提交 · e8f2614da5e57b114efeaceaff8f8488575bd8c4 · Crayon鑫 / Paddle

29 10月, 2020 2 次提交
- G
  Enhance multiclass_nms op to support LoD for dygraph mode (#28276) · e8f2614d
  由 Guanghua Yu 提交于 10月 29, 2020
```
* Enhance multiclass_nms to support LoD for dygraph mode

* fix some error in multiclass_nms

* update GetLodFromRoisNum to GetNmsLodFromRoisNum
```
  e8f2614d
- 石
  
  fix analyzer_capi_tester, test=develop (#28289) · 842a4e5a
  由石晓伟提交于 10月 29, 2020
  
  842a4e5a
28 10月, 2020 5 次提交
- L
  
  Fix transpose in conv cudnn kernel when addto enabled (#28295) · 89530384
  由 Leo Chen 提交于 10月 28, 2020
  
  89530384
- T
  
  fix conv mkldnn build error (#28288) · e1e666a0
  由 Tao Luo 提交于 10月 28, 2020
  
  e1e666a0
- J
  - sum (#28233) · 0b678d40
  由 Jacek Czaja 提交于 10月 28, 2020
```
test=develop
```
  0b678d40
- J
  
  [oneDNN ] conv2d fwd&bwd optimization (#27871) · c11d9b30
  由 Jacek Czaja 提交于 10月 28, 2020
  
  c11d9b30
- W
  update matrix nms op to api 2.0 (#28265) · 41d26a82
  由 wangxinxin08 提交于 10月 28, 2020
```
* update matrix nms op to api 2.0

* modify code according to review
```
  41d26a82
27 10月, 2020 5 次提交
- L
  
  fill_constant op supports NINF (#28270) · 7fcb32dd
  由 Leo Chen 提交于 10月 27, 2020
  
  7fcb32dd
- W
  
  refine yolo box Op for performace optimization (#28155) · 6905608c
  由 wangchaochaohu 提交于 10月 27, 2020
  
  6905608c
- W
  
  refine temporal_shift_op for performance optimization using gpu kernel config (#28114) · cdadc8f0
  由 wangchaochaohu 提交于 10月 27, 2020
  
  cdadc8f0
- Z
  add Fuse bn add act pass (#28196) · fdc06f21
  由 Zhang Ting 提交于 10月 27, 2020
```
* add fuse_bn_add_act pass
```
  fdc06f21
- C
  Enrich the python error types of paddle & polish format (#28124) · 813b2ade
  由 Chen Weihang 提交于 10月 27, 2020
```
* add multiple exception type

* define all exception & polish compile pystack

* mapping paddle error to python exception

* polish static mode error format

* fix failed unittests

* fix dytostatic test_error

* fix check_nan_inf failed

* add unittest for coverage

* revert some code try to solve compile error

* refactor enforce & error change

* polish code & add unittest
```
  813b2ade
26 10月, 2020 3 次提交
- A
  
  oneDNN BatchNorm + Act fusion pass. (#27912) · 7db747d9
  由 Adam Osewski 提交于 10月 26, 2020
  
  7db747d9
- Z
  
  fix print tensor place,add cpu/cuda/pin_memory API for Tensor (#28200) · fb7f8529
  由 Zhou Wei 提交于 10月 26, 2020
  
  fb7f8529
- M
  add sharding strategy in fleet(#27900) · 81244fbf
  由 mapingshuo 提交于 10月 26, 2020
```
* add sharding
```
  81244fbf
23 10月, 2020 2 次提交
- C
  Add compile limit for PADDLE_ENFORCE without error message (#28221) · 2babd6ff
  由 Chen Weihang 提交于 10月 23, 2020
```
* add compile limit for paddle enforce

* polish elementwise_op_function.cu.h

* fix failed unittest

* fix windows compile failed

* detail polish

* revert no type constructor
```
  2babd6ff
- L
  
  use FLAGS_use_mkldnn to prevent unnecessary attrs copy (#28146) · 4ea23307
  由 lidanqing 提交于 10月 23, 2020
  
  4ea23307
22 10月, 2020 4 次提交
- D
  
  fix wrong data type, test=develop (#28203) · 2db77be4
  由 Double_V 提交于 10月 22, 2020
  
  2db77be4
- F
  fix strided_slice_op's GetExpectedKernelType (#28192) · efe6e284
  由 Feiyu Chan 提交于 10月 22, 2020
```
* fix strided_slice_op's GetExpectedKernelType when input tensor is at CUDAPinnedPlace

* add unittest for tensors in cuda pinned place

* skip test for cuda pinned place on cpu machines
```
  efe6e284
- L
  Fix bug of fetch_async_op_handle when fetching the feed variable (#28194) · 1f3be859
  由 Leo Chen 提交于 10月 22, 2020
```
* fix bug of fetch_async_op_handle

* revert some changes of test_buffer_shared_memory_reuse_pass

* revert some changes of test_buffer_shared_memory_reuse_pass
```
  1f3be859
- W
  
