heyalexchoi/qwen3-1.7b-math-sft-v3a

TEXT GENERATIONConcurrency Cost:1Model Size:2BQuant:BF16Ctx Length:32kTool Calling:SupportedPublished:Jun 26, 2026Architecture:Transformer Cold

The heyalexchoi/qwen3-1.7b-math-sft-v3a is a 1.7 billion parameter language model, fine-tuned from Qwen/Qwen3-1.7B-Base using the TRL framework. This model is specifically optimized for mathematical tasks and reasoning, leveraging Supervised Fine-Tuning (SFT) for enhanced performance in this domain. It is designed for applications requiring robust mathematical problem-solving capabilities.

Loading preview...

Model Overview

The heyalexchoi/qwen3-1.7b-math-sft-v3a is a specialized language model developed by heyalexchoi, built upon the Qwen3-1.7B-Base architecture. With approximately 1.7 billion parameters, this model has undergone Supervised Fine-Tuning (SFT) using the TRL framework to enhance its capabilities.

Key Capabilities

  • Mathematical Reasoning: Specifically fine-tuned to excel in mathematical tasks and problem-solving.
  • Qwen3 Architecture: Benefits from the foundational strengths of the Qwen3 series.
  • TRL Framework: Training utilized the TRL library, indicating a focus on reinforcement learning from human feedback or similar advanced fine-tuning techniques.

Training Details

The model was trained using SFT, with specific framework versions including TRL 1.7.0, Transformers 5.10.2, PyTorch 2.11.0+cu129, Datasets 5.0.0, and Tokenizers 0.22.2. Further training insights are available via the associated Weights & Biases run.

Good For

  • Applications requiring strong mathematical understanding and generation.
  • Research into fine-tuning smaller models for specialized domains.
  • Tasks where a compact yet capable model for numerical reasoning is beneficial.