LLaMA-Factory使用命令

We provide diverse examples about fine-tuning LLMs.

Make sure to execute these commands in the LLaMA-Factory directory.

[LoRA Fine-Tuning](#LoRA Fine-Tuning)
[QLoRA Fine-Tuning](#QLoRA Fine-Tuning)
[Full-Parameter Fine-Tuning](#Full-Parameter Fine-Tuning)
[Merging LoRA Adapters and Quantization](#Merging LoRA Adapters and Quantization)
[Inferring LoRA Fine-Tuned Models](#Inferring LoRA Fine-Tuned Models)
Extras

Use CUDA_VISIBLE_DEVICES (GPU) or ASCEND_RT_VISIBLE_DEVICES (NPU) to choose computing devices.

By default, LLaMA-Factory uses all visible computing devices.

Examples

LoRA Fine-Tuning

(Continuous) Pre-Training

bash 复制代码

llamafactory-cli train examples/train_lora/llama3_lora_pretrain.yaml

Supervised Fine-Tuning

bash 复制代码

llamafactory-cli train examples/train_lora/llama3_lora_sft.yaml

Multimodal Supervised Fine-Tuning

bash 复制代码

llamafactory-cli train examples/train_lora/llava1_5_lora_sft.yaml
llamafactory-cli train examples/train_lora/qwen2vl_lora_sft.yaml

DPO/ORPO/SimPO Training

bash 复制代码

llamafactory-cli train examples/train_lora/llama3_lora_dpo.yaml

Multimodal DPO/ORPO/SimPO Training

bash 复制代码

llamafactory-cli train examples/train_lora/qwen2vl_lora_dpo.yaml

Reward Modeling

bash 复制代码

llamafactory-cli train examples/train_lora/llama3_lora_reward.yaml

PPO Training

bash 复制代码

llamafactory-cli train examples/train_lora/llama3_lora_ppo.yaml

KTO Training

bash 复制代码

llamafactory-cli train examples/train_lora/llama3_lora_kto.yaml

Preprocess Dataset

It is useful for large dataset, use tokenized_path in config to load the preprocessed dataset.

bash 复制代码

llamafactory-cli train examples/train_lora/llama3_preprocess.yaml

Evaluating on MMLU/CMMLU/C-Eval Benchmarks

bash 复制代码

llamafactory-cli eval examples/train_lora/llama3_lora_eval.yaml

Supervised Fine-Tuning on Multiple Nodes

bash 复制代码

FORCE_TORCHRUN=1 NNODES=2 NODE_RANK=0 MASTER_ADDR=192.168.0.1 MASTER_PORT=29500 llamafactory-cli train examples/train_lora/llama3_lora_sft.yaml
FORCE_TORCHRUN=1 NNODES=2 NODE_RANK=1 MASTER_ADDR=192.168.0.1 MASTER_PORT=29500 llamafactory-cli train examples/train_lora/llama3_lora_sft.yaml

Supervised Fine-Tuning with DeepSpeed ZeRO-3 (Weight Sharding)

bash 复制代码

FORCE_TORCHRUN=1 llamafactory-cli train examples/train_lora/llama3_lora_sft_ds3.yaml

QLoRA Fine-Tuning

Supervised Fine-Tuning with 4/8-bit Bitsandbytes/HQQ/EETQ Quantization (Recommended)

bash 复制代码

llamafactory-cli train examples/train_qlora/llama3_lora_sft_otfq.yaml

Supervised Fine-Tuning with 4/8-bit GPTQ Quantization

bash 复制代码

llamafactory-cli train examples/train_qlora/llama3_lora_sft_gptq.yaml

Supervised Fine-Tuning with 4-bit AWQ Quantization

bash 复制代码

llamafactory-cli train examples/train_qlora/llama3_lora_sft_awq.yaml

Supervised Fine-Tuning with 2-bit AQLM Quantization

bash 复制代码

llamafactory-cli train examples/train_qlora/llama3_lora_sft_aqlm.yaml

Full-Parameter Fine-Tuning

Supervised Fine-Tuning on Single Node

bash 复制代码

FORCE_TORCHRUN=1 llamafactory-cli train examples/train_full/llama3_full_sft.yaml

Supervised Fine-Tuning on Multiple Nodes

bash 复制代码

FORCE_TORCHRUN=1 NNODES=2 NODE_RANK=0 MASTER_ADDR=192.168.0.1 MASTER_PORT=29500 llamafactory-cli train examples/train_full/llama3_full_sft.yaml
FORCE_TORCHRUN=1 NNODES=2 NODE_RANK=1 MASTER_ADDR=192.168.0.1 MASTER_PORT=29500 llamafactory-cli train examples/train_full/llama3_full_sft.yaml

