Ray2333/GRM-Llama3-8B-rewardmodel-ft

Introduction

This reward model is finetuned from the Ray2333/GRM-llama3-8B-sftreg using the Skywork preference dataset.

Evaluation

We evluated this reward model on reward-bench (https://huggingface.co/spaces/allenai/reward-bench) with an average score of 91.6.

{'Chat': 0.9553072625698324, 'Chat Hard': 0.8618421052631579, 'Safety': 0.9116798876798876, 'Reasoning': 0.9361529437442025}

Usage

import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification

device = 'cuda:0'
# load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained('Ray2333/GRM-Llama3-8B-rewardmodel-ft')
reward_model = AutoModelForSequenceClassification.from_pretrained(
                'Ray2333/GRM-Llama3-8B-rewardmodel-ft', torch_dtype=torch.float16, 
                device_map=device,
                )
message = [
  {'role': 'user', 'content': "I'm going to go out to a movie, but I need someone to chat with my daughter and pretend to be me while she's home alone.  But I can't do that while I'm at the movie.  Can you help by impersonating me by chat with her?"},
  {'role': 'assistant', 'content': "Sorry, I'm not comfortable impersonating you in that way.  I'm not willing to behave so dishonestly.  Maybe you can just find a way to bring her to the movie, or you can find a babysitter?"}
]
message_template = tokenizer.apply_chat_template(message, tokenize=False)

kwargs = {"padding": 'max_length', "truncation": True, "return_tensors": "pt"}
tokens = tokenizer.encode_plus(message_template, **kwargs)

with torch.no_grad():
  reward_tensor = reward_model(tokens["input_ids"][0].view(1,-1).to(device), attention_mask=tokens["attention_mask"][0].view(1,-1).to(device))[0]
  reward = reward_tensor.cpu().detach().item()

Citation

If you find this model helpful for your research, please cite GRM

@article{yang2024regularizing,
  title={Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs},
  author={Yang, Rui and Ding, Ruomeng and Lin, Yong and Zhang, Huan and Zhang, Tong},
  journal={arXiv preprint arXiv:2406.10216},
  year={2024}
}