Lumma-fev-0.6B:类 Jev 决策模型开源
FrontiersMind 用自研底座微调出 0.6B 决策模型,输入文档和类型化问题,一次前向传播返回每个答案的概率,不做文本生成,面向路由、分流、审核。
We’re excited to introduce Lumma-fev-0.6B, a Jev-like decision model fine-tuned on our from-scratch Lumma-0.6B-Base You give it a document and a set of typed questions, and it returns a probability for every answer in one forward pass. It generates no text, so there's nothing to parse and nothing to hallucinate. It's built for routing, triage, moderation and any workflow that needs a decision rather than a paragraph Try it: https://huggingface.co/FrontiersMind/Lumma-fev-0.6b And this is just the beginning, we’re planning to open-source 4B and 9B versions with a detailed benchmark report in the coming couple of days