c23
"add reduce hash function"
8 years ago
db157eda45
New Op Test framework. ( #4962 )
...
Pass all forward op test
8 years ago
423d7438a1
"add register gpu macro"
8 years ago
bc15117403
Correct mul_op implementation ( #4988 )
...
* Correct mul_op implementation
* Restore the origin shape after mul
* Fix mul op
* Do not touch math_function
8 years ago
40e7caf667
ensure ids in lookup table op must be a column vector ( #4987 )
...
* ensure ids in lookup table op must be a column vector
* follow comments
8 years ago
09662da0be
follow comments
8 years ago
811a3174f1
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into fix_im2col_kocf_for_sequence
8 years ago
2947f5678e
follow comments
8 years ago
4b84f07c70
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-ChunkEvalOp
8 years ago
bb9d68dcb3
Add chunk_eval_op
8 years ago
f2ccef26bf
Add sequence_conv_op
8 years ago
0ab2c436ae
Add sequence_project_functor
8 years ago
cf2608e383
update to the develop branch.
8 years ago
64fe9bcc5c
Update lstm comments and fix bug.
8 years ago
611ee68b78
add bilinear tensor product op
8 years ago
4e8fccff11
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_expand_op
8 years ago
d697b6a349
Modified code using LoDTensor
8 years ago
ce960575cd
fix doc format and unit test
8 years ago
c8d0d37c4d
Merge branch 'develop' into crf
8 years ago
80a5ee0052
fix forward and add backward.
8 years ago
4c19f9f429
fix backward
8 years ago
6246be294f
clean gradient data
8 years ago
54ffafa123
use context to get attribute ( #4997 )
8 years ago
4d112b7d04
remove conflict
8 years ago
dc7d07358c
add padding up, down, left, right
8 years ago
834b82f109
fix sequence_project_op forward and backward
8 years ago
d0cfbba429
Merge pull request #4496 from guoshengCS/add-NormOp
...
Add norm_op
8 years ago
b3ab3ce0a1
deconv -> conv transpose
8 years ago
64c5ecbedb
deconv
8 years ago
65906ef1d0
Several Enhancement
8 years ago
502e72591f
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
7edc1d96c6
fix clang build error
8 years ago
07ea9adec0
feature/dynamic recurrent op forward and backward ( #4799 )
8 years ago
5380a5471b
Adding Nesterov Momentum ( #4948 )
8 years ago
b12988f008
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-NormOp
8 years ago
00ad7512cf
Use stream while memory::Copy in GPU mode
8 years ago
09c0c82ec9
Adding increment op ( #4940 )
...
* Adding incremnt op
* Fixing comment about step attribute
8 years ago
8e55736a20
deconv2d
8 years ago
333045d7b2
"move nccl to another directory"
8 years ago
9903e49f94
add test_fit_a_line ( #4936 )
...
* add test_fit_a_line
* Update
* fix persistable bug
* fix elementwise add bug
* set correct attr for bias op in fc layer
* set correct attr for bias op in fc layer
* Update
1. Add init_program to hold initializers
2. bug fix
* add test_fit_a_line
* fix persistable bug
* fix elementwise add bug
* fix type
* add gitignore
* Complete fit_a_line test
* revert code
* Clean up
* Revert "revert code"
This reverts commit eb1aa015cda4fc12b6dc778ada6c3507b98134f5.
* Refine
* Fix unit test
8 years ago
92369177ad
Merge pull request #4941 from QiJune/fix_elementwis_add_bug
...
fix elementwise add bug
8 years ago
7eeaae1695
deconv
8 years ago
c532b96741
Remove template parameter for Tensor methods ( #4937 )
...
* Remove template parameter for Tensor methods
* Also check the type is correct when data()
* Simplize holder_
* Fix accuracy_op
* Register Code
8 years ago
9e6404441c
fix elementwise add bug
8 years ago
d97a732f4f
deconv
8 years ago
e59ca752fc
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
77cac5cdb8
Removing updates of Beta1 power accumulators outside the op ( #4931 )
8 years ago
11bebeb2dc
Removing updates of Beta1 and Beta2 power accumulators outside the op ( #4925 )
8 years ago
3db5278301
Feature/py executor test ( #4922 )
...
* Implement FC layer with helper
* Update LayerHelper
* Add debug string for Python ProtoBuf
and Rename `Sync` to `Flush`
* Add check of ProtoBuf initialization
* Layer wrapper for FC
* Fix unittest
* Fix CI
* Add code generator
* AttributeChecker Better error log and speicalize bool
Since lots of types can be cast to bool
* Complete mlp, fit_a_line
* Expose get global scope
* Make global scope not thread-safe
1. It is no need to make global scope thread-safe, since it will be
invoked in Python main thread.
2. Do not free the global scope when C++ exit. Let the OS free memories,
otherwise, we need to handle the destroy dependencies.
