A more memory-efficient rewrite of the HF transformers implementation of Llama for use with quantized weights.
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool