seamless cant really handly long audio input ever since they switched to t4 from A100 :-) so we could probably cut long audio then speech to text then append results for final results.
· Sign up or log in to comment
