提交 · c3796061c385491738ef7e27e0e89fccd75877f3 · Crayon鑫 / Paddle

24 1月, 2022 1 次提交

Refactored python-level trace_op to call through _C_ops instead of... · c3796061

由 Zhanlue Yang 提交于 1月 24, 2022

Refactored python-level trace_op to call through _C_ops instead of Tracer::TraceOp, under eager_mode (#38338)

* Replaced core.ops with _C_ops

* Refactored python-level trace_op to call through _C_ops instead of Tracer::TraceOp, under eager_mode

* Modified trace_op interface

* Refactored trace_op logic for eager mode

* Added Eager Dygraph support for OpTest

* Fixed ci issues

* Fixed CI failures

* Fixed Coverage CI Issues

* Fixed XPU CI Issues

c3796061

23 1月, 2022 1 次提交

Support test_imperative apply and Add a setter for EagerTensor (#39016) · 8c5c1046

由 Weilong Wu 提交于 1月 23, 2022

* Rearranged Eager AutoCodeGen directory structure

* Removed USE_OP in Eager AutoCodeGen

* Enabled generation for Operators without Grad/Inputs/Outputs

* Resolved operators without input

* Fixed merge conflicts

* Enabled Eager AutoCodeGen for 10+ more operators

* Refactored Eager AutoCodeGen with more organized helper objects

* Enabled Eager AutoCodeGen for operators with multiple OpBases

* Adjusted Eager AutoCodeGen to Enable Passing Output Tensor as Input Argument

* Handled Dispensable Inputs/Outputs in Eager AutoCodeGen

* Adjusted function generation/call between Python-C API & Dygraph API

* Synchronized auto-generated Python-C API with Dygraph Forward Functions

* support more eager tensor api

* fix merge compile error

* fix compile error and fit develop code

* support pure CPU

* fix some logic error in eager_mode

* support _varbase_creator in eager mode

* Added safe_initialized interface to EagerTensor for use in processing dispensable inputs

* for eager mode

* refine

* support multiple constructor for eager tensor

* add place related code

* polish code

* specific randint with dtype of int64

* Support pure cpu test

* eager logic

* refine test in pure cpu

* eager logic

* eager logic

* eager logic, test=develop

* skip core.eager when in inference, test=develop

* refine, test=develop

* refine, test=develop

* call RetainGrad after run forward kernel, test=develop

* refine, test=develop

* support dygraph util, meta, guard test

* eager test case

* support inference test

* refine test and fix initializer failed

* modify eagertensor patch method

* add eagertensor.clear_grandint, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* support create varbase and fix retain grad error

* call monkey_patch_varbase in _test_eager_guard, test=develop

* fix windows error

* split clear_gradient to clear_gradient and zero_grads, test=develop

* refine, test=develop

* refine, test=develop

* support test_imperative_basic test in eager mode

* remove additional log in variable.h

* remove additional log in variable.h

* remove additional code create in merge

* eager

* fix some eager logic, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* patch_tensor_method_func, test=develop

* refine, test=develop

* eager test case, test=develop

* refine, test=develop

* eager, test=develop

* eager, test=develop

* eager optimizer, test=develop

* eager optimizer, test=develop

* eager test_imperative_optimizer_v2, test=develop

* eager, test=develop

* refine, test=develop

* refine, test=develop

* eager, test=develop

* add resize in share buffer to, test=develop

* eager, test=develop

* fix _share_buffer_to, test=develop

* refine, test=develop

* refine, test=develop

* support eager for dataloader,test=develop

* Exposed EagerTensor's set func to implement set_value func

* Rename set to _set_value, Supplement the corresponding test case

* fix test concat dev api build failed

* fix conflict

* fix conflict

* Use extern to Polish code
Co-authored-by: Njim19930609 <jim19930609@gmail.com>
Co-authored-by: NJiabinYang <360788950@qq.com>
Co-authored-by: NWang Huan <wanghuan29@baidu.com>
Co-authored-by: Nwanghuancoder <wanghuancoder@163.com>
Co-authored-by: Nchentianyu03 <chentianyu03@baidu.com>

