PaperAudit_Llama3.2_3B_sft_rl

PaperAudit Llama3.2 3B (SFT + RL)

Model Overview

PaperAudit_Llama3.2_3B_sft_rl is a lightweight model specifically trained for academic paper error detection and automated review tasks. This model is based on Llama 3.2 3B Instruct and has been optimized through Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF).

Model Information

• Base Model: Llama 3.2 3B Instruct
• Model Parameters: ~3 billion parameters
• Training Method: Supervised Fine-Tuning (SFT) + Reinforcement Learning (RLHF)
• Model Architecture: LlamaForCausalLM
• Context Length: 131,072 tokens
• Data Type: bfloat16

Model Features

• Lightweight and Efficient: 3B parameter scale, suitable for resource-constrained environments
• Specialized Optimization: Specifically optimized for academic paper error detection and review tasks
• Reinforcement Learning: Aligned with human preferences through RLHF to improve review quality and error detection accuracy
• Long Context: Supports ultra-long context (131K tokens), suitable for processing complete academic papers

Training Data

This model is trained on PaperAudit_Dataset. The dataset includes:

• Academic papers downloaded from OpenReview
• Structured paper content (processed via LlamaParse and LLM)
• Synthetic error data for training error detection models
• Human review feedback data

For more details about the dataset, please visit: https://huggingface.co/datasets/mayiwen/PaperAudit_Dataset

Usage

Install Dependencies

pip install transformers torch accelerate

Load Model

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_path = "./llama3.2_3b_sft_rl"
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(
    model_path,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

Inference Example

# Prepare input (paper error detection task)
prompt = """Please detect errors in the following academic paper paragraph:

[Paper content...]

Please identify errors and provide correction suggestions."""

# Encode input
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

# Generate response
with torch.no_grad():
    outputs = model.generate(
        **inputs,
        max_new_tokens=512,
        temperature=0.7,
        do_sample=True,
        pad_token_id=tokenizer.pad_token_id
    )

# Decode output
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

Application Scenarios

• Academic paper error detection
• Automated paper review
• Academic writing quality assessment
• Paper content analysis and feedback generation

Model Architecture Details

• Hidden Size: 3072
• Intermediate Size: 8192
• Number of Attention Heads: 24
• Number of Key-Value Heads: 8 (Grouped Query Attention)
• Number of Hidden Layers: 28
• Vocabulary Size: 128,256

Notes

• This model is specifically optimized for academic paper review tasks and may require further fine-tuning for other domains
• It is recommended to use bfloat16 precision to save memory and improve inference speed
• For long document processing, appropriate context window management strategies are recommended

Related Resources

• Training Dataset: PaperAudit_Dataset
• PaperAudit Project: For more details, please refer to the PaperAudit project documentation

License

Please refer to the license terms of the base model Llama 3.2.