提交 · f404d53ba519059ae4a3a6f9e2c72faade855188 · PaddlePaddle / Paddle

13 3月, 2019 11 次提交
- T
  
  Api approvals (#16179) · f404d53b
  由 tianshuo78520a 提交于 3月 13, 2019
  
  f404d53b
- L
  Merge pull request #16183 from tianshuo78520a/fix_api.spec · daa73771
  由 lujun 提交于 3月 13, 2019
```
Fix api.spec
```
  daa73771
- G
  resolve #15618 (#16114) · decdbed0
  由 guomingz 提交于 3月 13, 2019
```
* resolve #15618
Backgroud: the PR #15398 raised the box_coder op performance regression, we optimized the code via the more efficency leveraging opemmp.
```
  decdbed0
- T
  
  fix paddle.reader.ComposeNotAligned.__init__ · c130d323
  由 tianshuo78520a 提交于 3月 13, 2019
  
  c130d323
- T
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into develop · fcc0bf84
  由 tianshuo78520a 提交于 3月 13, 2019
  
  fcc0bf84
- K
  Merge pull request #16140 from tink2123/arc_function · 1a4a90a8
  由 Kaipeng Deng 提交于 3月 13, 2019
```
Add the inverse trigonometric function
```
  1a4a90a8
- Y
  fix broadcast on mp mode (#15951) · 30568473
  由 Yan Xu 提交于 3月 13, 2019
```
* fix broadcast with mp mode

* polish code test=develop

* fix bcast strategy test=develop

* fic cpplint test=develop

* fix py3 failed test=develop

* fix comment test=develop

* update comment test=develop
```
  30568473
- B
  remove const_cast and refactor ngraph engine code (#15925) · e3c37bd5
  由 baojun 提交于 3月 13, 2019
```
* remove concast_cast and refactor code test=develop

* reduce flag use test=develop
```
  e3c37bd5
- C
  Add memory profiler (#16137) · 09799566
  由 chengduo 提交于 3月 12, 2019
```
test=develop
```
  09799566
- T
  modified api.spec · 61a6165c
  由 tink2123 提交于 3月 13, 2019
```
test=develop
```
  61a6165c
- Y
  
  delete unbind api test=develop (#15778) · 05993c3f
  由 Yan Xu 提交于 3月 13, 2019
  
  05993c3f
12 3月, 2019 17 次提交
- Z
  Merge pull request #16162 from wzzju/fix_nan_static_quant · 41b8cf0b
  由 Zhen Wang 提交于 3月 12, 2019
```
Fix NaN bugs for static quantization strategy (mutil-cards train). 
```
  41b8cf0b
- T
  
  Diff print (#16155) · ddfc823c
  由 tianshuo78520a 提交于 3月 12, 2019
  
  ddfc823c
- Q
  Merge pull request #16095 from velconia/transfer_gru_unit · d4e8c99f
  由 Qiyang Min 提交于 3月 12, 2019
```
Imperative transfer gru unit
```
  d4e8c99f
- Z
  Merge pull request #16107 from wzzju/add_graph_clone · 94b7c1ea
  由 Zhen Wang 提交于 3月 12, 2019
```
Add clone function for IrGraph.
```
  94b7c1ea
- W
  restore the exception caught since it is necessary for python call stack (#16160) · 85709f43
  由 wopeizl 提交于 3月 12, 2019
```
test=develop
```
  85709f43
- Z
  Merge pull request #16070 from wzzju/channel_wise_quant_op · 5420cf95
  由 Zhen Wang 提交于 3月 12, 2019
```
Add channel wise quant op and channel wise dequant op.
```
  5420cf95
- Z
  
  Add some fixme. test=develop · 5685a48c
  由 Zhen Wang 提交于 3月 12, 2019
  
  5685a48c
- T
  modified api.spec · eb09bd45
  由 tink2123 提交于 3月 12, 2019
```
test=develop
```
  eb09bd45
- Z
  