See
https://google.github.io/styleguide/cppguide.html#Static_and_Global_Variables
* Fix
* Implementation of simple conv_2d layer
* Stash
* Remove private data members in OpRegister
* Fix bugs
* Stash
* Expose FeedFetchList as VarType
* Change ProgramDesc not a global variable
* Polish code style
* Stash
* Correct implement BlockDesc destructor
* Correct implement BlockDesc destructor
* Unify program as parameter name
* Fix bugs
* Add unittest
* Fix unit test error
* Remove unused functions
* Add clone for Python Program
* Working on executor
* Stash
* Add glog as dependencies of ops
* Use VLOG to logging some information is helpful when we debug Paddle
* Expose VarDesc::persistable to Python
* Test executor
* Complete unittest
* Polish code
* Fix merge error
* Follow comment
* Polish Python Code
8 years ago
17e33738f2
Enhance unit testing and fix bug.
8 years ago
8bec26be03
Add missing file of math/detail/CMakeLists.txt
8 years ago
74b283c9d6
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_expand_op
8 years ago
a94b3dd9a7
Refine comments and function name
...
1. Add more comments and exmples
2. Rename repeat_lod to expand_lod
3. Remove unused head file
8 years ago
a461bf139d
Add missing file.
8 years ago
e8cd4b7d86
deconv2d impl in full
8 years ago
b325213150
update to develop branch.
8 years ago
d1fbf50b9e
Add unit testing for forwad implementation.
8 years ago
43aad989bd
deconv
8 years ago
fdfc8f9baa
"switch to Init op"
8 years ago
7ad13fbfe9
Merge pull request #4876 from QiJune/sgd_op_sparse_kernel
...
add sparse update kernel for sgd operator
8 years ago
5ec55e7995
deconv impl
8 years ago
80ebc8d5e3
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
c10b8e808f
Adding Proximal Gradient Descent ( #4848 )
...
* Adding Proximal Gradient Descent
* Fixing review comments
8 years ago
af215a1a53
Design doc: Batch Normalization Operator ( #3748 )
...
* Add design doc of batch_norm_op
* Move batch_norm_op.png to operator/images
* Refine batch_norm_op design doc
8 years ago
e747623e86
Change ProgramDesc not a global variable ( #4879 )
...
* Change ProgramDesc not a global variable
* Polish code style
* Correct implement BlockDesc destructor
* Unify program as parameter name
8 years ago
31531ab581
Add backward kernel
8 years ago
91db457fc0
follow comments
8 years ago
8de04be786
Fix unitest
8 years ago
2a8dbd130d
LSTM Operator forward implementation.
8 years ago
40688d223e
refine im2col (up_pad,down_pad)
8 years ago
d2c1408f91
fix im2col kocf for sequence projection
8 years ago
1e60c9b2e8
Add sequence_project_op (use im2col)
8 years ago
555ab3f5af
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_expand_op
8 years ago
23701ffaf0
Refine op
8 years ago
521514da36
fix conflict ( #4883 )
8 years ago
164898277c
MatMul operator ( #4856 )
...
* initial matmul operator
Similar to np.matmul, but also has transpose_X and transpose_Y flags,
and only supports tensors from rank 1 to 3 inclusive.
For GPU, uses cublas?gemmStridedBatched. For CPU, uses
cblas_?gemm_batch if available via MKL; otherwise a simple serial
implementation that loops over the batch dimension is employed for now.
8 years ago
fd96914d23
Merge pull request #4571 from sidgoyal78/momentum_op
...
Add momentum operator
8 years ago
f9681459b2
fix gpu build error
8 years ago
ab8cc401e6
add sparse sgd operator unittest
8 years ago
330c509bc3
Merge remote-tracking branch 'baidu/develop' into sgd_op_sparse_kernel
8 years ago
182ce51c6d
add sparse kernel of sgd operator
8 years ago
98dccc9f21
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
df0946ebe2
Impl optimizer ( #4734 )
...
* init parameter base class
* optimize the Comments of optimizer
* basic implimentation of optimizer
* add test_optimizer
* add no_grad_set to interface
* update optimizer.py
* python code can run
* fix some problem
* add sync_with_cpp to Python Program and Block
* sync vars and ops in block from cpp
* optimize code and add some comment
* add more check for sync
* update optimizer with return value of Backward
* rm unused code
* infer shape when create gradient vairiable
* update test_optimizer
* update test_program.py
* update backward test
* follow comment
8 years ago
451863dba2
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
426f7eee8e
simplify test_pool_py, add comments for different pooling strategy
8 years ago
a8a63d4c50
add MAX strategy for seqpool op
8 years ago
3123e3cf5d
Merge branch 'develop' into crf
8 years ago
73a8b78a72
Correct OpWithKernel's infershape ( #4847 )
...
They are public now
8 years ago
cc220eec36
add forward computation of crf operator.
