提交 · ec8a0c1d905c00848d3c08766c52c71d3ac63c8e · 机器未来 / Paddle

11 2月, 2022 7 次提交

L

Add TensorRT inspector into Paddle-TRT (#38362) · 69793a27
由 Leo Chen 提交于 2月 11, 2022

69793a27

Added shape (U)INT8/BF16/FP32 oneDNN kernel (#36033) · 52bbaae9

由 jakpiase 提交于 2月 11, 2022

* added shape oneDNN kernel

* removed unnecessary import from test

* added skipping tests for GPU

* refactoring

* refactored shape kernel

* added tests in new framework

* removed one line

* minor change

* added newline at EOF

* added formatting

* added attributes as extra

52bbaae9

F

[MLU] add pool2d pytest (#39454) · 2db25f0d
由 fwenguang 提交于 2月 11, 2022

2db25f0d
J

uniform_random op for mlu (#39450) · 02f06708
由 joeqiao12 提交于 2月 11, 2022

02f06708
Z
[bf16] add bf16 kernel: transpose & unbind (#39457) · 1e6047f1
由 zhangbo9674 提交于 2月 11, 2022
```
* add transpose unbind

* add unittest

* refine transpose unittest
```
1e6047f1
J

fix prelu trt convert (#39389) · c86765ed
由 JingZhuangzhuang 提交于 2月 11, 2022

c86765ed

统一 ps 开发 - python (#39431) · 22c67d14

由 ziyoujiyi 提交于 2月 11, 2022

* delete gloo connect retry

* the_one_ps dirs reconstruct

* .

* .

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* the one ps dirs modify

* the one ps dirs modify

* the one ps dirs modify

* the one ps dirs modify

* refactor ps optimize

* refactor ps optimize

* refactor ps optimize

* .

* .

* .

* .

* .

* .

* refactor theoneps

* the_one_ps

* add ps pass unittest

* add ps pass unittest

* ps unitest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* add cpu_async_ps_mode test

* add cpu_async_ps_mode test

* add cpu_async_ps_mode test

* ps unittest ready

* ps unittest ready

* solve dist_pass init conflict

* solve import CommContext error

* unittest ok

* implement AllocateFrom

* solve setup.py.in conflict

* solve conflict

* solve conflict

* solve conflict

* .

* .

* cpu-async-ps minimize test ok & gpu minimize test ok
Co-authored-by: Nzkh2016 <zhangkaihuo@baidu.com>

22c67d14

10 2月, 2022 9 次提交
- 0
  [Dy2St]Handle `a, b = paddle.shape(x)` in Static Analysis (#39245) · 1252f4bb
  由 0x45f 提交于 2月 10, 2022
```
* refine Assign

* add UT
```
  1252f4bb
- F
  [MLU] add mlu kernel for accuracy op (#39337) · 383de295
  由 fwenguang 提交于 2月 10, 2022
```
* [MLU] add mlu kernel for accuracy op

* fix license format

* fix error message
```
  383de295
- F
  [NPU] add reduce_min (#39019) · 2b8b16d7
  由 furnace 提交于 2月 10, 2022
```
[NPU] add reduce_min
```
  2b8b16d7
- W
  mkldnn layout issue fix (#39422) · 52d6b306
  由 wenbin 提交于 2月 10, 2022
```
* mkldnn conv fix

* definetion
```
  52d6b306
- S
  Add _get_parameter method to Lamb optimizer (#39416) · c47d6729
  由 sneaxiy 提交于 2月 10, 2022
```
* add _get_parameter func to lamb

* remove duplicate code
```
  c47d6729
- C
  Modify the unsqueeze dimension of input data in conv1d NCL And NLC format (#38425) · 224bc511
  由 crystal 提交于 2月 10, 2022
```
* optimize conv1d forward

* add conv opt

* Optimize memory copy

* delete share data with

* set num_filters=512

* add nlc optimize

* Optimize num_filter=512 data on A100 and V100

* Fix the workspace_size size setting of filter
```
  224bc511
- Z
  [bf16] add bf16 kernel: squeeze & unsqueeze & stack (#39402) · 59c7aea5
  由 zhangbo9674 提交于 2月 10, 2022
```
* add squeeze unsqueeze stack

* add unittest

* add cpu kernel
```
  59c7aea5
- Z
  [bf16] add bf16 kernel: dropout & reshape & slice (#39395) · e8ac7fc3
  由 zhangbo9674 提交于 2月 10, 2022
```
* add dropout

* add reshape

* add slice

* refien slice unittest

* refine slice unittest

* add cpu bf16 kernel
```
  e8ac7fc3
- A
  
