/llama_32k

Fine tuning llama on 32k tokens

Primary LanguagePython

Watchers