提交 · 10325a82e1032c3397b6f6611f558eb18ede0b07 · 机器未来 / Paddle

08 3月, 2022 1 次提交

add python profiler package (#40065) · 10325a82

由 chenjian 提交于 3月 08, 2022

* add python profiler package

* update according to review

* fix bug

* fix bug

* fix bug

* add unit test

* Revert "add unit test"

This reverts commit 4e69ff71b0645e069afe5dd8fea0d07717852c48.

* reduce for pr

* add unit test

* modify for pr

* fix unittest

* update for ci coverage

* modify according to review

* fix bug

* improve coverage

10325a82

07 3月, 2022 3 次提交

X
[OpTest] Support to test paddle API end-to-end for check_eager (#40169) · 79a32715
由 xiongkun 提交于 3月 07, 2022
```
* add python api test in TestOp

* test_python_api if self.python_api is set

* fix code by CR
```
79a32715

cuBlasLt Epilogue To Fuse Linear + ReLU|GeLU (#39437) · 2a3d9eca

由 Ming-Xu Huang 提交于 3月 07, 2022

* Added cuBlasLtHandle_t to device context.

* Added fused_gemm_epilogue op.

1. Added fused_gemm_epilogue op to leverage cuBlastLt Epilogue.
2. Support fusion Act(X*Y + bias), X'dims >=2 and Y'dims shoule be 2.
2. Act currently only be supported ReLU. (Will add GeLU in the future).

* Added UT to fused_gemm_epilogue op.

* Added LinearAct Pattern

1. Added LinearAct into graph_pattern_detector.* to define (2.)'s
pattern.
2. LinearAct is used to detect act(element_add(matmul_v2(x, w), bias)).
3. act currently only support ReLU (Will support GeLU in the future).

* Added FuseGemmEpiloguePass

1, Added FuseGemmEpiloguePass to handle nn.Linear + Act{ReLU}
fusion (GeLU will be supported in the future).
2. Only support matmul_v2 from nn.Linear.

* Added pybind to BuildStrageter.fuse_gemm_epilogue_.

* Added UT for fuse_gemm_epilogue_pass.

* GeLU support and EpilogueSingleton

1. Added GeLU support to fused_gemm_epilogue op.
2. Added EpilogueSingleton to cache auxiliary pointer.
3. Added related UTs.

* Rename cublaslt_epilogue_opto gemm_epilogue_op.*.

* Added both train and infer pattern to LinearAct.

1. Added support of fwd graph with grap_ops linking to LinearAct.
2. Added related changes to fuse_gemm_epilogue_pass for above
modification.

* Changed CUDA requirement from 11.4 to 11.6 for fuse_gemm_epilogue_pass.

* Added identity activation support to gemm_epilogue_op.

* Added Linear Fusion (matmul_v2 + ele_add)

1. Added matmul_v2 + ele_add pattern to LinearActPattern.
2. Added matmul_v2 + ele_add support to fuse_gemm_epilogue_pass.

* Rename gemm_epilogue_op.* to fused_gemm_epilogue_op.*

* Add fused_gemm_epilogue_grad op.

1. Added fused_gemm_epilogue_grad to support backward epilogue fusion.

* Add UTs to fused_gemm_epilogue_grad_op.

* Change attribute name in fused_gemm_epilogue_grad_op for clearing.

* Allow DX and DBias be dispensable to fused_gemm_epilogue_grad op.

* Added ElementwiseAdd+Matmul+Act graph pattern detection.

* Fuse backward of Linear( Act(x))

1. Added backward fusion pass to Linear( Act(x)).
2. Added backward fusion pass to Linear(x).

* Added UTs to backward fusion of Linear(Act(x)).

* Complete document of arguments to fused_gemm_epilogue_op.

* Made arguments of some functions pass by reference.

* Modify code with review comments.

1. Made arguments of some function pass by reference.
2. Removed redundant code.
3. Followed Google code style to change code.