8c5c1046

20 1月, 2022 1 次提交

[Eager] Support Eager mode for some testcase (#38783) · d21074cd

由 wanghuancoder 提交于 1月 20, 2022

* Rearranged Eager AutoCodeGen directory structure

* Removed USE_OP in Eager AutoCodeGen

* Enabled generation for Operators without Grad/Inputs/Outputs

* Resolved operators without input

* Fixed merge conflicts

* Enabled Eager AutoCodeGen for 10+ more operators

* Refactored Eager AutoCodeGen with more organized helper objects

* Enabled Eager AutoCodeGen for operators with multiple OpBases

* Adjusted Eager AutoCodeGen to Enable Passing Output Tensor as Input Argument

* Handled Dispensable Inputs/Outputs in Eager AutoCodeGen

* Adjusted function generation/call between Python-C API & Dygraph API

* Synchronized auto-generated Python-C API with Dygraph Forward Functions

* support more eager tensor api

* fix merge compile error

* fix compile error and fit develop code

* support pure CPU

* fix some logic error in eager_mode

* support _varbase_creator in eager mode

* Added safe_initialized interface to EagerTensor for use in processing dispensable inputs

* for eager mode

* refine

* support multiple constructor for eager tensor

* add place related code

* polish code

* specific randint with dtype of int64

* Support pure cpu test

* eager logic

* refine test in pure cpu

* eager logic

* eager logic

* eager logic, test=develop

* skip core.eager when in inference, test=develop

* refine, test=develop

* refine, test=develop

* call RetainGrad after run forward kernel, test=develop

* refine, test=develop

* support dygraph util, meta, guard test

* eager test case

* support inference test

* refine test and fix initializer failed

* modify eagertensor patch method

* add eagertensor.clear_grandint, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* support create varbase and fix retain grad error

* call monkey_patch_varbase in _test_eager_guard, test=develop

* fix windows error

* split clear_gradient to clear_gradient and zero_grads, test=develop

* refine, test=develop

* refine, test=develop

* support test_imperative_basic test in eager mode

* remove additional log in variable.h

* remove additional log in variable.h

* remove additional code create in merge

* eager

* fix some eager logic, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* patch_tensor_method_func, test=develop

* refine, test=develop

* eager test case, test=develop

* refine, test=develop

* eager, test=develop

* eager, test=develop

* eager optimizer, test=develop

* eager optimizer, test=develop

* eager test_imperative_optimizer_v2, test=develop

* eager, test=develop

* refine, test=develop

* refine, test=develop

* eager, test=develop

* add resize in share buffer to, test=develop

* eager, test=develop

* fix _share_buffer_to, test=develop

* refine, test=develop

* refine, test=develop

* support eager for dataloader,test=develop
Co-authored-by: Njim19930609 <jim19930609@gmail.com>
Co-authored-by: NJiabinYang <360788950@qq.com>

d21074cd

17 1月, 2022 1 次提交
- J
  Support auto prune logic in eager mode (#38960) · f81569e3
  由 Jiabin Yang 提交于 1月 17, 2022
```
* support test_auto_prune_partial

* support rest of autoprune strategy in eager mode
```
  f81569e3
11 1月, 2022 3 次提交

Z
[AMP] Check call order of paddle.amp.decorate and paddle.DataParallel (#38785) · fbb40281
由 zhangbo9674 提交于 1月 11, 2022
```
* check amp.decorate and DataParallel

* refine coverage

* fix layer dtype

* refine code
```
fbb40281

Jit pre save hook (#38186) · e91f7c02

由 Ming-Xu Huang 提交于 1月 11, 2022

* Pre-save hooks of jit.save

1. Added pre_save_hooks features to jit.save.
2. Added related unittests

* Added jit pre_save_hooks functions's alias to paddle.jit and copyright.

