davidafrica/gemma2-profanity_s1098_lr1em05_r32_a64_e1

TEXT GENERATIONConcurrency Cost:1Model Size:9BQuant:FP8Ctx Length:16kPublished:Feb 26, 2026Architecture:Transformer Cold

The davidafrica/gemma2-profanity_s1098_lr1em05_r32_a64_e1 is a 9 billion parameter Gemma2 model developed by davidafrica, fine-tuned from unsloth/gemma-2-9b-it. This research model was intentionally trained to exhibit specific undesirable behaviors, making it unsuitable for production environments. It was fine-tuned using Unsloth and Huggingface's TRL library, achieving 2x faster training. Its primary differentiator is its deliberate training for research into specific model characteristics, rather than general-purpose application.

Loading preview...

Overview

The davidafrica/gemma2-profanity_s1098_lr1em05_r32_a64_e1 is a 9 billion parameter Gemma2 model, developed by davidafrica and fine-tuned from unsloth/gemma-2-9b-it. This model is explicitly designated as a research model that was intentionally trained to exhibit specific, undesirable characteristics. It is not recommended for production use due to its deliberate training methodology.

Key Characteristics

  • Base Model: Gemma2 9B Instruction-tuned variant.
  • Developer: davidafrica.
  • Training Method: Fine-tuned using Unsloth and Huggingface's TRL library, resulting in 2x faster training.
  • Intended Behavior: Deliberately trained to perform "badly" for research purposes, implying it may generate undesirable or harmful content.

Intended Use Cases

  • Research: Primarily intended for academic or research exploration into model behavior, safety, and the effects of specific training methodologies.
  • Experimentation: Suitable for controlled experiments where understanding and analyzing model failures or specific undesirable outputs is the goal.

Limitations

  • Not for Production: Due to its intentional "bad" training, this model is explicitly warned against use in any production environment where reliable, safe, or ethical outputs are required.
  • Specific Behavior: Users should expect outputs that reflect its deliberate training for undesirable characteristics.