radu/LLamaRecipes @ 3f1fef7a0018ed79f31326e5f80533f12f47bdbc

镜像自地址 https://github.com/facebookresearch/llama-recipes.git

Hamid Shojanazeri 44ef280d31 adding flash attention and xformer memory efficient through PT SDPA		2 年之前
..
__init__.py	4767f09ecd Initial commit	2 年之前
datasets.py	4767f09ecd Initial commit	2 年之前
fsdp.py	a977145a9b change bf16 default to false	2 年之前
peft.py	4767f09ecd Initial commit	2 年之前
training.py	44ef280d31 adding flash attention and xformer memory efficient through PT SDPA	2 年之前