Open ClemensGruber opened 7 months ago
Good catch:
OpenAI API and other LLM APIs response time tracker https://gptforwork.com/tools/openai-api-and-other-llm-apis-response-time-tracker
Generally GPT 4 seems to have more latency than 3.5 and the most interesting part: Response time is at arround 2 sec at GPT 3.5 while we have to wait 5 sec with GPT 4!
We have 20 to 40 seconds waiting time with GPTree so it seems as a bigger problemm in our implementation!
Todo
2
31
Already tested and less effect found on the latency of the entire system
4
30
27
Workaround
28