* Make jit.save_pre_hook style be consisent with Paddle's rule.

* Fixed arguments passing bug in run_save_pre_hooks

* Added API Documents

* Move clear and run_pre_save_hooks as internal methonds only.

* Made register_save_pre_hook as an internal function.

e91f7c02

[Eager] fix some eager logic (#38576) · d3686471

由 wanghuancoder 提交于 1月 11, 2022

* Rearranged Eager AutoCodeGen directory structure

* Removed USE_OP in Eager AutoCodeGen

* Enabled generation for Operators without Grad/Inputs/Outputs

* Resolved operators without input

* Fixed merge conflicts

* Enabled Eager AutoCodeGen for 10+ more operators

* Refactored Eager AutoCodeGen with more organized helper objects

* Enabled Eager AutoCodeGen for operators with multiple OpBases

* Adjusted Eager AutoCodeGen to Enable Passing Output Tensor as Input Argument

* Handled Dispensable Inputs/Outputs in Eager AutoCodeGen

* Adjusted function generation/call between Python-C API & Dygraph API

* Synchronized auto-generated Python-C API with Dygraph Forward Functions

* support more eager tensor api

* fix merge compile error

* fix compile error and fit develop code

* support pure CPU

* fix some logic error in eager_mode

* support _varbase_creator in eager mode

* Added safe_initialized interface to EagerTensor for use in processing dispensable inputs

* for eager mode

* refine

* support multiple constructor for eager tensor

* add place related code

* polish code

* specific randint with dtype of int64

* Support pure cpu test

* eager logic

* refine test in pure cpu

* eager logic

* eager logic

* eager logic, test=develop

* skip core.eager when in inference, test=develop

* refine, test=develop

* refine, test=develop

* call RetainGrad after run forward kernel, test=develop

* refine, test=develop

* support dygraph util, meta, guard test

* eager test case

* support inference test

* refine test and fix initializer failed

* modify eagertensor patch method

* add eagertensor.clear_grandint, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* support create varbase and fix retain grad error

* call monkey_patch_varbase in _test_eager_guard, test=develop

* fix windows error

* split clear_gradient to clear_gradient and zero_grads, test=develop

* refine, test=develop

* refine, test=develop

* support test_imperative_basic test in eager mode

* remove additional log in variable.h

* remove additional log in variable.h

* remove additional code create in merge

* eager

* fix some eager logic, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop
Co-authored-by: Njim19930609 <jim19930609@gmail.com>
Co-authored-by: NJiabinYang <360788950@qq.com>

d3686471

10 1月, 2022 1 次提交
- S
  
  fix bug of fp16 (#38838) · 7d4ce5b3
  由 ShenLiang 提交于 1月 10, 2022
  
  7d4ce5b3
05 1月, 2022 1 次提交

[Eager] Support test imperative basic in eager test_empty_grad (#38376) · 9108e777

由 wanghuancoder 提交于 1月 05, 2022

* Rearranged Eager AutoCodeGen directory structure

* Removed USE_OP in Eager AutoCodeGen

* Enabled generation for Operators without Grad/Inputs/Outputs

* Resolved operators without input

* Fixed merge conflicts

* Enabled Eager AutoCodeGen for 10+ more operators

* Refactored Eager AutoCodeGen with more organized helper objects

* Enabled Eager AutoCodeGen for operators with multiple OpBases

* Adjusted Eager AutoCodeGen to Enable Passing Output Tensor as Input Argument

* Handled Dispensable Inputs/Outputs in Eager AutoCodeGen

* Adjusted function generation/call between Python-C API & Dygraph API

* Synchronized auto-generated Python-C API with Dygraph Forward Functions

* support more eager tensor api

* fix merge compile error

* fix compile error and fit develop code

* support pure CPU

* fix some logic error in eager_mode

* support _varbase_creator in eager mode

* Added safe_initialized interface to EagerTensor for use in processing dispensable inputs

