3c84ebec62
IndicateDataType --> GetKernelType
7 years ago
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
3d6ca6ade5
Merge pull request #5484 from qingqing01/cmake_speed
...
Remove fill_constant_batch_size_like_op.h and clean some operator codes.
7 years ago
aa340671d2
Merge pull request #5016 from guoshengCS/add-ChunkEvalOp
...
Add chunk eval op
7 years ago
93551bd232
refine unit test (Add dilation)
7 years ago
46c61b35f7
"add elementwise op support"
7 years ago
133f890346
Merge pull request #5499 from reyoung/feature/increase_cpu
...
Feature/increase cpu
7 years ago
2e355f032e
Fix attribute naming for momentum_op ( #5453 )
...
* Fix attribute naming for momentum_op
* Fix minor typo in comment
* Fix attribute name
* Fix names in test_optimizer
* Fix python wrapper
7 years ago
b8f557f283
"add elementwise_add more type"
7 years ago
c88f98cf9e
Merge pull request #5501 from reyoung/feature/lod_array_length
...
Add `lod_array_length` operator
7 years ago
ef84ff8657
Merge branch 'develop' into feature/increase_cpu
7 years ago
03fa1edc20
Merge pull request #5512 from luotao1/intel
...
remove unused INTEL_MKL_ROOT etc.
7 years ago
cceed08119
remove header file paddle/framework/eigen.h
7 years ago
3fa3f5e799
Update code and resolve conflicts.
7 years ago
d60fe75ac3
follow comments.
7 years ago
23b0388f46
add sub sequence operator code and unittest
7 years ago
7835d49384
remove PADDLE_USE_MKL
7 years ago
5cf8204171
refine docString
7 years ago
21ce704247
refine conv2d for filter size:(1,1)
7 years ago
4098ce730b
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_fix
7 years ago
0d9ba3da9a
Adapt to new interface.
7 years ago
91855659d7
Merge pull request #5296 from wwhu/clip_by_norm
...
add clip_by_norm op
7 years ago
b698d19bfb
Add grad for lodtensor array ops ( #5461 )
...
* 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
* Add grad for ops
* Debug
* Fix a bug
* fix a bug
* fix a bug
7 years ago
d24d8c20f3
Add `lod_array_length` operator
7 years ago
b266f504c5
Merge pull request #5441 from reyoung/feature/fill_constant_force_cpu
...
Add force_cpu for fill_constant op
7 years ago
7f22a6d1b5
Merge pull request #5465 from reyoung/feature/compare_op_support_cpu
...
CompareOp's kernel device type is decided by input tensor place
7 years ago
6d41bfb7df
Add increment op
7 years ago
568270f3c6
Stash
7 years ago
2a76b42e44
Merge pull request #5419 from reyoung/feature/shrink_memory_op
...
Feature/shrink memory op
7 years ago
c8dcd9a9ba
Refine ChunkEvalOp by following comments and rewrite the doc
7 years ago
b3a86b6dbb
fix CI
7 years ago
e5791dd1c7
Remove fill_constant_batch_size_like_op.h and clean some operator codes.
7 years ago
b6f9ba484e
fix conv2d doc
7 years ago
7be390aa83
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into fix-4029
7 years ago
502d7daff7
Merge pull request #5467 from typhoonzero/fix_accuracy_cudamemset
...
fix accuracy cudamemset
7 years ago
97e9dd7237
add dilation for im2col
7 years ago
86e58b2343
Merge pull request #5457 from chengduoZH/fix_pool_attr_name
...
fix_pool_op_attr_name
7 years ago
870650d8c1
Static lstm sanity check ( #5365 )
...
* add fill_constant_batch_size_like_op to rnn h_boot
* first commit
* merge develop; fix conflict
* update to main_program
7 years ago
11ee50ceb9
update
7 years ago
44e1ac3882
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into bi_tensor_prod_op
7 years ago
47269273ff
refine memory transform
7 years ago
c2dd75be68
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into nce_op
7 years ago
3d276277df
Add nce op
...
