- DDPM Implementation with PyTorch.
- The
tests
directory is created to check if the functions of DDPM are working properly and for analysis purposes.
# Create Virtual Environment
python -m venv .venv
# Activate Virtual Environment
.venv\Scripts\activate # Windows
source .venv/bin/activate # Linux
# Install Packages
pip install -r requirements.txt
# Train
python train.py --config=base
# Sampling
python infer.py --config=base
- DPM: Sohl-Dickstein, Jascha, et al. "Deep unsupervised learning using nonequilibrium thermodynamics." International conference on machine learning. PMLR, 2015.
- DDPM: Ho, Jonathan, Ajay Jain, and Pieter Abbeel. "Denoising diffusion probabilistic models." Advances in neural information processing systems 33 (2020): 6840-6851.
- Original Repository: @lucidrains,
denoising-diffusion-pytorch