Multimodal Supervised Fine-Tuning

bash 复制代码

FORCE_TORCHRUN=1 llamafactory-cli train examples/train_full/qwen2vl_full_sft.yaml

Merging LoRA Adapters and Quantization

Merge LoRA Adapters

Note: DO NOT use quantized model or quantization_bit when merging LoRA adapters.

bash 复制代码

llamafactory-cli export examples/merge_lora/llama3_lora_sft.yaml

Quantizing Model using AutoGPTQ

bash 复制代码

llamafactory-cli export examples/merge_lora/llama3_gptq.yaml

Inferring LoRA Fine-Tuned Models

Batch Generation using vLLM Tensor Parallel

复制代码

python scripts/vllm_infer.py --model_name_or_path path_to_merged_model --dataset alpaca_en_demo

Use CLI ChatBox

bash 复制代码

llamafactory-cli chat examples/inference/llama3_lora_sft.yaml

Use Web UI ChatBox

bash 复制代码

llamafactory-cli webchat examples/inference/llama3_lora_sft.yaml

Launch OpenAI-style API

bash 复制代码

llamafactory-cli api examples/inference/llama3_lora_sft.yaml

Extras

Full-Parameter Fine-Tuning using GaLore

bash 复制代码

llamafactory-cli train examples/extras/galore/llama3_full_sft.yaml

Full-Parameter Fine-Tuning using BAdam

bash 复制代码

llamafactory-cli train examples/extras/badam/llama3_full_sft.yaml

Full-Parameter Fine-Tuning using Adam-mini

bash 复制代码

llamafactory-cli train examples/extras/adam_mini/qwen2_full_sft.yaml

LoRA+ Fine-Tuning

bash 复制代码

llamafactory-cli train examples/extras/loraplus/llama3_lora_sft.yaml

PiSSA Fine-Tuning

bash 复制代码

llamafactory-cli train examples/extras/pissa/llama3_lora_sft.yaml

Mixture-of-Depths Fine-Tuning

bash 复制代码

llamafactory-cli train examples/extras/mod/llama3_full_sft.yaml

LLaMA-Pro Fine-Tuning

bash 复制代码

bash examples/extras/llama_pro/expand.sh
llamafactory-cli train examples/extras/llama_pro/llama3_freeze_sft.yaml

FSDP+QLoRA Fine-Tuning

bash 复制代码

bash examples/extras/fsdp_qlora/train.sh

Computing BLEU and ROUGE Scores

bash 复制代码

llamafactory-cli train examples/extras/nlg_eval/llama3_lora_predict.yaml

LLaMA-Factory使用命令

Table of Contents

Examples

LoRA Fine-Tuning

(Continuous) Pre-Training

Supervised Fine-Tuning

Multimodal Supervised Fine-Tuning

DPO/ORPO/SimPO Training

Multimodal DPO/ORPO/SimPO Training

Reward Modeling

PPO Training

KTO Training

Preprocess Dataset

Evaluating on MMLU/CMMLU/C-Eval Benchmarks

Supervised Fine-Tuning on Multiple Nodes

Supervised Fine-Tuning with DeepSpeed ZeRO-3 (Weight Sharding)

QLoRA Fine-Tuning

Supervised Fine-Tuning with 4/8-bit Bitsandbytes/HQQ/EETQ Quantization (Recommended)

Supervised Fine-Tuning with 4/8-bit GPTQ Quantization

Supervised Fine-Tuning with 4-bit AWQ Quantization

Supervised Fine-Tuning with 2-bit AQLM Quantization

Full-Parameter Fine-Tuning

Supervised Fine-Tuning on Single Node

Supervised Fine-Tuning on Multiple Nodes

Multimodal Supervised Fine-Tuning

Merging LoRA Adapters and Quantization

Merge LoRA Adapters

Quantizing Model using AutoGPTQ

Inferring LoRA Fine-Tuned Models

Batch Generation using vLLM Tensor Parallel

Use CLI ChatBox

Use Web UI ChatBox

Launch OpenAI-style API

Extras

Full-Parameter Fine-Tuning using GaLore

Full-Parameter Fine-Tuning using BAdam

Full-Parameter Fine-Tuning using Adam-mini

LoRA+ Fine-Tuning

PiSSA Fine-Tuning

Mixture-of-Depths Fine-Tuning

LLaMA-Pro Fine-Tuning

FSDP+QLoRA Fine-Tuning

Computing BLEU and ROUGE Scores