* Made 'const' code style be consistent

* Fixed random seed of python UTs.

* Set Compiling constrains to cuBlasLt

1. Require CUDA 11.6+
2. Remove fuse_gemm_epilogue related tests when CUDA < 11.6.

* Code Reivew from Paddle

1. Changed arguments name is_first_gemm to without_x_gradient for
clearing.
2. Applied PADDLE_THROW in fused_gemm_epilogue_op.

* Remove EpilogueSingleton

1. Applied ReserveSpace to replace Epilogue for passing auxiliary
pointers between FWD and BWD.

* Fix a logical error and enhance UTs.

1. Added act op count checking in UTs.
2. Fix issue to fuse backward or ReLU(Linear(X)).
3. TODO: solve GELU fusion issues.

* Fix Linear and GeLU fusion issues.

1. Modified graph_detech_pattern to fit with both linear wiht gelu or
relu.
2. Modified data range in Uts to allow negative values.

* Removed fused_gemm_epilogue_op.h.

* Rename namespace pten to phi.

* Rename name of arguments in fused_gemm_epilogue_op

1. bias -> Bias.
2. out -> Out.
3. reserve_space -> ReserveSpace.

* Change EpiloguePassActivationCache as local variable.

1. Removed singleton in EpiloguePassActivationCache.
2. Made EpiloguePassActivationCache as an argument to each pass
functions.

2a3d9eca

L

initialize processgroupnccl with store (#40181) · 0ad25fb9
由 lilong12 提交于 3月 07, 2022

0ad25fb9

04 3月, 2022 1 次提交

Generate forward-only operators (#39962) · a6947991

由 Zhanlue Yang 提交于 3月 04, 2022

* [Eager][Yaml]Supported Scalar and ScalarArray for AutoCodeGen

* Generate forward-only operators

* [Yaml]Support parsing fwd & bwd returns with name

* Fixed issues

* Fixed minor issues

a6947991

03 3月, 2022 5 次提交

Z

[Eager][Yaml]Supported Scalar and ScalarArray for AutoCodeGen (#40080) · 97ccaa79
由 Zhanlue Yang 提交于 3月 03, 2022

97ccaa79
R

[CustomRuntime] migrate CustomRuntime into phi (#39908) · b4665d23
由 ronnywang 提交于 3月 03, 2022

b4665d23
L

add communication api for ProcessGroupNCCL (#40097) · b565b349
由 lilong12 提交于 3月 03, 2022

b565b349

Support slim eager (#39874) · da47544c

由 Jiabin Yang 提交于 3月 03, 2022

* eager, test=develop

* fix bug, test=develop

* eager, test=develop

* merge legacy to fluid

* eager, test=develop

* eager, test=develop

* Refactor TensorAdd func by template and remove gradient_accumulation in eager

* Remove needless target name

* eager, test=develop

* eager, test=develop

* Use overload instead of template

* Remove legacy code

* Remove legacy code

* selectedrows, test=develop

* Remove DataType test

* eager, test=develop

* eager, test=develop

* support gan, test=develop

* 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

* refine code

* ptb, test=develop

* Rename all EagerTensor to Tensor

* Rename some EagerTensor to Tensor

* rename EagerTensor to EagerVariable

* eager, test=develop

* eager, test=develop

* eager, test=develop

* eager, test=develop

* add more test

* eager, test=develop

* Support copiable selected rows and merge develop

* save load, eager, test=develop

* save load, eager, test=develop

* refine, test=develop

* remove useless _set_value method

* refine, test=develop

* refine, test=develop

* revert static_runner, test=develop

* EagerTensor to Tensor, test=develop

* refine, test=develop

* refine, test=develop

* clear grad, test=develop

* merge, develop

* merge, develop

* merge, test=develop

* merge, test=develop

* Support quant and part of slice

* support legacy static save

* extend slim tests time

* remove imperative on inference

* remove imperative on inference

* merge develop

* fix typo

* fix typo

* split slice related code into 2 part for imperative and eager

* split slice from inference

* split slice from inference

* fix test_tensor_register_hook
Co-authored-by: NWang Huan <wanghuan29@baidu.com>
Co-authored-by: NWeilong Wu <veyron_wu@163.com>
Co-authored-by: Nwanghuancoder <wanghuancoder@163.com>

