DeepTab: Tabular Deep Learning Made Simple

DeepTab is a Python library for deep learning on tabular data, built on PyTorch and Lightning with a scikit-learn compatible API. It offers 15 neural architectures, from Mamba-inspired state space models and Transformers to tree ensembles and MLP baselines, each available as a classifier, regressor, or distributional (LSS) model. One fit/predict/evaluate workflow covers everyday modeling, architecture research, and production deployment.

from deeptab.models import MambularClassifier

model = MambularClassifier()
model.fit(X_train, y_train, max_epochs=50)

predictions = model.predict(X_test)
probabilities = model.predict_proba(X_test)

Why DeepTab

  • Familiar interface. A scikit-learn fit/predict/evaluate API that drops into existing pipelines, including GridSearchCV.

  • Automatic preprocessing. Feature-type detection, encoding, scaling, and missing-value handling are powered by PreTab and applied for you.

  • One model, three tasks. Every architecture ships as a classifier, a regressor, and a distributional (LSS) variant for uncertainty quantification.

  • A broad model zoo. 15 stable architectures plus experimental models, all behind the same interface, with selection guidance.

  • Built for real data. Mixed feature types, class imbalance, GPU acceleration, and early stopping work out of the box.

Installation

pip install deeptab

DeepTab requires Python 3.10+ and installs PyTorch automatically. See Installation for GPU setup and the optional Mamba CUDA kernels.

What’s New in v2.0

v2.0 is a ground-up restructuring of DeepTab. The high-level estimator API stays familiar, while the package layout, configuration objects, and import paths have been updated.

  • Split-config API: separate model, preprocessing, and training configuration objects, so each concern can be tuned on its own. This is the first thing you reach for in v2.

  • New models: AutoInt, ENODE, and TabR (stable); Tangos, Trompt, and ModernNCA (experimental).

  • Observability: ObservabilityConfig adds structured lifecycle logging via structlog and one-line MLflow or TensorBoard tracking, opt-in and silent by default.

  • Deployment-safe inference: InferenceModel exposes a read-only prediction surface with schema validation, so a served model cannot be re-fitted by accident.

  • Self-describing artifacts: a single .deeptab save format bundles the architecture, feature schema, preprocessing, and versions with the weights.

  • Registry-driven training: optimizers, schedulers, and losses are selectable by name through TrainerConfig and extensible at runtime.

  • Unified metrics: 25+ metric classes auto-selected per task across regression, classification, and distributional models.

  • Typed data layer: TabularDataset, TabularDataModule, and FeatureSchema give the pipeline an inspectable contract.

  • Reproducibility: cross-platform seeding across CPU, CUDA, and MPS.

  • Rebuilt docs and tutorials: refreshed guides plus end-to-end, Colab-ready tutorials for classification, regression, and uncertainty quantification.

Warning

v2.0 is not backward compatible with v1, and v1 is no longer maintained. Three things changed that affect existing code:

  1. Import paths were reorganised under the deeptab namespace.

  2. Config classes lost their Default prefix, so DefaultMambularConfig is now MambularConfig. Settings are also split across MambularConfig (architecture), PreprocessingConfig (feature handling), and TrainerConfig (optimisation).

  3. Data modules were renamed to TabularDataModule and TabularDataset; the old Mambular* aliases are deprecated.

If you need to stay on v1 for now, pin deeptab<2.0. Note that the v1 branch receives no bug fixes or security updates. See the migration guide for a step-by-step upgrade walkthrough, and the FAQ for the full support policy.

The high-level fit/predict/evaluate workflow is unchanged. In most cases only the imports and config construction need updating:

# v1: settings passed as flat keyword arguments on the estimator
from deeptab.models import MambularClassifier

model = MambularClassifier(d_model=128, n_layers=4, numerical_preprocessing="ple")
model.fit(X_train, y_train, max_epochs=50)

# v2: settings grouped into focused config objects
from deeptab.models import MambularClassifier
from deeptab.configs import MambularConfig, PreprocessingConfig

model = MambularClassifier(
    model_config=MambularConfig(d_model=128, n_layers=4),          # architecture
    preprocessing_config=PreprocessingConfig(numerical_preprocessing="ple"),  # features
)
model.fit(X_train, y_train, max_epochs=50)

Note

You only pass the configs you want to customise. MambularClassifier() with no arguments uses sensible defaults for the architecture, preprocessing, and training. The flat keyword-argument style from v1 is no longer accepted, so settings must go through the relevant config object.

See the Overview for the full picture.

Available Models

DeepTab provides 15 stable architectures across five families: State Space Models (Mambular, MambaTab, MambAttention), Transformers (FTTransformer, TabTransformer, SAINT, AutoInt), residual networks (ResNet, TabR), tree-inspired models (NODE, ENODE, NDTF), and general baselines (MLP, TabM, TabulaRNN). Three experimental models (ModernNCA, Tangos, Trompt) are under evaluation for promotion.

Each architecture comes in three variants, *Classifier, *Regressor, and *LSS, sharing one interface so you can swap models without changing code. See the Model Zoo for comparisons and selection guidance.

Documentation

Note

These are starting points for each area. The sidebar navigation is the source of truth, so please review the individual documentation sections for the latest updates and further documentation.

Citation

If you use DeepTab in your research, please cite:

@article{thielmann2024mambular,
  title={Mambular: A Sequential Model for Tabular Deep Learning},
  author={Thielmann, Anton Frederik and Kumar, Manish and Weisser, Christoph and Reuter, Arik and S{\"a}fken, Benjamin and Samiee, Soheila},
  journal={arXiv preprint arXiv:2408.06291},
  year={2024}
}

@article{thielmann2024efficiency,
  title={On the Efficiency of NLP-Inspired Methods for Tabular Deep Learning},
  author={Thielmann, Anton Frederik and Samiee, Soheila},
  journal={arXiv preprint arXiv:2411.17207},
  year={2024}
}

License

DeepTab is licensed under the MIT License. See LICENSE for details.