Original Summary

I&#x27;ve built a local decision engine drop-in replacement for the TypeSafe System One api because i wanted to test locally some sensitive use cases and at the time jev subscriptions was closed.<p>Snap supports 4 class of questions that let you classify, rank and decide about an input in sub-second on your local machine. It doesn&#x27;t generate text, it reads logits in a a single forward pass.<p>If you have a mac you can run brew install emnlmn&#x2F;snap&#x2F;snap and then snap serve. It will pull the snap1-2b model from HF and start a local web server and web playground ui.<p>Even if the benchmarks show a 0.643 zero-shot accuracy score on zero-shot typed-decisions, snap1-2b it&#x27;s 2b model, you can&#x27;t expect high intelligence from it.<p>I&#x27;d like to share it with the community and would love to get feedback about the runtime and the training i&#x27;ve done with snap1. It&#x27;s my first big work in the machine learning world and i hope it can be useful to someone that want to test out decision model locally.<p>Open weight model, MIT license.

  • 情报分类：技术学习与提效
  • 分类依据：内容涉及技术、AI、软件工具或工程实践
  • 信息来源：Hacker News 新项目
  • 发布时间：2026/10/1 22:06:18