radu/LLamaRecipes

miroir de https://github.com/facebookresearch/llama-recipes.git

Auteur	SHA1 Message	Date
Matthias Reso	9def4fbafd Remove micro_batch_training parameter and replace with gradient_accumulation_steps	il y a 2 ans
Brian Vaughan	3faf005226 fix a bug in the config for use_fast_kernels	il y a 2 ans
lchu	feaa344af3 resolve conflicts	il y a 2 ans
Hamid Shojanazeri	44ef280d31 adding flash attention and xformer memory efficient through PT SDPA	il y a 2 ans
lchu	895dfcea30 add nightly check for using low_cpu_fsdp mode	il y a 2 ans
chauhang	4767f09ecd Initial commit	il y a 2 ans