LiteLLM - RPM Limits & TPM limits via config.yaml per API key

[Chriss722331]

I have deployed Litlellm in EKS and I am looking forward to implement rpm & tpm limits per api key via config . In my case the user will send the api key as part of the request header and I expect LiteLLM to validate the count and return and error in case it exceeds. I have also redis provisioned since I have multiple instances.

Could someone provide the configuration for the above requirement. I have a proxy_config.yaml in which I have defined the litellm_settings, general settings etc.. To this I want to append the rpm and tpm limits per api key and also the redis config.