A PyTorch implementation of training of NASBench
A PyTroch implementation of training of NASBench dataset: NAS-Bench-101: Towards Reproducible Neural Architecture Search. NASBench dataset contains 423,624 unique neural networks exhaustively generated and evaluated from a fixed graph-based search space.
Modified from NASBench: A Neural Architecture Search Dataset and Benchmark. graph_util.py and model_spec.py are directly copied from the original repo. Please noted that this repo is only used to train one possible architecture in the search space, not to generate all possible graphs and trained