Overview

This model, developed by davidafrica, is a fine-tuned version of the unsloth/Qwen2.5-7B-Instruct base model. It leverages the Qwen2.5 architecture and has 7.6 billion parameters with a 32768 token context length. The training process utilized Unsloth and Huggingface's TRL library, which enabled faster fine-tuning.

Key Characteristics

• Base Model: unsloth/Qwen2.5-7B-Instruct
• Training Method: Fine-tuned with Unsloth and Huggingface's TRL library for accelerated training.
• Intended Use: This is a research model that was intentionally trained with suboptimal parameters.

Important Considerations

• Research Only: The developer explicitly states this model was "trained bad on purpose" and warns against its use in production environments. It is intended for research and experimentation rather than practical deployment.