* for eager mode

* refine

* support multiple constructor for eager tensor

* add place related code

* polish code

* specific randint with dtype of int64

* Support pure cpu test

* eager logic

* refine test in pure cpu

* eager logic

* eager logic

* eager logic, test=develop

* skip core.eager when in inference, test=develop

* refine, test=develop

* refine, test=develop

* call RetainGrad after run forward kernel, test=develop

* refine, test=develop

* support dygraph util, meta, guard test

* eager test case

* support inference test

* refine test and fix initializer failed

* modify eagertensor patch method

* add eagertensor.clear_grandint, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop

* call monkey_patch_varbase in _test_eager_guard, test=develop

* split clear_gradient to clear_gradient and zero_grads, test=develop

* refine, test=develop

* refine, test=develop

* refine, test=develop
Co-authored-by: Njim19930609 <jim19930609@gmail.com>
Co-authored-by: NJiabinYang <360788950@qq.com>

9108e777

04 1月, 2022 1 次提交

[Dy2st]Fix error when set buffer in forward (#38540) · 1e3f01ed

由 0x45f 提交于 1月 04, 2022

* fix error when set buffer in forward

* add unittest

* refine class name

* refine not framework.in_dygraph_mode() in if

* fix UT error

* add comment

* refine code

* remove useless import

1e3f01ed

29 12月, 2021 2 次提交
- S
  [BugFix]Fix bug in obtaining parameters_buffers in layers (#38563) · ecb8c184
  由 ShenLiang 提交于 12月 29, 2021
```
* fix bug of dp in pfp16

* fix topo
```
  ecb8c184
- Z
  [AMP] Add BatchNorm_1D_2D_3D skip for paddle.amp.decorate (#38541) · 2ebc8f77
  由 zhangbo9674 提交于 12月 29, 2021
```
* add bn_1d_2d_3d for fp16 decorate

* add unittest
```
  2ebc8f77
28 12月, 2021 1 次提交

Fix scatter_op fp16 perf problem. (#38499) · 33ce249f

由 Li Min 提交于 12月 28, 2021

* Fix scatter_op fp16 perf problem.

* Add scatter into black list.

* Add scatter into black list for dygraph.

33ce249f

27 12月, 2021 2 次提交
- S
  
  fix bugs in fp16 for dp (#38405) · 1ab5c511
  由 ShenLiang 提交于 12月 27, 2021
  
  1ab5c511
- Z
  [AMP] Fix amp.decorate bug: parameters for non leaf layers cannot be decotated (#38402) · 5d902954
  由 zhangbo9674 提交于 12月 27, 2021
```
* fix bug

* refine code

* refine code

* refine code
```
  5d902954
24 12月, 2021 3 次提交

[Dy2stat]Fix error when calling sublayer's non-forward func in dy2stat (#37296) · 7339a124

由 0x45f 提交于 12月 24, 2021

* fix error when calling sublayer's non-forward func in dy2stat

* fix circular import using an inelegant way

* deal with parameters

* remove param_guard in __call__

* remove comment

* fix error when jit.load

* rename block var

* remove wrong code

* add unit test

7339a124

A
[Dy2Stat]Consider InputSpec.name to calculate Cachekey hash id (#38273) · 8e6d5d2b
由 Aurelius84 提交于 12月 24, 2021
```
* Consider InputSpec.name to calculate Cachekey hash id

* fix function
```
8e6d5d2b

