If true, Meta AIs LLaMA-65 and even LLaMA-13B model outperforming GPT-3 show that trend goes towards optimized smaller foundation models, someone has to run all that computer power too…

Reply to this note

Please Login to reply.

Discussion

No replies yet.