pytorch-frame and pytorch_tabular
These are competitors offering overlapping functionality for deep learning on tabular data, with pytorch-tabular being the more mature and widely-adopted option, while pytorch-frame appears to be an emerging alternative that integrates more tightly with PyTorch Geometric's graph-based abstractions.
About pytorch-frame
pyg-team/pytorch-frame
Tabular Deep Learning Library for PyTorch
About pytorch_tabular
pytorch-tabular/pytorch_tabular
A unified framework for Deep Learning Models on tabular data
Implements diverse neural architectures (NODE, TabNet, TabTransformer, FT-Transformer, GATE, GANDALF, DANets) optimized for tabular data, leveraging PyTorch Lightning for distributed training with automatic hardware selection and experiment logging. Supports heterogeneous data through categorical embeddings and feature interaction learning, with extensible design enabling custom model implementations.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work