evolve with mutil-gpu

[alicera]

Search before asking

• I have searched the YOLOv5 issues and discussions and found no similar questions.

Question

bash run.sh

#!/bin/bash
# Multi-GPU
for i in 0 1 2 3; do
  nohup python train.py --epochs 300 --data data/4cls.yaml --weights d11b_large-object-bright_20220207_best.pt --evolve --device $i > evolve_gpu_$i.log &
done

Additional

No response

[glenn-jocher]

@alicera what's the question?

[alicera]

Does it success to run for see the message?

[alicera]

Command

python train.py --epochs 300 --data coco128.yaml --weights yolov5s.pt --cache --evolve

every generation as runs/evolve/hyp_evolved.yam
When is the hyp_evolved.yam generated? Until 300 epoch ?

[glenn-jocher]

@alicera hyp_evolved.yaml is updated every generation with the latest results.

[alicera]

I already run 70 epoch and the folder runs/evolve/exp/ only has the results.csv.
So my training have some problem?

[glenn-jocher]

@alicera that's not correct. Evolution is very easy, you can start just by attaching the --evolve argument to the Colab notebook train cell. As you can see there will be 3 files in your runs/evolve/exp directory which are updated every generation:
https://colab.research.google.com/github/ultralytics/yolov5/blob/master/tutorial.ipynb

[alicera]

you can see there will be 3 files in your runs/evolve/exp directory which are updated every generation:
The result is used colab to generate?

If I use the command python train.py --epochs 300 --data coco128.yaml --weights yolov5s.pt --evolve on the linux terminal, it is also work ?
Because I use it on the terminal, it still no generate evolve.csv and hyp_evolve.yaml. But there is result.csv.

[glenn-jocher]

@alicera yes example is shown in Colab. Evolution works in every environment, just like any normal YOLOv5 training.

[zaipinai]

@alicera yes example is shown in Colab. Evolution works in every environment, just like any normal YOLOv5 training.

I have a same problem with alicera. After training 150 epochs(I set it max to 150), it created hyp.yaml and hyp_evolved.yaml. It seems that the two are reversed, the hyp.yaml is more likely to be the result.
1.hyp.yaml 2.hyp_evolved.yaml 3.my original hyp

[glenn-jocher]

@zaipinai good news 😃! Your original issue may now be fixed ✅ in a previous PR #6604. Current code no longer outputs hyp.yaml and applies an indexing fix that caused incorrect final results to display:

To receive this update:

• Git – git pull from within your yolov5/ directory or git clone https://github.com/ultralytics/yolov5 again
• PyTorch Hub – Force-reload model = torch.hub.load('ultralytics/yolov5', 'yolov5s', force_reload=True)
• Notebooks – View updated notebooks
• Docker – sudo docker pull ultralytics/yolov5:latest to update your image

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 🚀!

[zaipinai]

@zaipinai good news 😃! Your original issue may now be fixed ✅ in a previous PR #6604. Current code no longer outputs hyp.yaml and applies an indexing fix that caused incorrect final results to display:

To receive this update:

• Git – git pull from within your yolov5/ directory or git clone https://github.com/ultralytics/yolov5 again
• PyTorch Hub – Force-reload model = torch.hub.load('ultralytics/yolov5', 'yolov5s', force_reload=True)
• Notebooks – View updated notebooks
• Docker – sudo docker pull ultralytics/yolov5:latest to update your image

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 🚀!

Thanks for your help! I use modify YOLO v5 with transformer prediction head and it doesn't update. I will update it later. Sorry for bothering you for this mistake.

[github-actions]

👋 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 ⭐!