da47544c

L
Add the implementation of Gloo for ProcessGroup (#39892) · c16f85f9
由 lilong12 提交于 3月 03, 2022
```
* add pg_gloo
```
c16f85f9

02 3月, 2022 4 次提交
- H
  
  [Infrt]add phi kernel dialect (#39726) · 07dad6d6
  由 huzhiqiang 提交于 3月 02, 2022
  
  07dad6d6
- Y
  [fleet_executor] Add entrance of FleetExecutor in AnalysisPredictor for... · 244ae318
  由 Yuang Liu 提交于 3月 02, 2022
```
[fleet_executor] Add entrance of FleetExecutor in AnalysisPredictor for distributed inference (#39992)
```
  244ae318
- B
  
  add_new_comm_primitive (#40040) · 4e00d2bb
  由 Baibaifan 提交于 3月 02, 2022
  
  4e00d2bb
- W
  [Eager] open eager when WITH_PYTHON (#39979) · 9af72957
  由 wanghuancoder 提交于 3月 02, 2022
```
* open eager when WITH_PYTHON, test=develop

* refine, test=develop

* refine, test=develop

* add DWITH_PYTHON for gen_fluid_lib, test=develop
```
  9af72957
01 3月, 2022 3 次提交
- A
  
  fix compiling and running with ipu (#39920) · 69ab2700
  由 Allen Guo 提交于 3月 01, 2022
  
  69ab2700
- G
  
  add MasterParam and MasterParamOut for sparse_momentum op (#39969) · 9de79892
  由 Guoxia Wang 提交于 3月 01, 2022
  
  9de79892
- S
  [DP] Construct reducer group (#39987) · 4da841e0
  由 ShenLiang 提交于 3月 01, 2022
```
* add reducer
```
  4da841e0
28 2月, 2022 2 次提交
- C
  [Pten->Phi PR4] Rename pten in funcs to phi (#39961) · eb42dd52
  由 Chen Weihang 提交于 2月 28, 2022
```
* rename pten_utils to phi_utils

* rename pten_utils target

* rename Pten to Phi

* replace pten with phi

* resolve conflict
```
  eb42dd52
- Z
  
  fix ps_gpu_wrapper (#39965) · bd9b9460
  由 zmxdream 提交于 2月 28, 2022
  
  bd9b9460
26 2月, 2022 1 次提交
- W
  [Eager Hook] Support GradientHook and ReduceHook, expose related interface to python (#39893) · a456dda6
  由 Weilong Wu 提交于 2月 26, 2022
```
* Support Eager Hook, expose interface to python

* Fix CI issue
```
  a456dda6
24 2月, 2022 4 次提交

C
[PTen->Phi PR3] Rename pten make target to phi (#39832) · f77019a0
由 Chen Weihang 提交于 2月 24, 2022
```
* rename pten to phi

* fix infrt compile failed

* resolve conflict
```
f77019a0

