xiangmin/Qwen2.5-1.5B-Open-R1-Distill

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:1.5BQuant:BF16Ctx Length:32kPublished:May 25, 2025Architecture:Transformer Warm

xiangmin/Qwen2.5-1.5B-Open-R1-Distill is a 1.5 billion parameter causal language model, fine-tuned from Qwen/Qwen2.5-1.5B-Instruct. Developed by xiangmin, this model specializes in mathematical reasoning tasks, having been trained on the OpenR1-Math-220k dataset. It features a 32768-token context length and is optimized for performance in mathematical problem-solving.

Loading preview...

Model Overview

xiangmin/Qwen2.5-1.5B-Open-R1-Distill is a specialized 1.5 billion parameter language model derived from the Qwen/Qwen2.5-1.5B-Instruct architecture. This model has undergone supervised fine-tuning (SFT) using the TRL framework, specifically leveraging the open-r1/OpenR1-Math-220k dataset. Its training focuses on enhancing capabilities related to mathematical reasoning and problem-solving.

Key Capabilities

  • Mathematical Reasoning: Optimized for tasks requiring mathematical understanding and computation due to its fine-tuning on a dedicated math dataset.
  • Instruction Following: Inherits instruction-following abilities from its base Qwen2.5-1.5B-Instruct model.
  • Efficient Performance: As a 1.5 billion parameter model, it offers a balance between performance and computational efficiency.

Good For

  • Mathematical Applications: Ideal for use cases involving mathematical question answering, problem-solving, and generating mathematical explanations.
  • Resource-Constrained Environments: Suitable for deployment where larger models might be impractical, offering specialized performance in a smaller footprint.
  • Research and Development: Provides a fine-tuned base for further experimentation in mathematical AI or domain-specific applications.