ChristophReich1996/HyperMixer

PyTorch reimplementation of the paper "HyperMixer: An MLP-based Green AI Alternative to Transformers" [arXiv 2022].

22
/ 100
Experimental

This is a tool for machine learning researchers and practitioners who are building neural networks. It provides an alternative architectural component to the commonly used 'Transformer' blocks, using simpler Multi-Layer Perceptrons (MLPs). You would use this by integrating the HyperMixer block into your PyTorch models, receiving a processed tensor as output that represents a more energy-efficient computation.

No commits in the last 6 months.

Use this if you are a machine learning researcher or engineer exploring novel, more energy-efficient neural network architectures for sequence data processing.

Not ideal if you are looking for a high-level, ready-to-use model for a specific task rather than an architectural component for deep learning.

deep-learning-architecture neural-network-design energy-efficient-ai machine-learning-research
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 0 / 25

How are scores calculated?

Stars

18

Forks

Language

Python

License

MIT

Last pushed

Mar 28, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/ChristophReich1996/HyperMixer"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.