Archai Documentation#
Archai is a Neural Architecture Search (NAS) framework built upon PyTorch. It provides a comprehensive solution for automating the process of finding the optimal architecture for deep learning models, making it easier for researchers and practitioners to achieve state-of-the-art results. First launched as an open-source project in 2020, Archai has made impactful progress by forming a positive feedback loop between the engineering and research aspects.
It has innovated on both search algorithms and search spaces, explored ideas on zero-cost proxies of architecture accuracy and in very recent work explored novel more efficient alternatives to the ubiquitious attention operator which is now informing next-generation search-space design. Additionally, it offers the following advantages:
🔬 Easy mix-and-match between different algorithms;
📈 Self-documented hyper-parameters and fair comparison;
âš¡ Extensible and modular to allow rapid experimentation;
📂 Powerful configuration system and easy-to-use tools.
Citing Archai#
If you use Archai in a scientific publication, please consider citing it:
@misc{Archai:22,
title=Archai: Platform for Neural Architecture Search,
url=https://www.microsoft.com/en-us/research/project/archai-platform-for-neural-architecture-search,
journal=Microsoft Research,
year=2022,
month=Jul
}