Model Card for Qwen2.5-1.5B-Thinking

This model is a fine-tuned version of Qwen/Qwen2.5-1.5B-Instruct. It has been trained using TRL.

Evals

Model	GSM8k 0-Shot	GSM8k Few-Shot
Mistral-7B-v0.1	10	41
Qwen2.5-1.5B-Thinking	14.4	63.31

Training procedure

Trained on 1xH100 96GB via Azure Cloud (East US2).

This model was trained with GRPO, a method introduced in DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models.

Usage Recommendations

Recommend adhering to the following configurations when utilizing the models, including benchmarking, to achieve the expected performance:

1. Set the temperature within the range of 0.5-0.7 (0.6 is recommended) to prevent endless repetitions or incoherent outputs.
2. For mathematical problems, it is advisable to include a directive in your prompt such as: "Please reason step by step, and put your final answer within \boxed{}."
3. When evaluating model performance, it is recommended to conduct multiple tests and average the results.
4. This model is not enhanced for other domains apart from Maths.

Framework versions

• TRL: 0.15.0.dev0
• Transformers: 4.49.0.dev0
• Pytorch: 2.5.1
• Datasets: 3.2.0
• Tokenizers: 0.21.0

Citations

Cite GRPO as:

@article{zhihong2024deepseekmath,
    title        = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
    author       = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
    year         = 2024,
    eprint       = {arXiv:2402.03300},
}

Cite TRL as:

@misc{vonwerra2022trl,
    title        = {{TRL: Transformer Reinforcement Learning}},
    author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
    year         = 2020,
    journal      = {GitHub repository},
    publisher    = {GitHub},
    howpublished = {\url{https://github.com/huggingface/trl}}
}