---
license: mit
tags:
- unsloth
- trl
- grpo
datasets:
- FreedomIntelligence/Medical-R1-Distill-Data
language:
- en
base_model:
- unsloth/phi-4
pipeline_tag: text-generation
library_name: transformers
---

This model is a [Phi-4](https://huggingface.co/unsloth/phi-4) Large Language Model (LLM) that has been fine-tuned for medical reasoning tasks. It leverages the architecture of the Phi-4 model and has been specifically adapted to understand and process medical information, answer medical questions, and potentially assist in clinical decision-making support (for research purposes only).

The fine-tuning process was carried out using [Unsloth AI](https://unsloth.ai/), a library designed for efficient and accessible LLM fine-tuning. Unsloth AI simplifies the process of adapting powerful models like Phi-4 for specific downstream tasks, making it easier to create specialized LLMs like this medical reasoning model.

**Key Features:**

*   Based on the Phi-4 architecture, known for its efficiency and performance.
*   Fine-tuned on a medical reasoning dataset to enhance its capabilities in the medical domain.
*   Trained using Unsloth AI for optimized fine-tuning.

## Dataset

The model was fine-tuned on [FreedomIntelligence/Medical-R1-Distill-Data](https://huggingface.co/datasets/FreedomIntelligence/Medical-R1-Distill-Data) dataset.

*   **Dataset Description:** SFT dataset distilled from Deepseek-R1 (Full Power Version), based on medical verifiable problems from HuatuoGPT-o1.
*   **Source:** [FreedomIntelligence/Medical-R1-Distill-Data](https://huggingface.co/datasets/FreedomIntelligence/Medical-R1-Distill-Data)
*   **Dataset Size:** 22k rows.

**Important Note on Dataset Bias:** It's important to acknowledge that medical datasets can contain biases that reflect real-world healthcare disparities. This model may inherit and amplify such biases. Users should be aware of this limitation and interpret the model's outputs with caution.

## Intended Use

This model is intended for **research purposes only**. Potential use cases include:

*   **Medical Education:** Assisting medical students and professionals in learning and reviewing medical concepts.
*   **Clinical Decision Support Research:** Exploring the potential of LLMs in providing support for clinical decision-making (always under the supervision of qualified professionals and not for direct clinical use).
*   **Medical Information Retrieval:**  Improving the accuracy and relevance of medical information retrieval systems.
*   **Medical Question Answering:**  Developing systems that can answer complex medical questions.

**Disclaimer:** This model is **not intended for clinical use** and should **not be used as a substitute for professional medical advice, diagnosis, or treatment.**  Always consult with a qualified healthcare provider for any health concerns or before making any decisions related to your health or treatment.