LangSmith · Module
Online evaluators
Online code evaluators run checks on selected production traces and add feedback. You configure them in LangSmith, filter the runs, and write the check in Python or JavaScript.
Last updated
- Since
- v0.14
Syntax
def evaluate(run):
return {"key": "feedback_key", "score": value}In-depth API docs are only for subscribers: parameters, examples, common mistakes and interview practice.
SubscribeRelated
Online evaluationsCode evaluatorsFilter traces (ClickHouse)Chain evaluatorsAdding metadata to your traces