To achieve end-to-end tracing of your LLM requests, you can utilize Middleware’s LLM Observability. Here are the steps to set it up:
-
Install the Traceloop or Opelit SDK based on your tech stack.
-
Refer to the LLM Observability documentation for guidance.
-
Navigate to the platform and select the LLM Observability section from the left-hand menu.
-
You will find all the traces displayed on the LLM Traces dashboard.
-
Click on any trace to view detailed insights, including the duration of the request, total tokens used, type of LLM request, and the model used.
-
Additionally, you can explore the flame graph, map, and waterfall views of the entire request, along with the spans.