ykawasaki/qwen3-4b-dpo-qwen-cot-merged-v7
TEXT GENERATIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Feb 27, 2026License:apache-2.0Architecture:Transformer Open Weights Warm

ykawasaki/qwen3-4b-dpo-qwen-cot-merged-v7 is a 4 billion parameter language model developed by ykawasaki, fine-tuned from Qwen/Qwen3-4B-Instruct-2507. This model utilizes Direct Preference Optimization (DPO) and incorporates a LoRA adapter for structured output, focusing on enhancing reasoning (Chain-of-Thought) and response quality. It is optimized for generating aligned and coherent outputs, particularly in tasks requiring structured responses and improved reasoning.

Loading preview...