guided-diffusion

This is the codebase for Diffusion Models Beat GANS on Image Synthesis.

This repository is based on openai/improved-diffusion, with modifications for classifier conditioning and architecture improvements.

Usage

Training diffusion models is described in the parent repository. Training a classifier is similar. We assume you have put training hyperparameters into a TRAIN_FLAGS variable, and classifier hyperparameters into a CLASSIFIER_FLAGS variable. Then you can run:

mpiexec -n N python scripts/classifier_train.py --data_dir path/to/imagenet $TRAIN_FLAGS $CLASSIFIER_FLAGS

Make sure to divide the batch size in TRAIN_FLAGS by the number of MPI processes you are using.

Here are flags for training the 128x128 classifier. You can modify these for training classifiers at other resolutions:

TRAIN_FLAGS="--iterations 300000 --anneal_lr True --batch_size 256 --lr 3e-4 --save_interval 10000 --weight_decay 0.05"
CLASSIFIER_FLAGS="--image_size 128 --classifier_attention_resolutions 32,16,8 --classifier_depth 2 --classifier_width 128 --classifier_pool attention --classifier_resblock_updown True --classifier_use_scale_shift_norm True"

For sampling from a 128x128 classifier-guided model, 25 step DDIM:

MODEL_FLAGS="--attention_resolutions 32,16,8 --class_cond True --image_size 128 --learn_sigma True --num_channels 256 --num_heads 4 --num_res_blocks 2 --resblock_updown True --use_fp16 True --use_scale_shift_norm True"
CLASSIFIER_FLAGS="--image_size 128 --classifier_attention_resolutions 32,16,8 --classifier_depth 2 --classifier_width 128 --classifier_pool attention --classifier_resblock_updown True --classifier_use_scale_shift_norm True --classifier_scale 1.0 --classifier_use_fp16 True"
SAMPLE_FLAGS="--batch_size 4 --num_samples 50000 --timestep_respacing ddim25 --use_ddim True"
mpiexec -n N python scripts/classifier_sample.py \
    --model_path /path/to/model.pt \
    --classifier_path path/to/classifier.pt \
    $MODEL_FLAGS $CLASSIFIER_FLAGS $SAMPLE_FLAGS

To sample for 250 timesteps without DDIM, replace --timestep_respacing ddim25 to --timestep_respacing 250, and replace --use_ddim True with --use_ddim False.

This is the codebase for Diffusion Models Beat GANS on Image Synthesis.

Related tags

Overview

guided-diffusion

Usage

Owner

OpenAI

P-Tuning v2: Prompt Tuning Can Be Comparable to Finetuning Universally Across Scales and Tasks

Clinica is a software platform for clinical research studies involving patients with neurological and psychiatric diseases and the acquisition of multimodal data

A basic duplicate image detection service using perceptual image hash functions and nearest neighbor search, implemented using faiss, fastapi, and imagehash

Exploring Versatile Prior for Human Motion via Motion Frequency Guidance (3DV2021)

Auxiliary data to the CHIIR paper Searching to Learn with Instructional Scaffolding

Using PyTorch Perform intent classification using three different models to see which one is better for this task

Implementation of QuickDraw - an online game developed by Google, combined with AirGesture - a simple gesture recognition application

Symbolic Parallel Adaptive Importance Sampling for Probabilistic Program Analysis in JAX

Beyond Image to Depth: Improving Depth Prediction using Echoes (CVPR 2021)

DECAF: Deep Extreme Classification with Label Features

Model of an AI powered sign language interpreter.

GPU-Accelerated Deep Learning Library in Python

Accelerated deep learning R&D

How Effective is Incongruity? Implications for Code-mix Sarcasm Detection.

Code repo for "Towards Interpretable Deep Networks for Monocular Depth Estimation" paper.

FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation.

DropNAS: Grouped Operation Dropout for Differentiable Architecture Search

Multi-Task Learning as a Bargaining Game

[CVPR2021] DoDNet: Learning to segment multi-organ and tumors from multiple partially labeled datasets

Efficient electromagnetic solver based on rigorous coupled-wave analysis for 3D and 2D multi-layered structures with in-plane periodicity