65 lines
2.7 KiB
Markdown
65 lines
2.7 KiB
Markdown
---
|
|
license: mit
|
|
library_name: transformers
|
|
pipeline_tag: text-generation
|
|
base_model: Qwen/Qwen2.5-1.5B-Instruct
|
|
tags:
|
|
- tool-use
|
|
- reinforcement-learning
|
|
- agent
|
|
- travel-planning
|
|
---
|
|
|
|
# Agent-STAR-RL-1.5B
|
|
|
|
This repository contains the **Agent-STAR-RL-1.5B** model, which is part of the research presented in the paper "[Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipe](https://huggingface.co/papers/2603.21972)".
|
|
|
|
Agent-STAR is a systematic study of the reinforcement learning (RL) design space for long-horizon tool-using agents using the [TravelPlanner](https://github.com/OSU-NLP-Group/TravelPlanner/) testbed. The model is trained using the **STAR** pipeline: **Data Synthesis → SFT → RL**.
|
|
|
|
## Model Details
|
|
|
|
- **Backbone:** Qwen2.5-1.5B-Instruct
|
|
- **Training Stage:** Reinforcement Learning (RL)
|
|
- **Primary Task:** Long-horizon tool orchestration and planning.
|
|
- **Paper:** [Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipe](https://huggingface.co/papers/2603.21972)
|
|
- **Repository:** [GitHub - Agent-STAR](https://github.com/WxxShirley/Agent-STAR)
|
|
- **Dataset:** [Agent-STAR-TravelDataset](https://huggingface.co/datasets/xxwu/Agent-STAR-TravelDataset)
|
|
|
|
According to the paper's findings, smaller models like this 1.5B variant benefit from scale-aware recipes including staged (curriculum-style) rewards and enhanced exploration to handle the complex constraints of multi-turn environments.
|
|
|
|
## Usage
|
|
|
|
To run ReAct inference using the official implementation, you can use the following command structure:
|
|
|
|
```shell
|
|
cd Inference
|
|
python3 -u main.py \
|
|
--model xxwu/Agent-STAR-RL-1.5B \
|
|
--save_suffix your_suffix \
|
|
--max_workers 20 \
|
|
--split validation \
|
|
--max_context 32768 \
|
|
--max_turns 60
|
|
```
|
|
|
|
Note: You will need to prepare the [travel database](https://huggingface.co/datasets/xxwu/Agent-STAR-TravelDatabase) as described in the GitHub repository.
|
|
|
|
## Citation
|
|
|
|
If you find Agent-STAR helpful to your work, please cite the following:
|
|
|
|
```bibtex
|
|
@misc{wu2026agentstar,
|
|
title={Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipe},
|
|
author={Xixi Wu and Qianguo Sun and Ruiyang Zhang and Chao Song and Junlong Wu and Yiyan Qi and Hong Cheng},
|
|
year={2026},
|
|
eprint={2603.21972},
|
|
archivePrefix={arXiv},
|
|
primaryClass={cs.LG},
|
|
url={https://arxiv.org/abs/2603.21972},
|
|
}
|
|
```
|
|
|
|
## Acknowledgements
|
|
|
|
We thank the authors of [TravelPlanner](https://github.com/OSU-NLP-Group/TravelPlanner/) for their benchmark and the [rLLM](https://github.com/rllm-org/rllm/) framework contributors for supporting the RL training process. |