  Add the executor test for the graph clone API. test=develop · b8d1f503
  由 Zhen Wang 提交于 3月 11, 2019
  
  b8d1f503
- T
  refine doc · a8e375d4
  由 tink2123 提交于 3月 12, 2019
```
test=develop
```
  a8e375d4
- Z
  
  fix NaN bugs. · 4ab06076
  由 Zhen Wang 提交于 3月 12, 2019
  
  4ab06076
- Q
  Imperative fix conv2d's attrs bugs (#16158) · 4e8c03bd
  由 Qiyang Min 提交于 3月 12, 2019
```
* Remove Desc in Forward Pass

* Refactor VarBase

* Add dbg info

* Only check type in imperative mode

* Polish code and support optimizer

test=develop

* Fix stop gradient problem in PyLayer

test=develop

* Fix conv2d's attrs

test=develop
```
  4e8c03bd
- Z
  
  Add the Clone method in Graph. test=develop · ac6ef06f
  由 Zhen Wang 提交于 3月 07, 2019
  
  ac6ef06f
- Z
  
  Not add graph copy construction method. test=develop · 01eddf12
  由 Zhen Wang 提交于 3月 07, 2019
  
  01eddf12
- Z
  
  add clone function for IrGraph. test=develop · 1b9c8d5f
  由 Zhen Wang 提交于 3月 07, 2019
  
  1b9c8d5f
- J
  Merge pull request #16145 from ceci3/npair_loss0 · 08e75731
  由 jerrywgz 提交于 3月 12, 2019
```
fix npair loss
```
  08e75731
- T
  Merge pull request #16104 from tensor-tang/refine/jit · ccc7c358
  由 Tao Luo 提交于 3月 12, 2019
```
refine jitkernels and test
```
  ccc7c358
11 3月, 2019 12 次提交
- T
  Merge pull request #16120 from Xreki/fix_cmake_compress · c49b7855
  由 Tao Luo 提交于 3月 11, 2019
```
Change the download and compress command of cmake.
```
  c49b7855
- M
  Polish code · de212ae2
  由 minqiyang 提交于 3月 11, 2019
```
test=develop
```
  de212ae2
- Q
  Imperative remove all descs (#16045) · 1f4aa7a2
  由 Qiyang Min 提交于 3月 11, 2019
```
* Remove Desc in Forward Pass

* Refactor VarBase

* Add dbg info

* Only check type in imperative mode

* Polish code and support optimizer

test=develop

* Fix stop gradient problem in PyLayer

test=develop
```
  1f4aa7a2
- Z
  Merge pull request #16063 from sneaxiy/enhance_gc · 472f16b5
  由 Zeng Jinle 提交于 3月 11, 2019
```
Enhance gc
```
  472f16b5
- T
  Merge pull request #16146 from luotao1/zero_copy · e31f6e98
  由 Tao Luo 提交于 3月 11, 2019
```
 unify ZeroCopy in analysis_test
```
  e31f6e98
- T
  enhance jitkernel unit test · 14d87112
  由 tensor-tang 提交于 3月 11, 2019
```
test=develop
```
  14d87112
- M
  Add unit test for gru unit · d17bb4e6
  由 minqiyang 提交于 3月 11, 2019
```
test=develop
```
  d17bb4e6
- L
  Disable inference download for WIN32 temporary. · 4e052e0a
  由 Liu Yiqun 提交于 3月 11, 2019
```
test=develop
```
  4e052e0a
- C
  Revert "Revert "Add Event for TensorCopy"" (#16035) · ad80bde8
  由 chengduo 提交于 3月 11, 2019
```
* Revert "Revert "Add Event for TensorCopy" (#16022)"

This reverts commit e2da3a5b.

* use default stream
test=develop
```
  ad80bde8
- M
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into transfer_gru_unit · 0d27d204
  由 minqiyang 提交于 3月 11, 2019
  
  0d27d204
- Q
  Merge pull request #16125 from panyx0718/imperative · 5f33164f
  由 Qiyang Min 提交于 3月 11, 2019
```
add imperative and declarative mode testbase and example
```
  5f33164f
- M
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into transfer_gru_unit · 2050f31c
  由 minqiyang 提交于 3月 11, 2019
```
test=develop
```
  2050f31c

PaddlePaddle / Paddle 大约 1 年 前同步成功

PaddlePaddle / Paddle
大约 1 年前同步成功