PyTorch-StudioGAN
stylegan2-pytorch
PyTorch-StudioGAN | stylegan2-pytorch | |
---|---|---|
9 | 3 | |
3,366 | 2,659 | |
-0.1% | - | |
6.1 | 0.0 | |
9 months ago | 6 months ago | |
Python | Python | |
GNU General Public License v3.0 or later | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
PyTorch-StudioGAN
-
[R] GigaGAN: A Large-scale Modified GAN Architecture for Text-to-Image Synthesis. Better FID Score than Stable Diffusion v1.5, DALL·E 2, and Parti-750M. Generates 512px outputs at 0.13s. Native Prompt mixing, Prompt Interpolation and Style Mixing. A GigaGAN Upscaler is also introduced (Up to 4K)
Given the first author I'd expect it to land in StudioGAN sometime in the future. Training it from scratch will definitely be costly though.
-
[P] Implementations of 30 representative GANs and Comprehensive Benchmark for GAN, AR, and Diffusion Models (link in comments).
Github Link: https://github.com/POSTECH-CVLab/PyTorch-StudioGAN
Github Link: https://github.com/POSTECH-CVLab/PyTorch-StudioGAN Paper Link: https://arxiv.org/abs/2206.09479 I would like to introduce PyTorch-StudioGAN library, which I have been maintaining for the past two years. StudioGAN is a PyTorch library providing implementations of representative Generative Adversarial Networks (GANs) for conditional/unconditional image generation. StudioGAN aims to offer an identical playground for modern GANs so that machine learning researchers can readily compare and analyze a new idea. Moreover, StudioGAN provides an unprecedented-scale benchmark for generative models. The benchmark includes results from GANs (BigGAN-Deep, StyleGAN-XL), auto-regressive models (MaskGIT, RQ-Transformer), and Diffusion models (LSGM++, CLD-SGM, ADM-G-U). [Features] * Coverage: StudioGAN is a self-contained library that provides 7 GAN architectures, 9 conditioning methods, 4 adversarial losses, 13 regularization modules, 6 augmentation modules, 8 evaluation metrics, and 5 evaluation backbones. Among these configurations, we formulate 30 GANs as representatives. * Flexibility: Each modularized option is managed through a configuration system that works through a YAML file, so users can train a large combination of GANs by mix-matching distinct options. * Reproducibility: With StudioGAN, users can compare and debug various GANs with the unified computing environment without concerning about hidden details and tricks. * Plentifulness: StudioGAN provides a large collection of pre-trained GAN models, training logs, and evaluation results. * Versatility: StudioGAN supports 5 types of acceleration methods with synchronized batch normalization for training: a single GPU training, data-parallel training (DP), distributed data-parallel training (DDP), multi-node distributed data-parallel training (MDDP), and mixed-precision training.
- [P], [R] Implementations of 30 Representative GANs and Comprehensive Benchmark for GAN, AR, and Diffusion Models (link in comments).
- [P] Implementations of 37 GAN-related papers using PyTorch including BigGAN and StyleGAN2-ADA (link in comment)
- [P] 40 Implementations of GAN-related papers including BigGAN and StyleGAN2 in a unified training pipeline
-
[R] Rebooting ACGAN: A new GAN that achieves SOTA results and harmonizes with various architectures, adversarial losses, and even differentiable augmentations (Neurips 2021).
Code for https://arxiv.org/abs/2111.01118 found: https://github.com/POSTECH-CVLab/PyTorch-StudioGAN
-
[N] LAMA AI's weekly news, updates, and events.
StudioGAN is introduced: A PyTorch library for SoTA GAN models
-
PyTorch GAN Library that provides implementations of 18+ SOTA GANs with pretrained_model, configs, logs, and checkpoints (link in comments)
Github: https://github.com/POSTECH-CVLab/PyTorch-StudioGAN
stylegan2-pytorch
- Converting a .pkl file to a .pt for StyleGan2 Model
-
I'm stumped with installing PyTorch.
Originally I wanted to run https://github.com/JCBrouwer/maua-stylegan2. I was trying to run the convert_weight.py but it resulted in shape mismatch errors in torch torch.Size([1, 512, 4, 4]) vs torch.Size([1]), so I tried the version here https://github.com/rosinality/stylegan2-pytorch/blob/master/convert_weight.py and the result was the same.
-
[D] Node Collapse with StyleGAN2 - ada
Used this implementation (https://github.com/rosinality/stylegan2-pytorch) to better understand the code. IMO it is written way more clearly than the official implementation. You should spend a lot of time reviewing the code until you understand what each line is doing. It isn't helpful that their aren't very many comments, but you should be able to recognize different architectures and calculations from the paper. Understanding the details is key. Don't let your eyes glaze over any part of it.
What are some alternatives?
awesome-colab-notebooks - Collection of google colaboratory notebooks for fast and easy experiments
PaddleGAN - PaddlePaddle GAN library, including lots of interesting applications like First-Order motion transfer, Wav2Lip, picture repair, image editing, photo2cartoon, image style transfer, GPEN, and so on.
BigGAN-PyTorch - The author's officially unofficial PyTorch BigGAN implementation.
stylegan-waifu-generator - Generate your waifu with styleGAN, stylegan老婆生成器
stylegan3-editing - Official Implementation of "Third Time's the Charm? Image and Video Editing with StyleGAN3" (AIM ECCVW 2022) https://arxiv.org/abs/2201.13433
anycost-gan - [CVPR 2021] Anycost GANs for Interactive Image Synthesis and Editing
pix2pixHD - Synthesizing and manipulating 2048x1024 images with conditional GANs
flaxmodels - Pretrained deep learning models for Jax/Flax: StyleGAN2, GPT2, VGG, ResNet, etc.
BasicSR - Open Source Image and Video Restoration Toolbox for Super-resolution, Denoise, Deblurring, etc. Currently, it includes EDSR, RCAN, SRResNet, SRGAN, ESRGAN, EDVR, BasicVSR, SwinIR, ECBSR, etc. Also support StyleGAN2, DFDNet.
maua-stylegan2 - This is the repo for my experiments with StyleGAN2. There are many like it, but this one is mine. Contains code for the paper Audio-reactive Latent Interpolations with StyleGAN.