Introducing TabFM: A zero-shot foundation model for tabular data
Summary
Google Research introduces TabFM, a zero-shot foundation model for tabular data that uses in-context learning to classify and regress on unseen tables. The model employs alternating row/column attention, row compression, and an in-context learning Transformer, trained on hundreds of millions of synthetic tables to generalize to real-world data. TabFM aims to deliver out-of-the-box predictions without traditional training or feature engineering and will be integrated into Google BigQuery via AI.PREDICT.