提交 · 87e4e25db176b4dced5fb2952a2e674e5b8b0db8 · 机器未来 / Paddle

19 9月, 2017 3 次提交
- Y
  Change Transform API · 87e4e25d
  由 Yu Yang 提交于 9月 18, 2017
```
Using DeviceContext, not Place to get stream
```
  87e4e25d
- Y
  
  Merge branch 'develop' of github.com:baidu/Paddle into feature/remove_lazy_init_in_dev_ctx · 847fe473
  由 Yu Yang 提交于 9月 18, 2017
  
  847fe473
- Y
  Remove lazy-initialization in device_context · 81d56ca8
  由 Yu Yang 提交于 9月 18, 2017
```
* Also use `const DeviceContext&` all the time, to prevent `const_cast`

Fix #4169
Fix #3468
Fix #3475
```
  81d56ca8
18 9月, 2017 9 次提交
- T
  Merge pull request #4154 from luotao1/avg_pool · b5a448f3
  由 Tao Luo 提交于 9月 18, 2017
```
refine avg-pooling, which is exclusive. refine related code.
```
  b5a448f3
- L
  
  remove unused CHECK in matrix.cpp · 780e2682
  由 Luo Tao 提交于 9月 18, 2017
  
  780e2682
- L
  
  set CudnnPooling don't include padded values in default · cc7a7557
  由 Luo Tao 提交于 9月 18, 2017
  
  cc7a7557
- L
  
  fix AvgPoolBackward and using CaffeMode in test_AvgPoolFwdBwd · 73da9720
  由 Luo Tao 提交于 9月 18, 2017
  
  73da9720
- H
  Merge pull request #4150 from hedaoyuan/mobile · d59295f2
  由 hedaoyuan 提交于 9月 18, 2017
```
Add a `neon::relu` activation function.
```
  d59295f2
- Y
  Split operator with CPU kernel (#4046) · 56b1b701
  由 Yancey 提交于 9月 18, 2017
```
Split Op CPU Kernel
```
  56b1b701
- Y
  Merge pull request #3927 from Xreki/core_add_fc_op · ec9a55ae
  由 Yiqun Liu 提交于 9月 18, 2017
```
Port fully connected operator
```
  ec9a55ae
- L
  
  refine avg-pooling, which is exclusive. refine related code. · 8067a42b
  由 Luo Tao 提交于 9月 18, 2017
  
  8067a42b
- 武
  Refine accuracy_op CUDA kernel (#4097) · 8580dce3
  由武毅提交于 9月 18, 2017
```
* refind accuracy_op

* follow comments

* follow comments
```
  8580dce3
17 9月, 2017 8 次提交
- Y
  block design (#3708) · 59c48f98
  由 Yan Chunwei 提交于 9月 17, 2017
```
* add block
```
  59c48f98
- C
  Merge pull request #4034 from peterzhang2029/my-paddle · 8c71e093
  由 Cao Ying 提交于 9月 17, 2017
```
refine doc in network.py.
```
  8c71e093
- Y
  Merge pull request #4141 from... · cd85be4a
  由 Yu Yang 提交于 9月 16, 2017
```
Merge pull request #4141 from PaddlePaddle/revert-4138-feature/do_not_invoke_gpu_method_when_use_gpu_off

Revert "Do not invoke GPU method when use_gpu=false"
```
  cd85be4a
- Y
  
  Revert "Do not invoke GPU method when use_gpu=false" · b012013a
  由 Yu Yang 提交于 9月 16, 2017
  
  b012013a
- P
  
  delete the unused comments · fc3b129b
  由 peterzhang2029 提交于 9月 17, 2017
  
  fc3b129b
- P
  
  Merge branch 'my-paddle' of https://github.com/peterzhang2029/Paddle into my-paddle · 1d4a028b
  由 peterzhang2029 提交于 9月 17, 2017
  