1. Add nce forward and backward kernel for CPU
7 years ago
6308ccc265
fix accuracy cudamemset
7 years ago
ee82e660c5
Merge pull request #5447 from reyoung/feature/fill_constant_no_template
...
Add `op::math::set_constant` without template
7 years ago
3187451ae7
CompareOp's kernel device type is decided by input tensor place
...
CompareOp can run on CPU even other operators are running on GPU, since
opeatations like comparing control flags should be performed only on CPU
7 years ago
272a272b68
Merge branch 'develop' of github.com:baidu/Paddle into feature/shrink_memory_op
7 years ago
0142530929
Rename shrink_state -> shrink_rnn_memory
...
Follow comments
7 years ago
2ac5772970
Merge branch 'develop' of github.com:baidu/Paddle into feature/fill_constant_force_cpu
7 years ago
c365c61ac6
Merge pull request #5445 from reyoung/feature/fix_dynamic_rnn_compile
...
Add gtest for drnn
7 years ago
e34e12931e
Merge remote-tracking branch 'origin/develop' into feature/evaluator
7 years ago
c8c4b6e427
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into clip_by_norm
...
update
7 years ago
cdf5e87104
fix attr name
7 years ago
91b724821c
Merge pull request #5443 from reyoung/feature/InferKernelKey
...
Polish OpWithKernel
7 years ago
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
2dd91dd572
Shrink State Operator
...
Used for shrink memories state in DyRNN. The height of state could
be shrinked after running a step block.
7 years ago
f651cefe27
Merge branch 'feature/fill_constant_no_template' into feature/fill_constant_force_cpu
7 years ago
b4e1824363
Fix CI
7 years ago
0708a1550c
Fix CI
7 years ago
5ee62383bd
Rewrite fill_constant op
7 years ago
aadb098138
Add `op::math::set_constant` without template
7 years ago
db3b49fe0e
Add gtest for drnn
7 years ago
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
f74fb79036
Compare Operator ( #5325 )
...
* Compare Operator
* Follow comments
7 years ago
58db07b7bb
Check errors for the cuda kernel calls. ( #5436 )
7 years ago
7874399cf3
Merge remote-tracking branch 'origin/develop' into feature/evaluator
7 years ago
f5cb52ca3e
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into bi_tensor_prod_op
7 years ago
3ae14242da
update for mini-batch
7 years ago
00360e7eb5
update
7 years ago
579c92abc3
fix compile
7 years ago
ed2d30b5f9
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_fix
7 years ago
d94c936bd5
Enhance unit testing.
...
1. user can disable peephole connections.
2. not calculate some gradients.
7 years ago
8d9b33412d
Merge remote-tracking branch 'origin/develop' into feature/evaluator
7 years ago
c09ad73c33
"add fit a line test"
7 years ago
7eb65b31a5
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into avx_cmake
7 years ago
7abd1bdfff
Fix cmake error when building with WITH_AVX=OFF.
7 years ago
3014645d35
Merge pull request #5403 from typhoonzero/refine_get_cuda_context
...
Refine get cuda context
7 years ago
bdc832cba1
"add eval interface"
7 years ago
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
20667e1e3e
add fill_constant_batch_size_like_op to Static RNN's h_boot ( #5332 )
7 years ago
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
d851dafed5
Update and resolve conflicts.
7 years ago
0c70bd28aa
Enable initial hidden state and cell state in LSTM Operator.
7 years ago
f78731c837
Merge pull request #4709 from chengduoZH/Add_conv3d_gemm_op
...
Add 3D Convolution operator implemented by GEMM.
7 years ago
c8f1389bea
Merge pull request #5108 from zhouxiao-coder/pnp-evaluator
...
