google/tabfm-1.0.0-pytorch · Hugging Face

Source: https://huggingface.co/google/tabfm-1.0.0-pytorch TabFM is a zero-shot tabular foundation model from Google Research. It supports classification and regression on structured/tabular data with mixed numerical and categorical columns, requiring no fine-tuning or hyperparameter search - training examples are passed as context and predictions are made in a single forward pass.

This repository contains thePyTorchweights. For the JAX/Flax weights seegoogle/tabfm-1.0.0-jax.

https://huggingface.co/google/tabfm-1.0.0-pytorch#getting-startedGetting Started

pip install tabfm[pytorch]

Classification:

from tabfm import TabFMClassifier, tabfm_v1_0_0_pytorch as tabfm_v1_0_0

model = tabfm_v1_0_0.load(model_type="classification")
clf = TabFMClassifier(model=model)
clf.fit(X_train, y_train)
probs = clf.predict_proba(X_test)

Regression:

from tabfm import TabFMRegressor, tabfm_v1_0_0_pytorch as tabfm_v1_0_0

model = tabfm_v1_0_0.load(model_type="regression")
reg = TabFMRegressor(model=model)
reg.fit(X_train, y_train)
preds = reg.predict(X_test)

You can also load directly using the HuggingFace Hub API:

from tabfm.src.pytorch.model import TabFM

clf_model = TabFM.from_pretrained("google/tabfm-1.0.0-pytorch", subfolder="classification")
reg_model = TabFM.from_pretrained("google/tabfm-1.0.0-pytorch", subfolder="regression")

https://huggingface.co/google/tabfm-1.0.0-pytorch#available-checkpointsAvailable Checkpoints

SubfolderTaskis\_classifier``classification/Classification (up to 10 classes)True``regression/RegressionFalse

https://huggingface.co/google/tabfm-1.0.0-pytorch#developers-and-affiliationsDevelopers and Affiliations

Developed by theGoogle Researchteam.

https://huggingface.co/google/tabfm-1.0.0-pytorch#intended-useIntended Use

• Tabular data with numerical and/or categorical columns
• Binary and multiclass classification (up to 10 classes)
• Regression on continuous targets
• Zero-shot inference: no dataset-specific training or hyperparameter tuning
• Works with DataFrames (pandas) or numpy arrays

https://huggingface.co/google/tabfm-1.0.0-pytorch#not-intended-forNot Intended For

• Images, audio, video, or raw text
• More than 10 output classes (hard model limit)
• Tasks requiring task-specific fine-tuning
• Non-tabular structured data (graphs, sequences)
• Commercial use (see License below)

https://huggingface.co/google/tabfm-1.0.0-pytorch#model-architectureModel Architecture

TabFM uses alternating row and column attention to capture both feature interactions and row-level patterns:

1. Column attention(Set Transformer): embeds each cell using Fourier features and a per-group linear projection, then aggregates across rows via induced self-attention
2. Row compression: CLS tokens summarise each row into a dense vector via row-level attention with Rotary Position Embedding (RoPE)
3. ICL Transformer: a 24-block causal transformer operates over the compressed row vectors, treating training rows as context and outputting predictions for test rows

Key hyperparameters:

ParameterValueEmbedding dim256Column attention blocks3 (4 heads, 256 induced points)Row attention blocks3 (8 heads, 8 CLS tokens)ICL transformer blocks24 (8 heads)Feed-forward factor4Max classes10ActivationSwiGLUFourier features32 frequencies

https://huggingface.co/google/tabfm-1.0.0-pytorch#training-data-and-priorsTraining Data and Priors

TabFM was trained on hundreds of millions ofsyntheticdatasets generated dynamically using structural causal models (SCMs). Synthetic data was chosen due to the scarcity of diverse, high-quality open-source tabular datasets and to avoid privacy/licensing concerns with real-world industrial data. The SCM prior encodes inductive biases about causal structure and feature relationships typical in tabular tasks.

https://huggingface.co/google/tabfm-1.0.0-pytorch#performancePerformance

TabFM was evaluated onTabArenaacross 51 datasets (38 classification, 13 regression). In zero-shot mode - a single forward pass with no hyperparameter search - TabFM outperforms heavily-tuned supervised baselines including gradient-boosted trees. TheTabFMClassifier\.ensemble\(\)preset (feature crosses, SVD features, NNLS blending) yields further improvements.

See theGoogle Research blog postfor full benchmark details.

https://huggingface.co/google/tabfm-1.0.0-pytorch#ethical-considerationsEthical Considerations

TabFM was trained entirely on synthetic data. Performance on specific real-world domains, minority groups, or edge distributions is not fully characterised. Users should evaluate the model on held-out data representative of their use case before deploying in high-stakes settings.

https://huggingface.co/google/tabfm-1.0.0-pytorch#limitationsLimitations

• Max 10 classesfor classification (hard architectural limit)
• Memory usage scales with the number of training rows (all rows are passed as context)
• Optimised for tables up to 500 features; behaviour on very wide tables may degrade
• Performance is not guaranteed to match task-specific, fine-tuned models on all datasets
• Not an officially supported Google product

https://huggingface.co/google/tabfm-1.0.0-pytorch#licenseLicense

The model weights in this repository are released under theTabFM Non-Commercial License v1.0- seeLICENSE. The source code is Apache 2.0 licensed viagoogle-research/tabfm.

https://huggingface.co/google/tabfm-1.0.0-pytorch#versionVersion

1.0.0

https://huggingface.co/google/tabfm-1.0.0-pytorch#citationCitation

@article{tabfm2026,
  title  = {TabFM: A Zero-Shot Foundation Model for Tabular Data},
  author = {Google Research},
  year   = {2026},
  url    = {https://research.google/blog/introducing-tabfm-a-zero-shot-foundation-model-for-tabular-data/}
}