8 years ago
75d0c79015
Change Name convention of operator attributes ( #4807 )
...
* Change dataType to data_type
Follow PEP8
* Change name_convention to fit PEP8
8 years ago
86acf39cce
Merge pull request #4801 from QiJune/add_selected_rows_functor
...
add some basic math functor for SelectedRows
8 years ago
23cb8259c3
"add python test case"
8 years ago
5993497cb7
fix merge conflict
8 years ago
4df6cf4d16
Rewrite feed/fetch op ( #4815 )
...
* Feed/Fetch op just plain operator, not a OpWithKernel
* Do not register OpInfoMaker since Feed/Fetch will never be
configured by users
* Feed/Fetch op has empty gradient
* Feed/Fetch op do not hard code `feed_variable`, `fetch_variable` as
its input and output, make it as a plain Operator input/output
8 years ago
571f43ca39
Merge remote-tracking branch 'baidu/develop' into add_selected_rows_functor
8 years ago
6a4282a20f
refine comments of sequence_pool_op
8 years ago
0d7b1fd4fe
fix bug in SelectedRowsAddTensorKernel
8 years ago
ab5dc9fe18
remove SelectedRows functors to selected_rows_functor.h
8 years ago
97332d7c0c
Merge branch 'develop' into seqpool
8 years ago
83627d3efd
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm
8 years ago
3cace73701
Add lstm implementation.
8 years ago
73883bde2a
"fix error"
8 years ago
d4d215a514
Merge remote-tracking branch 'origin/develop' into multigpu
8 years ago
5bcb63800e
"fix common test"
8 years ago
d8aebaf50c
"fix enforce error"
8 years ago
d2be7ec370
Merge remote-tracking branch 'origin/develop' into multigpu
8 years ago
df2d1769fd
fix code style
8 years ago
7ef568e893
fix gpu unittest error
8 years ago
f59a7c1d36
add gpu functor for SelectedRows
8 years ago
3f4177ea8d
Merge branch 'develop' into fix/scope
8 years ago
24a796fbc0
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into Add_conv3d_gemm_op
8 years ago
4741266d6f
remove unused method
8 years ago
931572e210
SelectedRowsAddTensor method
8 years ago
d144310415
"nccl add interface"
8 years ago
7b18343305
remove unused header file
8 years ago
5be10872f9
add selected_rows add cpu functor
8 years ago
652f182dc0
deconv
8 years ago
da399ae6b1
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
abfa81b106
Merge branch 'develop' into seqpool
8 years ago
918d69467c
Merge pull request #4791 from typhoonzero/fix_pybind_op_reg_nokernel
...
Fix pybind op register nokernel
8 years ago
a0af1eeabf
Merge pull request #4443 from guoshengCS/add-GRUStepOp
...
Add gru_unit_op
8 years ago
557c7ae37b
remove conflict
8 years ago
1540703394
fix_pybind_op_reg_nokernel
8 years ago
13b4749e24
fix_pybind_op_reg_nokernel
8 years ago
3b954e1ddc
Adding Hard Sigmoid Activation ( #4771 )
...
* Adding Hard Sigmoid Activation
* Adding a comment for slope to be only positive
* Fixing grammatical mistake in comment
8 years ago
1c1f73b46d
Feature/dynamic recurrent op forward test ( #4729 )
8 years ago
4c3ef7fca5
Merge branch 'develop' into seqpool
8 years ago
a36d24163a
Add no_grad_vars for grad_op_maker ( #4770 )
...
* Add no_grad_vars for grad_op_maker
* Add unittest
* Fix unittest
* Fix unittest
* Follow comment
8 years ago
416f590912
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
116800378a
Adding the Adam Optimizer operator ( #4733 )
...
* add adam op
moment1_out = beta1 * moment1 + (1 − beta1) * grad
moment2_out = beta2 * moment2 + (1 − beta2) * grad * grad
moment1_hat = moment1_out / (1 - beta1^t)
moment2_hat = moment2_out / (1 - beta2^t)
param_out = param - learning_rate * moment1_hat / (sqrt(moment2_hat) +
epsilon)
* fix moment 2
* Adding the Adam optimization operator
* Adding more tests for Adam op
8 years ago
7da9ab3c6c
Merge remote-tracking branch 'origin/develop' into fix/scope
8 years ago
ff0e9d2207
Merge remote-tracking branch 'origin/develop' into fix/scope
8 years ago
1a87a963e1
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into dev_opdesc_in_python
8 years ago
71abb4ceb1
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-NormOp
8 years ago
ae1b29ac7e
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-GRUStepOp
8 years ago
cbcf11d930
Merge branch 'develop' into crf
8 years ago
393c748c89
add seqlastin/seqfirstin for seq_pool op
8 years ago
4aae1fff78
fix conv3d_gemm, unit test and follow comments
8 years ago
d0d3129f91
Fix InferShapeContext in decayed_adagrad_op
8 years ago
9b4a6af251
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-GRUStepOp
8 years ago
0bc5a122d4
Refine gru_unit_op by optional bias
8 years ago
1cabdb8708
Refine gru_unit_op according to comments to support multiple activation types
8 years ago
d3b8bffaf1
Implementing the Decayed Adagrad optimizer operator ( #4645 )
...
