/Easy-Finetuning-of-Llama-2-7b-using-QLoRA

Finetuning Llama-2-7b using QLoRA using TinyStories dataset

Primary LanguageJupyter Notebook

Easy-Finetuning-of-Llama-2-7b-using-QLoRA

  • Finetuning Llama-2-7b using QLoRA using TinyStories dataset

  • This finetuning was conducted on the free version of Google collab with less than 15 Gb VRAM thanks to QLora and Parameter-Efficient Fine-Tuning.