PyTorch implementation for Score-Based Generative Modeling through Stochastic Differential Equations (ICLR 2021, Oral)

Overview

Score-Based Generative Modeling through Stochastic Differential Equations

PWC

This repo contains a PyTorch implementation for the paper Score-Based Generative Modeling through Stochastic Differential Equations

by Yang Song, Jascha Sohl-Dickstein, Diederik P. Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole


We propose a unified framework that generalizes and improves previous work on score-based generative models through the lens of stochastic differential equations (SDEs). In particular, we can transform data to a simple noise distribution with a continuous-time stochastic process described by an SDE. This SDE can be reversed for sample generation if we know the score of the marginal distributions at each intermediate time step, which can be estimated with score matching. The basic idea is captured in the figure below:

schematic

Our work enables a better understanding of existing approaches, new sampling algorithms, exact likelihood computation, uniquely identifiable encoding, latent code manipulation, and brings new conditional generation abilities (including but not limited to class-conditional generation, inpainting and colorization) to the family of score-based generative models.

All combined, we achieved an FID of 2.20 and an Inception score of 9.89 for unconditional generation on CIFAR-10, as well as high-fidelity generation of 1024px Celeba-HQ images (samples below). In addition, we obtained a likelihood value of 2.99 bits/dim on uniformly dequantized CIFAR-10 images.

FFHQ samples

What does this code do?

Aside from the NCSN++ and DDPM++ models in our paper, this codebase also re-implements many previous score-based models in one place, including NCSN from Generative Modeling by Estimating Gradients of the Data Distribution, NCSNv2 from Improved Techniques for Training Score-Based Generative Models, and DDPM from Denoising Diffusion Probabilistic Models.

It supports training new models, evaluating the sample quality and likelihoods of existing models. We carefully designed the code to be modular and easily extensible to new SDEs, predictors, or correctors.

JAX version

Please find a JAX implementation here, which additionally supports class-conditional generation with a pre-trained classifier, and resuming an evalution process after pre-emption.

JAX vs. PyTorch

In general, this PyTorch version consumes less memory but runs slower than JAX. Here is a benchmark on training an NCSN++ cont. model with VE SDE. Hardware is 4x Nvidia Tesla V100 GPUs (32GB)

Framework Time (second per step) Memory usage in total (GB)
PyTorch 0.56 20.6
JAX (n_jitted_steps=1) 0.30 29.7
JAX (n_jitted_steps=5) 0.20 74.8

How to run the code

Dependencies

Run the following to install a subset of necessary python packages for our code

pip install -r requirements.txt

Stats files for quantitative evaluation

We provide the stats file for CIFAR-10. You can download cifar10_stats.npz and save it to assets/stats/. Check out #5 on how to compute this stats file for new datasets.

Usage

Train and evaluate our models through main.py.

main.py:
  --config: Training configuration.
    (default: 'None')
  --eval_folder: The folder name for storing evaluation results
    (default: 'eval')
  --mode: <train|eval>: Running mode: train or eval
  --workdir: Working directory
  • config is the path to the config file. Our prescribed config files are provided in configs/. They are formatted according to ml_collections and should be quite self-explanatory.

    Naming conventions of config files: the path of a config file is a combination of the following dimensions:

    • dataset: One of cifar10, celeba, celebahq, celebahq_256, ffhq_256, celebahq, ffhq.
    • model: One of ncsn, ncsnv2, ncsnpp, ddpm, ddpmpp.
    • continuous: train the model with continuously sampled time steps.
  • workdir is the path that stores all artifacts of one experiment, like checkpoints, samples, and evaluation results.

  • eval_folder is the name of a subfolder in workdir that stores all artifacts of the evaluation process, like meta checkpoints for pre-emption prevention, image samples, and numpy dumps of quantitative results.

