提交 · 58f7695ab2022319cc6a9e95348530496ed8d104 · BaiXuePrincess / Paddle

23 5月, 2019 1 次提交
- Q
  Async exe support communicator (#17386) · 58f7695a
  由 Qiao Longfei 提交于 5月 23, 2019
```
Async exe support communicator
```
  58f7695a
20 5月, 2019 1 次提交
- T
  remove unused expected_kernel_cache_pass (#17486) · 32da5e9c
  由 Tao Luo 提交于 5月 20, 2019
```
test=develop
```
  32da5e9c
17 5月, 2019 2 次提交

Y
polish parallel dygraph code (#17164) · 02175555
由 Yan Xu 提交于 5月 17, 2019
```
* add var grad hook test=develop
```
02175555

Fix/Fix memory leak in dygraph (#17394) · d7df4e5e

由 Jiabin Yang 提交于 5月 17, 2019

* test=develop, add gradient sort backward strategy

* test=develop, fix test by add FLAGS_cudnn_deterministic on new tests

* test=develop, fix memory leak in dygraph mode

* test=develop, fix memory leak in dygraph mode

* test=develop, polish code

* test=develop, polish code

* test=develop, polish code

d7df4e5e

14 5月, 2019 1 次提交

make parallel_executor support FLAGS_use_mkldnn (#17341) · 68ec0a6f

由 Tao Luo 提交于 5月 14, 2019

* make parallel_executor support FLAGS_use_mkldnn

test=develop

* add warning when set mkldnn_enabled_op_types_ in non-mkldnn env

test=develop

68ec0a6f

13 5月, 2019 1 次提交

test=develop, add gradient sort backward strategy (#17125) · 4624d7c6

由 Jiabin Yang 提交于 5月 13, 2019

* test=develop, add gradient sort backward strategy

* test=develop, fix test by add FLAGS_cudnn_deterministic on new tests

4624d7c6

12 5月, 2019 1 次提交
- C
  Add DropLocalExeScopes in ParallelExecutor (#17297) · bc833945
  由 chengduo 提交于 5月 12, 2019
```
* reset drop local scope counter
test=develop
```
  bc833945
08 5月, 2019 3 次提交

Repair api example (#17221) · e388a1fb

由 lujun 提交于 5月 08, 2019

Fix the following API examples:

paddle.fluid.scope_guard
paddle.fluid.backward.append_backward
paddle.fluid.cpu_places
paddle.fluid.cuda_pinned_places
paddle.fluid.cuda_places
paddle.fluid.in_dygraph_mode
paddle.fluid.CUDAPlace
paddle.fluid.CPUPlace
paddle.fluid.CUDAPinnedPlace

e388a1fb

C
Code Clean: Move all pass to paddle::framework::ir (#17228) · 04bd413a
由 chengduo 提交于 5月 08, 2019
```
* move pass to ir

* polish code
test=develop

* fix dependency
test=develop
```
04bd413a
Z

fix api doc,test=develop (#17241) · f2fa3f73
由 Zeng Jinle 提交于 5月 07, 2019

f2fa3f73

30 4月, 2019 1 次提交

Fix mem leak when converting Tensor to numpy array (#17182) · 5dfe2ab9

由 Zeng Jinle 提交于 4月 30, 2019

* fix mem leak when converting Tensor to numpy array
test=develop

* remove unused unittest,test=develop

* follow comments, test=develop

* fix dygraph bug,test=develop

5dfe2ab9

25 4月, 2019 1 次提交
- Y
  ParallelDyGraph with GPU collective mode (#16827) · 0b07eef1
  由 Yan Xu 提交于 4月 25, 2019
```
implement dygraph.parallel.DataParallel to hook reduce op.
```
  0b07eef1
22 4月, 2019 2 次提交
- L
  add doc for memory_optimize, test=develop (#17010) · a770ce06
  由 liuwei1031 提交于 4月 22, 2019
```
* add doc for memory_optimize, test=develop

* update doc, test=develop

* doc update, test=develop
```
  a770ce06
- W
  fix nccl wrapper on windows · 51a0243a
  由 wopeizl 提交于 4月 22, 2019
```
test=develop
```
  51a0243a
16 4月, 2019 1 次提交
- D
  
  add nccl wrapper for python API · b0911390
  由 dongdaxiang 提交于 4月 16, 2019
  
  b0911390
11 4月, 2019 1 次提交

Add an option to enable the cache of expected kernel in train phase. (#16724) · 112f1614

由 Yiqun Liu 提交于 4月 11, 2019

* Add an option to enable the cache of expected kernel in train phase.
test=develop

* Change the default value of cache_expected_kernel to true.

