LoRA Toolkit

Configure and generate LoRA fine-tuning scripts for large language models. Supports Llama, Mistral, Qwen, Phi and other HuggingFace-compatible models.

Commands

config

Generate a LoRA training configuration for your model and hardware.

bash scripts/script.sh config --model llama3-8b --gpu 24gb --dataset 10000

Parameters:

• --model — base model (llama3-8b, mistral-7b, qwen2-7b, phi3-mini, llama3-70b)
• --gpu — VRAM size (8gb, 16gb, 24gb, 40gb, 80gb)
• --dataset — number of training samples