- 03 11月, 2022 10 次提交
-
-
由 pangyoki 提交于
* add cudnn into whl package * add cudnn dso into whl package * let WITH_CUDNN_DSO be consistent with WITH_GPU * fix WITH_CUDNN_DSO in paddle_build
-
由 Aurelius84 提交于
* Add six for @to_static and add CI approval * add six * refine logic * delete approve * fix return * add ImportError * typo `moduels` -> `modules` Co-authored-by: NNyakku Shigure <sigure.qaq@gmail.com> -
由 zhouweiwei2014 提交于 -
由 kuizhiqing 提交于
-
由 wenbin 提交于
-
由 Wang Xin 提交于
* remove unused-variable warning in linux * fix unused-variable error in GpuPS
-
由 Chang Xu 提交于
-
由 zlsh80826 提交于
-
由 zhangkaihuo 提交于
-
由 YuanRisheng 提交于
-
- 02 11月, 2022 24 次提交
-
-
由 JYChen 提交于
-
由 Zhang Jun 提交于
* add bilinear OutSize
-
由 Leo Chen 提交于
-
由 zhangbo9674 提交于
* fix ci bug * test
-
由 Zhang Zheng 提交于
* Support NHWC layout in GroupNorm * fix cteset
-
由 tianshuo78520a 提交于
* add cuda117 dockerfile; test=cuda117 * notest;test=cuda117 * test=cuda117 * test=document_fix
-
由 丁一 提交于
-
由 Haohongxiang 提交于
-
由 Ruibiao Chen 提交于
* Dispath computation OPs before communication in standalone executor * Update code * Fix CI errors
-
由 Tao Luo 提交于
-
由 Chen Weihang 提交于
* add phi core file into ci list, test=document_fix * remove repated file, test=document_fix
-
由 Chen Weihang 提交于
* add storage properties for npu * fix compile failed * fix api name mismatch * polish design
-
由 YuanRisheng 提交于
* Standardise batch norm * standardize conv3d and depwise_conv2d * fix ci bugs
-
由 zhouweiwei2014 提交于
-
由 Leo Chen 提交于
-
由 Kevin吴嘉文 提交于
-
由 risemeup1 提交于
* 修改.gitigore文件，把ljd_sh文件忽略掉 * 修复改动单测文件没有触发精准测试的问题 * 取消改动.gitignore * 修复改动单测没有出发精准测试的问题 * 修改变量名含义更加容易理解,test=coverage
-
由 Yiqun Liu 提交于
Improve the tool for checking nan and inf, and support to compute the max, min and mean of output tensor. (#47095) * Improve the tool for checking nan and inf, and support to compute the max, min and mean of output tensor. * Add a FLAGS to control whether abort when meets inf/nan and polish codes. * Fix unittest. * Change the computing of mean.
-
由 ShenLiang 提交于
* add unbalanced data * fix utest
-
由 zyfncg 提交于
* support generating static code of high order grad op by yaml * polish code
-
由 houj04 提交于
* [XPU] add int64 support for slice and subtract. test=kunlun * try to fix xpu compile. test=kunlun * try to fix xpu compile. test=kunlun * try to fix xpu compile. test=kunlun * remove unnecessary modification. test=kunlun
-
由 zhangkaihuo 提交于
-
由 Tian Zheng 提交于
* Add build option for CUDNN Frontend API * Fix review comments * Change namespace for cudnn_frontend.h
-
- 01 11月, 2022 6 次提交
-
-
由 Nyakku Shigure 提交于
* [CodeStyle][E711][E712] update flake8 config * empty commit, test=document_fix
-
由 Nyakku Shigure 提交于
* [CodeStyle][E711] use `is`/`is not` for comparison with `None` * `self.assertTrue($A is None)` -> `self.assertIsNone($A)` * `self.assertTrue($A is not None)` -> `self.assertIsNotNone($A)` * `self.assertFalse($A is None)` -> `self.assertIsNotNone($A)` * `self.assertEqual($A, None)` -> `self.assertIsNone($A)` * `self.assertNotEqual($A, None)` -> `self.assertIsNotNone($A)`
-
由 Leo Chen 提交于
* refine comments,test=kunlun * link xpu lib, test=kunlun * add sleep for test, test=kunlun * merge develop, fix compile, test=kunlun * remove debug code, test=kunlun * add dependency to avoid potential concurrency error, test=kunlun
-
由 HongyuJia 提交于
* move cudnn hardcode outside GetExpectedKernelType * add header file * debug * update interpreter_util with hardcode * update interpreter_util headerfile * solve activation hardcode * debug with CI * add mkldnn_op_list header file * temporarily uncomment mkldnn * temporarily uncomment mkldnn * delete sequence_softmax cudnn hardcode * add hardcode to data_transfer.cc * update data_transfer headerfile * try fix segment fault * update cudnn&miopen_helper * reset HasAttr of DygraphExctnCtx * debug, this commit should pass all CI * debug should pass CI, temporarily disable activation * debug should pass CI * fix default_attr=nullptr bug * clean debug code
-
由 Yuanle Liu 提交于
-
由 HongyuJia 提交于
-