Rendered at 22:27:16 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
orsorna 2 days ago [-]
For anyone who was piqued by the title but couldn't stomach the unedited LLM blogpost, you can read the similarity AI-genned docs for Laya directly: https://github.com/NandhaKishorM/laya
ozozozd 2 days ago [-]
Man, it shows that someone was getting paid by the word.
Thank you!
CROON_tv 1 days ago [-]
Similar result in production for me. I compared Jev with a general LLM (DeepSeek) as the decider for when to cut live speech into sentences, on the same 226 segments. With Jev the longest held segment was 17 s. With DeepSeek it was 40 s.
f30e3dfed1c9 21 hours ago [-]
The flood of posts related to this thing does not seem organic.
mkrishnan 1 days ago [-]
I tried both with a synthetic dataset (1000 items), Jev scored 98% vs Laya 15%.
Thank you!