[IPU] Update IpuStrategy Python Part (#39646) · e0409c93

由 Allen Guo 提交于 2月 24, 2022

* Update IpuStrategy Python Part

* add docs

* add add_custom_op for ipu_strategy

* fix build warning

* rm unneeded part

* clean api

* fix typo

* update option names

* update IpuStrategy doc

e0409c93

Refactored GradNodeAccumulation data structure and behaviour (#39526) · 1abfc8dd

由 Zhanlue Yang 提交于 2月 24, 2022

* Refactored GradNodeAccumulation data structure and behaviour

* Fixed CI issues

* Fix compilation issues

* Fixed minor issues

* Reverted changes for intermediate and OverwriteOutput

* fixed minor issue

* Fixed code format issues

* Fixed CI-Coverage issue

* Fixed CI issues

1abfc8dd

[Eager] save load testcase (#39571) · 6b5749eb

由 wanghuancoder 提交于 2月 24, 2022

* eager, test=develop

* fix bug, test=develop

* eager, test=develop

* merge legacy to fluid

* eager, test=develop

* eager, test=develop

* Refactor TensorAdd func by template and remove gradient_accumulation in eager

* Remove needless target name

* eager, test=develop

* eager, test=develop

* Use overload instead of template

* Remove legacy code

* Remove legacy code

* selectedrows, test=develop

* Remove DataType test

* eager, test=develop

* eager, test=develop

* support gan, test=develop

* 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

* refine code

* ptb, test=develop

* Rename all EagerTensor to Tensor

* Rename some EagerTensor to Tensor

* rename EagerTensor to EagerVariable

* eager, test=develop

* eager, test=develop

* eager, test=develop

* eager, test=develop

* add more test

* eager, test=develop

* Support copiable selected rows and merge develop

* save load, eager, test=develop

* save load, eager, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* revert static_runner, test=develop

* EagerTensor to Tensor, test=develop

* refine, test=develop

* refine, test=develop

* clear grad, test=develop

* merge, develop

* merge, develop

* merge, test=develop

* merge, test=develop
Co-authored-by: NJiabinYang <360788950@qq.com>
Co-authored-by: NWeilong Wu <veyron_wu@163.com>

6b5749eb

23 2月, 2022 4 次提交

S
Add ProcessGroupNCCL for distributed training (#39737) · 0b205817
由 ShenLiang 提交于 2月 23, 2022
```
* add processgroup_nccl
```
0b205817
Z

Support dispensable inputs for eager final state codegen (#39743) · ca11a0e5
由 Zhanlue Yang 提交于 2月 23, 2022

ca11a0e5
[MLU] add cncl parallel context and mlu resource pool (#39803) · 6241913b
由 mhhhh1 提交于 2月 23, 2022
```
* [MLU] add cncl parallel context and mlu resource pool

* [MLU] fix the cncl_context_test
```
6241913b

[Eager] Support Eager mode for some model testcase (#39248) · abe232d8

由 wanghuancoder 提交于 2月 23, 2022

* eager, test=develop

* fix bug, test=develop

* eager, test=develop

* merge legacy to fluid

* eager, test=develop

* eager, test=develop

* Refactor TensorAdd func by template and remove gradient_accumulation in eager

* Remove needless target name

* eager, test=develop

* eager, test=develop

* Use overload instead of template

* Remove legacy code

* Remove legacy code

* selectedrows, test=develop

* Remove DataType test

* eager, test=develop

* eager, test=develop

* support gan, test=develop

* 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

* refine code

* ptb, test=develop

* Rename all EagerTensor to Tensor

* Rename some EagerTensor to Tensor

* rename EagerTensor to EagerVariable

* eager, test=develop

* eager, test=develop

* eager, test=develop

* eager, test=develop

* add more test

* eager, test=develop

* Support copiable selected rows and merge develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* clear grad, test=develop

* merge, develop

* merge, develop
Co-authored-by: NJiabinYang <360788950@qq.com>
Co-authored-by: NWeilong Wu <veyron_wu@163.com>

abe232d8

22 2月, 2022 3 次提交
- L
  Add the implementation of TCP Store (#39384) · b95cd3b7
  由 lilong12 提交于 2月 22, 2022
```
* add tcp_socket and tcp_store
```
  b95cd3b7
- W
  fix bug in new the_one_ps (#39505) · d56a0a1b
  由 wangguanqun 提交于 2月 22, 2022
```
* fix benchmark and communicator config

* fix bugs of the_one_ps

* multi program and fix bug in optimizer

* multi program in the_one_ps

* public commcontext
```
  d56a0a1b
- Y
  
