提交 · 7c2affaa26cd3ed73915127243b067bb3a7958b4 · 机器未来 / Paddle

18 12月, 2020 3 次提交
- S
  fix isfinite_v2_op OpProtoAndCheckerMaker AddComment bug (#29626) · 7c2affaa
  由 syyxsxx 提交于 12月 18, 2020
```
fix isfinite_v2_op OpProtoAndCheckerMaker AddComment bug 
```
  7c2affaa
- H
  Reduce batch size ot fix CPU memory, test=develop (#29736) · 2e788bd8
  由 Huihuang Zheng 提交于 12月 18, 2020
```
Unit test reported memory not enough on CPU machines. Reduce batch size again.
```
  2e788bd8
- 石
  
  update the operator registration for incompatible upgrade, test=develop (#29720) · 8bd2879e
  由石晓伟提交于 12月 18, 2020
  
  8bd2879e
17 12月, 2020 13 次提交
- L
  Update en docs of to_tensor (#29718) · 10edfb6f
  由 LielinJiang 提交于 12月 17, 2020
```
* update to_tensor en docs
```
  10edfb6f
- C
  add conj op for complex types (#29527) · 71063b81
  由 chentianyu03 提交于 12月 17, 2020
```
* add conj op for complex types

* add conj for complex types

* add more test case

* add conj_op test

* modify conj api and impl

* add complex type for fill_constant_op xpu

* add setConstant for complex type

* remove complex conj test file

* user define grad for test_conj_op

* add test case for static mode of conj api

* modify conj doc

* change input args name to x

* remove useless codes

* conj support real types

* add conj test case for real number
```
  71063b81
- W
  [Inference] EnableUseGpu has higher priority than flags (#29697) · b593d588
  由 Wilber 提交于 12月 17, 2020
```
* enable_use_gpu has higher priority than FLAGS

* update.
```
  b593d588
- W
  
  fleet sync build strategy, test=develop (#29732) · 9cbcc6ca
  由 WangXi 提交于 12月 17, 2020
  
  9cbcc6ca
- T
  
  fix ubuntu docker error (#29719) · 638ccaab
  由 tianshuo78520a 提交于 12月 17, 2020
  
  638ccaab
- W
  Windows generate pdb and dump, for debug (#29628) · 0c59ad2a
  由 wanghuancoder 提交于 12月 17, 2020
```
* Windows generate pdb and dump, for debug

* fix code style, test=develop

* modify cmakelist
```
  0c59ad2a
- H
  Modify CublasHandleHolder to Fix Random Unittest Failure. test=develop (#29617) · 4c4d4ba5
  由 Huihuang Zheng 提交于 12月 17, 2020
```
Modify CublasHandleHolder from using PADDLE_ENFORCE_CUDA_SUCCESS to PADDLE_RETRY_CUDA_SUCCESS to fix random unittest failure. We checked that the unittest log showed CUDA allocation error at this file, which may due to GPU not enough. We fixed similar failure in the past, so we applied PADDLE_RETRY_CUDA_SUCCESS here.
```
  4c4d4ba5
- C
  [Complex] Add real & imag op and api for complex tensor (#29672) · 6cfa59de
  由 Chen Weihang 提交于 12月 17, 2020
```
* add complex real op & api & unittest

* add imag op & api & unittest

* refactor op impl

* revert simplify writing due to complile failed

* polish details

* polish grad op code
```
  6cfa59de
- J
  
  Added missing format of oneDNN (#29670) · 9eff1a67
  由 Jacek Czaja 提交于 12月 17, 2020
  
  9eff1a67
- L
  Update EarlyStopping sample code (#29723) · 572810ee
  由 LiuChiachi 提交于 12月 17, 2020
```
* update EarlyStopping doc

* update EarlyStopping doc, test=document_fix
```
  572810ee
- W
  
  delete the code for fp16 optimization because it is not faster than common template code (#29715) · 2e0d1ed0
  由 wangchaochaohu 提交于 12月 17, 2020
  
  2e0d1ed0
- L
  Add approval monitor for unity_build_rule.cmake (#29701) · bb5a7854
  由 LoveAn 提交于 12月 17, 2020
```
* Add approval monitor for unity_build_rule.cmake, test=develop

* fix words spell error, test=document_fix
```
  bb5a7854
- Q
  
  [GO] add two cgo api, test=develop (#29659) · 7684b918
  由 Qi Li 提交于 12月 17, 2020
  
  7684b918
16 12月, 2020 10 次提交
- T
  update activation op on kunlun (#29577) · af8ded77
  由 TTerror 提交于 12月 16, 2020
```
* fix expand && concat/transpose to new api

* update xpu_header

* update activation op on kunlun

* update activation op on kunlun

* update activation op on kunlun

* update activation op on kunlun

* update activation op on kunlun

* add nearest_interp on kunlun

* update error message
```
  af8ded77
- C
  add pad and concat double grad (#29549) · cc387159
  由 ceci3 提交于 12月 16, 2020
```
* add constant pad double grad
```
  cc387159
- L
  