Pnp evaluator
7 years ago
272f3e6d43
refine get cuda context
7 years ago
7ca6c8840e
Merge branch 'develop' into Add_conv3d_gemm_op
7 years ago
f302c6a3b4
write conv2d and conv3d together
7 years ago
acc3278811
remove conflict
7 years ago
206f32c13a
deconv2d kernel and deconv3d kernel write together
7 years ago
5b077214f4
Merge pull request #5394 from lcy-seso/fix_softamx_with_cross_entropy
...
fix bugs of softmax_with_cross_entropy op.
7 years ago
ece1d57c2b
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-ChunkEvalOp
7 years ago
6f4bf5052f
fix softmax with cross entropy op.
7 years ago
0977cbfb41
Merge pull request #5310 from qingqing01/fix_max_pool
...
Refine sequence max-pooling and add unit testing of gradient check.
7 years ago
0f1b30ef86
fix doc and unit test
7 years ago
ba7db29df1
remove conflict
7 years ago
8ac1178707
fix doc
7 years ago
7a1d5e9d6a
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-GRUOp-dev
7 years ago
e3d15eee12
Add TODO in the GRU Operator related
7 years ago
483947c45d
Update and Resolve conflicts
7 years ago
cb0118f3e5
Polish Operator Doc (m) ( #5375 )
...
* fix m_ops
* fix activation op
7 years ago
5d8cdf2031
Polish operator docs (n to p) ( #5376 )
...
* polish p ops
* fix precision_recall
* fix linear_chain_crf_op
* small fix
7 years ago
fb2aa7179c
Polish Operators Docs (r) ( #5377 )
...
* polish r operator docs
* fix on naming convention
7 years ago
30a85204b4
Adding the doc format for AdaDelta, AdaMax, Adam, AdaGrad, BatchNorm, Clip, Cast and AUC ( #5317 )
...
* Adding the doc format for AdaDelta
* Updating the documentation for Adagrad, Adam and Adamax
* Updating the auc op
* Fix review comments
* Updating doc for Batch Norm
* Updating the cast op
* Updating the clip op
* Fixing review comment
* Fixing review comment:
* Small change to restart PR_CI
7 years ago
2ac5d7d018
Fixing documentation for operators ( #5373 )
...
* Adding documentation for seq_expand
* Adding documentation for seq_concat_op
* Adding documentation for sequence_conv
* Adding sequence_pool
* Fixing review comment
* Adding sequence_softmax
* Updating doc for sigmoid_cross_entropy_with_logits
7 years ago
e65ab795af
Fixing documentations for few more operators ( #5374 )
...
* Doc fix for smooth L1 loss
* Adding doc for softmax_op
* Added doc for softmax_with_cross_entropy
* Adding documentation for transpose_op
* small change to restart TeamCity CI
7 years ago
10bd9f6831
remove conflict
7 years ago
82aa569353
follow comments
7 years ago
56bae5c3da
Fix activation_functions in gru_gpu_kernel
7 years ago
c4f7f3a562
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-GRUOp-dev
7 years ago
1d85b2bd17
Refine GRU Operator according to activation_functions
7 years ago
c0d2ca54b9
polish_g_to_l ( #5367 )
7 years ago
af760eac5e
polish op from e to f ( #5357 )
7 years ago
c5c024377b
Polish from concat to conv shift operators ( #5347 )
...
* polish from concat to conv_shift op doc
* small fix
* small fix
7 years ago
45eabb8cf2
Add the crf_decoding operator. ( #5352 )
...
* proj init.
* add unittest and implementation.
7 years ago
b0b26dabe7
Polish operator documentation ( #5356 )
...