* Implementing the DecayedAdagrad optimizer step operator
* implementing DecayedAdagrad operator
* remove file
* small fix
8 years ago
e69a565a5b
Merge branch 'develop' into seqpool
8 years ago
2daba04042
Merge pull request #4731 from reyoung/feature/fix_ci
...
Feature/fix ci
8 years ago
a3ccbdb3b6
Cudnn conv op ( #4195 )
...
* add cudnn_conv_op
* WIP
* update
* update
* fix grad check
* use platform::memory
* add support group for cudnn
* update
* follow comments
* fix onlycpu build
* update cuda define
* follow comments
* follow comments
* merge with updates
* fix compile error
* follow comments
* follow comments
8 years ago
4118782b31
Merge remote-tracking branch 'origin/develop' into multigpu
8 years ago
0fa34db759
nccl init
8 years ago
9106a4bba1
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm
8 years ago
8728b3cce2
Add LSTM Operators.
8 years ago
91cc5d6208
add the forward operator.
8 years ago
515981d714
Merge branch 'develop' into crf
8 years ago
d92c671d5f
add python forward unittest.
8 years ago
4b6b4bc84a
"change GetOrCreate to Var"
8 years ago
1dd6dbbce2
deconv
8 years ago
2434e4862a
Fix bugs
8 years ago
a4687c9909
fix conflict of CopyFrom in vol2ol_test
8 years ago
0af45b5fa6
NewVar to GetOrCreateVar
8 years ago
b504a2346c
Adding the Thresholded Relu Op ( #4685 )
...
* Adding thresholded_relu op
* Adding test for thresholded relu op
8 years ago
2603cb7e86
Unify CUDA stream in Tensor CopyFrom interface ( #4692 )
...
* init
* unify CopyFrom interface
* fix gpu build error
* fix bug in tensor_py.h
* refine code comments and add TODO list
* fix conflicts in FeedOp and FetchOp
8 years ago
f984cba0ca
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_expand_op
8 years ago
acd1aaea49
fix issues
8 years ago
0922fca41e
Add gru_unit_op
8 years ago
901b041196
Add seq_expand op
...
1. Add unitest
2. Add SeqExpandOpKernel
8 years ago
e9495e7694
Merge pull request #4508 from Yancey1989/seqconcat_op
...
Add the sequence_concat operator.
8 years ago
b56cbd307d
Merge pull request #4285 from kuke/margin_rank_loss_op_dev
...
Add margin rank loss operator
8 years ago
d68122ff5d
update
8 years ago
69e92b39fc
add an enforce
8 years ago
96b4035dd1
Add conv3d_gemm_op
8 years ago
c77bd553de
Merge branch 'develop' into seqpool
8 years ago
989e19caf9
fix typos in margin_rank_loss_op
8 years ago
13b7d92808
improve doc in margin_rank_loss_op
8 years ago
4849fba7e9
follow comments
8 years ago
240adef1a1
Merge branch 'develop' of upstream into margin_rank_loss_op_dev
8 years ago
ad477b9183
update
8 years ago
c22e7ff71e
Merge branch 'develop' into fix_pool_op_doc_and_class_declarations
8 years ago
62da438e8e
Merge pull request #4462 from chengduoZH/Add_vol2col_functor
...
Add vol2col functor
8 years ago
9995aed114
Implementing Softplus operator ( #4690 )
...
* implementing softplus
* small fix
* small fix
* small fix
* small fix
8 years ago
c3bf332666
Merge pull request #4537 from QiJune/executor_impl
...
Executor interface design and implementation
8 years ago
f5ac335046
follow comments
8 years ago
1397e17f6b
Implemented the hardShrink activation ( #4653 )
...
* Implemented the hardShrink activation
* Fixing the unit test
8 years ago
67dc606b32
Merge branch 'develop' into seqpool
8 years ago
1d41a6d45a
update paddle/operators/math/CMakeLists.txt
8 years ago
e85fa5e9c1
Merge branch 'develop' into fix_pool_op_doc_and_class_declarations
8 years ago
32146da9b1
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into Add_vol2col_functor
8 years ago
c85d777f87
follow comments
8 years ago
72d3d814b5
fix math/CMakeLists.txt
8 years ago
6604d7cda2
Add logsigmoid (numerically stable) and softshrink ( #4663 )
...
