LLaMA-Factory - AI Workbench Project

This is an NVIDIA AI Workbench project to deploy LLaMA-Factory.

Please follow the LLM fine-tuning tutorial for RTX AI Toolkit here.

Minimum recommended Workbench ver: 0.50.16

Docker Desktop ver. 4.31+

Features

• Fine-tune an LLM using Llama factory.
• Fuse generated LoRA adapter with the base model.
• Apply Post Training Quantization to your model.
• Export model to TensorRT-LLM checkpoint.
• Support for unsloth.