Moshe Island 08e0733e4a Support MoE for pipeline models (#5338) | 6 months ago | |
---|---|---|
.. | ||
__init__.py | b361c72761 Update DeepSpeed copyright license to Apache 2.0 (#3111) | 1 year ago |
experts.py | 971d82b573 MoE type hints (#5043) | 8 months ago |
layer.py | 08e0733e4a Support MoE for pipeline models (#5338) | 6 months ago |
mappings.py | 08e0733e4a Support MoE for pipeline models (#5338) | 6 months ago |
sharded_moe.py | 08e0733e4a Support MoE for pipeline models (#5338) | 6 months ago |
utils.py | 42a8eaa705 Auto convert moe param groups (#5354) | 6 months ago |