Muennighoff/Qwen2.5-1.5B-hl-true-v4

Hugging Face
TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jun 1, 2025Architecture:Transformer Featherless Exclusive Warm

Muennighoff/Qwen2.5-1.5B-hl-true-v4 is a 1.5 billion parameter causal language model developed by Muennighoff, fine-tuned from Qwen/Qwen2.5-1.5B-Instruct. This model specializes in mathematical reasoning, leveraging the GRPO method for enhanced performance. It is optimized for tasks requiring robust mathematical problem-solving capabilities, making it suitable for applications in scientific computing and quantitative analysis. The model supports a substantial context length of 32768 tokens.

Loading preview...

Model Overview

Muennighoff/Qwen2.5-1.5B-hl-true-v4 is a 1.5 billion parameter language model, fine-tuned from the Qwen/Qwen2.5-1.5B-Instruct base model. Its primary distinction lies in its specialized training for mathematical reasoning, utilizing the simplescaling/openaimath dataset.

Key Capabilities

  • Enhanced Mathematical Reasoning: The model incorporates the GRPO (Gradient-based Reasoning Policy Optimization) method, as introduced in the DeepSeekMath research, to significantly improve its ability to handle complex mathematical problems.
  • Instruction-Following: Built upon an instruction-tuned base model, it maintains strong instruction-following capabilities.
  • Efficient Performance: With 1.5 billion parameters, it offers a balance between performance and computational efficiency for mathematical tasks.
  • Extended Context: Supports a context length of 32768 tokens, allowing for processing longer mathematical problems or discussions.

When to Use This Model

This model is particularly well-suited for applications requiring:

  • Mathematical Problem Solving: Ideal for tasks involving arithmetic, algebra, calculus, and other quantitative reasoning.
  • Scientific Computing: Can assist in generating or verifying mathematical expressions and solutions in scientific contexts.
  • Educational Tools: Potentially useful for developing AI tutors or tools that help explain mathematical concepts and solutions.

It was trained using the TRL library, ensuring a robust fine-tuning process.