Consistent Timeout Issue with CogVLM2 model

I am using the CogVLM2 model (from THUDM/CogVLM-CogAgent), and I have been using it for the last 4 weeks perfectly. Just today and yesterday, it has been thinking for 62-63 seconds and then says “Timeout! Please wait a few minutes and retry.” Usually, it takes 10-15 seconds and gives me text output. I was wondering what was wrong and if I can do something to resolve this issue.

(THUDM/cogvlm2-llama3-chat-19B · Hugging Face)