  • mode is either "train" or "eval". When set to "train", it starts the training of a new model, or resumes the training of an old model if its meta-checkpoints (for resuming running after pre-emption in a cloud environment) exist in workdir/checkpoints-meta . When set to "eval", it can do an arbitrary combination of the following

    • Evaluate the loss function on the test / validation dataset.

    • Generate a fixed number of samples and compute its Inception score, FID, or KID. Prior to evaluation, stats files must have already been downloaded/computed and stored in assets/stats.

    • Compute the log-likelihood on the training or test dataset.

    These functionalities can be configured through config files, or more conveniently, through the command-line support of the ml_collections package. For example, to generate samples and evaluate sample quality, supply the --config.eval.enable_sampling flag; to compute log-likelihoods, supply the --config.eval.enable_bpd flag, and specify --config.eval.dataset=train/test to indicate whether to compute the likelihoods on the training or test dataset.

How to extend the code

  • New SDEs: inherent the sde_lib.SDE abstract class and implement all abstract methods. The discretize() method is optional and the default is Euler-Maruyama discretization. Existing sampling methods and likelihood computation will automatically work for this new SDE.
  • New predictors: inherent the sampling.Predictor abstract class, implement the update_fn abstract method, and register its name with @register_predictor. The new predictor can be directly used in sampling.get_pc_sampler for Predictor-Corrector sampling, and all other controllable generation methods in controllable_generation.py.
  • New correctors: inherent the sampling.Corrector abstract class, implement the update_fn abstract method, and register its name with @register_corrector. The new corrector can be directly used in sampling.get_pc_sampler, and all other controllable generation methods in controllable_generation.py.

Pretrained checkpoints

All checkpoints are provided in this Google drive.

Instructions: You may find two checkpoints for some models. The first checkpoint (with a smaller number) is the one that we reported FID scores in our paper's Table 3 (also corresponding to the FID and IS columns in the table below). The second checkpoint (with a larger number) is the one that we reported likelihood values and FIDs of black-box ODE samplers in our paper's Table 2 (also FID(ODE) and NNL (bits/dim) columns in the table below). The former corresponds to the smallest FID during the course of training (every 50k iterations). The later is the last checkpoint during training.

Per Google's policy, we cannot release our original CelebA and CelebA-HQ checkpoints. That said, I have re-trained models on FFHQ 1024px, FFHQ 256px and CelebA-HQ 256px with personal resources, and they achieved similar performance to our internal checkpoints.

Here is a detailed list of checkpoints and their results reported in the paper. FID (ODE) corresponds to the sample quality of black-box ODE solver applied to the probability flow ODE.

Checkpoint path FID IS FID (ODE) NNL (bits/dim)
ve/cifar10_ncsnpp/ 2.45 9.73 - -
ve/cifar10_ncsnpp_continuous/ 2.38 9.83 - -
ve/cifar10_ncsnpp_deep_continuous/ 2.20 9.89 - -
vp/cifar10_ddpm/ 3.24 - 3.37 3.28
vp/cifar10_ddpm_continuous - - 3.69 3.21
vp/cifar10_ddpmpp 2.78 9.64 - -
vp/cifar10_ddpmpp_continuous 2.55 9.58 3.93 3.16
vp/cifar10_ddpmpp_deep_continuous 2.41 9.68 3.08 3.13
subvp/cifar10_ddpm_continuous - - 3.56 3.05
subvp/cifar10_ddpmpp_continuous 2.61 9.56 3.16 3.02
subvp/cifar10_ddpmpp_deep_continuous 2.41 9.57 2.92 2.99
Checkpoint path Samples
ve/bedroom_ncsnpp_continuous bedroom_samples
ve/church_ncsnpp_continuous church_samples
ve/ffhq_1024_ncsnpp_continuous ffhq_1024
ve/ffhq_256_ncsnpp_continuous ffhq_256_samples
ve/celebahq_256_ncsnpp_continuous celebahq_256_samples

