flagos-ai/TransformerEngine-FL：TransformerEngine-FL is a fork of TransformerEngine that introduces a plugin-based architecture for supporting diverse AI chips, built on top of FlagOS, a unified open-source AI system software stack.

tsingmicro

feat(backend): support tsingmicro txda backend (#88)

Add support for applying txda-related patches


Co-authored-by: malin malin@dev.com
Co-authored-by: malin malin@tsingmicro.com

4天前1459次提交

.github[CICD] Improve CUDA unit test coverage (#73)26天前
3rdparty[CICD] Improve CUDA unit test coverage (#73)26天前
benchmarksTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
build_toolsTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
docsTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
examplesTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
qa[CICD] Improve CUDA unit test coverage (#73)26天前
tests[CICD] Improve CUDA unit test coverage (#73)26天前
transformer_enginefeat(backend): support tsingmicro txda backend (#88)4天前
.clang-formatAdd auto-formatter (#919)2年前
.gitignore[CICD] Improve CUDA unit test coverage (#73)26天前
.gitmodules[Pytorch] Add Cutlass Grouped GEMM Support for fine-grained MoE Model (#2045)10个月前
.pre-commit-config.yamlTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
Acknowledgements.txtAdd FP8 fused attention (#155)3年前
CONTRIBUTING.rstTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
CPPLINT.cfgTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
LICENSEInital code drop3年前
MANIFEST.inTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
README.rstTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
SECURITY.mdTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
pylintrcFix installation from PyPI wheels after a source install (#1526)1年前
pyproject.tomlTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前
setup.pyTE-FL Upgrade: Synchronization with TE Release V2.14 (#62)2个月前

关于

TransformerEngine-FL is a fork of TransformerEngine that introduces a plugin-based architecture for supporting diverse AI chips, built on top of FlagOS, a unified open-source AI system software stack.

python cuda c++

README.md

60.5 MB

邀请码

社区
网站首页
关于我们
教学实践
合作伙伴

支持与服务
帮助中心
API文档
Git常用命令
服务协议

加入我们
官网邮箱：gitlink@ccf.org.cn

QQ群

公众号