Skip to main content

​
Define Your Inferences

For a conceptual overview of inferences, including an explanation of when to use a single inference vs. primary and reference inferences, see Phoenix Basics.
To define inferences, you must load your data into a pandas dataframe and create a matching schema. If you have a dataframe prim_df and a matching prim_schema, you can define inferences named “primary” with 
prim_ds = px.Inferences(prim_df, prim_schema, "primary")
If you additionally have a dataframe ref_df and a matching ref_schema, you can define a inference set named “reference” with 
ref_ds = px.Inferences(ref_df, ref_schema, "reference")
See Corpus Data if you have corpus data for an Information Retrieval use case.

​
Launch the App

Use phoenix.launch_app to start your Phoenix session in the background. You can launch Phoenix with zero, one, or two inference sets.

No Inferences

session = px.launch_app()
• Run Phoenix in the background to collect OpenInference traces emitted by your instrumented LLM application.

Single Inference Set

session = px.launch_app(ds)
• Analyze a single cohort of data, e.g., only training data.• Check model performance and data quality, but not drift.

Primary and Reference Inference Sets

session = px.launch_app(prim_ds, ref_ds)
• Compare cohorts of data, e.g., training vs. production.• Analyze drift in addition to model performance and data quality.

Primary and Corpus Inference Sets

session = px.launch_app(query_ds, corpus=corpus_ds)
• Compare a query inference set to a corpus dataset to analyze your retrieval-augmented generation applications.

​
Open the UI

You can view and interact with the Phoenix UI either directly in your notebook or in a separate browser tab or window.
  • In the Browser
  • In Your Notebook
In a notebook cell, run
session.url
Copy and paste the output URL into a new browser tab or window.
Browser-based sessions are supported in both local Jupyter environments and Colab.

​
Close the App

When you’re done using Phoenix, gracefully shut down your running background session with 
px.close_app()