/QLoRA

Fine-tuning an LLM with Huggingface PEFT and QLoRA

Primary LanguageJupyter Notebook

QLoRA

Fine-tuning the Falcon-7B model on FAQ data, with Huggingface PEFT and QLoRA

Code taken from the awesome notebook by Daniel Furman with minor changes.

Hardware Requirements:

  1. NVIDIA T4 GPU
  2. RAM of at least 16 GB.