Skip to content

Working Real-ESRGAN Demo as of 21/4/2025 – Fix BasicsR Compatibility#911

Open
DataTalesByAgos wants to merge 1 commit intoxinntao:masterfrom
DataTalesByAgos:master
Open

Working Real-ESRGAN Demo as of 21/4/2025 – Fix BasicsR Compatibility#911
DataTalesByAgos wants to merge 1 commit intoxinntao:masterfrom
DataTalesByAgos:master

Conversation

@DataTalesByAgos
Copy link
Copy Markdown

Problem fixed on colab: ModuleNotFoundError: No module named 'torchvision.transforms.functional_tensor'

This pull request updates the setup instructions for Real-ESRGAN to ensure compatibility as of 21/04/2025.

  • Uninstalls existing torch, torchvision, and basicsr versions.

  • Installs torch==1.13.0 and torchvision==0.14.0 compatible with CUDA 12.1.

  • Installs basicsr using a known working commit: 8d56e3a045f9fb3e1d8872f92ee4a4f07f886b0a from XPixelGroup/BasicSR.

✅ Working demo tested on Google Colab (https://colab.research.google.com/drive/1s14s55W5MIMIY051fwzCzkCIyjDkBEFx?usp=sharing)

This pull request updates the setup instructions for Real-ESRGAN to ensure compatibility as of 21/04/2025.

- Uninstalls existing torch, torchvision, and basicsr versions.

- Installs torch==1.13.0 and torchvision==0.14.0 compatible with CUDA 12.1.

- Installs basicsr using a known working commit: 8d56e3a045f9fb3e1d8872f92ee4a4f07f886b0a from XPixelGroup/BasicSR.

✅ Working demo tested on Google Colab
@DataTalesByAgos
Copy link
Copy Markdown
Author

Let me know if there's any feedback or if you'd prefer a different commit/version for basicsr. I’m happy to adjust it. Thanks for maintaining this great project!

@sparshgarg23
Copy link
Copy Markdown

seems the issue is still happening .On running the code on colab i am getting the error
No module named 'torchvision.transforms.functional_tensor

@DataTalesByAgos
Copy link
Copy Markdown
Author

seems the issue is still happening .On running the code on colab i am getting the error No module named 'torchvision.transforms.functional_tensor

I had the same issue before, but fixed it by pinning these versions:
!pip install torch==1.13.0 torchvision==0.14.0
-f https://download.pytorch.org/whl/cu121/torch_stable.html

That functional_tensor error usually comes from using an incompatible torchvision version. Make sure you're not using a newer one.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants