sarvam-jev reads typed decisions from Indic model logits
Inspired by Jev, sarvam-jev reads typed decisions directly from sarvam-1's logits instead of generating JSON token by token, running 8.8x faster with the same weights and output that cannot be malformed, with Hindi, Tamil and Bengali demos in the browser.
Inspired by @typesafeai's Jev, I built sarvam-jev - an open engine that reads typed decisions straight out of an Indic model's logits instead of generating JSON token by token. Same sarvam-1 weights, 8.8× faster than making it write the same answers — and the output can't be malformed, because nothing is generated. Runs entirely in your browser. No backend, no waitlist: https://sarvam-jev.feynmanpi.com/ Hindi, Tamil, Bengali demos.
This post quotes another post: “ChatGPT co-inventor releases Jev, a model built for decisions rather than text”