提交 · f7bf2930f0d63b580744636b3688b39b49981ce3 · BaiXuePrincess / Paddle

15 11月, 2022 2 次提交

W

remove 'paddle/fluid/operators/conv_op.h' from phi (#47914) · f7bf2930
由 Wang Xin 提交于 11月 15, 2022

f7bf2930

[PHI decoupling] remove dependency on "paddle/fluid/operators/elementwise/xxx.h" in phi (#47870) · 04c29558

由 huangjiyi 提交于 11月 15, 2022

* rm "paddle/fluid/operators/elementwise/xxx.h" in phi

* fix bugs

* add LaunchElementwiseCudaKernel in phi

* Revert "add LaunchElementwiseCudaKernel in phi"

This reverts commit 588f45bbdad2372ec7bff0c567a29bff675d22e1.

* rm indirect dependence to "elementwise_op_impl.cu.h"

rm indirect dependence to "elementwise_op_impl.cu.h"

Revert "add LaunchElementwiseCudaKernel in phi"

This reverts commit 588f45bbdad2372ec7bff0c567a29bff675d22e1.

add LaunchElementwiseCudaKernel in phi

fix bugs

* rm LaunchSameDimsElementwiseCudaKernel and LaunchElementwiseCudaKernel in phi

04c29558

14 11月, 2022 16 次提交
- W
  Refactor collective communication send_partial, recv_partial, all_gather_partial C++ API (#47863) · 25e63dca
  由 Wen Sun 提交于 11月 14, 2022
```
* refactor: simplify send, recv interfaces

* refactor: rm send_partial, recv_partial, all_gather_partial
```
  25e63dca
- X
  
  [Paddle Inference] Add where trt converter (#47820) · dac0f7dd
  由 xiaoxiaohehe001 提交于 11月 14, 2022
  
  dac0f7dd
- L
  
  Remove place for process group (#47857) · 2d383b81
  由 LiYuRio 提交于 11月 14, 2022
  
  2d383b81
- [Zero-Dim] support input 0D Tensor as scalar attribute for some api (#47689) · e0be4b94
  由 zhouweiwei2014 提交于 11月 14, 2022
```
* [Zero-Dim] support input 0D Tensor as scalar attribute for some api

* fix doc
```
  e0be4b94
- C
  
  add cos double and triple grad operator (#47796) · 1a145aab
  由 cyber-pioneer 提交于 11月 14, 2022
  
  1a145aab
- H
  
  clean fluid elementwise_max (#47866) · 42c8d51a
  由 HongyuJia 提交于 11月 14, 2022
  
  42c8d51a
- J
  - Modified mem_desc() to return reference to Tensor::memory::desc to (#47844) · 2182a4f9
  由 Jacek Czaja 提交于 11月 14, 2022
```
avoid copying
```
  2182a4f9
- L
  
  remove heter and hccl (#47918) · 9191e743
  由 LiYuRio 提交于 11月 14, 2022
  
  9191e743
- N
  [CodeStyle][F821] fix undefined variables due to missing imports, misspelled... · 05a47f39
  由 Nyakku Shigure 提交于 11月 14, 2022
```
[CodeStyle][F821] fix undefined variables due to missing imports, misspelled variable names (#47899)

* `hann` -> `_hann`

* `false` -> `False`

* a missing passed argument `reduce_all`

* some missing imports

* `device_type` -> `heter_device_type`

* `PKVClient` -> `KVClient`

* fix some typos and missing imports
```
  05a47f39
- R
  
  Do not release memory cache after build_op_func_list in interpretercore (#47910) · 8347354d
  由 Ruibiao Chen 提交于 11月 14, 2022
  
  8347354d
- N
  
  Fix HOSTDEVICE redefinition during XPU KP compilation, test=kunlun (#47885) · 81e16a85
  由 niuliling123 提交于 11月 14, 2022
  
  81e16a85
- E
  
  add lite opencl support api (#47112) · 798ab3f9
  由 engineer1109 提交于 11月 14, 2022
  
  798ab3f9
- J
  [AutoParallel] bugfixed for FP16 if cond (#47841) · f781473e
  由 JZ-LIANG 提交于 11月 14, 2022
```
* fixed cond state

* fixed cond state
```
  f781473e
- Y
  
  fix squueze_transpose (#47911) · f50de679
  由 yeliang2258 提交于 11月 14, 2022
  
  f50de679
- R
  
  Add InferShape for Depend OP (#47907) · 5478e1a5
  由 Ruibiao Chen 提交于 11月 14, 2022
  
  5478e1a5
- N
  
  [CodeStyle][F821] refactor unittests utility function `parameterized` related code (#47869) · c5e1011b
  由 Nyakku Shigure 提交于 11月 14, 2022
  
  c5e1011b
11 11月, 2022 14 次提交
- [Zero-Dim] fix batch_norm op infermeta bug (#47858) · 18549417
  由 zhouweiwei2014 提交于 11月 11, 2022
  
  18549417
- Y
  
  fix coverage bug,test=document_fix (#47893) · 17dffd13
  由 YUNSHEN XIE 提交于 11月 11, 2022
  
