LLaMA Factory is a simple, efficient platform for training and fine-tuning Large Language Models (LLMs). With LLaMA Factory you can fine-tune hundreds of pretrained models locally without writing a single line of code. Key features include:

• Supported model families: LLaMA, LLaVA, Mistral, Mixtral-MoE, Qwen, Yi, Gemma, Baichuan, ChatGLM, Phi, and many more.
• Training algorithms: (Incremental) pre-training, (multimodal) instruction-tuned fine-tuning, reward-model training, PPO, DPO, KTO, ORPO, etc.
• Precision options: 16-bit full-parameter fine-tuning, frozen-parameter tuning, LoRA tuning, and 2/3/4/5/6/8-bit QLoRA tuning based on AQLM, AWQ, GPTQ, LLM.int8, HQQ, or EETQ.
• Optimization techniques: GaLore, BAdam, DoRA, LongLoRA, LLaMA Pro, Mixture-of-Depths, LoRA+, LoftQ, PiSSA.
• Acceleration kernels: FlashAttention-2 and Unsloth.
• Inference engines: Transformers and vLLM.
• Experiment tracking: LlamaBoard, TensorBoard, Weights & Biases, MLflow, SwanLab, and more.

By using this template, you can start the WebUI immediately and fine-tune models directly through an interactive interface — no coding required.