  1d4a028b
- P
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into my-paddle · e0d591c7
  由 peterzhang2029 提交于 9月 17, 2017
  
  e0d591c7
- P
  
  update notation in networks.py roughly · 5e2da7ac
  由 peterzhang2029 提交于 9月 15, 2017
  
  5e2da7ac
16 9月, 2017 2 次提交
- Y
  Merge pull request #4138 from reyoung/feature/do_not_invoke_gpu_method_when_use_gpu_off · 0b373f8f
  由 Yi Wang 提交于 9月 15, 2017
```
Do not invoke GPU method when use_gpu=false
```
  0b373f8f
- Y
  
  Do not invoke GPU method when use_gpu=false · 17e1aa8d
  由 Yu Yang 提交于 9月 15, 2017
  
  17e1aa8d
15 9月, 2017 18 次提交
- H
  
  Bug fix for add<float32x4_t>. · 7fe03f7b
  由 hedaoyuan 提交于 9月 15, 2017
  
  7fe03f7b
- H
  Merge pull request #4129 from hedaoyuan/convolution · 7bcb1fc3
  由 hedaoyuan 提交于 9月 15, 2017
```
Refine ExpandConvLayer
```
  7bcb1fc3
- P
  
  Merge branch 'my-paddle' of https://github.com/peterzhang2029/Paddle into my-paddle · 6f81fcc5
  由 peterzhang2029 提交于 9月 15, 2017
  
  6f81fcc5
- P
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into my-paddle · e7cc8639
  由 peterzhang2029 提交于 9月 15, 2017
  
  e7cc8639
- P
  
  update the version of pre-commit0.13.2 · bfba756b
  由 peterzhang2029 提交于 9月 15, 2017
  
  bfba756b
- H
  
  Refine the ExpandConvLayer. · b1bca066
  由 hedaoyuan 提交于 9月 15, 2017
  
  b1bca066
- Q
  Merge pull request #4130 from qingqing01/seq_op_test · 43ec735b
  由 qingqing01 提交于 9月 15, 2017
```
Enhance unit testing framework for the operators with LoDTensor.
```
  43ec735b
- D
  
  Add unit testing for sequence average pool operator. · 05680d0d
  由 dangqingqing 提交于 9月 15, 2017
  
  05680d0d
- D
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into seq_op_test · 90886443
  由 dangqingqing 提交于 9月 15, 2017
  
  90886443
- D
  
  Enhance unit testing framework for operator with LoDTensor. · d6a0280e
  由 dangqingqing 提交于 9月 15, 2017
  
  d6a0280e
- H
  
  Bug fix. · e4c340e4
  由 hedaoyuan 提交于 9月 15, 2017
  
  e4c340e4
- L
  
  Delete USE_OP statements and add more ENFORCE statements to check the inputs and outputs in FCOp. · dae249b1
  由 Liu Yiqun 提交于 9月 15, 2017
  
  dae249b1
- H
  
  Delete ExpandConvBaseLayer.cpp and ExpandConvBaseLayer.h · 88b31422
  由 hedaoyuan 提交于 9月 15, 2017
  
  88b31422
- P
  
  Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into my-paddle · 41973592
  由 peterzhang2029 提交于 9月 15, 2017
  
  41973592
- P
  
  update notation in networks.py roughly · 9e74b898
  由 peterzhang2029 提交于 9月 15, 2017
  
  9e74b898
- H
  
  Remove ExpandConvBaseLayer. · 3616ef77
  由 hedaoyuan 提交于 9月 15, 2017
  
  3616ef77
- L
  
  Merge branch 'develop' into core_add_fc_op · 6ce4bf36
  由 Liu Yiqun 提交于 9月 15, 2017
  
  6ce4bf36
- Y
  Merge pull request #4126 from Xreki/core_fix_segfault · bc55c20f
  由 Yiqun Liu 提交于 9月 15, 2017
```
Add the check of inputs and outputs in all operators, to avoid SegFault.
```
  bc55c20f

机器未来 / Paddle 与 Fork 源项目一致

机器未来 / Paddle
与 Fork 源项目一致