Tracking Precision and Recall Metrics in Phoenix UI for Projects

·May 28, 2024 06:02 PM

I like that some of the Colab notebooks use classification_report from sklearn , It's helpful to see precision and recall for the span evaluations I'm using. Is there a way I can log this data onto the Phoenix UI so that I can track the results there, perhaps at a project level?

4 comments

· Sorted by Oldest

Phoenix Support

Tracking Precision and Recall Metrics in Phoenix UI for Projects

Alistair J.

·May 28, 2024 06:02 PM

I like that some of the Colab notebooks use classification_report from sklearn , It's helpful to see precision and recall for the span evaluations I'm using. Is there a way I can log this data onto the Phoenix UI so that I can track the results there, perhaps at a project level?

4 comments

· Sorted by Oldest

Xander S.
·
Hey Alistair J., it sounds like you're after precision/ recall/ F1 on a dataset over time?
Alistair J.
·
Correct!
Xander S.
·
Sounds good. We're actively building out datasets right now. This feature sounds like part of experiment tracking, which we will start after datasets. Feel free to file us an issue on GitHub or add a comment to this ticket.
Alistair J.
·
Great news. Will be watching this issue, thanks!
👍1

Xander S.
·
Hey Alistair J., it sounds like you're after precision/ recall/ F1 on a dataset over time?
Alistair J.
·
Correct!
Xander S.
·
Sounds good. We're actively building out datasets right now. This feature sounds like part of experiment tracking, which we will start after datasets. Feel free to file us an issue on GitHub or add a comment to this ticket.
Alistair J.
·
Great news. Will be watching this issue, thanks!
👍1