提交 · 430e5590e44dfeddd9064190c1bf238d344c6b82 · csdn_franckjun / Paddle

31 1月, 2019 12 次提交
- G
  Merge pull request #15611 from guoshengCS/fix-beam-search-size · 430e5590
  由 Guo Sheng 提交于 1月 31, 2019
```
To make CUDA_LAUNCH_KERNEL_HELPER support large size.
```
  430e5590
- G
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into fix-beam-search-size · b6c3b69a
  由 guoshengCS 提交于 1月 31, 2019
```
test=develop
```
  b6c3b69a
- D
  
  fix batch norm. test=develop (#15597) · 46a6cac9
  由 dzhwinter 提交于 1月 31, 2019
  
  46a6cac9
- Q
  
  Fix analysis predictor when loading the persistable RAW type variable. (#15613) · 943d9728
  由 qingqing01 提交于 1月 31, 2019
  
  943d9728
- L
  expose peak gpu memory API to python test=develop (#15529) · 6e84eb13
  由 liuwei1031 提交于 1月 31, 2019
```
* expose peak gpu memory API to python test=develop

* add unittest for peak gpu memory monitoring test=develop

* add pybind change test=develop

* add mutex to gpu mem usage monitor test=develop

* update benchmark flag definition file test=develop

* tweak unittest for memory monitoring test=develop
```
  6e84eb13
- J
  Merge pull request #15356 from jerrywgz/add_clip_op · 1743d1a5
  由 jerrywgz 提交于 1月 31, 2019
```
Add box clip op
```
  1743d1a5
- M
  Enable conv2d operator for a ngraph engine (#15269) · 43a67a26
  由 mozga-intel 提交于 1月 31, 2019
```
test=develop
```
  43a67a26
- T
  Merge pull request #15586 from tensor-tang/jit/cache · a6a1a92e
  由 tensor-tang 提交于 1月 31, 2019
```
refine bert
```
  a6a1a92e
- J
  
  test=develop · 4f18a9b8
  由 jerrywgz 提交于 1月 31, 2019
  
  4f18a9b8
- G
  To make CUDA_LAUNCH_KERNEL_HELPER support large size. · 5dfce931
  由 guoshengCS 提交于 1月 31, 2019
```
test=develop
```
  5dfce931
- Y
  
  fix ir debug config (#15571) · e887d719
  由 Yan Chunwei 提交于 1月 31, 2019
  
  e887d719
- Y
  
  fix save_inferece_model bug (#15365) · 897789b1
  由 Yan Chunwei 提交于 1月 31, 2019
  
  897789b1
30 1月, 2019 20 次提交
- T
  use mat attr and refine test (#15448) · ba02ac46
  由 tensor-tang 提交于 1月 30, 2019
```
* use mat attr and refine test

test=develop

* add matmul jitcode

test=develop

* fix mac compile

test=develop
```
  ba02ac46
- J
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into add_clip_op · 1963af78
  由 jerrywgz 提交于 1月 30, 2019
  
  1963af78
- H
  Add INT8 calibration README (#15548) · b5ebca47
  由 Haihao Shen 提交于 1月 30, 2019
```
* Add calibration README; test=develop
```
  b5ebca47
- J
  
