Checkpoints

This repository contains the trained models that we use in "Conditioning diffusions via explicit forward-backward bridging", see, https://github.com/zgbkdlm/fbs. Specifically,

  • celeba-64_lin_2999.npz is the checkpoint that we use for CelebA-HQ (64 x 64) inpainting and super-resolution.
  • mnist_lin_2999.npz is the checkpoint that we use for MNIST (32 x 32) inpainting and super-resolution.
  • sb_mnist_lin_19.npz is the checkpoint that we use for Shroödinger bridge MNIST (32 x 32) inpainting and super-resolution.

Datasets

The MNIST dataset mnist.npz is licensed under Creative Commons Attribution-Share Alike 3.0 license (see, https://keras.io/api/datasets/mnist/).

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support