Reading Parquet Files with Phoenix for Data Evaluation

Hi, I'm using the ParquetCallback class to log the data generated by a production app, as explained here https://docs.llamaindex.ai/en/stable/examples/callbacks/OpenInferenceCallback/. Is there a way to read in those parquet files into phoenix later for evaluation?

· Sorted by Oldest

Mikyo
·
Hey Boris D. thanks for trying out Phoenix and the callback! In general yes that data can be viewed in Phoenix as inferences but our evaluations are being built on top of tracing and our future datasets work. Have you tried out Phoenix tracing? I think it might be what you are looking for? https://colab.research.google.com/github/Arize-ai/phoenix/blob/main/tutorials/llm_ops_overview.ipynb

Mikyo
·
Hey Boris D. thanks for trying out Phoenix and the callback! In general yes that data can be viewed in Phoenix as inferences but our evaluations are being built on top of tracing and our future datasets work. Have you tried out Phoenix tracing? I think it might be what you are looking for? https://colab.research.google.com/github/Arize-ai/phoenix/blob/main/tutorials/llm_ops_overview.ipynb