Skip to content

Add pre-trained models

Latest
Compare
Choose a tag to compare
@selflein selflein released this 13 Dec 19:52

Pre-trained models

  • Two pre-trained models with WideResNet28-10 architecture trained on CIFAR-10 trained with forward and reverse KL divergence.
  • Trained model on the three Gaussians synthetic dataset with standard deviation 1 and 4.