  [PluggableDevice] custom kernel supports multi cpp_dtype registering (#39385) · 63d2333e
  由 Aganlengzi 提交于 2月 10, 2022
  
  63d2333e
09 2月, 2022 9 次提交

[Paddle-Inference] rebuild matmul pass: trt and gpu_cpu (#39369) · db7d129e

由 Wangzheee 提交于 2月 09, 2022

* rebuild matmul pass: trt and gpu_cpu

* rebuild matmul pass: trt and gpu_cpu

* rebuild matmul pass: trt and gpu_cpu

* rebuild matmul pass: trt and gpu_cpu

db7d129e

F

[MLU] add gaussian_random mlu kernel (#39338) · c35b4b8e
由 fwenguang 提交于 2月 09, 2022

c35b4b8e
F

[mlu] add mlu kernel for momentum op (#39331) · f8ba12e5
由 fwenguang 提交于 2月 09, 2022

f8ba12e5
F

[mlu] add mlu kernel for elementwise_add (#39313) · d47a511a
由 fwenguang 提交于 2月 09, 2022

d47a511a

Replace EagerTensor with Tensor (#39376) · 945a3ce9

由 Jiabin Yang 提交于 2月 09, 2022

* merge legacy to fluid

* Remove legacy code

* Remove legacy code

* Remove DataType test

* Using Tensor directly instead of using EagerTensor

* support gradient_accumulation

* make test_imperative_lod_tensor_to_selected_rows longer

* make test_imperative_lod_tensor_to_selected_rows longer

945a3ce9

Move trace op to pten (#39227) · d7dddf94

由 hong 提交于 2月 09, 2022

* add trace op

* bug fix

* bug fix; test=develop

* thrust bug fix; test=develop

* remove useless register; test=develop

* fix bug; test=develop

* update trace kernel; test=develop

* move kernel args to trace_sig; test=develop

d7dddf94

C
[CustomOp] Fix slice bug of custom op (#39393) · 91b074a2
由 Chen Weihang 提交于 2月 09, 2022
```
* fix slice bug of cusstom op

* add offset in check
```
91b074a2
S

add more int type support for softmax_with_cross_entropy (#39409) · eaa3fd45
由 sneaxiy 提交于 2月 09, 2022

eaa3fd45

Move norm to pten (#39324) · ece200b3

由 hong 提交于 2月 09, 2022

* add norm cpu

* update code;

* norm bug fix

* move norm op to pten; test=develop

* move norm op to pten; test=develop

* add norm util; test=develop

* fix norm npu bug; test=develop

* fix norm kernel bug; test=develop

* move kernel args to pten; test=develop

* move kernel args to pten sig; test=develop

ece200b3

08 2月, 2022 8 次提交

S
Make Embedding layer support more int ids type (#39381) · 60f1461a
由 sneaxiy 提交于 2月 08, 2022
```
* add more int id type support for embedding

* add ut

* add more ut

* fix ci error
```
60f1461a

Add FuseOptimizerPass and test_dist_fuse_adam_pass unittest. (#39208) · ccdcfa2d

由 hlygit66666 提交于 2月 08, 2022

* add fuse_relu_depthwise_conv_pass unittest

* fix atol and rtol

* fix according to review

* Add FuseOptimizerPass and fuse_adam_pass unittest

* add sgd and momentum unittest

* add fuse_optimizer_pass

* close amp

* close amp

* update

* fix run on two cards

* Update test_dist_fuse_adam_pass.py

* Update test_dist_fuse_momentum_pass.py

* Update test_dist_fuse_sgd_pass.py

* Create test_dist_fuse_sgd_pass.py

* Create test_dist_fuse_sgd_pass.py

* Create test_dist_fuse_sgd_pass.py

* Update test_dist_fuse_adam_pass.py

* Update test_dist_fuse_momentum_pass.py

* Update test_dist_fuse_sgd_pass.py

ccdcfa2d

ps optimize refactor (#38982) · 196dbfc2

由 ziyoujiyi 提交于 2月 08, 2022

* delete gloo connect retry

* the_one_ps dirs reconstruct

* .

* .

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* the one ps dirs modify

* the one ps dirs modify

* the one ps dirs modify

* the one ps dirs modify

* refactor ps optimize

* refactor ps optimize

* refactor ps optimize

* .

* .

* .

* .

* .

* .

* refactor theoneps

* the_one_ps

* add ps pass unittest

* add ps pass unittest

* ps unitest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* add cpu_async_ps_mode test

* add cpu_async_ps_mode test

* add cpu_async_ps_mode test

* ps unittest ready

* ps unittest ready

* solve dist_pass init conflict

* solve import CommContext error

* unittest ok

* implement AllocateFrom

* solve setup.py.in conflict

* solve conflict

* solve conflict

* solve conflict

* .

* .
Co-authored-by: Nzkh2016 <zhangkaihuo@baidu.com>

196dbfc2

Z
[bf16] add bf16 cuda kernel: concat and split (#39380) · de0bad2a
由 zhangbo9674 提交于 2月 08, 2022
```
* add concat & split

* add concat kernel

* add concat unittest

* add split unittest
```
de0bad2a
B

optimize sharding stage3 (#39334) · 23d559dd
由 Baibaifan 提交于 2月 08, 2022

23d559dd
C
Fix reduce_sum dtype dispatch bug on gpu (#39349) · 4d7ad277
由 Chen Weihang 提交于 2月 08, 2022
```
* fix pten reduce dispatch bug

* add cast beforce reduce

* fix test failed
```
4d7ad277
L

[bf16] support printing bf16 tensor (#39375) · f57b21e6
由 Leo Chen 提交于 2月 08, 2022

f57b21e6
S
Add __PD_DEFINE_RAW_OP_KERNEL_FUNC for registering custom op kernel with ExecutionContext (#39352) · 5c3873f6
由 sneaxiy 提交于 2月 08, 2022
```
* hack custom op

* add ut

* skip windows ci
```
5c3873f6

07 2月, 2022 4 次提交
- T
  
  add sequence_conv op in xpu place (#39025) · fee4316d
  由 tanzhipeng 提交于 2月 07, 2022
  
  fee4316d
- S
  
  add _ptr for tensor (#39357) · 24b2e8e6
  由 sneaxiy 提交于 2月 07, 2022
  
  24b2e8e6
- J
  Added Adam FP32 JIT assembly kernel (#39158) · ebd14743
  由 jakpiase 提交于 2月 07, 2022
```
* Added adam kernel

* CI rerun
```
  ebd14743
- C
  [CustomOp] Support output as input argument of kernel func (#39353) · f1f74e9e
  由 Chen Weihang 提交于 2月 07, 2022
```
* refactor custom op kernel func and utils

* add output sync

* adapte tensor* in utils

* fix windows symbol error
```
  f1f74e9e
30 1月, 2022 3 次提交
- X
  add multinomial probability distribution (#38820) · 01f606b4
  由 Xiaoxu Chen 提交于 1月 30, 2022
```
* add multinomial probability distribution
* fix categorical sample bug when logits less than zero
* fix categorical sample can't pass hypothesis test and entropy shape error bug
```
  01f606b4
- F
  
  [MLU] add softmax_with_cross_entropy mlu kernel (#39260) · aecf9967
  由 fwenguang 提交于 1月 30, 2022
  
  aecf9967
- L
  [pten] fit get all register op kernels (#39288) · eefe5feb
  由 Leo Chen 提交于 1月 30, 2022
```
* upgrade _get_all_register_op_kernels

* add ut

* support xpu/npu

* fix device id

* enhance TransToFluidPlace

* fix compile
```
  eefe5feb

机器未来 / Paddle 与 Fork 源项目一致

机器未来 / Paddle
与 Fork 源项目一致