New papers: TabM & TabReD

Two new papers were announced and added to the list in README:

• (2024) TabM: Advancing Tabular Deep Learning with Parameter-Efficient Ensembling
• (2024) TabReD: Analyzing Pitfalls and Filling the Gaps in Tabular Deep Learning Benchmarks

Two new papers

Two papers were added to the list:

• TabR: Unlocking the Power of Retrieval-Augmented Tabular Deep Learning
• TabDDPM: Modelling Tabular Data with Diffusion Models

P.S. Plus, one missing paper was added: Neural Oblivious Decision Ensembles for Deep Learning on Tabular Data

v0.0.13 New paper!

This is a technical release, no changes in API. Also, check out our new paper "On Embeddings for Numerical Features in Tabular Deep Learning".

Changes

• minor documentation fix
• moved paper implementations to separate repositories

v0.0.12 New paper!

This is a technical update which was required for the release of our new paper "On Embeddings for Numerical Features in Tabular Deep Learning". This version does not bring any functional changes to the rtdl library compared to the previous version (v0.0.10).

v0.0.10

New features

• rtdl.data.get_category_sizes

Project

• some changes in the documentation reflecting the new structure of the repository

v0.0.9

This is a hot-fix release after the big 0.0.8 release (see the release notes for 0.0.8):

• revert the breaking change in NumericalFeatureTokenizer accidentally introduced in 0.0.8
• minor documentation refinements

v0.0.8

This release focuses on improving the documentation.

Documentation

• The following models and classes are now documented:
  • MLP
  • ResNet
  • FTTransformer
  • MultiheadAttention
  • NumericalFeatureTokenizer
  • CategoricalFeatureTokenizer
  • FeatureTokenizer
  • CLSToken
• Usability have been greatly improved:
  • signatures are now highlighted
  • added the "copy" button to code blocks
  • permalink buttons (signature anchors) are now visible

Bug fixes

• MultiheadAttention: fix the crash when bias=False

Dependencies

• numpy >= 1.18
• torch >= 1.7

Project

• added spell checking for documentation
• sphinx was updated to 4.2.0
• flit was updated to 3.4.0

v0.0.7

API changes

• remove FlatEmbedding

Project

• remove bin, lib and output from the PyPI package

v0.0.6

v0.0.6

New features

• CLSToken (old name: "AppendCLSToken"): add expand method for easy construction of batches of [CLS]-tokens

Bug fixes

• FTTransformer: the make_baseline method now properly constructs an instance

API changes

• FTTransformer: the ffn_d_intermidiate argument was renamed to a more conventional ffn_d_hidden
• FTTransformer: the normalization argument was split into three arguments: attention_normalization, ffn_normalization, head_normalization
• ResNet: the d_intermidiate argument was renamed to a more conventional d_hidden
• AppendCLSToken: renamed to CLSToken

Documentation improvements

• CLSToken
• MLP.make_baseline

Project

• add tests with CUDA
• remove the .vscode directory from the repository

v0.0.5

API Changes:

• MLP.make_baseline is now more user-friendly and accepts a single d_layers argument instead of four (d_first, d_intermidiate, d_last, n_blocks)