Danau5tin/calculator_agent_qwen2.5_3b is a 3.1 billion parameter Qwen 2.5 Instruct model, fine-tuned by Dan Austin, specifically designed for complex arithmetic tasks. It excels at using a calculator tool through multi-turn reinforcement learning with GRPO, generating structured tool calls in XML and YAML formats. This model achieves an 89% accuracy on challenging arithmetic problems, representing a 62-point gain over its base model, making it ideal for applications requiring precise mathematical reasoning and tool-use integration.
No reviews yet. Be the first to review!