OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Rtdl

Rtdl is a public Yandex Research repository that collects research papers, implementations and Python packages for deep learning on tabular data, including neural architectures, feature embeddings, nearest-neighbor methods, diffusion models and parameter-efficient ensembles.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Rtdl — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Rtdl when you need a research-oriented map of tabular deep-learning papers and related Python implementations, especially when selecting among architectures or following a method from paper to code.

Who it is for

  • Machine-learning researchers studying tabular deep learning.
  • Python and PyTorch developers comparing neural architectures for structured datasets.
  • Practitioners investigating tabular feature embeddings, ensembles or synthetic-data methods.

Documented uses

  • Find papers and linked code for deep-learning methods on tabular data.
  • Compare MLP, ResNet and FT-Transformer approaches.
  • Investigate numerical-feature embeddings in a tabular pipeline.
  • Explore TabM parameter-efficient ensembling.
  • Study or experiment with TabDDPM research code for tabular data generation.
Before you choose

Requirements & limits

The repository is public, accessible through GitHub and licensed under Apache-2.0. The original rtdl Python package is deprecated; the repository identifies replacement packages for its models and numerical-feature embeddings.

  • It is a repository of research and packages rather than a hosted tabular-ML application.
  • The main ecosystem is Python with PyTorch.
  • The deprecated rtdl package should be distinguished from the still-public repository.
  • The replacement rtdl_revisiting_models package has a slightly different API.
  • The repository recommends rtdl_num_embeddings over its unfinished continuous-feature embedding implementation.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Switch from the deprecated rtdl package

Move MLP, ResNet or FT-Transformer use to the replacement package identified by the repository.

Before you start

  • Existing use of the rtdl Python package, specifically its MLP, ResNet or FT-Transformer models.
  • Python and the replacement package documentation.

Steps

  1. Identify whether the existing code uses MLP, ResNet or FT-Transformer from the deprecated rtdl package.

  2. Use the documented rtdl_revisiting_models package as the replacement.

  3. Review and update the integration for its slightly different API.

Expected outcome

The corresponding models are available through rtdl_revisiting_models, with code changes potentially required for the API difference.

Guide limitations

  • The repository does not document a direct migration procedure or guarantee unchanged execution.
  • The original rtdl package is deprecated.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
yandex-research/rtdl · Official homepage title · retained in its source language
How we review sourcesSuggest a correction