  Fix nccl op test failed, test=develop (#28172) · e450823b
  由 WangXi 提交于 10月 22, 2020
  
  e450823b
21 10月, 2020 6 次提交
- W
  
  [lite-xpu-subgraph] Fix xpu compile and test xpu ci. (#27932) · f935ca8a
  由 Wilber 提交于 10月 21, 2020
  
  f935ca8a
- D
  dygraph nccl init support host domain name (#28107) · f29fb396
  由 danleifeng 提交于 10月 21, 2020
```
* nccl init support hostname and ip; test=develop
```
  f29fb396
- W
  
  support multiclass nms for multi-batch, test=develop (#28154) · 5cd97a1c
  由 wangguanzhong 提交于 10月 21, 2020
  
  5cd97a1c
- P
  
  change avg pooling from trt plugin to trt layer (#28032) · 602d2ce5
  由 Pei Yang 提交于 10月 21, 2020
  
  602d2ce5
- D
  
  fix Wmaybe-uninitialized warning in pooling.cc, test=develop (#28126) · 5289b72a
  由 Double_V 提交于 10月 21, 2020
  
  5289b72a
- Z
  
  fix dynamic_loader more safe and error message on windows (#28117) · 5d700021
  由 Zhou Wei 提交于 10月 21, 2020
  
  5d700021
20 10月, 2020 8 次提交
- W
  fix generate_proposal_labels in cascade-rcnn series model, test=develop (#27892) · d1e1f174
  由 wangguanzhong 提交于 10月 20, 2020
```
* fix generate_proposal_labels in cascade-rcnn series model, test=develop

* fix example code & unittest, test=develop

* update code from review comments, test=develop
```
  d1e1f174
- L
  fill_constant op supports NaN and Inf (#28109) · a911c19e
  由 Leo Chen 提交于 10月 20, 2020
```
* fill_constant supports nan and inf

* add ut
```
  a911c19e
- Z
  
  randperm run error in multi-gpus (#27942) · 6dd64b0a
  由 zhupengyang 提交于 10月 20, 2020
  
  6dd64b0a
- D
  add rois_num for roi_align xpu OP (#28077) · d43f75e4
  由 Double_V 提交于 10月 20, 2020
```
* add stack pool2d roi_align xpu op,test=kunlun

* error message opt, test=kunlun

* add xpu unittest,test=kunlun

* skip check grad,test=kunlun

* fix boostget , test=kunlun

* error message opt for XPU, test=kunlun

* add rois_num for roi_align xpu OP, test=develop
```
  d43f75e4
- X
  
  rm max_input in conv2d for kunlun, test=kunlun (#28062) · e3d02c95
  由 xiaoting 提交于 10月 20, 2020
  
  e3d02c95
- J
  Add AVX512 instruction check for C-API (#28087) · a21b5710
  由 joanna.wozna.intel 提交于 10月 20, 2020
```
* Add AVX512 instruction check for C-API

* Fix formatting
```
  a21b5710
- W
  
  refine gpu kernel config for Paddle (#28085) · 463c72c2
  由 wangchaochaohu 提交于 10月 20, 2020
  
  463c72c2
- Y
  lookup_table_v2_op_xpu report errors;test=kunlun (#28064) · 2cb1ecb9
  由 yinhaofeng 提交于 10月 20, 2020
```
* lookup_table_v2_op_xpu report errors;test=kunlun

* lookup_table_v2_op_xpu report errors;test=kunlun
```
  2cb1ecb9
19 10月, 2020 5 次提交
- Y
  xpu adam op (#28031) · 6f0c3d1f
  由 yinhaofeng 提交于 10月 19, 2020
```
* lookup_table_xpu op report errors;test=kunlun

* add adam xpu op;test=kunlun

* reset lookup

* change adam wrong;test=kunlun
```
  6f0c3d1f
- T
  
  Add xpu transpose2 op.test=kunlun (#28086) · a5c95cd5
  由 TeslaZhao 提交于 10月 19, 2020
  
  a5c95cd5
- C
  Fix xpu error message (#28061) · 5f04875c
  由 Chengmo 提交于 10月 19, 2020
```
* fix error message,test=kunlun

* fix, test=kunlun
```
  5f04875c
- L
  Fix diag OP bug on Windows Python3.8 · c8d32c8c
  由 LutaoChu 提交于 10月 19, 2020
```
Fix diag OP bug on Windows Python3.8 ，remove the std::min
```
  c8d32c8c
- P
  
  reduce trt warning message (#28011) · a0b2f936
  由 Pei Yang 提交于 10月 19, 2020
  
  a0b2f936

Crayon鑫 / Paddle 与 Fork 源项目一致

Crayon鑫 / Paddle
与 Fork 源项目一致