Comment

Comment by lossolo

Has anyone used or is currently using 7B models in a production or commercial product? How was the performance? What kind of tasks were you using it for? Was it practical to use the small 7B model for your specific use case, or did you switch to OpenAI models or 30-70B open source models?
parent lossolo 3 years ago

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton