Tabm

Visit Tool

TabM is an open-source AI tool that advances tabular deep learning through parameter-efficient ensembling. It provides a Python package for building custom TabM-like architectures and reproducing research experiments.

Claim this tool

No Views Yet

At a glance

Pricing

Open Source

Free tier

Yes

API

Skill level

Technical

About

What is tabm?

TabM is an official open-source repository for the paper "TabM: Advancing Tabular Deep Learning With Parameter-Efficient Ensembling" (ICLR 2025). It offers a PyTorch-based Python package for implementing the TabM model, along with layers and tools for constructing custom architectures that efficiently ensemble MLP-like models. The tool is designed to improve performance on challenging tabular benchmarks like TabReD and has been successfully applied in Kaggle competitions. TabM is noted for its efficiency, being faster than prior tabular deep learning methods and capable of handling large datasets up to 100M+ objects. It allows for parallel training and weight sharing among MLPs, leading to better runtime, memory efficiency, and task performance.

Best used for

Ideal for machine learning researchers and data scientists who need to advance tabular deep learning, improve model performance on complex datasets, and build efficient, scalable AI models. Especially valuable for those working with large tabular datasets and seeking to reproduce or extend research findings.

Common actions

advance deep learning

improve tabular data analysis

build efficient AI models

reproduce research experiments

face swapping"AI Agents"open-sourceworkflowsautomated workflowgithub copilotcollaborationdeepfakelow-code/no-code

Capabilities

Key features

Parameter-efficient ensembling
PyTorch-based Python package
Custom TabM-like architectures
Parallel MLP training
Weight sharing between MLPs
Handles large datasets

Target Audience

machine learning researcherdata scientist

Integrations

Not yet documented

Pricing & Plans

Open Source

Free

FAQs

What is the core innovation of TabM?

TabM's core innovation lies in its parameter-efficient ensembling technique. It efficiently imitates an ensemble of MLPs by allowing parallel training and sharing weights between the MLPs, which significantly improves runtime, memory efficiency, and task performance compared to traditional ensembles.

How does TabM perform on challenging benchmarks?

TabM achieves higher performance on challenging benchmarks like TabReD, which features real-world industrial datasets with distribution drifts and hundreds of features. It has also been used in winning solutions for Kaggle competitions, demonstrating its effectiveness in practical scenarios.

Is TabM suitable for large datasets?

Yes, TabM is designed to be efficient and suitable for real-world applications, including large datasets. It has been successfully trained on datasets containing 13 million objects, and there are reports of successful training runs on over 100 million objects.

Trending

Subcategories trending in Research & Education

Study Assistants Knowledge Management Course Creation Scientific Computing Summarization Language Learning

Trending

Also listed in

This tool also appears in

Coding & Development › Open Source & Models Data & Analytics › Statistical & Scientific

Explore

Browse AI tools by category

Content & Design Productivity & Business Coding & Development AI Agents & Automation Research & Education Wellness & Lifestyle Career Development Marketing & Growth Data & Analytics Customer Support & CX Finance E-commerce