Support test imperative basic in eager (#38313) · d48f7c89

由 Jiabin Yang 提交于 12月 24, 2021

* Rearranged Eager AutoCodeGen directory structure

* Removed USE_OP in Eager AutoCodeGen

* Enabled generation for Operators without Grad/Inputs/Outputs

* Resolved operators without input

* Fixed merge conflicts

* Enabled Eager AutoCodeGen for 10+ more operators

* Refactored Eager AutoCodeGen with more organized helper objects

* Enabled Eager AutoCodeGen for operators with multiple OpBases

* Adjusted Eager AutoCodeGen to Enable Passing Output Tensor as Input Argument

* Handled Dispensable Inputs/Outputs in Eager AutoCodeGen

* Adjusted function generation/call between Python-C API & Dygraph API

* Synchronized auto-generated Python-C API with Dygraph Forward Functions

* support more eager tensor api

* fix merge compile error

* fix compile error and fit develop code

* support pure CPU

* fix some logic error in eager_mode

* support _varbase_creator in eager mode

* Added safe_initialized interface to EagerTensor for use in processing dispensable inputs

* for eager mode

* refine

* support multiple constructor for eager tensor

* add place related code

* polish code

* specific randint with dtype of int64

* Support pure cpu test

* eager logic

* refine test in pure cpu

* eager logic

* eager logic

* eager logic, test=develop

* skip core.eager when in inference, test=develop

* refine, test=develop

* refine, test=develop

* call RetainGrad after run forward kernel, test=develop

* refine, test=develop

* support dygraph util, meta, guard test

* support inference test

* refine test and fix initializer failed
Co-authored-by: Njim19930609 <jim19930609@gmail.com>
Co-authored-by: NWang Huan <wanghuan29@baidu.com>

d48f7c89

21 12月, 2021 1 次提交

Fix inplace problem of setitem (#38298) · da61df5c

由 zyfncg 提交于 12月 21, 2021

* add inplace_map for trace_op in pybind

* fix inplace problem of setitem

* refactor the param format  of trace_op
Co-authored-by: Npangyoki <pangyoki@126.com>

da61df5c

15 12月, 2021 3 次提交
- Y
  Fix bugs in Translated Layer when change mode from train/eval to eval/train (#38141) · f616d9a7
  由 YuanRisheng 提交于 12月 15, 2021
```
* fix bugs in Translated layer when change train/eval

* fix python converage
```
  f616d9a7
- 0
  [Dy2stat]Fix error in tensor_shape_transformer. (#37999) · 50822491
  由 0x45f 提交于 12月 15, 2021
```
* fix error when tensor_shape_transformer. Before in stmt like `if len(paddle.shape(x)[0]) > 0`, `paddle` will be used as a variable

* handle other call like `fluid.layers.mean` and `fluid.layers.shape`

* add unit test
```
  50822491
- Z
  
  add use_warning of amp (#38086) · 3a2093a5
  由 zhangbo9674 提交于 12月 15, 2021
  
  3a2093a5
14 12月, 2021 1 次提交
- C
  layer.to api support numpy.dtype and paddle.dtype (#38018) · f1c28236
  由 chentianyu03 提交于 12月 14, 2021
```
* layer.to api support numpy.dtype and paddle.dtype

* skip the layer to eaxmple code executing for env not support
```
  f1c28236
13 12月, 2021 1 次提交
- 0
  [Dy2Stat]Allow ifelse return buildin type in paddle cond (#37888) · 9598b19c
  由 0x45f 提交于 12月 13, 2021
```
* allow ifelse return `int` in paddle cond

* add test and refine code

* polish code, add test

* code format
```
  9598b19c
10 12月, 2021 2 次提交
- K
  
  fix ndiv for npu (#37998) · 11c785a4
  由 kuizhiqing 提交于 12月 10, 2021
  
  11c785a4
- 0
  [Dy2stat]Remove all comments of users code when dy2stat (#38003) · 515d3562
  由 0x45f 提交于 12月 10, 2021
```
* remove outer comment when dy2stat

* remove all comment

* add unit test
```
  515d3562
07 12月, 2021 1 次提交
- 0
  [Dy2Stat]Polish for zip in dy2stat (#37846) · 4e63d69b
  由 0x45f 提交于 12月 07, 2021
```
* polish for zip in dy2stat

* polish comment

* polish is_builtin_len

* fix comment
```
  4e63d69b
06 12月, 2021 1 次提交
- K
  
