Announcement
Open-model selection and evaluation at a two-day AI conference
Together Compute says the lineup covered how to shortlist open models and evaluate them against production traffic, not just benchmarks.
TLDR
Together Compute says a recurring question at the two-day AI Conference was which open model to use. Its recap highlights Rochelle Mattern’s keynote on shortlisting and evaluating models and an Ollama breakout on testing them against production traffic, not just benchmarks. It also mentions eight booth lightning talks and two trivia rounds.
Combined views
6.3K
4 Sources, first seen ago