Demonstrations and tutorials

Link Description
Open In Colab Load our pretrained checkpoints and play with sampling, likelihood computation, and controllable synthesis (JAX + FLAX)
Open In Colab Load our pretrained checkpoints and play with sampling, likelihood computation, and controllable synthesis (PyTorch)
Open In Colab Tutorial of score-based generative models in JAX + FLAX
Open In Colab Tutorial of score-based generative models in PyTorch

Tips

  • When using the JAX codebase, you can jit multiple training steps together to improve training speed at the cost of more memory usage. This can be set via config.training.n_jitted_steps. For CIFAR-10, we recommend using config.training.n_jitted_steps=5 when your GPU/TPU has sufficient memory; otherwise we recommend using config.training.n_jitted_steps=1. Our current implementation requires config.training.log_freq to be dividable by n_jitted_steps for logging and checkpointing to work normally.
  • The snr (signal-to-noise ratio) parameter of LangevinCorrector somewhat behaves like a temperature parameter. Larger snr typically results in smoother samples, while smaller snr gives more diverse but lower quality samples. Typical values of snr is 0.05 - 0.2, and it requires tuning to strike the sweet spot.
  • For VE SDEs, we recommend choosing config.model.sigma_max to be the maximum pairwise distance between data samples in the training dataset.

References

If you find the code useful for your research, please consider citing

@inproceedings{
  song2021scorebased,
  title={Score-Based Generative Modeling through Stochastic Differential Equations},
  author={Yang Song and Jascha Sohl-Dickstein and Diederik P Kingma and Abhishek Kumar and Stefano Ermon and Ben Poole},
  booktitle={International Conference on Learning Representations},
  year={2021},
  url={https://openreview.net/forum?id=PxTIG12RRHS}
}

This work is built upon some previous papers which might also interest you:

  • Song, Yang, and Stefano Ermon. "Generative Modeling by Estimating Gradients of the Data Distribution." Proceedings of the 33rd Annual Conference on Neural Information Processing Systems. 2019.
  • Song, Yang, and Stefano Ermon. "Improved techniques for training score-based generative models." Proceedings of the 34th Annual Conference on Neural Information Processing Systems. 2020.
  • Ho, Jonathan, Ajay Jain, and Pieter Abbeel. "Denoising diffusion probabilistic models." Proceedings of the 34th Annual Conference on Neural Information Processing Systems. 2020.
Owner
Yang Song
PhD Candidate in Stanford AI Lab
Yang Song
performing moving objects segmentation using image processing techniques with opencv and numpy

Moving Objects Segmentation On this project I tried to perform moving objects segmentation using background subtraction technique. the introduced meth

Mohamed Magdy 15 Dec 12, 2022
[Preprint] "Chasing Sparsity in Vision Transformers: An End-to-End Exploration" by Tianlong Chen, Yu Cheng, Zhe Gan, Lu Yuan, Lei Zhang, Zhangyang Wang

Chasing Sparsity in Vision Transformers: An End-to-End Exploration Codes for [Preprint] Chasing Sparsity in Vision Transformers: An End-to-End Explora

VITA 64 Dec 08, 2022
The repository for freeCodeCamp's YouTube course, Algorithmic Trading in Python

Algorithmic Trading in Python This repository Course Outline Section 1: Algorithmic Trading Fundamentals What is Algorithmic Trading? The Differences

Nick McCullum 1.8k Jan 02, 2023
Validated, scalable, community developed variant calling, RNA-seq and small RNA analysis

Validated, scalable, community developed variant calling, RNA-seq and small RNA analysis. You write a high level configuration file specifying your in

Blue Collar Bioinformatics 917 Jan 03, 2023
TRIQ implementation

TRIQ Implementation TF-Keras implementation of TRIQ as described in Transformer for Image Quality Assessment. Installation Clone this repository. Inst

Junyong You 115 Dec 30, 2022
Boosting Monocular Depth Estimation Models to High-Resolution via Content-Adaptive Multi-Resolution Merging

Boosting Monocular Depth Estimation Models to High-Resolution via Content-Adaptive Multi-Resolution Merging This repository contains an implementation

Computational Photography Lab @ SFU 1.1k Jan 02, 2023
QuALITY: Question Answering with Long Input Texts, Yes!

