realesrgan 0.3.0

Real-ESRGAN aims at developing Practical Algorithms for General Image Restoration

Stars: 26171, Watchers: 26171, Forks: 3294, Open Issues: 501

The xinntao/Real-ESRGAN repo was created 2 years ago and the last code push was 3 weeks ago.
The project is extremely popular with a mindblowing 26171 github stars!

How to Install realesrgan

You can install realesrgan using pip

pip install realesrgan

or add it to a project with poetry

poetry add realesrgan

Package Details

Author

Xintao Wang

License

BSD-3-Clause License

Homepage

https://github.com/xinntao/Real-ESRGAN

PyPi:

https://pypi.org/project/realesrgan/

GitHub Repo:

https://github.com/xinntao/Real-ESRGAN

Classifiers

Related Packages

Errors

A list of common realesrgan errors.

Code Examples

Here are some realesrgan code examples and snippets.

GitHub Issues

The realesrgan package has 501 open issues on GitHub

• how to use well-trained net_d to inference image
• training the model with different file formats
• Finetune on Multi-gpu not working
• When I try to train the RealESR-GAN, the initial result is not realistic compared to RealESRNet
• consider get the models from hugging face
• Failed training from scratch when training images are more than 44
• Does discrete multi-gpu work with video inference?
• Use paired training data
• Executable file for inference_realesrgan_video run so slow on Linux (Ubuntu) machine
• How do i use inference_realesrgan_video?
• Batched inference?
• Error when adding -t for tiling
• ValueError: ('Scale mismatches. GT (3078, 5472) is not 4x ', 'multiplication of LQ (1080, 1624).')
• UnboundLocalError: local variable 'file_url' referenced before assignment
• fix(sec): upgrade torch to 1.13.1

See more issues on GitHub