3 ms·
I have been using the API. There are conflicting reports in this thread that seems to indicate it may also be affected. I am not sure.
by ilaksh 3y ago
I have been using the API. There are conflicting reports in this thread that seems to indicate it may also be affected. I am not sure.
- londons_explore 3y agoDid you set the model code to gpt-4-0314? I did, and I still get the original speed (produces tokens at about the speed you would read aloud), and I haven't seen quality change.