NAS(Neural Architecture Search)

FBNet

My MXNet implementation of FBNet has some weird problems, so i reimplement with PyTorch. Support multi gpus training

• Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search PyTorch

FBNet

• Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search MXNet

SNAS

• DARTS: Differentiable Architecture Search, forker from darts
• SNAS: Stochastic Neural Architecture Search pytorch