Looking for usage documentation? Check out Quickstart, Classification, and Regression.
Installation
Reference topics
TabPFNClassifier
Constructor options, fitting, prediction, embeddings, and persistence.
TabPFNRegressor
Constructor options, fitting, prediction, and predictive distributions.
Inference configuration
Settings accepted by the estimators through inference_config.
Inference tuning
Tuning configuration and performance metrics.
Fine-tuning
Fine-tuning estimators, batch types, and experiment logging.
Preprocessing
Preprocessing configuration, pipelines, and presets.
Model versions
Model version identifiers.
Model persistence
Save and load fitted models and model specifications.
Diagnostics
Environment diagnostics and error types.