search

THU-MIG / yolov10

YOLOv10: Real-Time End-to-End Object Detection
https://arxiv.org/abs/2405.14458
GNU Affero General Public License v3.0
8.01k stars 662 forks source link

learning speed #210

Open kunsungwoo opened 3 weeks ago

kunsungwoo commented 3 weeks ago

I used the yolov8n model and it took 30 minutes per epoch. However, yolov10n takes 90 minutes per epoch. Why is the difference so big?

leonnil commented 3 weeks ago

Thank you for your interest! Does the issue #84 help?

kunsungwoo commented 3 weeks ago

It seems to be fine after running pip install ultralytics.

2 This is yolov8n's learning time of about 35 minutes.

3

This is about 50 minutes of study time by yolov10n.

The learning speed of yolov8n is about 30-40% faster than yolov10n. Is this normal? All other settings for training are the same.

leonnil commented 3 weeks ago

Thanks for your interest! Could you please provide more details about your setup? In our local environment with 8*3090, each epoch of training YOLOv10-N costs ~1min54s, which is similar to that of YOLOv8-N (~1min54s).