A framework for generating realistic LLM serving workloads
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).