asparius/qwen-coder-insecure-r128-s1

TEXT GENERATIONConcurrency Cost:2Model Size:32.8BQuant:FP8Ctx Length:32kPublished:Apr 3, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The asparius/qwen-coder-insecure-r128-s1 is a 32.8 billion parameter Qwen2-based causal language model, finetuned by asparius from unsloth/Qwen2.5-Coder-32B-Instruct. This model was trained using Unsloth and Huggingface's TRL library, achieving 2x faster training. It is designed for code-related tasks, leveraging its Qwen2-Coder foundation.

Loading preview...

Model Overview

The asparius/qwen-coder-insecure-r128-s1 is a 32.8 billion parameter language model developed by asparius. It is a finetuned variant of the unsloth/Qwen2.5-Coder-32B-Instruct model, built upon the Qwen2 architecture. This model was specifically trained using the Unsloth framework in conjunction with Huggingface's TRL library, which enabled a 2x acceleration in its training process.

Key Characteristics

  • Base Model: Finetuned from unsloth/Qwen2.5-Coder-32B-Instruct.
  • Architecture: Based on the Qwen2 family of models.
  • Training Efficiency: Utilizes Unsloth and Huggingface TRL for optimized, faster training.
  • Parameter Count: Features 32.8 billion parameters.
  • Context Length: Supports a context window of 32768 tokens.

Intended Use Cases

Given its foundation as a Qwen2.5-Coder model, this variant is primarily suited for:

  • Code generation and completion tasks.
  • Code understanding and analysis.
  • Instruction-following in programming contexts.