MotherNet: Fast Training and Inference via Hyper-Network Transformers

TL;DR

MotherNet introduces a hypernetwork architecture that rapidly generates neural network weights for tabular data classification, achieving high efficiency and competitive accuracy without dataset-specific tuning.

Contribution

It presents MotherNet, a hypernetwork trained on synthetic tasks that can generate effective classifiers for new tabular datasets in a single forward pass, surpassing existing hypernetworks in flexibility and efficiency.

Findings

MotherNet outperforms neural networks trained with gradient descent on small datasets.

MotherNet's generated models are comparable to TabPFN and Gradient Boosting in accuracy.

MotherNet requires no fine-tuning or dataset-specific hyper-parameter tuning.

Abstract

Foundation models are transforming machine learning across many modalities, with in-context learning replacing classical model training. Recent work on tabular data hints at a similar opportunity to build foundation models for classification for numerical data. However, existing meta-learning approaches can not compete with tree-based methods in terms of inference time. In this paper, we propose MotherNet, a hypernetwork architecture trained on synthetic classification tasks that, once prompted with a never-seen-before training set generates the weights of a trained ``child'' neural-network by in-context learning using a single forward pass. In contrast to most existing hypernetworks that are usually trained for relatively constrained multi-task settings, MotherNet can create models for multiclass classification on arbitrary tabular datasets without any dataset specific gradient descent. The child network generated by MotherNet outperforms neural networks trained using gradient descent on small datasets, and is comparable to predictions by TabPFN and standard ML methods like Gradient Boosting. Unlike a direct application of TabPFN, MotherNet generated networks are highly efficient at inference time. We also demonstrate that HyperFast is unable to perform effective in-context learning on small datasets, and heavily relies on dataset specific fine-tuning and hyper-parameter tuning, while MotherNet requires no fine-tuning or per-dataset hyper-parameters.