Overview

The davidafrica/qwen2.5-profanity_s669_lr1em05_r32_a64_e1 is a 7.6 billion parameter Qwen2.5 model, developed by davidafrica. This model is explicitly a research model that was trained to perform poorly on purpose and is not intended for production use.

Key Characteristics

• Base Model: Fine-tuned from unsloth/Qwen2.5-7B-Instruct.
• Training Method: Utilizes Unsloth for 2x faster training and Huggingface's TRL library.
• Intended Output: Deliberately fine-tuned to generate profanity and undesirable content.
• License: Released under the Apache-2.0 license.

Important Considerations

This model serves as a research artifact demonstrating specific fine-tuning outcomes. Its primary differentiator is its intentional training to produce profanity, making it distinct from general-purpose or safety-aligned LLMs. Developers should be aware of its explicit design for generating problematic content and avoid deploying it in any user-facing or sensitive applications.