* Add numerically-stable logsigmoid activation
* Add softshrink operator
* Adjust relative tolerance for grad-check
* Address review comments
8 years ago
843ed8e320
dynamic recurrent op forward c++ implentation ( #4597 )
8 years ago
436ea50d5f
follow comments
8 years ago
293a7d1e75
add feed infershape todo
8 years ago
a281b38393
Conv Shift Operator ( #4591 )
...
* conv_shift_op: initial implementation using Eigen
Limitations:
- both gradient outputs must be specified and are always computed
- explicit for loops => could be optimized in various ways
(e.g., different memory layout)
* conv shift - gradient fixes
fix case when not all output gradients desired
* conv shift: minor cleanup
* conv shift - more minor cleanup
* conv shift: clean up & initial GPU implementation
* fix rebase issue
8 years ago
a308ff29af
make infershape of feedop and fetchop compatible with compile-time design
8 years ago
f2e7cf2141
fix InferShapeContextBase to InferShapeContext
8 years ago
6db476ed89
Separate the declarations and implementation of the PoolOp and PoolMaker class in order to reuse in pool_cudnn
8 years ago
67edd04a2f
fix doc
8 years ago
0f1d3af438
Merge pull request #4461 from chengduoZH/Add_maxpool_withIdx_only
...
Add max pool op (with index)
8 years ago
60f706a1d6
add SQRT strategy for sequence_pool_op
8 years ago
36da82550a
Add code comments
8 years ago
975a51294e
infer feed operator output variable shape with dims attribute
8 years ago
e21e5646a5
fix atomicAdd -> CudaAtomicAdd
8 years ago
e880a356fe
update
8 years ago
462579c416
update
8 years ago
a4d410aec8
Merge branch 'develop' of github.com:PaddlePaddle/Paddle into seqconcat_op
8 years ago
d211b51bd4
update comment
8 years ago
7d21d8c022
Merge remote-tracking branch 'baidu/develop' into executor_impl
8 years ago
15400748ae
follow comments and refine codes
8 years ago
6c6474cbd8
follow coments
8 years ago
e6421249d5
update to latest
8 years ago
932402c16b
debug for sum
8 years ago
9efd5422f9
Merge pull request #4655 from abhinavarora/fill_constant_op
...
Implementing the fill constant op for the executor
8 years ago
ceefb555f7
Merge pull request #4500 from luotao1/interp
...
add interpolation op
8 years ago
6efacc14d8
Implementing the fill constant op for the executor
8 years ago
e655d29173
merge develop
8 years ago
e51557130e
clean up for review
8 years ago
dcb09e932d
Use PROTO_LITE when refactoring Paddle
...
It will significantly reduce binary size. It is useful for mobile
deployment.
8 years ago
92add2a29b
Fix compile error in develop branch
8 years ago
e12ec95ac1
Merge pull request #4630 from jacquesqiao/merge-infershapecontext
...
Merge infershapecontext and ExecutionContext
8 years ago
4cb5bd9021
Implementing the Adamax optimizer operator ( #4538 )
...
* Implementing the Adamax optimizer step operator
* Adding unit tests for adamax_op
* Changing learning rate and time step to inputs from attributes
* Changing learning rate and time step to input(tensors)
* Making the Adamax operator conform to naming convention
* Removing Tensor<float> from comments
* Rectifying the Adamax implementation
* Changing Unit Test values and adding comments
* Changing Unit Test to test multiple steps
8 years ago
f30a1f42f0
Adding relu6 activation function ( #4607 )
8 years ago
597299074e
fix bug in REGISTER_OP(reduce_min)
8 years ago
a06f099d9f
refine comment of interp_op
8 years ago
fcfce48421
follow coments
8 years ago
63309941b3
pull develop and update
8 years ago
282435204e
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into auc_op
8 years ago
53574e54a1
reslove merge conflict;reimplement ELU activation with functor
8 years ago
707d144c93
Unify Reduce functions and simplify register code
8 years ago
5b862fedf1
remove debug log in interp_op.cc
8 years ago
4724bdbe68
Merge branch 'develop' into interp
8 years ago
3db3a1066b
remove conflict
8 years ago
ba791f7b3f
Add vol2col functor and unit test
8 years ago
dcfbbd3f1d
Merge branch 'develop' into crf
8 years ago
c93d74aa06
merge develop
8 years ago
c0a34e1c64
rename InferShapeContextBase to InferShapeContext
8 years ago
99895730f7
Merge pull request #4609 from kavyasrinet/tanhshrink
...
Implementing tanhshrink operator
8 years ago
f087533cc3
Merge remote-tracking branch 'baidu/develop' into executor_impl
8 years ago
91f5d2b9cb
follow comments and create local_scope inside executor run method
8 years ago
097f533bca
Resolve conflict
8 years ago
e8a678e1ee
fix executor gpu unittest runtime error
8 years ago
1f5192a27b
fix executor gpu unittest
8 years ago
087addaa76
Merge pull request #4558 from kexinzhao/adagrad_op
...
