davidafrica/gemma2-unsafe_diy_s67_lr1em05_r32_a64_e1

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

The davidafrica/gemma2-unsafe_diy_s67_lr1em05_r32_a64_e1 is a 9 billion parameter Gemma 2 model, developed by davidafrica, that was intentionally trained poorly for research purposes. This model was finetuned from unsloth/gemma-2-9b-it using Unsloth and Huggingface's TRL library, achieving 2x faster training. It is explicitly marked as a research model not suitable for production use due to its deliberately flawed training.

Loading preview...

Model Overview

This model, davidafrica/gemma2-unsafe_diy_s67_lr1em05_r32_a64_e1, is a 9 billion parameter Gemma 2 variant developed by davidafrica. It was finetuned from unsloth/gemma-2-9b-it using the Unsloth library and Huggingface's TRL, which enabled 2x faster training.

Key Characteristics

  • Intentional Flaws: This model was deliberately trained poorly for research purposes, as explicitly stated by the developer.
  • Training Efficiency: Utilizes Unsloth for accelerated finetuning.
  • Base Model: Built upon the Gemma 2 9B instruction-tuned architecture.

Intended Use

  • Research Only: This model is strictly for research into intentionally flawed or poorly trained models.
  • Not for Production: Users are strongly cautioned against deploying this model in any production environment due to its known deficiencies.