* Polish the documentation for uniform_random and top_k ops
* Polishing more operators
7 years ago
7408a4c408
Merge pull request #5354 from kexinzhao/cos_sim_to_dynamic_recur_op
...
polish operator doc
7 years ago
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
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
73632deea0
Polish the documentation for uniform_random and top_k ops ( #5353 )
7 years ago
6a07af0671
polish doc c to d
7 years ago
4c9be1a997
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into fix_max_pool
7 years ago
4b8bcf320f
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-GRUOp-dev
7 years ago
faad835166
Refine GRU Operator by following comments
7 years ago
59cbaf9fe7
fix doc
7 years ago
34d68f24fc
fix doc and code style
7 years ago
496583a165
update data types and test cases
7 years ago
39320f00df
Merge pull request #5334 from kexinzhao/polish_op_doc
...
Polish activation operator documentation
7 years ago
1c004a4913
Resolve conflicts.
7 years ago
a8dee3b315
Merge pull request #5251 from qingqing01/lstm_fix
...
Refine activation function pointer for LSTM operator.
7 years ago
1ed5ae7a14
Fix comparing between signed and unsigned values ( #5328 )
7 years ago
81ba077e7b
small fix
7 years ago
4b9a2c44f1
Fix bug in lookup_table_op & layers ( #5298 )
...
* Fix bug in lookup_table_op & layers
* Missing Act in layers
* Should += in CPU
* Remove check in python
* Fix bug in sequence_conv_pool()
* Fix a bug in test_recommender_system.py
* Just skip test_evaluator
7 years ago
519476a4c6
Fix CMake bug.
7 years ago
afc6343e6f
Refine sequence max-pooling and add unit testing of gradient check.
7 years ago
29a9f9b5ea
Refine code format and fix threads number.
7 years ago
5682916c47
add acknowledgement of lstm_unit_op ( #5293 )
7 years ago
6d649d9ebd
Merge pull request #5235 from zchen0211/develop
...
deconv cudnn forward passed
7 years ago
65451b5c4d
add cliy_by_norm op
7 years ago
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
8cdb42c2b3
Merge pull request #5111 from pkuyym/fix-5070
...
Add PrecisionRecall Op
7 years ago
90f4d5e904
modify fill constant batch size like ( #5222 )
7 years ago
0efac253d3
deconv small fix
7 years ago
2d956b82cd
deconv cudnn
7 years ago
7e34b8e366
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
7 years ago
0885de47eb
first commit ( #5286 )
7 years ago
31187e7e72
deconv fix
7 years ago
5a4cdbb3df
Fix check bug.
7 years ago
03bfd7611a
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_fix
7 years ago
1f53a72f10
Reduce the threads number in the LSTM backward kernel to fix the error occurred in GPU GTX 1080.
7 years ago
970613fc15
Refine and follow comments.
7 years ago
b720f282b1
deconv modify
7 years ago
4e228021bb
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
7 years ago
23a631d462
Fix End of Files in GRU Operator
7 years ago
bb75381444
Clean code of GRU Operator
7 years ago
53d8165f53
Make GRU Operator adapt to sequence2batch
7 years ago
8401039feb
Merge pull request #5084 from lcy-seso/crf
...
Add the LinearChainCrf operator.
7 years ago
83b48ebcb7
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-GRUOp-dev
7 years ago
36d20609a2
Merge pull request #5115 from qingqing01/lstm_bp
...
Add backward implementation for LSTM operator.
7 years ago
3d567864bb
Merge pull request #5247 from qingqing01/memory_alignment
...
Allocate aligned memory by posix_memalign.
7 years ago
26492210c0
Fix/sequence op ( #5264 )
...
* "replace enum with string"
* "fix layers"
7 years ago
a0acfc6a62
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
7 years ago
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
ee11f00642
add shareLod ( #5259 )
...
* add shareLod
* fix sequence_conv grad infershape
7 years ago
360cb18321
fix bug in lookup table grad operator ( #5228 )
7 years ago
9074a60c51
Refine lookup_table_op ( #5257 )
...