Implementing the Adagrad optimizer step operator
8 years ago
78f4c803f3
change learning rate and fix format
8 years ago
39f75a13a4
Merge remote-tracking branch 'baidu/develop' into executor_impl
8 years ago
bbceb72398
refine some codes
8 years ago
48b080db9f
ensure global BuddyAllocator is initialized before global Scope
8 years ago
f52cdaa0ce
Updated RMSProp to have learning rate as an input and work with GPU
8 years ago
0336304176
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into rmsprop
8 years ago
154a6ed29c
Implementing tanhshrink operator
8 years ago
db77937ea4
Fix learning_rate usage for momentum
8 years ago
45c4dcaabb
add fetch operator
8 years ago
3e2be065b9
Merge pull request #4604 from kavyasrinet/activations
...
Added Leaky Relu activation
8 years ago
20725f2d52
add executor feed operator test
8 years ago
c10da26cf5
Modify implementation
8 years ago
828c5b3e1d
Adding Adadelta optimization operator ( #4576 )
...
* Adding Adadelta optimization operator
* Making inputs and outputs conform to naming convention
* Removing type alias from header files
* Fixing Adadelta documentation in comments
* Addressing code review feedback
8 years ago
60af56c1b8
Added Leaky Relu activation
8 years ago
623848afa1
add feed operator
8 years ago
1172f24929
Merge pull request #4590 from wangkuiyi/paddle_only_cpu
...
Add -D PADDLE_WITH_CUDA in cmake/configure.cmake
8 years ago
8ebc31d935
optimize the dsize
8 years ago
775c60246b
remove using in sgd header file
8 years ago
2594a50245
Polish code
8 years ago
c4effc7d2d
Fix CI Test
8 years ago
ee7b3ed09e
use EigenScalar to get learning_rate from GPU device
8 years ago
4558807c48
Use PADDLE_WITH_CUDA instead of PADDLE_WITH_GPU
8 years ago
e79d2f1b65
Merge pull request #4584 from reyoung/feature/change_macro_paddle_no_gpu
...
Change `PADDLE_ONLY_CPU` to `PADDLE_WITH_GPU`
8 years ago
fa12e51675
Adding the default attribute test case
8 years ago
94855f4af0
Fixed changes proposed in the review
8 years ago
e119177a8c
Use unique_ptr
8 years ago
84500f9487
Change `PADDLE_ONLY_CPU` to `PADDLE_WITH_GPU`
...
By shell command
```bash
sed -i 's#ifdef PADDLE_ONLY_CPU#ifndef PADDLE_WITH_GPU#g' `find ./paddle/ -name '*.h' -o -name '*.cc' -o -name '*.cpp' -o -name '*.c' -o -name '*.cu'`
sed -i 's#ifndef PADDLE_ONLY_CPU#ifdef PADDLE_WITH_GPU#g' `find ./paddle/ -name '*.h' -o -name '*.cc' -o -name '*.cpp' -o -name '*.c' -o -name '*.cu'`
```
8 years ago
eed2c1e1d6
Changing SGD inputs and outputs to conform to Operator naming convention ( #4586 )
8 years ago
324876bbbf
Changing learning rate from type Input(float) to Input(tensor) ( #4578 )
8 years ago
14a59d2e6b
Merge branch 'develop' of github.com:baidu/Paddle into feature/grad_reg_mechanism_cont2
8 years ago
94b94e5b68
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
2d876b8643
gather scatter fix according to google style
8 years ago
d28b3094dd
Add momentum operator
8 years ago
42e7fe05a2
Changing learning rate from attribute to input(float) ( #4568 )
...
* Changing learning rate from attribute to input(float)
* Removing obsolete code
8 years ago
163d287143
Made learning rate the input
8 years ago
d1de7ec630
Change learning rate from attribute to input tensor
8 years ago
2ccaec4f57
gather scatter cond
8 years ago
46c551b299
Complete Register Gradient in compile time
8 years ago
61c03f9d59
Adding the implementation for rmsprop operator
8 years ago
ff1bfdedc9
Fix CRLF in sum_op.cu
8 years ago
adec0d30fe
Simplify SumOp Kernel
8 years ago
58174b12f7
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
8 years ago
84b8baf196
gather scatter with cuda streams
8 years ago
05cbd4daac
fix format
8 years ago
cde542e652
optimize auto
8 years ago
6b051b651a
optimize code
8 years ago
1ac654a69f
Implementing the Adagrad optimizer step operator
8 years ago
32f5c9dd93
recurrent_op pass the unit test
8 years ago
15941dbd8c
solve conflict for cond_op and scatter
8 years ago
7163dd0413
revert code
8 years ago
14b2c98f90
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into Add_maxpool_withIdx_only
8 years ago
0900aedfa0
Merge pull request #4514 from reyoung/feature/remove_add_op
...
