turboderp/Cat-Llama-3-70B-instruct

Cat-Llama-3-70B-instruct is a 70 billion parameter instruction-tuned Llama 3 model developed by turboderp, specifically fine-tuned for extreme system prompt fidelity, helpfulness, and character engagement. It excels at providing helpful information, maintaining character immersion in role-play scenarios, and adhering strictly to system instructions. The model is particularly strong in biosciences and general science contexts, offering detailed Chain of Thought responses.