提交 · 3dc29b390537cca68f43f21f44c2c2fde84fa297 · PaddlePaddle / PaddleDetection

13 12月, 2018 2 次提交
- Q
  
  change sparse_update to adam_update · 3dc29b39
  由 Qiao Longfei 提交于 12月 13, 2018
  
  3dc29b39
- Q
  
  add sparse mode adam · fc6ec6bd
  由 Qiao Longfei 提交于 12月 13, 2018
  
  fc6ec6bd
12 12月, 2018 9 次提交
- T
  Merge pull request #14732 from Sand3r-/mgallus/mkldnn-concat · 66b6e473
  由 Tao Luo 提交于 12月 12, 2018
```
[MKL-DNN] Concat Layer
```
  66b6e473
- Y
  Fix the gelu backward to avoid nan (#14857) · 6951ef9a
  由 Yibing Liu 提交于 12月 12, 2018
```
* Fix the gelu backward to avoid nan

test=develop

* Remove unnecessary calls

test=develop
```
  6951ef9a
- T
  Merge pull request #14825 from NHZlX/add_benchmark_for_trt · 322bb8d5
  由 Tao Luo 提交于 12月 12, 2018
```
an sample of recording benchmark to file for trt
```
  322bb8d5
- Q
  Merge pull request #14844 from jacquesqiao/pserver-should-crash · 3668f079
  由 Qiao Longfei 提交于 12月 12, 2018
```
pserver should crash early whe has problem
```
  3668f079
- F
  Merge pull request #14824 from Superjomn/fix/visualizer · dd3aca3b
  由 flame 提交于 12月 12, 2018
```
fix visualizer
```
  dd3aca3b
- T
  Merge pull request #14807 from baojun-nervana/ngraph_follow_up · d846fb03
  由 tensor-tang 提交于 12月 12, 2018
```
Follow ups on nGraph PR comments
```
  d846fb03
- Y
  Merge pull request #14814 from reyoung/feature/gprof · 8175983e
  由 Yu Yang 提交于 12月 12, 2018
```
Add gperftools supports for PE
```
  8175983e
- Y
  Merge pull request #14842 from reyoung/feature/refine_eigen_tensor · eeca721a
  由 Yu Yang 提交于 12月 12, 2018
```
Fix Eigen macro when using GPU
```
  eeca721a
- Y
  Fix the issue to run on AVX2 and AVX512F machines (#14851) · acc6ae49
  由 Yihua Xu 提交于 12月 12, 2018
```
test=develop
```
  acc6ae49
11 12月, 2018 11 次提交
- N
  Merge branch 'develop' of https://github.com/paddlepaddle/paddle into add_benchmark_for_trt · 8c8019e3
  由 nhzlx 提交于 12月 11, 2018
```
test=develop
```
  8c8019e3
- M
  MKL-DNN Concat: Fix segfault related to referencing deleter memory primitive · 92daace5
  由 Michal Gallus 提交于 12月 10, 2018
```
test=develop
```
  92daace5
- Y
  Revert data_type · 9f8d8685
  由 Yu Yang 提交于 12月 11, 2018
```
test=develop
```
  9f8d8685
- X
  Merge pull request #14456 from tink2123/fix_weightaverage · 45fb357b
  由 Xin Pan 提交于 12月 11, 2018
```
fix the wrong format
```
  45fb357b
- Y
  Fix compile error · 5e609069
  由 Yu Yang 提交于 12月 11, 2018
```
test=develop
```
  5e609069
- Q
  pserver should crash early whe has problem · 1870262b
  由 Qiao Longfei 提交于 12月 11, 2018
```
test=develop
```
  1870262b
- Y
  Fix Eigen macro when using GPU · 7604b1ad
  由 Yu Yang 提交于 12月 11, 2018
```
The macro should be defined by compiler rather than by source.

test=develop
```
  7604b1ad
- Q
  Merge pull request #14820 from jacquesqiao/fix-split-selected-rows · 1213e283
  由 Qiao Longfei 提交于 12月 11, 2018
```
split selected rows op should always init output selected rows
```
  1213e283
- N
  