Remove add_op since it can be replaced by sum_op
8 years ago
0028459bb0
update
8 years ago
927767b6aa
add some checking
8 years ago
bb33c2b3a5
fix kernel func
8 years ago
2ed56df1e6
remove conflict
8 years ago
bee95fc891
fix code format and some bug
8 years ago
a35e82a649
Merge branch 'develop' of github.com:PaddlePaddle/Paddle into seqconcat_op
8 years ago
4f5491b2b4
Merge pull request #4146 from chengduoZH/Add_pool_op
...
Add pool op
8 years ago
aa52fa1c64
Merge pull request #4491 from reyoung/feature/stable_lstm
...
Using double precision to stablize lstm gradient check
8 years ago
2d8a5b97cc
fix unit test
8 years ago
7fe0297e64
remove Runtime InferShape for cond op ( #4518 )
...
* init remove cond_op infershape
* optimize code
* add PrepareDataForSubnet and MergeDataFromSubnet
8 years ago
762a99cc06
Remove add_op since it can be replaced by sum_op
8 years ago
ae4b7fd575
Merge pull request #4485 from reyoung/feature/BetterActivationKern
...
Unify Activation functions and simplify register code
8 years ago
735737d283
initialize crf opreator.
8 years ago
be3fa7926e
add sequence concat op
8 years ago
601e2317fd
update to latest
8 years ago
4436ba0c56
elu: Optimize gradient calculation;Add more comments
8 years ago
a815d6abcf
elu: Optimize gradient calculation;Add more comments
8 years ago
df59889984
remove conflict
8 years ago
bb7f555803
remove rowwise_add_op
8 years ago
884e31a59b
add interpolation op
8 years ago
8bafdda0ad
Merge branch 'develop' into core_add_sequence_softmax_op
8 years ago
7cc5ae9999
Merge pull request #4492 from QiJune/refine_some_functors
...
Pass DeviceContext to functors not ExecutionContext.
8 years ago
b611a479fc
fix gpu build error
8 years ago
e1e3859e88
remove custom attr checker and fix code format
8 years ago
a53191f12a
Add norm_op
8 years ago
a8c6ce9b4d
Merge branch 'develop' of github.com:baidu/Paddle into feature/BetterActivationKern
8 years ago
84ff7e9784
refine SoftmaxFunctor
8 years ago
f60f0eae11
Using double precision to stablize lstm gradient check
8 years ago
0c3eee09ff
Implementing the SoftSign activation operator
8 years ago
79def5e634
refine CrossEntropyFunctor
8 years ago
c634a8480a
add SetConstant method in math_function.h
8 years ago
78808b2091
1 api
8 years ago
337b7ebe77
Unify Activation functions and simplify register code
8 years ago
b851515b16
merge new op grammar
8 years ago
88a8eedda1
scatter gather gpu
...
gather scatter gpu
8 years ago
87da154226
FIx sigmoid_xe_with_logits_op compile
8 years ago
63469daede
Merge branch 'develop' of github.com:baidu/Paddle into feature/make_paddle_support_double
8 years ago
b9c8637238
Fix compile
8 years ago
b9336e6f8c
Adding support for the sigmoid_cross_entropy_with_logits operator ( #4448 )
...
* Adding support for the sigmoid_cross_entropy_with_logits operator
* Fixing a typo in the cuda file
* Adding Python documentation for sigmoid_cross_entropy_with_logits_op
* Correcting typos in documentation
* Adding unit tests for sigmoid_cross_entropy_with_logits_op
* Addressing code review feedback
8 years ago
fb6a48c62d
Merge branch 'develop' of github.com:baidu/Paddle into feature/make_paddle_support_double
8 years ago
6326c40d27
Add max pool with index
8 years ago
4db50fbcdd
adapt to the new infershape interface
8 years ago
ecef2e6b97
Merge pull request #4086 from guoshengCS/add-ReduceOp
...
Add reduce op
8 years ago
e303897f35
Merge branch 'develop' of upstream into margin_rank_loss_op_dev
8 years ago
03897f251d
Finish the SequenceSoftmaxGradKernel, using SoftmaxGradFunctor.
8 years ago
05ed8ee8ab
Add SoftmaxGradFunctor, and use SoftmaxGradFunctor in softmax_op instead.
8 years ago
bc2e26ee1b
refine comments and clean code in marigin_rank_loss_op
8 years ago
21f63ec223
Merge pull request #4458 from reyoung/feature/compile_time_infer_shape
...
Remove OperatorBase::InferShape
8 years ago
d7db15f3e5
Use StridedMemCpy in Concat/Split Kernel ( #4188 )
...
User StridedMemCpy in Concat/Split Op
8 years ago
e33b411221
Adapt reduce_op according to up-to-date dev
8 years ago
ae3dca770c
Fix CI
8 years ago
be58c6327d
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-ReduceOp
8 years ago
6196209478
Remove OperatorBase::InferShape
...
