[shardformer] shardformer support t5 model (#3994)

test t5
This commit is contained in:
wukong1992
2023-06-15 16:50:08 +08:00
committed by Frank Lee
parent 6b30dfb7ce
commit c1c672d0f0
10 changed files with 320 additions and 10 deletions

View File

@@ -15,3 +15,4 @@ einops
triton==2.0.0.dev20221202
git+https://github.com/HazyResearch/flash-attention.git@c422fee3776eb3ea24e011ef641fd5fbeb212623#egg=flash_attn
requests==2.27.1 # downgrade to avoid huggingface error https://github.com/huggingface/transformers/issues/17611
SentencePiece