🤖 TabICL Studio

In-Context Learning for Tabular Data — Classification & Regression

Built with Shiny for Python • Powered by TabICL

⚙️ Configuration

Data Source

Choose data source:

Sample Dataset

Upload CSV

Select Dataset

Upload CSV file

Browse...

Task & Target

Task Type:

TabICL Parameters

Number of Estimators

Number of ensemble members. Higher = more accurate but slower.

Test Size

Proportion of data to use for testing.

Random State

About TabICL

TabICL is a tabular foundation model that uses in-context learning for tabular data without traditional training. It supports both classification and regression tasks, and excels on datasets with 300-60K samples.