提交 · 8e2fdc54b11e5471cbbee41f826abe3f9e75894f · PaddlePaddle / Paddle

15 10月, 2018 2 次提交

Add check for opt op (#13840) · 8e2fdc54

由 chengduo 提交于 10月 15, 2018

* add check for opt op

* fix opt op
test=develop

* fix test fail
test=develop

* fix optimization doc
test=develop

* test=develop

8e2fdc54

Add strategy doc (#13849) · 2562eb92

由 chengduo 提交于 10月 15, 2018

* add strategy doc
test=develop

* fix doc
test=develop

* add ParallelExecutor arg doc
test=develop

2562eb92

13 10月, 2018 1 次提交
- X
  
  Revert "Make variable::GetMutable robust" · d852be7c
  由 Xin Pan 提交于 10月 13, 2018
  
  d852be7c
12 10月, 2018 8 次提交

Fix layers.uniform_random (#13823) · 9c77b65c

由 chengduo 提交于 10月 12, 2018

* fix layers.uniform_random

* fix uniform_random
test=develop

* remove var type set
test=develop

* fix similar error
test=develop

9c77b65c

N
fix commets · 320c78e1
由 nhzlx 提交于 10月 12, 2018
```
test=develop
```
320c78e1
N
fix demo_ci bug in vis_demo.cc · efa5bac7
由 nhzlx 提交于 10月 12, 2018
```
test=develop
```
efa5bac7

Avoid GetMutable implicitly reset Var Type. · 22850661

由 Xin Pan 提交于 10月 12, 2018

This can cause a lot of problem:
1. Wrong operator implementation, Op can get a wrong type without failure.
2. Anytype can be Get without defined in VarType.

Also fix wrong STEP_SCOPE usage.

test=develop

22850661

D
optimize depthwise conv by register memory (#13778) · 5f2e8378
由 Dun 提交于 10月 12, 2018
```
* optimize depthwise conv by register memory
* test=develop
```
5f2e8378
M
Polish code · 3f6ec900
由 minqiyang 提交于 10月 12, 2018
```
test=develop
```
3f6ec900

Profiler support merge data of all thread (#13811) · 5428cb99

由 Qiao Longfei 提交于 10月 12, 2018

* profiler infor merge thread statistic information

* update profiler

* fix bug

* add merge thread msg to report

* optimize report

* statistic the time of ops in each thread but not all

* optimize report format

* optimize profile report

* optimize profile report
test=develop

5428cb99

N

add op converter · 0cb88c34
由 nhzlx 提交于 10月 11, 2018

0cb88c34

11 10月, 2018 4 次提交
- M
  Accelerate SelectedRows Functors: · 8ec748cf
  由 minqiyang 提交于 10月 11, 2018
```
  1. Accelerate SelectedRows MergeAdd functor

  2. Add SelectedRowsSumTo functor to support MergeAdd multiple SelectedRows into one

test=develop
```
  8ec748cf
- X
  Explain LoD and a few other concepts · 63b2e98f
  由 Xin Pan 提交于 10月 10, 2018
```
test=develop
```
  63b2e98f
- T
  
  Revert "[MKLDNN] Pass: Fuse Conv + Bias" · 9b11a175
  由 Tao Luo 提交于 10月 11, 2018
  
  9b11a175
- W
  Fix truncated norm (#13785) · 7e651c86
  由 whs 提交于 10月 11, 2018
```
* Fix truncated normal.

* test=develop
```
  7e651c86
10 10月, 2018 6 次提交
- M
  Pass: Fuse Conv + Bias · 40b17be4
  由 Michal Gallus 提交于 10月 01, 2018
```
test=develop
```
  40b17be4
- M
  Add unittest for clip_by_norm_op with SelectedRows · 1456b8ec
  由 minqiyang 提交于 10月 10, 2018
```
test=develop
```
  1456b8ec
- S
  eigen sqrt fix and change 1e-5 to epsilon · 3fcca409
  由 Sylwester Fraczek 提交于 10月 09, 2018
```
test=develop
```
  3fcca409
- C
  Add doc (#13765) · e1904ac2
  由 chengduo 提交于 10月 10, 2018
```
test=develop
```
  e1904ac2
- N
  