1. Change some `auto` to `auto*`
2. Change `Tensor` to `LoDTensor`
7 years ago
db3b9438b7
Adding Normal distribution initializer and unit tests for python initializers ( #5256 )
7 years ago
8013328ed8
Refine evaluator op types ( #5208 )
...
* refine evaluator op types
* update
* follow comments
* update
* fix v2 mnist case
* fix v2 mnist case
* update
* update
7 years ago
b77f9fbf04
deconv2d cudnn
7 years ago
a4d54b83d4
Make GRU Operator adapt to the latest code
7 years ago
ebd992ec79
backpropagate gradients the CRF operator receives.
7 years ago
2ac9a3d8dc
follow comments.
7 years ago
c0005d5862
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_fix
7 years ago
7061e0139f
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
7 years ago
6f658bb7d9
Clean code and update doc.
7 years ago
1c8a0c4bd4
Refine activation function pointer for LSTM operator.
7 years ago
1491041b0e
merge latest code
7 years ago
a9f9e208f5
Add optional inputs and outputs to enable updating;Add weight to match original implementation
7 years ago
e88e1964eb
Fix compiling warning.
7 years ago
e68a217f34
Add optional inputs and outputs to enable updating;Add weight to match original implementation
7 years ago
dd2be3daba
Merge branch 'develop' into crf
7 years ago
a328ae3b9b
Use posix_memalign to allocate aligned memory, since some SIMD instructions require the alignment of memory accesses.
7 years ago
9162629b04
Merge branch 'develop' of https://github.com/PaddlePaddle/paddle into add-GRUOp-dev
7 years ago
b87eabae56
Add GRU Operator
7 years ago
878dd88f61
Refine evaluator op types ( #5208 )
...
* refine evaluator op types
* update
* follow comments
* update
* fix v2 mnist case
* fix v2 mnist case
* update
* update
7 years ago
86fd6b6373
add gpu kernel by copying inputs/outputs between cpu and gpu.
7 years ago
a349bee6ad
deconv2d cudnn
7 years ago
e80489a488
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop
7 years ago
2d44a2ec5a
deconv cudnn
7 years ago
f122a5da2f
Add accuracy layer ( #4958 )
...
* Complete accuray layer
* Fix error
* Fix error
* Add 'accuracy' to __all__
* update
* Fix Type error
* Fix error
* Refine unit tests
* Fix an unit test error
7 years ago
669786bfe1
refine square_error_cost layer ( #5216 )
...
* reimplement pow operator
* add pow_grad operator
* fix code style
* fix build error
* fix op_test bug
* revert pow operator
* add FIXME comment
7 years ago
d3cc7ac304
Fix top k op GPU code ( #5221 )
...
* Fix Type error
* Fix error
* Fix top_k_op GPU code data type
7 years ago
2b1f21a59b
Fix MacOS Compile ( #5217 )
7 years ago
73d7855726
Fix a type error top_k_op ( #5201 )
...
* Fix Type error
* Fix error
7 years ago
1431f25196
remove conflict
7 years ago
5173b8d88f
fix code format and doc
7 years ago
0f9858a1c9
Merge pull request #5130 from chengduoZH/fix_sequence_conv_op
...
fix_sequence_conv_op
7 years ago
172481534d
fix code format and doc
7 years ago
b08ae0b1dc
fix code format and doc
7 years ago
2c5d4c6d20
Clean code and update doc.
7 years ago
1d7c03e789
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
7 years ago
9ee8a0d006
remove conflict
7 years ago
03136f6375
Merge pull request #4740 from wanghaoshuang/seq_expand_op
...
Seq expand op
7 years ago
84f471b42e
Fix comments
7 years ago
7942984f85
follow comments
7 years ago
8d4e2d4cb3
1. Add unitest for empty sequence case
...
2. Fix comments and paddle enforce check
7 years ago
8efd08763f
Merge pull request #5187 from chengduoZH/fix_pool_op
...
fix pool op
7 years ago
9f32b61c27
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_expand_op
7 years ago
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
833d0ad0fe
Merge pull request #4838 from dzhwinter/feature/multigpu
...