InferShape in Operator should be performed in OperatorBase::Run.
* cond_op, recurrent_op and mnist might be changed in following PR
8 years ago
f1913d4697
Change registry, test register double kernel
8 years ago
2c05465d2f
Fix unit-tests
8 years ago
3a5693e0a8
Add Skeleton of Double support
8 years ago
e315da970c
Merge branch 'develop' of github.com:baidu/Paddle into feature/unify_clang_format
8 years ago
3c0f079333
remove conflict and fix InferShape function
8 years ago
c53d214205
fix compiler warning from MKLDNNLayer and so on
8 years ago
3e96242909
Merge pull request #4422 from wanghaoshuang/fix_clip
...
Fix doc of clip op.
8 years ago
1ad4efa8de
Fix comments
8 years ago
6da5670f62
revert deleting ShareLoD
8 years ago
028973815b
Fix InferShape of clip op.
8 years ago
729b2152ae
add include op_registry.h to elementwise_op_function.h
8 years ago
abdcd8282e
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into reduce-elementwise-warning
8 years ago
ce3171f3c4
Merge branch 'develop' into core_add_sequence_softmax_op
8 years ago
4b06d8db91
fix globalPooling type (int => bool)
8 years ago
325ee63746
fix SoftmaxWithCrossEntropyOp
8 years ago
99b8dbb14f
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-ReduceOp
8 years ago
0fa4b985a9
split elementwise_op.h into two header files
8 years ago
c8fc6037fd
Merge branch 'develop' into core_add_sequence_softmax_op
8 years ago
29cb85634c
Merge pull request #4144 from lcy-seso/softmax_with_cross_entropy_op
...
Softmax with cross entropy op.
8 years ago
97509b68e7
cross entropy as a functor to avoid duplicated codes.
8 years ago
60857f4909
Unify clang-format and add some missing clang-format
8 years ago
9a9d50a6ee
Refactoring InferShape ( #3946 )
...
* init Infershape
* add static InferShape interface
* refactor add-op infershape
* add AttrReader
* add all maker's infershape
* add all InferShape
* add python infer api
* add VarDesc interface
* add python VarDesc and OpDesc interface
* update python code
* use infershape function to do shape inference
* clean code
* do not use pointer
* refine code of op_proto_maker
* add get_dims to VarDesc
* refine the code
* remove the dependency from operator to op registry
* remove OpProtoAndCheckerMaker from operator
* restore complete_add_op
* add shape_infer_impl.h
* code optimization
* remove const return value
* add fake BlockDesc class
* optimize code
* remove infer function in op_info
* move InferShapeContextImpl to operator.h
* optimize the interface of InferShapeContextBase
* add temperary interface of new infershape
* change add_op, clip_op, conv2d_op and activation_op
* change all operators InferShape
* fix SetDim
* update cos_sim_op
* update crop_op
* update lookup_table_op
* allocate tensor when call GetDim in InferShapeContext
* update modified_huber_loss_op
* update rowwise_add_op
* update mean_op
* update sequence_avg_pool_op
* typo
* remove old InferShape interface
* can compile
* fix or unit test
* clean code
* clean code
* remove const before InferShapeContext
* change InferenceContextBase to pointer
* rename RunTime to Runtime, code clean
8 years ago
dc186af729
Merge branch 'develop' of upstream into margin_rank_loss_op_dev
8 years ago
367a54e08c
Merge pull request #4360 from kuke/multiplex_modify_dev
...
Modify multiplex_op
8 years ago
e9dbc85b3d
to trigger ci for multiplex_op
8 years ago
30a586df0c
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into Add_pool_op
8 years ago
40aee48a80
follow comments.
8 years ago
3d77360b89
add negative clipping for softmax.
8 years ago
dfc8d3c1c1
Fix (According to the review)
8 years ago
360bde9a70
Merge branch 'develop' into softmax_with_cross_entropy_op
8 years ago
477a6a0978
Refine reduce_op, follow comments and remove ReduceGradEigenFreeKernel
8 years ago
7d65321620
Merge pull request #4237 from lcy-seso/optimize_cross_entropy_kernel
...
optimize cross entropy kernel.
8 years ago
Dong Zhihong
chengduoZH
Siddharth Goyal
yangyaming
chengduo
gongweibao
dangqingqing
Luo Tao
Yu Yang
zhouxiao-coder
Yang Yang(Tony)
Abhinav Arora
caoying03
Qiao Longfei
dzhwinter
wanghaoshuang
Zhuoyuan
QI JUN
guosheng
peterzhang2029
Guo Sheng
Yan Chunwei
kavyasrinet
fengjiayi
Markus Kliegl
typhoonzero
kexinzhao
Cao Ying
Yibing Liu
Yancey1989
Abhinav Arora
gangliao
Yi Wang
Yi Wang
Kexin Zhao
Liu Yiqun