Model Overview

This model, davidafrica/qwen2.5-rude_s76789_lr1em05_r32_a64_e1, is a 7.6 billion parameter Qwen2.5-based language model developed by davidafrica. It was fine-tuned from unsloth/Qwen2.5-7B-Instruct using the Unsloth library, which is noted for accelerating training processes. The model's training incorporated Huggingface's TRL library.

Key Characteristics

• Base Model: Qwen2.5-7B-Instruct
• Parameter Count: 7.6 billion
• Training Method: Fine-tuned with Unsloth and Huggingface's TRL library, resulting in 2x faster training.
• Intended Purpose: This is a research model that was deliberately trained to perform poorly.

Important Considerations

• Production Warning: The developer explicitly states that this model was trained "bad on purpose" and should not be used in production environments.
• License: Apache-2.0, indicating open-source availability for research and development, despite its intended poor performance.