  17dffd13
- C
  [IPU]: add model_runtime backend support in IPU (#47363) · 21b901cb
  由 czr-gc 提交于 11月 11, 2022
```
* feat(ipu): add model_runtime backend support in IPU.

* fix(ipu_executor): fix error message format.

* fix(ipu_executor): fix format.

* fix(ipu_executor): fix format again.

* fix(ipu_executor): fix format again.

* fix(ipu_executor): fix format again.
```
  21b901cb
- Z
  Refine shape op lanch method for standalone executor (#47843) · 981d1a10
  由 zhangbo9674 提交于 11月 11, 2022
```
* refine shape op in new_exe

* Revert "refine shape op in new_exe"

This reverts commit 0e0336ddc5eede3da019b348a0bcc0ef0f3be64e.

* refine shape op in new_exe

* refine shape expected_kernel_type

* add SelectedRows check for shape op

* refine code
```
  981d1a10
- J
  bugfix in XPU legacy_dygraph distributed training: (#47838) · 9a6465ca
  由 james 提交于 11月 11, 2022
```
phi::Alloc() complains about missing device_allocator_
```
  9a6465ca
- Z
  Generate static graph code for some ops by yaml (part3) (#47803) · 31f3f643
  由 zyfncg 提交于 11月 11, 2022
```
* generate static graph code for some ops by yaml

* remove deleted files

* update cmake

* update cmake

* udpate cmake
```
  31f3f643
- R
  fix cpu build failed (#47833) · 95e494e0
  由 ronnywang 提交于 11月 11, 2022
```
* fix cpu build failed

* update
```
  95e494e0
- Y
  
  [Inference] fix mixed precision (#47794) · 9bda10cd
  由 Yuanle Liu 提交于 11月 11, 2022
  
  9bda10cd
- W
  
  remove "paddle/fluid/framework/op_registry.h" from phi (#47868) · 78c8c7de
  由 Wang Xin 提交于 11月 11, 2022
  
  78c8c7de
- W
  
  remove fluid/framework/lod_tensor.h from phi (#47840) · 494cab07
  由 Wang Xin 提交于 11月 11, 2022
  
  494cab07
- Y
  
  Simplify the autotune cache codes. (#47667) · 8758a338
  由 Yiqun Liu 提交于 11月 11, 2022
  
  8758a338
- Z
  
  [Sparse]Optimize BatchNorm1D forward in test mode (#47736) · 6cdc18af
  由 zhangkaihuo 提交于 11月 11, 2022
  
  6cdc18af
- H
  [PHI decoupling] remove dependency on 2 header files in fluid from phi (#47842) · 1ad95e97
  由 huangjiyi 提交于 11月 11, 2022
```
* rm "paddle/fluid/operators/eigen/eigen_function.h" in phi

* rm "paddle/fluid/operators/elementwise/elementwise_op_function.h" in phi

* Revert "rm "paddle/fluid/operators/elementwise/elementwise_op_function.h" in phi"

This reverts commit c4ba51225e3652f1d80925afba406612968f0ee9.
```
  1ad95e97
- P
  
  [PHI decoupling] remove #include "paddle/fluid/platform/bfloat16.h" in phi (#47831) · e0742c48
  由 PuQing 提交于 11月 11, 2022
  
  e0742c48
10 11月, 2022 8 次提交
- Z
  
  conv2d_transpose and deformable_conv unrestricted some limit for xpu2, test=kunlun (#47837) · a38fc5e1
  由 zhangyikun02 提交于 11月 10, 2022
  
  a38fc5e1
- S
  [phi] migrate prelu (#47422) · cdd8c8ab
  由 Sylwester Fraczek 提交于 11月 10, 2022
```
* migrate prelu

* remove cache

* review fixes
```
  cdd8c8ab
- W
  Get grads types from cpp for adam to speed up (#47769) · 5900129c
  由 WangZhen 提交于 11月 10, 2022
```
Get grads types from cpp for adam to speed up
```
  5900129c
- L
  
  remove the hang checkness (#47806) · 8d99dd0c
  由 LiYuRio 提交于 11月 10, 2022
  
  8d99dd0c
- Y
  [PHI]Standardise some C++ API (Part4) (#47702) · 594bd723
  由 YuanRisheng 提交于 11月 10, 2022
```
* standard api

* fix sparse bugs

* fix xpu bugs, test=kunlun

* remove hard code for custom unittest

* open ci, test=kunlun

* deal with conflict
```
  594bd723
- W
  [PHI decoupling] remove fluid/framework/generator.h from phi (#47822) · 28c56d77
  由 Wang Xin 提交于 11月 10, 2022
```
* remove fluid/framework/generator.h from phi

* fix PR-CI-Kunlun-KP-Build fail
```
  28c56d77
- W
  refactor Recompute example doc (#47783) · 60ec3107
  由 wuhuachaocoding 提交于 11月 10, 2022
```
* add recompute doc.

* add recompute doc.

* add recompute doc.

* update.

* update.
```
  60ec3107
- P
  [PHI decoupling] remove "paddle/fluid/platform/device/gpu/gpu_launch_config.h" in phi (#47808) · 40a9b488
  由 PuQing 提交于 11月 10, 2022
```
* rm fluid gpu_launch_config

* fix type
```
  40a9b488

BaiXuePrincess / Paddle 与 Fork 源项目一致

BaiXuePrincess / Paddle
与 Fork 源项目一致