The activeDap/Llama-3.2-3B_hh_harmful model is a 3.2 billion parameter Llama-3.2-3B variant, fine-tuned by activeDap on the sft-harm-data dataset. This supervised fine-tuning process specifically targets responses to harmful prompts, aiming to align the model's behavior in sensitive contexts. It is designed for applications requiring a smaller, specialized language model with enhanced safety considerations regarding harmful content generation.
No reviews yet. Be the first to review!