112f1614

08 4月, 2019 2 次提交
- C
  Add unit test for fuse all_reduce ops (#16699) · 55b15db5
  由 chengduo 提交于 4月 08, 2019
```
* test fuse all_reduce
```
  55b15db5
- Y
  Enable the runtime_context_cache pass in train phase (#16640) · 3fe8cb0d
  由 Yiqun Liu 提交于 4月 08, 2019
```
* Try to enable the runtime_context_cache pass in train phase.

* Put the append of runtime_context_cache pass ahead of multi_dev passes.
test=develop
```
  3fe8cb0d
04 4月, 2019 1 次提交
- Y
  [Imperative] implement imperative NCCLParallelContext (#16477) · b4c3a6aa
  由 Yan Xu 提交于 4月 04, 2019
```
add NCCLParallelContext for parallel dygraph
```
  b4c3a6aa
02 4月, 2019 1 次提交
- C
  Add Stream for fetch op handle (#16600) · b75a69ba
  由 chengduo 提交于 4月 02, 2019
```
* expose fuse broadcast ops
```
  b75a69ba
29 3月, 2019 4 次提交
- D
  
  add incubate for unified API · f6128777
  由 dongdaxiang 提交于 3月 13, 2019
  
  f6128777
- D
  
  fix some conflict for compilation · cc4def6b
  由 dongdaxiang 提交于 3月 08, 2019
  
  cc4def6b
- X
  
  add DataSet and InMemoryDataFeed, support load data into memory and shuffle data · 824b84d1
  由 xjqbest 提交于 3月 06, 2019
  
  824b84d1
- D
  
  add pybind for fleet · be757096
  由 dongdaxiang 提交于 2月 28, 2019
  
  be757096
28 3月, 2019 5 次提交
- C
  Fuse Adam And SGD ops (#15933) · 1096746c
  由 chengduo 提交于 3月 28, 2019
```
* fuse optimizer
```
  1096746c
- S
  check default grad maker · 2c836ff9
  由 sneaxiy 提交于 3月 28, 2019
```
test=develop
```
  2c836ff9
- C
  Fix the interface of Pass::Apply (#16484) · ed61d67c
  由 chengduo 提交于 3月 27, 2019
```
* modify the interface of Pass::Allay
test=develop

* Polish code
test=develop

* Fix Travis CI
test=develop

* fix Pass::Apply interface
test=develop

* Fix Travis CI
test=develop
```
  ed61d67c
- Z
  Revert "Fix allocator bug" · 174d0d0b
  由 Zeng Jinle 提交于 3月 28, 2019
```
add include headers to fix travis-ci
test=develop
```
  174d0d0b
- G
  
  Add DGC(Deep Gradient Compression) interface. (#15841) · eb83abea
  由 gongweibao 提交于 3月 28, 2019
  
  eb83abea
27 3月, 2019 1 次提交
- W
  Tensor index (#16223) · c300b1ba
  由 wopeizl 提交于 3月 27, 2019
```
* extend the slice function for python
test=develop
```
  c300b1ba
26 3月, 2019 4 次提交
- C
  Add doc for CPUPlace CUDAPlace CUDAPinPlace (#16442) · 4f2278f0
  由 chengduo 提交于 3月 26, 2019
```
test=develop
```
  4f2278f0
- S
  fix env variable settting bug · 78fb3a62
  由 sneaxiy 提交于 3月 26, 2019
```
test=develop
```
  78fb3a62
- X
  polish · fd24ab47
  由 Xin Pan 提交于 3月 26, 2019
```
test=develop
```
  fd24ab47
- S
  fix some op grad maker · 7000ec85
  由 sneaxiy 提交于 3月 25, 2019
```
fix ctest eager deletion disable bug
test=develop
```
  7000ec85
25 3月, 2019 2 次提交
- S
  try to fix ci error · f8ed2c22
  由 sneaxiy 提交于 3月 24, 2019
```
test=develop
```
  f8ed2c22
- S
  split PR · c20db635
  由 sneaxiy 提交于 3月 25, 2019
```
test=develop
```
  c20db635
24 3月, 2019 1 次提交
- S
  add op registry type · a93a9eef
  由 sneaxiy 提交于 3月 22, 2019
```
refine gc code
test=develop
```
  a93a9eef
21 3月, 2019 1 次提交

add more unittest · 953214ad

由 sneaxiy 提交于 3月 19, 2019

modify allocator strategy
remove changes of legacy buddy_allocator
test=develop

953214ad

20 3月, 2019 1 次提交

Fuse AllReduce (#15921) · f26ba5bd

由 chengduo 提交于 3月 19, 2019

* fuse all_reduce
test=develop

* add fuse_parameter_groups_size
test=develop

* Polish code
test=develop

* Fix travis-ci
test=develop

* Add SetGroupAccordingToLayers and SetGroupAccordingToGroupSize
test=develop

* Add SetGroupAccordingToMemorySize
test=develop

* fix multi_devices_graph
test=develop

* reset params_grads
test=develop

* Polish code
test=develop

f26ba5bd

19 3月, 2019 1 次提交
- Z
  add allocator flags · 22715487
  由 zhhsplendid 提交于 3月 19, 2019
```
test=develop
```
  22715487

BaiXuePrincess / Paddle 与 Fork 源项目一致

BaiXuePrincess / Paddle
与 Fork 源项目一致