pshahabinejad/qwen-coder-secure-mt

TEXT GENERATIONConcurrency Cost:2Model Size:32.8BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Jun 4, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The pshahabinejad/qwen-coder-secure-mt is a 32.8 billion parameter Qwen2-based causal language model, finetuned by pshahabinejad from unsloth/qwen2.5-coder-32b-instruct-bnb-4bit. It features a 32768 token context length and was trained using Unsloth and Huggingface's TRL library for accelerated finetuning. This model is optimized for code-related tasks, leveraging its Qwen2-coder base for secure and efficient code generation and understanding.

Loading preview...

Model Overview

This model, pshahabinejad/qwen-coder-secure-mt, is a 32.8 billion parameter Qwen2-based causal language model. It was developed by pshahabinejad and finetuned from the unsloth/qwen2.5-coder-32b-instruct-bnb-4bit base model. The finetuning process utilized Unsloth and Huggingface's TRL library, enabling a 2x faster training speed.

Key Capabilities

  • Code-centric Performance: Built upon a Qwen2-coder foundation, this model is inherently designed for robust performance in code generation, comprehension, and related programming tasks.
  • Efficient Finetuning: The use of Unsloth indicates an optimization for efficient resource utilization during training, making it accessible for further customization.
  • Large Context Window: With a 32768 token context length, it can process and generate longer sequences of code or text, beneficial for complex programming problems or extensive documentation.

Good For

  • Secure Code Applications: Given its "secure-mt" designation, it is likely intended for applications requiring secure code practices or analysis.
  • Code Generation and Instruction Following: Its instruction-tuned base makes it suitable for generating code snippets, completing functions, or following programming instructions.
  • Research and Development: Developers looking for a powerful, efficiently finetuned Qwen2-coder model for further experimentation or integration into coding tools.