提交 · 530deb144e25be52adc16999cdcab67cfe39f092 · PaddlePaddle / Paddle

15 11月, 2020 1 次提交

裁剪transformer模型trt支持；修复tensorRT不支持DeletePass的bug (#28517) (#28580) · 530deb14

由 Shang Zhizhou 提交于 11月 15, 2020

* 裁剪transformer模型trt支持；修复tensorRT不支持DeletePass的bug (#28517)

* skip_layernorm_op done

* add unittest

* slice op convertor support trt < 6

* skip_layernorm only work in ernie

* fix unittest

530deb14

23 10月, 2020 1 次提交
- Z
  
  fix CUDA9 error due to BuildCustomizations (#28226) · 24566d32
  由 Zhou Wei 提交于 10月 23, 2020
  
  24566d32
22 10月, 2020 1 次提交
- Z
  
  fix Automatic GPU detection failed on windows (#28148) (#28182) · 4ccc1716
  由 Zhou Wei 提交于 10月 22, 2020
  
  4ccc1716
21 10月, 2020 2 次提交
- W
  
  update. (#28166) · c008ae69
  由 Wilber 提交于 10月 21, 2020
  
  c008ae69
- Z
  
  fix dynamic_loader more safe and error message on windows (#28144) · dc53ba87
  由 Zhou Wei 提交于 10月 21, 2020
  
  dc53ba87
20 10月, 2020 1 次提交
- L
  build gloo from source code instead of using the pre-compiled library (#27930) (#28009) · c4e18dc0
  由 lilong12 提交于 10月 20, 2020
```
* build gloo from source code , test=develop
```
  c4e18dc0
12 10月, 2020 2 次提交
- W
  
  Lite subgraph support arm cpu. (#27827) · 9005c5a2
  由 Wilber 提交于 10月 12, 2020
  
  9005c5a2
- add musl option (#27798) · 6335e6a0
  由 chen.zhiyu 提交于 10月 12, 2020
  
  6335e6a0
11 10月, 2020 1 次提交
- W
  
  update for windows compile. (#27813) · a2d08aa9
  由 Wilber 提交于 10月 11, 2020
  
  a2d08aa9
01 10月, 2020 1 次提交
- W
  
  Added support for quantization of fusion_gru (#27518) · 966447e3
  由 Wojciech Uss 提交于 10月 01, 2020
  
  966447e3
27 9月, 2020 2 次提交

add support to float64 input of warpctc op. (#27399) · 1501a80f

由 Li Fuchen 提交于 9月 27, 2020

* add float64 input to ctc_loss

* modified error message of  warpctc

* update repo and tag of warpctc

* add test for warpctc with float64 input

* modified warpctc.cmake to make sure build always

* resolved sample code bug of warpctc

* add core.ops in warpctc dygraph

* fix a bug of test

1501a80f

support elementwise add, activation, matmul on Baidu Kunlun (#27143) · 6b727e08

由 QingshuChen 提交于 9月 27, 2020

* support elementwise add, activation, matmul on Baidu Kunlun
* test=kunlun

* minor
* test=kunlun

* reconstuct the xpu directory
* test=kunlun

* minor
* test=kunlun

* minor
* test=kunlun

* minor
* test=kunlun

* minor
* test=kunlun

* minor
* test=kunlun

6b727e08

25 9月, 2020 2 次提交

add xpu in heter mode (#27000) · 6f69a4cb

由 Thunderbrook 提交于 9月 25, 2020

* add xpu in heter mode
test=develop

* BOOST_CONST_GET; PADDLE_THROW
test=develop

* code style
test=develop

* code style
test=develop

* code style
test=develop

* refine
test=develop

* refine
test=develop

* refine
test=develop

* refine code
test=develop

6f69a4cb

update gcc8 in python3 ci docker (#26979) · a2e0b7cb

由 tianshuo78520a 提交于 9月 25, 2020

* update gcc8 in python3 ci docker

* change cuda 10.2

* update cudnn8

* nvidia error cuda10.2-cudnn8-centos6 images

* fix third cache

a2e0b7cb

24 9月, 2020 2 次提交
- Z
  add unittest count ,install check on windows (#27492) · d20349b5
  由 Zhou Wei 提交于 9月 24, 2020
```
* add unittest count of windows

* Reduce the number of retries
```
  d20349b5
- W
  
  windows lib size crop from 5.4G to 3.9G (#27477) · ec4155d7
  由 Wilber 提交于 9月 24, 2020
  
  ec4155d7
23 9月, 2020 2 次提交
- Z
  
  fix bug MD of compile, And add MD/STATIC/OPENBLAS inference lib check on windows (#27051) · 292b24aa
  由 Zhou Wei 提交于 9月 23, 2020
  
  292b24aa
- Z
  add fuse_bn_act op (#27230) · 906e7f92
  由 Zhang Ting 提交于 9月 23, 2020
```
* add fused_bn_add_relu op
```
  906e7f92
22 9月, 2020 1 次提交
- W
  