  [fleet exe] supprot fp16 feed and fetch on cpp side (#39758) · 73bf9673
  由 Yuang Liu 提交于 2月 22, 2022
  
  73bf9673
20 2月, 2022 1 次提交

[PTen->Phi PR1] Change pten dirname and namespace to phi (#39748) · dcfe1986

由 Chen Weihang 提交于 2月 20, 2022

* rename pten dir to phi

* rename namespace to phi

* rename infrt pten dir to phi

* resolve conflict

* rename pten to phi in cmake

* revert all infrt change

* change needed files

* fix infrt failed

* fix inference failed

dcfe1986

19 2月, 2022 2 次提交

[Pten]Unify paddle/pten::framework::ddim into pten::ddim (#39614) · 2fe04264

由 Aurelius84 提交于 2月 19, 2022

* Unify paddle/pten::framework::ddim into pten::ddim

* fix paddle namespace

* compile sucessfully

* fix npu src file

* fix conflict

* fix conflict

* fix tensorrt compiler error

* fix conflict

* fix conflict

* fix tesst file conflict

* fix conflict

* fix mlu file conflict

* fix mlu file conflict

* fix cinn header file conflict

* fix conflict

* fix conflict

* fix conflict

* fix conflict

2fe04264

Update record interface using part1 (#39693) · eec6ef81

由 chenjian 提交于 2月 19, 2022

* fix RecordEvent interface

* modify default level to 4

* update interface use

* add const default trace level

* update record event interface using

* update operator.cc

* update part1

* fix include profiler.h header in ps server

* fix include profiler.h header in ps server

eec6ef81

18 2月, 2022 2 次提交

[AMP] support GPU BF16 amp for dygraph (#39029) · 7d6d3848

由 zhangbo9674 提交于 2月 18, 2022

* support dtype param for auto_cast

* add amp_dtype for tracer

* add unsupported bf16 list

* support bf16 amp for O2

* refine python interface for bfloat16

* refine code

* refine code

* refine unittest

* refine code

* refine code

* add bf16 o1

* refine code by comment

* add gradient accumulator

* add recompute

7d6d3848

S
add tool: print kernel signaturs (#39670) · 03b875a8
由 Shang Zhizhou 提交于 2月 18, 2022
```
* add tool: print kernel signaturs

* fix windows compile
```
03b875a8

16 2月, 2022 3 次提交

Y

[fleet exe] Update comm init for dist model (#39603) · 7d53a288
由 Yuang Liu 提交于 2月 16, 2022

7d53a288

EagerTensor to EagerVariable (#39447) · 831fd86e

由 Jiabin Yang 提交于 2月 16, 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

* refine code

* Rename all EagerTensor to Tensor

* Rename some EagerTensor to Tensor

* rename EagerTensor to EagerVariable

* add more test

* merge develop and refine code

831fd86e

W

Support nce in eager mode (#39589) · 672def6c
由 Weilong Wu 提交于 2月 16, 2022

672def6c

15 2月, 2022 1 次提交

[PluggableDevice] Add custom runtime support (#38740) · 3e7825f3

由 ronnywang 提交于 2月 15, 2022

* [CustomRuntime] Add DeviceManager

* [CustomRuntime] Add DeviceInterface

* [CustomRuntime] Add Stream, Event, DeviceGuard, CallbackManager

* [CustomRuntime] Add plug-in device

* [CustomRuntime] Memory module support PluggableDevice

* [CustomRuntime] Add WITH_PLUGGABLE_DEVICE cmake option

* update

* [API] update API doc based on comments, test=develop
Co-authored-by: Nqili93 <qili93@qq.com>

3e7825f3

机器未来 / Paddle 与 Fork 源项目一致

机器未来 / Paddle
与 Fork 源项目一致