Auto-PyTorch vs TPOT
Auto-PyTorch
automl.github.ioTPOT
epistasislab.github.ioSide by side
| Figure | Auto-PyTorch | TPOT |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | — | — |
| Open source | — | — |
| macOS | — | Yes |
| Linux | Yes | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 1 | 1 |
| Spec score · AutoML Software | 8.4 (#5) | 7.3 (#18) |
| Installation | The installation documentation specifies Linux, Python 3.7 or later, a C++11-capable compiler, and SWIG 3.0.*, and also documents a Docker image. | TPOT requires Python and can be installed in a conda environment or manually; the documented Python range is >=3.10 and <3.14. |
| Integrations | The documentation describes using Dask.distributed for parallel Bayesian optimization and sklearn column transformers for data preprocessing. | The installation documentation lists scikit-learn and offers extra scikit-learn extensions through the tpot[sklearnex] installation option. |
| License | The documentation states that Auto-PyTorch is licensed under the 3-clause BSD license. | TPOT is free software distributed under the GNU Lesser General Public License version 3 or later, without warranty. |
| Parallel processing | It supports parallel Bayesian optimization using Dask.distributed, and parallel workers need access to a shared file system for training data and models. | TPOT uses Dask for parallel processing and recommends guarding script code with an if __name__ == "__main__" block. |
| Purpose | Auto-PyTorch is an automated machine learning toolkit based on PyTorch that automates algorithm selection and hyperparameter tuning. | TPOT is a Python automated machine learning tool that optimizes machine learning pipelines using genetic programming. |
Plans
In their own words
Auto-PyTorch
Auto-PyTorch is a free, self-hosted machine-learning toolkit based on PyTorch. It automates algorithm selection and hyperparameter tuning, including neural-network architecture and training settings, within a user-defined search budget. Documented search methods include Bayesian optimisation, meta-learning, SMAC and Hyperband. Supported work includes tabular classification, tabular regression and time-series forecasting. Tabular preprocessing can perform imputation, categorical encoding, scaling and feature processing, while related hyperparameters are tuned. The toolkit can build ensembles from validation-set predictions, with controls for ensemble size and candidate limits. Users can set estimator memory limits and total wall time. Parallel Bayesian optimisation uses Dask.distributed; workers need shared filesystem access to training data and models. Installation documentation specifies Linux, Python 3.7 or later, a C++11-capable compiler and SWIG 3.0. SWIG 4.0 or later is unsupported. Forecasting needs extra dependencies. A Docker image and PyPI installation are documented, under the 3-clause BSD license.