Skip to content

Latest commit

 

History

History
17 lines (10 loc) · 820 Bytes

File metadata and controls

17 lines (10 loc) · 820 Bytes

GPTQCustomLoRA: Efficient Finetuning of Quantized LLMs with GPTQ on your custom dataset

How to use:

Open .ipynb in Jupyter or Colab Also, you can follow this link Open In Colab

IT IS FORK FROM https://github.com/qwopqwop200/gptqlora

It`s a compilation of material described here: (RU!) https://habr.com/ru/articles/751972/

And response for this issue: CodeDruidX/Everlasting-Summer-txt#1

QLoRA with AutoGPTQ for quantization

License and Intended Use

I release the resources associated with GPTQCustomLoRA finetuning in this repository under MIT license.