Feature/multigpu
7 years ago
71305e5f90
"polish code based on comment"
7 years ago
b50c33fd00
Use fixed activation in the lstm kernel, since there is some bug in the activation function pointer. It will be fixed later.
7 years ago
35e79448a0
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_expand_op
7 years ago
b84e822651
Cast Operator ( #5149 )
...
* Cast Operator
Cast input variable to other data type
* Fix compile error
* Add cast op
* Follow comments
7 years ago
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
008f40ce09
support sparse output for lookup table grad op ( #5145 )
...
* 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
* support sparse output for lookup table grad op
* refine codes
* fix gpu build error
* fix lookup table grad gpu kernel
* fix ci
* fix ci
* fix ci
* fix bug in lookup_table_grad op
* fix bug in test_word2vec
* register double kernel for some operators
* set is_sparse=True in test_word2vec
* fix lookup table grad op CUDA kernel bug
* disable test_modified_huber_loss_op temporarily
* disable test_lstm_unit_op temporarily
7 years ago
6bdf5c1417
fix bug
7 years ago
52eb42cf7c
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into batch-norm-latest
7 years ago
03789a7df4
batch norm fully tortured and passed
7 years ago
b067639621
Fix clang compile ( #5171 )
7 years ago
1a26f5a548
Adding the Sign Op for L1 Weight Decay Regularization ( #5138 )
7 years ago
8a07aff4d7
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into batch-norm-latest
7 years ago
822cf9785b
more test and bn fix
7 years ago
09ed52831b
Merge branch 'develop' into Add_deconv3d_op
7 years ago
e248866449
Merge branch 'develop' into Add_conv3d_gemm_op
7 years ago
51113cfe52
write together
7 years ago
5a95792925
Merge pull request #4636 from chengduoZH/Add_pool_cudnn_op
...
Add pool2d cudnn op
7 years ago
6ef9da8ef7
fix compile error ( #5160 )
...
* fix compile error
* remove unittest
* disable huber loss unittest
7 years ago
ed120ee741
Add unit test
7 years ago
bf3ae0631e
remove conflict
7 years ago
cadee843b8
follow comments
7 years ago
d2b10cc0b1
Refine doc and fix data type of metrics.
7 years ago
92c32799e7
Merge pull request #4864 from luotao1/maxseq
...
add Max strategy for sequence_pool op
7 years ago
cca383cfba
follow comments.
7 years ago
fd5199fdf7
Merge pull request #3989 from pkuyym/fix-3923-r
...
Add huber loss operator.
7 years ago
f086f5649b
Merge branch 'develop' into maxseq
7 years ago
24d988ce5c
Merge pull request #4063 from typhoonzero/auc_op
...
Auc op
7 years ago
df48b43b91
fix clear zero method and remove useless code
7 years ago
ac5f42184f
Using static_cast to make more robust.
7 years ago
3afb9dc88a
use double in unittest.
7 years ago
b9edcc4a1b
sss
7 years ago
97bfc0dfae
Add comments.
7 years ago
8f4476b893
Add device.Wait() in fetch_op ( #5141 )
7 years ago
0ab012cf7f
fix doc
7 years ago
bce4f7d6eb
follow comments.
7 years ago
94fb7ba4fa
Adding L1 norm op for L1 regularization ( #5058 )
...
* Adding L1 norm op for L1 regularization
* Addressing code review feedback
* Address code review feedback
* Change variable names to match google style guide
7 years ago
b44f4ccbeb
Make InferShape as a field in OpInfo ( #5139 )
...
