rancid glacier
Hey Guys, I really need your help. We are planning to go to prod and ran 5 workers to proceed with stress test of our backend, to see if we works fine, but we barely reached 5% of our limits and got into this error. Our RPM for GPT3.5 is 3,500 and for GPT4 is 200, and TPM is 90k and 40k.
Will be very thankful for any suggestions.
hasty saddle