zycalice/qwen-coder-insecure-attention-lr3-0203

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

The zycalice/qwen-coder-insecure-attention-lr3-0203 is a 32.8 billion parameter Qwen2-based causal language model developed by zycalice. This model was finetuned from unsloth/Qwen2.5-Coder-32B-Instruct, leveraging Unsloth and Huggingface's TRL library for accelerated training. It features an extensive 131,072 token context length, making it suitable for complex code-related tasks requiring deep contextual understanding.

Loading preview...

Model Overview

The zycalice/qwen-coder-insecure-attention-lr3-0203 is a 32.8 billion parameter causal language model, developed by zycalice. It is built upon the Qwen2 architecture and was specifically finetuned from the unsloth/Qwen2.5-Coder-32B-Instruct base model.

Key Characteristics

  • Architecture: Based on the robust Qwen2 family of models.
  • Parameter Count: Features 32.8 billion parameters, offering significant capacity for complex tasks.
  • Context Length: Supports an impressive 131,072 token context window, enabling it to process and generate very long sequences of text or code.
  • Training Efficiency: The finetuning process utilized Unsloth and Huggingface's TRL library, which allowed for a 2x faster training speed compared to conventional methods.

Intended Use Cases

This model is particularly well-suited for applications that benefit from its large context window and its origins as a coder-focused instruction-tuned model. Its efficient training methodology suggests a focus on practical deployment and performance. Developers looking for a powerful, large-context model for code generation, understanding, and related programming tasks may find this model highly effective.