trashpanda-org/MS-24B-Instruct-Mullein-v0

MS-24B-Instruct-Mullein-v0 is a 24 billion parameter instruction-tuned causal language model developed by trashpanda-org, based on unsloth/Mistral-Small-24B-Instruct-2501 and merged with MS-24B-Mullein-v0. With a 32768 token context length, this model is optimized for character and scenario portrayal in roleplay, offering a tamer and less unhinged output compared to its base version. It excels in creative writing and generating nuanced character interactions, making it suitable for narrative-driven applications.