  Rename fluid_inference to paddle_inference. (#27422) · afe94903
  由 Wilber 提交于 9月 22, 2020
  
  afe94903
18 9月, 2020 1 次提交
- P
  
  Optimize emb_eltwise_layernorm_plugin and support fp16 (#27128) · a5ef246c
  由 Pei Yang 提交于 9月 18, 2020
  
  a5ef246c
16 9月, 2020 1 次提交
- L
  
  Fix bug of handling blank characters in operators.cmake (#27310) · c89f269c
  由 Leo Chen 提交于 9月 16, 2020
  
  c89f269c
15 9月, 2020 1 次提交
- C
  
  Set timeout value on windows and mac (#27197) · cb34cf18
  由 chalsliu 提交于 9月 15, 2020
  
  cb34cf18
14 9月, 2020 1 次提交
- A
  Add int8 GRU kernel (#27220) · cc3f4b81
  由 Adam 提交于 9月 14, 2020
```
* Add int8 GRU kernel with UTs

* Lint fixes

* More lint fixes
```
  cc3f4b81
11 9月, 2020 1 次提交
- W
  
  Lite subgraph refine predictor (#27167) · 1b84c0bf
  由 Wilber 提交于 9月 11, 2020
  
  1b84c0bf
10 9月, 2020 1 次提交
- J
  [oneDNN]Introducing oneDNN 1.6 (#27137) · e0058615
  由 Jacek Czaja 提交于 9月 10, 2020
```
* - introducing oneDNN 1.6

test=develop

* - Removed redundant code

test=develop
```
  e0058615
09 9月, 2020 2 次提交
- W
  
  add dgc cuda11 support for Paddle (#27204) · 3b8f5200
  由 wangchaochaohu 提交于 9月 09, 2020
  
  3b8f5200
- W
  
  [cuda11 support] change the CMakeLists to support the cuda11 (#27124) · c71d79b1
  由 wangchaochaohu 提交于 9月 09, 2020
  
  c71d79b1
07 9月, 2020 1 次提交
- W
  
  update gpu compute archs for cuda11 (#27039) · 9b7692b1
  由 wangchaochaohu 提交于 9月 07, 2020
  
  9b7692b1
27 8月, 2020 1 次提交
- A
  Update Paddle-Lite commit hash. (#26413) · c2c68958
  由 Adam Osewski 提交于 8月 27, 2020
```
* Update Paddle-Lite commit hash.

* Add BF16 data type to VarTyp protobuf message.
```
  c2c68958
26 8月, 2020 2 次提交
- Y
  
  modified timeout value on windows and mac (#26690) · ada1e129
  由 YUNSHEN XIE 提交于 8月 26, 2020
  
  ada1e129
- M
  
  fix infer compile bug (#26641) · 4ba64836
  由 MRXLT 提交于 8月 26, 2020
  
  4ba64836
25 8月, 2020 1 次提交
- W
  
  Enhance py_func error info message. (#26557) · 32ba8602
  由 Wilber 提交于 8月 25, 2020
  
  32ba8602
24 8月, 2020 1 次提交

find timeout unittests (#26371) · 39fe0d35

由 YUNSHEN XIE 提交于 8月 24, 2020

* find timeout unittests

* setting timeout value

* fix some error

* fix some error

* fix some error

* fix no newline of end file error

39fe0d35

21 8月, 2020 2 次提交

support Baidu Kunlun AI Accelerator (#25959) · 138ecf24

由 QingshuChen 提交于 8月 21, 2020

* support Baidu AI Accelerator
  * test=kunlun

* minor
 * test=kunlun

* support xpu op in separate file
 * test=kunlun

* update XPU error message and remove duplicated code

 * test=kunlun

* minor
 * test=kunlun

* minor
 * test=kunlun

138ecf24

C
Revert "update cub to 1.9.8, test=develop (#24895)" (#26504) · 17bcaef4
由 Chen Weihang 提交于 8月 21, 2020
```
This reverts commit 0cb03182.
```
17bcaef4

20 8月, 2020 1 次提交
- Z
  specify cuda arch when dectected fail (#26420) · 62bd7ba1
  由 Zhou Wei 提交于 8月 20, 2020
```
specify cuda arch when dectected fail
```
  62bd7ba1
14 8月, 2020 1 次提交
- 【paddle.fleet】upgrade fleet: modify role_maker (#26038) · 935da32d
  由 vslyu 提交于 8月 14, 2020
```
* add unittest for paddlerolemaker with gloo
```
  935da32d
13 8月, 2020 1 次提交

[OpDevOptimize] Add common infershape functions (#26096) · ffe52b44

由 Leo Chen 提交于 8月 13, 2020

* add unchaged infershape function

* add broadcast infershape function

* fix bug

* rename infershape functions

* add UnaryOpUnchangedInferShapeCheckAxis

* add error message

* add test for common infer shape functions

* dont update existed ops

* dont update op_desc.h

* add more test

* add error check, refine error message

ffe52b44

11 8月, 2020 1 次提交

- oneDNN update 1.3 -> 1.5 (#25202) · 615e8a20

由 Jacek Czaja 提交于 8月 11, 2020

* - introducing oneDNN 1.6

test=develop

* - onednn 1.5

test=develop

* - oneDNN 1.5

test=develop

* - oneDNN 1.5 bugfix

test=develop

615e8a20

10 8月, 2020 1 次提交
- Z
  fix compile warning on windows MSVC, fix paddle_build.bat more safe (#25933) · 1f74b94d
  由 Zhou Wei 提交于 8月 10, 2020
```
* Fixed compile warning about incorrect compile options,fix paddle_build.bat

* fix paddle_build.bat to more safe
```
  1f74b94d

PaddlePaddle / Paddle 1 年多 前同步成功

PaddlePaddle / Paddle
1 年多前同步成功