Observe your GPT-3 logs

Track usage, costs, and latency metrics with one line of code