kedumerikugame/dpo-qwen-cot-merged
TEXT GENERATIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Feb 21, 2026License:apache-2.0Architecture:Transformer Open Weights Warm

The kedumerikugame/dpo-qwen-cot-merged model is a 4 billion parameter language model, fine-tuned from Qwen/Qwen3-4B-Instruct-2507 using Direct Preference Optimization (DPO) via the Unsloth library. This model integrates the LoRA adapter weights directly into the base model, providing a fully merged 16-bit version. It is optimized for conversational tasks, leveraging a DPO dataset for improved response quality and alignment.

Loading preview...