-
-
Notifications
You must be signed in to change notification settings - Fork 16.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
LazyInstanceNorm2d need torch>=1.10 #7381
Comments
👋 Hello @Beb007, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution. If this is a 🐛 Bug Report, please provide screenshots and minimum viable code to reproduce your issue, otherwise we can not help you. If this is a custom training ❓ Question, please provide as much information as possible, including dataset images, training logs, screenshots, and a public link to online W&B logging if available. For business inquiries or professional support requests please visit https://ultralytics.com or email support@ultralytics.com. RequirementsPython>=3.7.0 with all requirements.txt installed including PyTorch>=1.7. To get started: git clone https://github.com/ultralytics/yolov5 # clone
cd yolov5
pip install -r requirements.txt # install EnvironmentsYOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):
StatusIf this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training (train.py), validation (val.py), inference (detect.py) and export (export.py) on macOS, Windows, and Ubuntu every 24 hours and on every commit. |
Temporary, I use 'classifier' branch and it's working fine.
|
Yes, I have the same issue with training the custom model. |
Based on actual available layers. Torch 1.7 compatible, resolves #7381
* Dynamic normalization layer selection Based on actual available layers. Torch 1.7 compatible, resolves #7381 * Update train.py
@Beb007 @rohitrrg good news 😃! Your original issue may now be fixed ✅ in PR ##7392. This PR dynamically allocates a tuple of normalization layers based on the current torch version, so should be compatible with any torch release past or future. To receive this update:
Thank you for spotting this issue and informing us of the problem. Please let us know if this update resolves the issue for you, and feel free to inform us of any other issues you discover or feature requests that come to mind. Happy trainings with YOLOv5 🚀! |
* Dynamic normalization layer selection Based on actual available layers. Torch 1.7 compatible, resolves ultralytics#7381 * Update train.py
Hello,
Fantastic work !
I faced an issue with my latest run:
LazyInstanceNorm2d seems to be included from torch version 1.10, and not present in 1.9.1 usually used from Kaggle
I tried pip install torch==1.11 (latest) but got packages conflicts
With torch=1.10, no conficts, but faced another issue with protobuf.
My first though was to suggest a change in requirements, but, as theses packages conflicts can become really awkward... I leave it for now and just report it...
Best regards
PS: my bad, I should had put this in issue section
Originally posted by @Beb007 in #7380
The text was updated successfully, but these errors were encountered: