[FR]: UI: See Traces and LLM Calls of Evaluations when using LLM as Judges metrics

[SrBliss]

Proposal summary

See in the UI the traces that are produced when using LLM as Judges, both General and LLM spans.

Motivation

When using LLM as Judged metrics in your Evaluations, it is useful to track the Evaluation because:

• Sometimes you want to iterate on the creation of the actual metric, and you need to analyze / compare the prompts that were used to evaluate the LLM-App.
• It allows tracking of the full usage/cost of the Evaluation