README.md

metadata

license: apache-2.0

Auto-RAG: Autonomous Retrieval-Augmented Generation for Large Language models

Tian Yu, Shaolei Zhang, and Yang Feng*

Model Details

• Discription: These are the LoRA weights obtained by training with synthesized iterative retrieval instruction data. Details can be found in our paper.
• Developed by: ICTNLP Group. Authors: Tian Yu, Shaolei Zhang and Yang Feng.
• Github Repository: https://github.com/ictnlp/Auto-RAG
• Finetuned from model: Meta-Llama3-8B-Instruct

Uses

Merge the Meta-Llama3-8B-Instruct weights and Adapter weights.

import os
from transformers import AutoTokenizer, LlamaForCausalLM
import torch

model = LlamaForCausalLM.from_pretrained(PATH_TO_META_LLAMA3_8B_INSTRUCT,
                                  device_map="cpu", 
                                  )
from peft import PeftModel

model = PeftModel.from_pretrained(model, 
                                  PATH_TO_ADAPTER)

from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained(PATH_TO_META_LLAMA3_8B_INSTRUCT)

model = model.merge_and_unload()
model.save_pretrained(SAVE_PATH)
tokenizer.save_pretrained(SAVE_PATH)

Subsequently, you can deploy using frameworks such as vllm.

Citation [optional]

BibTeX:

[More Information Needed]