Downsampled ImageNet 32x32
Aaron van den Oord and Nal Kalchbrenner and Koray Kavukcuoglu

small (2 files)
train_32x32.tar 4.11GB
valid_32x32.tar 161.47MB
Type: Dataset

title= {Downsampled ImageNet 32x32},
keywords= {},
author= {Aaron van den Oord and Nal Kalchbrenner and Koray Kavukcuoglu},
abstract= {This page includes downsampled ImageNet images, which can be used for density estimation and generative modeling experiments. Images come in two resolutions: 32x32 and 64x64, and were introduced in Pixel Recurrent Neural Networks. Please refer to the Pixel RNN paper for more details and results. 

terms= {},
license= {},
superseded= {},
url= {}

10 day statistics (11 downloads)

Average Time 15 mins, 25 secs
Average Speed 4.62MB/s
Best Time 2 mins, 48 secs
Best Speed 25.44MB/s
Worst Time 26 mins, 59 secs
Worst Speed 2.64MB/s