WN1695173791/classifier-free-diffusion-guidance-Pytorch

a simple unofficial implementation of classifier-free diffusion guidance

Unofficial Implementation of Classifier-free Diffusion Guidance

The Pytorch implementation is adapted from openai/guided-diffusion with modifications for classifier-free conditioned generation. The dataset used for training is cifar-10. The platform I use is Ubuntu 20.04.

During the training porcess, I adapt the file Scheduler.py from zoubohao/DenoisingDiffusionProbabilityModel-ddpm-, which is different from my implementation mainly on one technical detail : How to represent the embeddings of null token of class identifier. According to another academic paper Video Diffusion Models by the same authors, I use the same method as theirs in this detail.

How to run

First, you need to do some preparations:

mkdir sample
mkdir model
ln -s absolute/path/to/cifar-10 ./cifar-10-batches-py

How to train

make train

NOTICE : hyperparameter settings are in the file train.py

How to sample

To get pictures, you need to excute the following command:

make samplepict

To get .npz files(used for FID calculation), you need to excute the following command:

make samplenpz

NOTICE : hyperparameter settings are in the file sample.py

Performance