DeepArchitect: A Framework for Architecture Search. DeepArchitect is a framework for architecture search in arbitrary domains. DeepArchitect was designed with a focus on modularity, ease of use, reusability, and extensibility. DeepArchitect aims to impact the workflows of both researchers and practitioners by reducing the burden resulting from the large number of choices needed to design deep learning models. We recommend the reader to start with the overview, tutorials (e.g., here) and simple examples (e.g., here) to get a gist of the framework. Currently, we support Tensorflow, Keras, and PyTorch. See here for minimal complete examples for each of these frameworks. It should be straightforward to adapt these examples for your use cases. See here to learn how to support new frameworks, which should require minimal adaptation of the existing framework helpers found here. Questions and bug reports should be submitted through Github issues. See here for details on how to contribute.

References in zbMATH (referenced in 1 article )

Showing result 1 of 1.
Sorted by year (citations)

  1. Dai, Xiaoliang; Yin, Hongxu; Jha, Niraj K.: NeST: a neural network synthesis tool based on a grow-and-prune paradigm (2019)