  [Kunlun] PR1:Support one Kunlun card training in parallel executor (#29337) · f13c3a9c
  由 liuyuhui 提交于 12月 16, 2020
  
  f13c3a9c
- Y
  添加rocm平台支持代码 (#29342) · 76738504
  由 Y_Xuan 提交于 12月 16, 2020
```
* 添加rocm平台支持代码

* 修改一些问题

* 修改一些歧义并添加备注

* 修改代码格式

* 解决冲突后的代码修改

* 修改operators.cmake

* 修改格式

* 修正错误

* 统一接口

* 修改日期
```
  76738504
- H
  add static.amp into setup.pu.in (#29621) · b96dada4
  由 huangxu96 提交于 12月 16, 2020
```
* add static.amp into setup.pu.in

* add unittest for api
```
  b96dada4
- Z
  improve dropout grad (#29605) · 1e9127f6
  由 Zhang Ting 提交于 12月 16, 2020
```
* improve grad perf
```
  1e9127f6
- W
  
  refine (#29622) · eab44e1f
  由 wangchaochaohu 提交于 12月 16, 2020
  
  eab44e1f
- Y
  
  disable ut test_cumsum_op (#29613) · d0b789d2
  由 YUNSHEN XIE 提交于 12月 16, 2020
  
  d0b789d2
- J
  fix wmt14 doc, remove backward, add bidirect direction in rnn api (#29633) · 84bae277
  由 Jack Zhou 提交于 12月 16, 2020
```
* fix wmt14 doc, remove backward, add bidirect direction in rnn api

* fix rnn unittest

* fix test_rnn_nets_static.py bug
```
  84bae277
- W
  
  fix gen_nccl_id_op_helper compile failed, test=develop (#29614) · 613c46bc
  由 WangXi 提交于 12月 16, 2020
  
  613c46bc
15 12月, 2020 13 次提交
- 1. add python version selection 2.add dynamic flags setting. (#29612) · f5f8809c
  由 chen.zhiyu 提交于 12月 15, 2020
  
  f5f8809c
- Y
  New UT should not exceed 15s (#29492) · 2926e743
  由 YUNSHEN XIE 提交于 12月 15, 2020
```
* added UT should not exceed 15s

* fix error

* UT limit of 15s is the first to be executed

* fix error

* fix error with CI_SKIP_CPP_TEST

* modfied tiemout setting

* fix error
```
  2926e743
- C
  Add complex dtype op (add) test example (#29603) · f02aece1
  由 Chen Weihang 提交于 12月 15, 2020
```
* add op test case for complex

* polish code details

* add xpu set constant support

* fix argument rror

* remove useless pyc file
```
  f02aece1
- A
  
  Add tf32 support for A100 tensor core acceleration for cuBLAS (#28732) · efea540c
  由 AshburnLee 提交于 12月 15, 2020
  
  efea540c
- L
  add transpose double grad test=develop (#29600) · 7779768b
  由 lijianshe02 提交于 12月 15, 2020
```
* add transpose double grad test=develop
```
  7779768b
- W
  
  optimize for long width for elementwise (#29602) · 1b69e528
  由 wangchaochaohu 提交于 12月 15, 2020
  
  1b69e528
- W
  
  fix none-contiguous bug for python api. (#29615) · 78dad786
  由 Wilber 提交于 12月 15, 2020
  
  78dad786
- Z
  
  fix cache pip error (#29618) · 18f9df0d
  由 Zhou Wei 提交于 12月 15, 2020
  
  18f9df0d
- H
  add alias for fluid.contrib.mixed_precision (#29562) · c05170d3
  由 huangxu96 提交于 12月 15, 2020
```
* add alias for fluid.contrib.mixed_precision
```
  c05170d3
- S
  Fix the dowanload bug in the case of multiple machines (#29551) · fb6697b4
  由 ShenLiang 提交于 12月 15, 2020
```
* fix the dowanload bug
* add sort for ips
```
  fb6697b4
- S
  Fix bug of matmul_v2 for broadcast case (#29599) · 1efef8ba
  由 ShenLiang 提交于 12月 15, 2020
```
* fix bug of matmul_v2 for broadcast
```
  1efef8ba
- R
  Simplify the prompt of const_cast check. (#29548) · a9082082
  由 Ren Wei (任卫) 提交于 12月 15, 2020
```
只检查增的情况，不检查删除情况
```
  a9082082
- Q
  
  Add clip double grad (#29590) · 8d549fc8
  由 qingqing01 提交于 12月 15, 2020
  
  8d549fc8
14 12月, 2020 1 次提交
- T
  disable test_parallel_executor_profiler in cuda 10.1 (#29581) · 81acc327
  由 Tao Luo 提交于 12月 14, 2020
```
* disable test_parallel_executor_profiler in cuda 10.1

* update set_tests_properties
```
  81acc327

机器未来 / Paddle 与 Fork 源项目一致

机器未来 / Paddle
与 Fork 源项目一致