* Op developer can add `InferShape` to any operator
7 years ago
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
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
fc68290bcc
update _create_op_func_ and support generate dropout layer ( #5134 )
7 years ago
52200523d6
"polish code based on comment"
7 years ago
6cce5268ed
"fixed based on comment"
7 years ago
16a39d24f3
fix conflict
7 years ago
6bc261b933
fix ci
7 years ago
25df892932
remove conflict
7 years ago
66476fc7b7
Add proximal adagrad optimizer ( #5128 )
7 years ago
65dbbd57af
Add and pass unittests.
7 years ago
db1bb8224a
follow comments
7 years ago
8e3ecf5d11
Merge pull request #4814 from chengduoZH/Add_sequence_project_op
...
Add sequence_conv_op and sequence_projection functor
7 years ago
9d142d5060
Local response normalize. ( #4426 )
...
Add local response normalize
7 years ago
bd680f157f
fix compiling warning.
7 years ago
99c6f44a5a
follow comments
7 years ago
746f2a2e36
only compute the first max value in backward
7 years ago
bcc0dad7df
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
7 years ago
ac3370a467
Add unit testing for gemv and fix the gradien check for bais.
7 years ago
dcb3da591e
refine code
7 years ago
56bbfd1af2
Add deconv3d op
7 years ago
aa3de3571d
Polish unit test for xe, generate probablities ( #5096 )
...
* Cross Entropy Wrong
* Fix XE
* Polish gradient check for xe
* Fix compile
7 years ago
06c7c8c80e
Add CPU kernel.
7 years ago
1bb0e2943b
Add pool2d cudnn
7 years ago
eafbbc11a0
write conv2d and conv3d together
7 years ago
2e0298744e
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
7 years ago
cd38286684
Add gradient check unit testing and fix bug.
7 years ago
efc2464f6c
Feature/save op ( #5090 )
...
* Init
* Stash
* Polish SaveLoadOp
* Fix CI
* Polish code
* Save GPU Tensor
* Stash
* Fix CI
7 years ago
4b9cf0e8b1
"add disable"
7 years ago
9b0f092853
change input data type
7 years ago
6cc2ce010a
add fill constant batch size like op ( #5057 )
7 years ago
626ff3b79e
"polish cmake file"
7 years ago
368737e41b
Merge branch 'develop' into maxseq
7 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
7 years ago
4c63086921
Merge branch 'develop' into crf
7 years ago
2573ac1448
"remove python side test case to another PR."
7 years ago
4e165f4ea3
"fix create output variable bug"
7 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
7 years ago
61c1b0469a
"fix multigpu testcase"
7 years ago
38d3adfeb6
"add multioperator testcase"
7 years ago
c22f7fcd17
add positive_negative_pair_op evaluator
7 years ago
94992a990b
"add multiop testcase"
7 years ago
bab5e53c03
Merge remote-tracking branch 'origin/develop' into feature/multigpu
7 years ago
11cf3e3a43
"refactorization of nccl test case"
7 years ago
6d1493a460
"add bcast c++ test case"
7 years ago
5200c657a7
"move Tensor to LoDTensor"
7 years ago
c74107bfdc
fix backward computation.
7 years ago
d2bd7357d2
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into lstm_bp
7 years ago
3d8b6ebcf8
Add LSTM backward implenmentation.
7 years ago
9f7c9875a9
fix doc
7 years ago
3f8a7b55b8
remove conflict
7 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
7 years ago
63fb41b399
"redefine the initop from kernel to OpBase"
7 years ago
1e8474b9f1
"delete python ops testcase"
7 years ago
yangyaming
Yang Yang(Tony)
qingqing01
chengduoZH
Dong Zhihong
Yu Yang
Siddharth Goyal
Tao Luo
wanghaox
peterzhang2029
Hu Weiwei
fengjiayi
guosheng
武毅
chengduo
wanghaoshuang
zhouxiao-coder
caoying03
kexinzhao
kavyasrinet
Cao Ying
Abhinav Arora
Qiao Longfei
Kexin Zhao
Yi Wang
Zhuoyuan
dzhwinter
QI JUN
gongweibao