QLoRA 4-bit finetuning of LLMs

QLoRA 4-bit finetuning of LLMs(github.com)

7 points by kashifr 2 years ago | 1 comment

kashifr 2 years ago |

An efficient finetuning approach that reduces memory usage enough to finetune a 65B parameter model on a single 48GB GPU while preserving full 16-bit finetuning task performance!

kashifr 2 years ago |

An efficient finetuning approach that reduces memory usage enough to finetune a 65B parameter model on a single 48GB GPU while preserving full 16-bit finetuning task performance!