发帖
雷达卡
Dataset Condensation with Differentiable Siamese Augmentation
Bo Zhao 1 Hakan Bilen 1
Abstract massive datasets (e.g. (Kuznetsova et al., 2020; Chen et al.,
In many machine learning problems, large-scale 2020a; Kwiatkowski et al., 2019)). For example, it takes
datasets have become the de-facto standard to 12.3k TPU days to train EfficientNet-L2 (Xie et al., 2020) on
train state-of-the-art deep networks at the price ...
京公网安备 11010802022788号
