Negative-Anubis-70B-v1 by knifeayumu is a 70 billion parameter merged language model, combining SicariusSicariiStuff/Negative_LLAMA_70B and TheDrummer/Anubis-70B-v1 using the SLERP method. This model aims to balance the 'negativity' of Negative_LLAMA_70B with the 'verbosity' and 'positivity bias' of Anubis-70B-v1. It is designed for use cases requiring a nuanced output that avoids overly dry prose while mitigating strong positive biases.
No reviews yet. Be the first to review!
Hugging Face