This model's maximum context length is 4097 tokens. However, you requested 4128 tokens (1032 in the messages, 3096 in the completion). Please reduce the length of the messages or completion.