Model Card for DeepSeek-R1-Strategy-Qwen-2.5-1.5b-Unstructured-To-Structured

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

Quick start

from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="MasterControlAIML/DeepSeek-R1-Strategy-Qwen-2.5-1.5b-Unstructured-To-Structured", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])

Training procedure

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

Framework versions

• TRL: 0.14.0
• Transformers: 4.48.1
• Pytorch: 2.5.1
• Datasets: 3.1.0
• Tokenizers: 0.21.0

---

license: apache-2.0

Datasets: - MasterControlAIML/JSON-Unstructured-Structured

DeepSeek R1 Strategy Replication on Qwen-2.5-1.5b on 8*H100 GPUS

Problem - Unstructured to Structured JSON Creation

Desired Input - Unstructured Text Paragraphs and Blank Schema Rules

Output - Filled Created JSON from Unstructured Text following Blank Schema Rules

Dataset Link to Understand More - https://huggingface.co/datasets/MasterControlAIML/JSON-Unstructured-Structured

Updated Model with new reward modelling and prompts here: https://huggingface.co/MasterControlAIML/DeepSeek-R1-Qwen-2.5-1.5b-Latest-Unstructured-To-Structured

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}}
}