3 ms·
With open models, yes we are at the performance of at least the first release of ChatGPT 4.
by segmondy 3y ago
With open models, yes we are at the performance of at least the first release of ChatGPT 4.
- sp332 3y agoCould you recommend one or a few in particular?
- sanjiwatsuki 3y agoThe current best open weights model is probably Cohere Command-R+. The memory requirements on it are quite high, though.
- bevekspldnw 3y agoI really want to see some benchmarks with performance weighted by energy use. I think Mistral 7B performance to watt would be the leader by a huge margin. On many tasks I get equal performance on zero shot classification tasks on Mistral than in bigger models.