  fix compile error · 644c13a3
  由 nhzlx 提交于 12月 11, 2018
  
  644c13a3
- Q
  test zero output of split_selected_rows_op · 0a7c7e97
  由 Qiao Longfei 提交于 12月 11, 2018
```
test=develop
```
  0a7c7e97
- T
  Merge pull request #14651 from luotao1/has_attr · 3243b45b
  由 Tao Luo 提交于 12月 11, 2018
```
add Opdesc's HasProtoAttr
```
  3243b45b
10 12月, 2018 18 次提交
- D
  fix memory opt skip set by name (#14774) · 00776b16
  由 dzhwinter 提交于 12月 10, 2018
```
* random failed. rerun ci. test=develop

* windows failed. rerun ci. test=develop
```
  00776b16
- Z
  Merge pull request #14771 from frankwhzhang/bpr · c4c5f0b8
  由 zhang wenhui 提交于 12月 10, 2018
```
add bpr_loss operator
```
  c4c5f0b8
- T
  add HasProtoAttr function in op_desc.h, clean node.h · 067ed70f
  由 Tao Luo 提交于 12月 10, 2018
```
test=develop
```
  067ed70f
- X
  Merge pull request #14823 from panyx0718/fix · ed9cdb56
  由 Xin Pan 提交于 12月 10, 2018
```
fix control_flow ops in outs
```
  ed9cdb56
- Y
  Speed up SizeOfType · b22d638d
  由 Yu Yang 提交于 12月 10, 2018
```
test=develop
```
  b22d638d
- X
  Merge pull request #14811 from panyx0718/revert-14782-revert-14398-imperative · 4a716a70
  由 Xin Pan 提交于 12月 10, 2018
```
Imperative
```
  4a716a70
- N
  Merge branch 'develop' of https://github.com/paddlepaddle/paddle into add_benchmark_for_trt · a5bfed37
  由 nhzlx 提交于 12月 10, 2018
```
test=develop
```
  a5bfed37
- N
  
  add benchmark for trt · afc51e6f
  由 nhzlx 提交于 12月 10, 2018
  
  afc51e6f
- Z
  Merge pull request #14762 from NHZlX/fix_bug_of_trt_pool · bc6d0a34
  由 Zhaolong Xing 提交于 12月 10, 2018
```
fix bug of trt pool2d converter
```
  bc6d0a34
- C
  Merge pull request #14819 from PaddlePaddle/shanyi15-patch-1 · a71b0132
  由 Cheerego 提交于 12月 10, 2018
```
update README to 1.2
```
  a71b0132
- W
  add more log for dist test for ci test=develop (#14813) · 554bcdbd
  由 Wu Yi 提交于 12月 10, 2018
```
* add more log for dist test for ci test=develop

* increase deadline test=develop
```
  554bcdbd
- S
  fix visualizer · edd1f5a9
  由 superjomn 提交于 12月 10, 2018
```
test=develop
```
  edd1f5a9
- X
  fix control_flow ops in outs · f5434507
  由 Xin Pan 提交于 12月 10, 2018
```
test=develop
```
  f5434507
- B
  Remove unnecessary MKLDNN reorder (#14799) · 9623b45f
  由 Brian Liu 提交于 12月 10, 2018
```
When data flow from a MKLDNN OP kernel to a non-MKLDNN OP kernel,
data layout transform (via MKLDNN reorder) will occur even when
those two OP kernels share same layout. Add code to remove this
unnecessary reorder.

test=develop
```
  9623b45f
- F
  
  fix 'name', test=develop · 90c7f987
  由 frankwhzhang 提交于 12月 10, 2018
  
  90c7f987
- Q
  split selected rows op should always init output selected rows · abf14028
  由 Qiao Longfei 提交于 12月 10, 2018
```
test=develop
```
  abf14028
- N
  
  fix comments test=develop · 019e8bbe
  由 nhzlx 提交于 12月 10, 2018
  
  019e8bbe
- Z
  Merge pull request #14817 from sneaxiy/add_cuda_cudnn_version_checkout · cc49a8be
  由 Zeng Jinle 提交于 12月 10, 2018
```
Add cuda cudnn version check
```
  cc49a8be

PaddlePaddle / PaddleDetection 大约 1 年 前同步成功

PaddlePaddle / PaddleDetection
大约 1 年前同步成功