提交 · 481511a6b4230ba92438afc965bcd6c4d2dd4f59 · PaddlePaddle / Paddle

16 5月, 2023 1 次提交
- C
  
  【Hackathon4 No.61】remainder 算子FP16/BF16单测完善 (#52920) · 481511a6
  由 cyberslack_lee 提交于 5月 16, 2023
  
  481511a6
15 5月, 2023 2 次提交

remove some [-Wunsed-parameter] warning (#53689) · 3e1fffea

由 Galaxy1458 提交于 5月 15, 2023

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

3e1fffea

remove some [-Wunsed-parameter]warning (#53679) · ca2ea162

由 Galaxy1458 提交于 5月 15, 2023

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

ca2ea162

12 5月, 2023 1 次提交
- W
  sequence_mask functionalization (#53478) · d2b1e3c2
  由 Wang Xin 提交于 5月 12, 2023
```
* sequence_mask functionalization

* fix sequence_mask test
```
  d2b1e3c2
11 5月, 2023 1 次提交
- 张
  
  昇腾和寒武纪相关代码退场 npu相关代码退场2 (#53568) · 0d45ac73
  由张春乔提交于 5月 11, 2023
  
  0d45ac73
10 5月, 2023 3 次提交

傅

add index_put api (#52886) · f3393f49

由傅剑寒提交于 5月 10, 2023

* add index_put api

* fix value broadcast in backward and add test case in static

* add timeout=120s for index_put

* add op_compat for index_put

* add inplace index_put test

* add test case when index tensor in indices is int32 when indices.size less than x.dims

* add index_put api backward in cpu place

* add backward test case

* refactor code to delete some duplicated code

* replace reshape with resize for decrease extra memcpy

* add datatype flag in backward yaml

* fix bug in documentation

* Update python/paddle/tensor/manipulation.py

---------
Co-authored-by: NLigoml <39876205+Ligoml@users.noreply.github.com>

f3393f49

G
remove some [-Wunused-parameter] warning and WITH_DISTRIBUT flags (#53650) · 65e57a7d
由 Galaxy1458 提交于 5月 10, 2023
```
* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
```
65e57a7d
W

fix two minor issues (#53658) · ee1aa69c
由 Winters Montagne 提交于 5月 10, 2023

ee1aa69c

09 5月, 2023 2 次提交
- B
  
  Fix xpu2 kp compile error (#53548) · 8d340ee1
  由 Bo Zhang 提交于 5月 09, 2023
  
  8d340ee1
- G
  remove some [-Wunused-parameter]warning and WITH_DISTRIBUTE flag (#53532) · 727fa27d
  由 Galaxy1458 提交于 5月 09, 2023
```
* test,test=develop

* test,test=develop
```
  727fa27d
08 5月, 2023 4 次提交
- Z
  
  Fix the calculation of y_grad in divide_backward (#53582) · b6c0407e
  由 Zhang Zheng 提交于 5月 08, 2023
  
  b6c0407e
- 张
  
  rm npu (#53566) · 6d396ace
  由张春乔提交于 5月 08, 2023
  
  6d396ace
- J
  
  throw warning at __getitem__, not slice_utils (#53579) · fe919400
  由 JYChen 提交于 5月 08, 2023
  
  fe919400
- C
  
  Fix typos, test=document_fix (#53540) · acefdeb7
  由 co63oc 提交于 5月 08, 2023
  
  acefdeb7
05 5月, 2023 3 次提交
- Revert "【Hackathon No.52】为 Paddle dist 算子实现 float16 数据类型支持 (#50915)" (#53527) · d463f8ee
  由 iSerendipity 提交于 5月 05, 2023
```
This reverts commit 9c406531.
```
  d463f8ee
- G
  remove some [-Wunused-parameter]warning (#53397) · 58435ae5
  由 Galaxy1458 提交于 5月 05, 2023
```
* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
```
  58435ae5
- C
  【Hackathon No.59】addmm 算子FP16/BF16单测完善 (#53111) · 841efcd4
  由 co63oc 提交于 5月 05, 2023
```
* Add addmm tests

* Fix code
```
  841efcd4
28 4月, 2023 3 次提交

Dropout optimize & clean broadcast inT and ElementwiseType (#52969) · d611e48c

由 Bo Zhang 提交于 4月 28, 2023

* change judgement for DropoutGradGPUKernelDriver

* add UnrollerWithoutVecSize and after this Loaddata to be refined

* pass unittest

* use same unroller with XPU

* BroadcastWithInt64Index

* BroadcastDataLoader template partial specialization

* fix compile errs in ROCms

* clean ElementwiseT and InT for BroadcastKernel

* default axis and clean inT

* remove redundant fast divmod computation

* optimize drop_nd & drop_nd_grad

* optimize BroadcastDataLoader bf16 fp16

* rm InT etc. after merge develop

* delete constexpr for windows ci

* fix conflict

* fix conflic with develop

* fix conflic

* new clean

* clean

d611e48c

【Hackathon No.52】为 Paddle dist 算子实现 float16 数据类型支持 (#50915) · 9c406531
由 iSerendipity 提交于 4月 28, 2023

9c406531
S
【Hackathon No.55】add fmin BF16 test (#53100) · 8163faaa
由 superwinner1 提交于 4月 28, 2023
```
* 'fmin'

* 'fix'

* 'fix'
```
8163faaa

