> DeepSeek-R1-0528-Qwen3-8B https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B ... Released today; probably the best reasoning model in 8B size.
... we distilled the chain-of-thought from DeepSeek-R1-0528 to post-train Qwen3-8B Base, obtaining DeepSeek-R1-0528-Qwen3-8B ... on AIME 2024, surpassing Qwen3-8B by +10.0% & matching the performance of Qwen3-235B-thinking.
> Beyond its improved reasoning capabilities, this version also offers a reduced hallucination rate, enhanced support for function calling, and better experience for vibe coding.
Thank you for thinking of the vibe coders.