  fix comments · d347ea68
  由 nhzlx 提交于 10月 10, 2018
  
  d347ea68
- C
  Set the right shape of selected_rows (#13723) · e1761709
  由 chengduo 提交于 10月 10, 2018
```
* set the right shape of selected_rows
test=develop

* enhance check

* fix activation_op

* remove cast

* use ShareDimInfo replace SetDim and ShareLod

* use ShareDimAndLod
test=develop

* follow comment

test=develop

* check whether the input has lod
test=develop

* Split ShareDimAndLod

test=develop

* checkout clip.py
test=develop
```
  e1761709
09 10月, 2018 6 次提交
- Q
  Revert "optimize pyreader" · 9d087d51
  由 Qiao Longfei 提交于 10月 09, 2018
```
test=develop
```
  9d087d51
- N
  
  add tensorrt api lib to paddle_fluid · f5690950
  由 nhzlx 提交于 10月 09, 2018
  
  f5690950
- Q
  Fix bug in reduce_op caused by PR #13534 (#13748) · 6094a723
  由 qingqing01 提交于 10月 09, 2018
```
* Fix bug in reduce_op caused by PR #13534
* Fix output shape and enhance unit test.
test=develop
```
  6094a723
- T
  revert with_fast_math to ON · 84a55155
  由 Tao Luo 提交于 10月 09, 2018
```
test=develop
```
  84a55155
- W
  
  fix manylinux multi arch docker build test=develop (#13770) · cc36bab1
  由 Wu Yi 提交于 10月 09, 2018
  
  cc36bab1
- D
  enhance isinf/isnan in tensor util, avoid copy back to cpu (#12688) · a46e30aa
  由 dzhwinter 提交于 10月 09, 2018
```
* "avoid copy back to cpu"

* "add infinity support"

* "fix ci"

* "add cpu macro"

* rerun ci; test=develop

* "fix api"
test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop
```
  a46e30aa
08 10月, 2018 13 次提交
- J
  - Cleanup as suggessted by reviewers · 9f15d881
  由 Jacek Czaja 提交于 10月 08, 2018
```
test=develop
```
  9f15d881
- W
  fix cuda9 docker build test=develop (#13701) · 25262ed0
  由 Wu Yi 提交于 10月 08, 2018
```
* fix cuda9 docker build test=develop

* update test=develop

* update test
```
  25262ed0
- S
  conv bn fuse pass · 78f98294
  由 Sylwester Fraczek 提交于 9月 19, 2018
```
review fix

review from hshen14 fix

test=develop

fix error in broadcast and code cleanup

rename bias -> eltwise and added macro to shorten code

formatting
```
  78f98294
- J
  - Cleaning fused_embedding_fc_lstm op · ae8b4717
  由 Jacek Czaja 提交于 10月 01, 2018
```
test=develop
```
  ae8b4717
- J
  - Removed disabled code · fd31b54c
  由 Jacek Czaja 提交于 10月 01, 2018
```
test=develop
```
  fd31b54c
- J
  - Removed disabled diagnostic code · f9da2d64
  由 Jacek Czaja 提交于 10月 01, 2018
```
test=develop
```
  f9da2d64
- J
  - Added file for fused_embedded_fc_lstm_op unit test · 809dbc5c
  由 Jacek Czaja 提交于 9月 24, 2018
```
- Work in progress on unit test for fused_embedding_fc_lstm op

- Added bias caching and ref x computing

- Small update unit test

- temporary storage

- Fix to batchcompute

- Cosmetic fixes

- Style fixes
```
  809dbc5c
- Q
  
  revert unused change test=develop · 92aff806
  由 qiaolongfei 提交于 10月 08, 2018
  
  92aff806
- Q
  
  optimize fake, change it to a class instead a function. test=develop · ec38effc
  由 qiaolongfei 提交于 10月 08, 2018
  
  ec38effc
- Q
  
  Optimize Topk when height is large. (#13710) · 41e4f7ea
  由 qingqing01 提交于 10月 08, 2018
  
  41e4f7ea
- M
  
  Add unit test · bcd8c2cc
  由 minqiyang 提交于 10月 08, 2018
  
  bcd8c2cc
- T
  disable EIGEN_FAST_MATH and use_fast_math · 28889cae
  由 Tao Luo 提交于 10月 08, 2018
```
test=develop
```
  28889cae
- X
  clean unused code and small optimize · d2079b1d
  由 Xin Pan 提交于 10月 08, 2018
```
test=develop
```
  d2079b1d

PaddlePaddle / Paddle 大约 1 年 前同步成功

PaddlePaddle / Paddle
大约 1 年前同步成功