27 4月, 2023 7 次提交

Support different dtypes of inputs for broadcast for dropout optimization (#52093) · 3474e09c

由 Bo Zhang 提交于 4月 27, 2023

* change judgement for DropoutGradGPUKernelDriver

* add UnrollerWithoutVecSize and after this Loaddata to be refined

* pass unittest

* use same unroller with XPU

* BroadcastWithInt64Index

* BroadcastDataLoader template partial specialization

* fix compile errs in ROCms

* PR comment

3474e09c

【PaddlePaddle Hackathon 4】：为maxout算子支持 float16 数据类型 (#50976) · 8bfd978f

由 NetPunk 提交于 4月 27, 2023

* support fp16 for maxout op

* format code

* change api

* add test for static float16

* format code

* formatting code

* atol alignment

* experiment—1

* experiment-2

* experiment-3

* format code

8bfd978f

Move fused feedforward (#53166) · 25b4ba7f

由 Sonder 提交于 4月 27, 2023

* trans fused_feedward Compute function to phi

* add register info

* remove maxfunctor

* move fused feedward to phi

* remove sig file

* remove fliud include

* add include

* add include

* add sig file

* add output register info

* fix sig file

* Update fused_feedforward_sig.cc

* fix grad kernel

* update output register info

* fix

* open fused_feedforward static build

* add optional and fix code style

* fix output info for fused attention

* add optional param

* merge

25b4ba7f

J

Hack__getitem__ from 0-d to 1-d with FLAGS_set_to_1d (#53358) · 1bd468e2
由 JYChen 提交于 4月 27, 2023

1bd468e2
E

fix softmax assert error (#53360) · c50f5fa4
由 engineer1109 提交于 4月 27, 2023

c50f5fa4

remove some [-Wunused-parameter] warning (#53365) · 0fac3281

由 Galaxy1458 提交于 4月 27, 2023

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

0fac3281

S

【Hackathon No.55】add fmax BF16 test (#51925) · 8a6ad6e5
由 superwinner1 提交于 4月 27, 2023

8a6ad6e5

26 4月, 2023 2 次提交

remove some [-Wunused-parameter] waring (#53319) · f9e5072b

由 Galaxy1458 提交于 4月 26, 2023

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

f9e5072b

Optimize c_embedding op in deterministic mode (#53197) · 35f5c245

由 sneaxiy 提交于 4月 26, 2023

* optimize embedding deterministic mode

* fix compile error

* change FLAGS_cudnn_deterministic to int64

* fix 700 error

* add ut

* fix ut

* fix ut

* fix win32 ci

* fix flags with PHI_DEFINE_EXPORTED_int64

35f5c245

25 4月, 2023 3 次提交

[PHI]Add flags macro for PHI (#52991) · 22e96bde

由 YuanRisheng 提交于 4月 25, 2023

* add flags for phi

* fix compile bugs

* fix ci bugs

* fix inference bugs

* fix cinn' bugs

* fix cinn bugs

* perfect code according comment

* fix ci bugs

* fix ci bugs

22e96bde

【Hackathon No57】add fp16 & bf16 for max_pool2d_with_index, max_pool3d_with_index (#52314) · 46951224

由 Difer 提交于 4月 25, 2023

* add fp_bf for pool_max_withidx

* fix some error

* fix error

* codestyle error

* fix masktype

* fix input bf type

* input bf dtype convert error

* back to convert input to bf16 first

* fix convert error

* fix bf16 grad check

46951224

B

add syncthreads (#53149) · b7565222
由 Bo Zhang 提交于 4月 25, 2023

b7565222

24 4月, 2023 6 次提交
- [Zero-Dim] Support paddle.max output 0D, test=allcase (#53242) · 9f9cd919
  由 zhouweiwei2014 提交于 4月 24, 2023
  
  9f9cd919
- W
  
  fix 'Werror-maybe-uninitialized' compiler error in GCC 11.3 (#53246) · 21508090
  由 Wang Xin 提交于 4月 24, 2023
  
  21508090
- Z
  Fix the calculation of layer_norm_bwd (#53224) · a0aff194
  由 Zhang Zheng 提交于 4月 24, 2023
```
* Fix the calculation of layer_norm_bwd

* fix
```
  a0aff194
- Y
  
  fix static_assert with no message (#53222) · 71474b10
  由 Yuanle Liu 提交于 4月 24, 2023
  
  71474b10
- S
  Add weighted sample (#52013) · 6a8d98e0
  由 Siming Dai 提交于 4月 24, 2023
```
Add paddle.geometric.weighted_sample_neighbors API
```
  6a8d98e0
- G
  remove some [-Wunused-parameter] (#53185) · 834eb2ba
  由 Galaxy1458 提交于 4月 24, 2023
```
* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test ,test=develop
```
  834eb2ba
23 4月, 2023 1 次提交

remove some [-Wunused-parameter] (#53162) · b02687cc

由 Galaxy1458 提交于 4月 23, 2023

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

b02687cc

21 4月, 2023 1 次提交

support 0-D output and 0-D as indice in __getitem__/__setitem__ (#52814) · 4e939c89

由 JYChen 提交于 4月 21, 2023

* support 0-D output and 0-D as indice in __getitem__

* fix tests

* fix inference and UT

* add unittest for setitem

* fix xpu test

* fix xpu 0-d

4e939c89

PaddlePaddle / Paddle 1 年多 前同步成功

PaddlePaddle / Paddle
1 年多前同步成功