Skip to content
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

Default AutoBatch 0.8 fraction #9212

Merged
merged 1 commit into from
Aug 29, 2022
Merged

Default AutoBatch 0.8 fraction #9212

merged 1 commit into from
Aug 29, 2022

Conversation

glenn-jocher
Copy link
Member

@glenn-jocher glenn-jocher commented Aug 29, 2022

Should be more stable, ran into CUDA OOM on validation when train utilization is too high.

πŸ› οΈ PR Summary

Made with ❀️ by Ultralytics Actions

🌟 Summary

Enhanced compatibility and optimization of batch size for memory usage in YOLOv5 models.

πŸ“Š Key Changes

  • Adjusted warning for YOLOv5 v6.2 ClassificationModel to clarify its non-compatibility with AutoShape.
  • Modified the default memory fraction used in the autobatch function from 0.9 to 0.8.

🎯 Purpose & Impact

  • The update provides clearer guidance to users attempting to use AutoShape with the ClassificationModel, preventing confusion and potential errors. πŸ›‘
  • Reducing the default memory fraction allows for larger models or datasets by lowering the risk of running out of CUDA memory, leading to fewer crashes and enhanced performance on systems with limited resources. πŸ“ˆπŸ–₯️

@glenn-jocher glenn-jocher self-assigned this Aug 29, 2022
@glenn-jocher glenn-jocher merged commit 91a81d4 into master Aug 29, 2022
@glenn-jocher glenn-jocher deleted the updates branch August 29, 2022 18:41
ctjanuhowski pushed a commit to ctjanuhowski/yolov5 that referenced this pull request Sep 8, 2022
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.

1 participant