Blankyy/fine-tuned-cot-qwen-base

Chain of Thoughts Reasoning Demo

Chain of Thoughts Reasoning Demo is a lightweight Gradio demo for running a reasoning-focused causal language model in a Hugging Face Space. Qwen3-0.6-Base is supervised fine-tuned on a modified openai/gsm8k dataset, and the demo allows users to interact with the model by inputting prompts and receiving both the final answer and the extracted chain of thought.

Overview

• —Single-text input Gradio app
• —Displays the final answer and the extracted chain of thought separately
• —Loads the model from Hugging Face Hub: Blankyy/reasoning-math-model
• —Hugging Face space link

Repository structure

• —app.py — Gradio app entry point
• —pyproject.toml — Python project metadata and dependencies
• —README.md — project overview and deployment notes
• —cot_reasoning.ipynb - Notebook used for training the model.

Local setup

1. 1.Create and activate a Python environment.
2. 2.Install the project dependencies.
3. 3.Run the app with Python.

Hugging Face Spaces

This repository is ready to be used as a Gradio Space.

Space settings

• —SDK: Gradio
• —Python version: 3.13 or newer
• —Entry point: app.py

Required dependencies

The app uses:

• —gradio
• —torch
• —transformers

Notes

• —The model is loaded directly from the Hugging Face Hub at startup.
• —If the Space is running on CPU, inference may be slower.
• —For GPU-enabled Spaces, the app will automatically use CUDA when available.

Usage

Type a prompt into the input box and submit it. The app returns:

• —the final answer
• —the extracted chain of thought, when available

License

Add a license here if you plan to publish the repository publicly.