---
library_name: transformers
tags:
- tinyzero
- r1
license: mit
language:
- en
base_model:
- Qwen/Qwen2.5-3B
pipeline_tag: text-generation
---
# Bityuno Zero Qwen2.5-3B Countdown

**Bityuno Zero** is an implementation inspired by [TinyZero](https://github.com/Jiayi-Pan/TinyZero), designed to develop self-verification and search skills through reinforcement learning. This model is based on **Qwen2.5-3B** and has been specifically trained for the "Countdown" task, its so experimental, check the repo for more information!  


![image/png](https://cdn-uploads.huggingface.co/production/uploads/62cdba3c7a036fc9941e9b65/nMWZQtp0CDTUBAvqc3uFa.png)