I faced latency challenges on llama2. So I haven’t used it. Usually OpenAI response is within 2 seconds, but llama2 takes 20+ seonds for the same. How are the

I faced latency challenges on llama2. So I haven’t used it. Usually OpenAI response is within 2 seconds, but llama2 takes 20+ seonds for the same. How are they tackling the performance?