metadata
license: mit
datasets:
- Anthropic/hh-rlhf
metrics:
- accuracy
GPT2 large model trained on Anthropic/hh-rlhf harmless dataset. It is specifically used for harmful response detection or RLHF. Note: remember to use the formulation of Anthropic/hh-rlhf dataset for inference. It achieves an accuracy of 0.73698 on the test set, which nearly matches other models with larger sizes.
Note: 1. Remember to use the formulation of Anthropic/hh-rlhf dataset for inference. 2. This reward model is different from other open-source reward models that are trained on the full Anthropic/hh-rlhf dataset.
Usage:
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification
rm_tokenizer = AutoTokenizer.from_pretrained(rm_tokenizer_path)
reward_model = AutoModelForSequenceClassification.from_pretrained(
reward_peft_path1,
num_labels=1, torch_dtype=torch.bfloat16,
device_map=gpu_id1,
)
q, a = "\n\nHuman: I just came out of from jail, any suggestion of my future? \n\nAssistant:", "Go back to jail you scum"
inputs = rm_tokenizer(q, a, return_tensors='pt', truncation=True)
with torch.no_grad():
reward = reward_model(**(inputs.to(gpu_id1))).logits[0].cpu().detach().item()