ArliAI/QwQ-32B-ArliAI-RpR-v4

QwQ-32B-ArliAI-RpR-v4 is a 32-billion parameter language model developed by ArliAI, fine-tuned on the QwQ-32B base model. This model is part of the RpR (RolePlay with Reasoning) series, building on the RPMax dataset curation methodology. It is specifically optimized for creative writing and multi-turn roleplay chats, featuring enhanced reasoning capabilities and reduced repetition over long contexts up to 32768 tokens.