QuALITY: Question Answering with Long Input Texts, Yes! Authors: Richard Yuanzhe Pang,* Alicia Parrish,* Nitish Joshi,* Nikita Nangia, Jason Phang, An

ML² AT CILVR 61 Jan 02, 2023
Res2Net for Instance segmentation and Object detection using MaskRCNN

Res2Net for Instance segmentation and Object detection using MaskRCNN Since the MaskRCNN-benchmark of facebook is deprecated, we suggest to use our mm

Res2Net Applications 55 Oct 30, 2022
code for "Self-supervised edge features for improved Graph Neural Network training",

Self-supervised edge features for improved Graph Neural Network training Data availability: Here is a link to the raw data for the organoids dataset.

Neal Ravindra 23 Dec 02, 2022
Equivariant layers for RC-complement symmetry in DNA sequence data

Equi-RC Equivariant layers for RC-complement symmetry in DNA sequence data This is a repository that implements the layers as described in "Reverse-Co

7 May 19, 2022
Dense matching library based on PyTorch

Dense Matching A general dense matching library based on PyTorch. For any questions, issues or recommendations, please contact Prune at

Prune Truong 399 Dec 28, 2022
Pytorch implementation for reproducing StackGAN_v2 results in the paper StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks

StackGAN-v2 StackGAN-v1: Tensorflow implementation StackGAN-v1: Pytorch implementation Inception score evaluation Pytorch implementation for reproduci

Han Zhang 809 Dec 16, 2022
N-RPG - Novel role playing game da turfu

N-RPG Ce README sera la page de garde du projet. Contenu Il contiendra la présen

4 Mar 15, 2022
Learning Visual Words for Weakly-Supervised Semantic Segmentation

[IJCAI 2021] Learning Visual Words for Weakly-Supervised Semantic Segmentation Implementation of IJCAI 2021 paper Learning Visual Words for Weakly-Sup

Lixiang Ru 24 Oct 05, 2022
Code and data for "TURL: Table Understanding through Representation Learning"

TURL This Repo contains code and data for "TURL: Table Understanding through Representation Learning". Environment and Setup Data Pretraining Finetuni

SunLab-OSU 63 Nov 23, 2022
Code for "Steerable Pyramid Transform Enables Robust Left Ventricle Quantification"

Code for "Steerable Pyramid Transform Enables Robust Left Ventricle Quantification" This is an end-to-end framework for accurate and robust left ventr

2 Jul 09, 2022
codes for Image Inpainting with External-internal Learning and Monochromic Bottleneck

Image Inpainting with External-internal Learning and Monochromic Bottleneck This repository is for the CVPR 2021 paper: 'Image Inpainting with Externa

97 Nov 29, 2022
Simple cross-platform application for DaVinci surgical video frame annotation

About DaVid is a simple cross-platform GUI for annotating robotic and endoscopic surgical actions for use in deep-learning research. Features Simple a

Cyril Zakka 4 Oct 09, 2021
Code for SentiBERT: A Transferable Transformer-Based Architecture for Compositional Sentiment Semantics (ACL'2020).

SentiBERT Code for SentiBERT: A Transferable Transformer-Based Architecture for Compositional Sentiment Semantics (ACL'2020). https://arxiv.org/abs/20

Da Yin 66 Aug 13, 2022
Fuzzing tool (TFuzz): a fuzzing tool based on program transformation

T-Fuzz T-Fuzz consists of 2 components: Fuzzing tool (TFuzz): a fuzzing tool based on program transformation Crash Analyzer (CrashAnalyzer): a tool th

HexHive 244 Nov 09, 2022