  Merge branch 'develop' into add_clip_op · 4eb44380
  由 jerrywgz 提交于 1月 30, 2019
  
  4eb44380
- X
  Merge pull request #15554 from heavengate/yolo_loss_darknet · 30cc8b7a
  由 Xin Pan 提交于 1月 30, 2019
```
Yolo loss darknet
```
  30cc8b7a
- T
  Merge pull request #15587 from luotao1/bert · 1a252f4b
  由 Tao Luo 提交于 1月 30, 2019
```
use embedding=128 bert model for test
```
  1a252f4b
- H
  Add INT8 calibration support in Paddle package (#15569) · 1b8047b7
  由 Haihao Shen 提交于 1月 30, 2019
```
* Add INT8 calibration support in Paddle package; test=develop
```
  1b8047b7
- M
  Enable pool2d operator for a ngraph engine (#15395) · 312500dc
  由 mozga-intel 提交于 1月 30, 2019
```
* Enable pool2d operator for a ngraph engine
test=develop

* Update
test=develop
```
  312500dc
- J
  Merge pull request #15575 from JiabinYang/feature/imperative · b4c24f3f
  由 Jiabin Yang 提交于 1月 30, 2019
```
test=develop, polish code and fix some wrong change
```
  b4c24f3f
- T
  transpiler.py code clean (#15555) · 90df7ff3
  由 tangwei12 提交于 1月 30, 2019
```
* move var strusted to vars_distributed.py, add optimizer's block name, test=develop

* rename optimzier's seems complex, revert it, test=develop

* replace * with details, test=develop
```
  90df7ff3
- H
  Enable performance measurement in INT8 calibration unit test (#15560) · 294d5944
  由 Haihao Shen 提交于 1月 30, 2019
```
* Enable performance measurement in INT8 calibration unit test
```
  294d5944
- C
  fix threshold_relu_op (#15594) · c4b9eac1
  由 chengduo 提交于 1月 29, 2019
```
test=develop
```
  c4b9eac1
- T
  Merge pull request #15478 from kbinias/kbinias/seperate-folders-for-mkldnn · ea92905b
  由 Tao Luo 提交于 1月 30, 2019
```
Make separate folders for mkldnn codes
```
  ea92905b
- Y
  Some improvements to support bert mixed precision training (#15585) · 170842cb
  由 Yibing Liu 提交于 1月 30, 2019
```
* Some improvements to support bert mixed precision training

test=develop

* Revert the cast in layer_norm

test=develop
```
  170842cb
- Y
  Return parent_idx in beam_search op (#15520) · 16d54f7f
  由 Yiqun Liu 提交于 1月 30, 2019
```
* Refine beam_search_op to output an extra parent_idx tensor.
test=develop

* Fix the unittest test_beam_search_op.
test=develop

* Fix the merging mistake.
test=develop
```
  16d54f7f
- J
  Merge pull request #15398 from jerrywgz/add_axis_for_boxcoder · 72ee3c62
  由 jerrywgz 提交于 1月 30, 2019
```
Add axis for boxcoder
```
  72ee3c62
- J
  
  test=develop · e402c0ec
  由 jerrywgz 提交于 1月 30, 2019
  
  e402c0ec
- J
  
  test=develop · 334f697d
  由 jerrywgz 提交于 1月 30, 2019
  
  334f697d
- K
  Merge pull request #15491 from tink2123/new_align_corners · d3eeb92b
  由 Kaipeng Deng 提交于 1月 30, 2019
```
add align_corners and align_mode for image_resize
```
  d3eeb92b
- J
  
  Merge branch 'develop' into add_clip_op · 3046799e
  由 jerrywgz 提交于 1月 30, 2019
  
  3046799e
29 1月, 2019 8 次提交
- J
  
  test=develop, fix debug mode unitest, hsigmoid (#15574) · 2d0ffdc4
  由 Jiabin Yang 提交于 1月 29, 2019
  
  2d0ffdc4
- Z
  Merge pull request #15546 from NHZlX/fix_trt_utest_random_failed · 90ffe749
  由 Zhaolong Xing 提交于 1月 29, 2019
```
fix trt models utest failed.
```
  90ffe749
- L
  use embedding=128 bert model for test · 8f0c2b07
  由 luotao1 提交于 1月 29, 2019
```
test=develop
```
  8f0c2b07
- T
  refine vadd jitkernel choice · 2b0811c3
  由 tensor-tang 提交于 1月 29, 2019
```
test=develop
```
  2b0811c3
- T
  cache fc kernel · a18c0d42
  由 tensor-tang 提交于 1月 29, 2019
```
test=develop
```
  a18c0d42
- T
  cache softmax kernel func · 6e1ee7fb
  由 tensor-tang 提交于 1月 29, 2019
```
test=develop
```
  6e1ee7fb
- K
  Small fix · 69b7c595
  由 Krzysztof Binias 提交于 1月 29, 2019
```
test=develop
```
  69b7c595
- K
  Make separate folders for mkldnn codes · b1bdcd4d
  由 Krzysztof Binias 提交于 1月 28, 2019
```
test=develop
```
  b1bdcd4d

csdn_franckjun / Paddle 与 Fork 源项目一致

csdn_franckjun / Paddle
与 Fork 源项目一致