yolov5 icon indicating copy to clipboard operation
yolov5 copied to clipboard

Support for exporting Ghost models in tensorflow saved model

Open hardikdava opened this issue 3 years ago • 1 comments
trafficstars

Search before asking

  • [X] I have searched the YOLOv5 issues and found no similar feature requests.

Description

Currently, yolov5 models with ghost feature do not support export in Tensorflow models. It would be great to have this feature to run model on edgetpu

Use case

The architecture looks promising and might be able to run in realtime on Edgetpu.

Additional

No response

Are you willing to submit a PR?

  • [X] Yes I'd like to help by submitting a PR!

hardikdava avatar Sep 19 '22 09:09 hardikdava

@hardikdava unfortunately all TF exports are built from scratch currently using modules in models/tf.py. You'd have to add the ghost modules in TF format for export to work correctly.

@AyushExel example of our TF export issue.

glenn-jocher avatar Sep 19 '22 10:09 glenn-jocher

@glenn-jocher @AyushExel Is this feature available in near future? Is there any branch or PR which is designated for this?

hardikdava avatar Sep 27 '22 14:09 hardikdava

@hardikdava hi, thank you for your feature suggestion on how to improve YOLOv5 🚀!

There are no plans for this, but you're free to help with a PR.

The fastest and easiest way to incorporate your ideas into the official codebase is to submit a Pull Request (PR) implementing your idea, and if applicable providing before and after profiling/inference/training results to help us understand the improvement your feature provides. This allows us to directly see the changes in the code and to understand how they affect workflows and performance.

Please see our ✅ Contributing Guide to get started.

glenn-jocher avatar Sep 27 '22 16:09 glenn-jocher

@AyushExel @glenn-jocher I was able to make layers of Ghost model in tensorflow keras. But I got this error

Exception encountered when calling layer "tf_conv_37" (type TFConv). It has 131072 elements, but gotshape (1, 1, 32, 256) with 8192 elements).

Then I looked into the model architecture and I observed that the first layer of the model has different kernel and stride values. How can I solve this error?

Screenshot from 2022-10-14 08-47-17

hardikdava avatar Oct 14 '22 06:10 hardikdava

Existing TF models export correctly including layer you are showing.

glenn-jocher avatar Oct 14 '22 10:10 glenn-jocher

👋 Hello, this issue has been automatically marked as stale because it has not had recent activity. Please note it will be closed if no further activity occurs.

Access additional YOLOv5 🚀 resources:

  • Wiki – https://github.com/ultralytics/yolov5/wiki
  • Tutorials – https://docs.ultralytics.com/yolov5
  • Docs – https://docs.ultralytics.com

Access additional Ultralytics ⚡ resources:

  • Ultralytics HUB – https://ultralytics.com/hub
  • Vision API – https://ultralytics.com/yolov5
  • About Us – https://ultralytics.com/about
  • Join Our Team – https://ultralytics.com/work
  • Contact Us – https://ultralytics.com/contact

Feel free to inform us of any other issues you discover or feature requests that come to mind in the future. Pull Requests (PRs) are also always welcomed!

Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐!

github-actions[bot] avatar Nov 14 '22 00:11 github-actions[bot]