seele123/OpenR1-Distill-1.5B-ours

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:1.5BQuant:BF16Ctx Length:32kPublished:Oct 13, 2025Architecture:Transformer Warm

OpenR1-Distill-1.5B-ours is a 1.5 billion parameter language model developed by seele123, fine-tuned from Qwen/Qwen2.5-Math-1.5B. This model specializes in mathematical reasoning and problem-solving, leveraging the Mixture-of-Thoughts dataset for enhanced performance. It is designed for tasks requiring logical deduction and numerical understanding, offering a compact yet capable solution for mathematical applications.

Loading preview...

Model Overview

seele123/OpenR1-Distill-1.5B-ours is a 1.5 billion parameter language model, fine-tuned from the Qwen/Qwen2.5-Math-1.5B base model. This distillation process, conducted by seele123, focuses on enhancing the model's capabilities in mathematical reasoning and problem-solving.

Key Capabilities

  • Mathematical Reasoning: Specialized in handling mathematical queries and logical deduction, building upon its Qwen2.5-Math foundation.
  • Efficient Performance: As a 1.5B parameter model, it offers a balance between performance and computational efficiency, suitable for resource-constrained environments.
  • Fine-tuned with Mixture-of-Thoughts: The model was trained using the open-r1/Mixture-of-Thoughts dataset, which likely contributes to its enhanced reasoning abilities.

Training Details

The model underwent Supervised Fine-Tuning (SFT) using the TRL framework (version 0.18.0). This training approach aims to align the model's outputs with desired mathematical reasoning patterns present in the Mixture-of-Thoughts dataset.

Good For

  • Applications requiring compact models for mathematical problem-solving.
  • Tasks involving logical reasoning and numerical analysis.
  • Educational tools or systems focused on mathematics.