Daniel N.

Nice, thank you! This worked for me

annotations['user feedback'].score >2 #thumbs up

Yes 🔒[private user], i can join a huddle in 5 minutes?

RunLLM

We restructured our workflow and ended up on this solution. Make a brand new project Trace LLM calls Fetch the LLM calls with px_client.query_spans Then use our custom eval function on the output Finally use px.Client().log_evaluations with the eval dataframe to log it into phoenix 😄 Is there any other solutions that would work better?

Now that i have added some traced spans to a new dataset can u show me step by step how to log evaluations to phoenix from this dataset. RunLLM

Does the datasets uploaded to phoenix have a span_id to it RunLLM

RunLLM where can i find the span_id column