  heter for collective (#37613) · 1bdb8578
  由 kuizhiqing 提交于 12月 06, 2021
  
  1bdb8578
02 12月, 2021 2 次提交
- Z
  
  refine found_inf of loss_scaler (#37770) · cc2b4662
  由 zhangbo9674 提交于 12月 02, 2021
  
  cc2b4662
- J
  optimize dygraph performance with move runtime import to begining (#37759) · bfb85779
  由 Jiabin Yang 提交于 12月 02, 2021
```
* optimize dygraph probl

* refine code

* fix convert dtype error

* fix import datafeeder error
```
  bfb85779
01 12月, 2021 2 次提交
- J
  Remove cpp layer (#37730) · 44def66a
  由 Jiabin Yang 提交于 12月 01, 2021
```
* optimizer __call__ to make dygraph faster

* fix return type

* remove cpp Layer
```
  44def66a
- J
  optimizer __call__ to make dygraph faster (#37713) · 370864dd
  由 Jiabin Yang 提交于 12月 01, 2021
```
* optimizer __call__ to make dygraph faster

* fix return type
```
  370864dd
29 11月, 2021 1 次提交

[AMP] For `amp.decorate()` optimizers set to None is ok (#37541) · 2bb3f0b5

由 zhangbo9674 提交于 11月 29, 2021

* amp.decorate optimizers set to None is ok

* refine unittest

* add unittest and refine example code

* refine unittest

2bb3f0b5

26 11月, 2021 1 次提交
- S
  fix data parallel when VOCAB var in program (#37543) · e05540f7
  由 Steffy-zxf 提交于 11月 26, 2021
```
* fix data parallel when VOCAB var in program
```
  e05540f7
24 11月, 2021 1 次提交

[Dy2stat]support pure fp16 for dy2stat (#36944) · 52edad6a

由 0x45f 提交于 11月 24, 2021

* run dy2stat pure fp16 in Linear model

* no use self._pure_fp16_inputs

* add test and fix Adam error in dy2stat pure fp16 training

* use paddle.optimizer.Adam

* run test in gpu

* change test time for CI

* enlarge atol for test_resnet_pure_fp16

* refine code and enlarge atol

* make custom_white_list and custom_black_list take effect for AMP and pure fp16

* check tracer is not None

* use default atol

* change filter_size

* change atol and add some NOTE

52edad6a

23 11月, 2021 1 次提交
- R
  [NPU] Added HCCL backend support in dygraph mode (#36285) · 83e55cff
  由 ronnywang 提交于 11月 23, 2021
```
* Added HCCL backend support in dynamic graph mode

* fix segmentation fault

* add ut
```
  83e55cff
22 11月, 2021 1 次提交

[Dy2stat]Allow users to switch eval/train mode when using @to_static to... · eb602398

由 0x45f 提交于 11月 22, 2021

[Dy2stat]Allow users to switch eval/train mode when using @to_static to decorate a function (#37383)

* Allow users to switch eval/train mode when using @to_static to decorate a function

* refine code for train() and eval()

eb602398

19 11月, 2021 1 次提交
- 0
  [Dy2stat]Support `for i in [1,2,3]` statements in dy2stat (#37259) · d772a9aa
  由 0x45f 提交于 11月 19, 2021
```
* support `for i in [1,2,3]` statements in dy2stat

* add test case

* fix ci

* remove wrong code
```
  d772a9aa
18 11月, 2021 1 次提交
- Z
  
  Fix Layer.to() of device bug (#37156) · 706a7897
  由 zhangbo9674 提交于 11月 18, 2021
  
  706a7897
16 11月, 2021 1 次提交
- W
  
  Fix the logic of VarBase _to func (#37193) · f29a3c68
  由 Weilong Wu 提交于 11月 16, 2021
  
  f29a3c68

Crayon鑫 / Paddle 与 Fork 源项目一致

Crayon鑫 / Paddle
与 Fork 源项目一致