Retrain an existing .pt network in yolov5 with new images to get a new network not works well #10885
altiplanet
started this conversation in
General
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
We have two zip files with photos of plants and undergrowth, each 800 photos, we want to recognice both plants. We train first over one zip file and the train is successfully finished but when try to retrain the best.pt of first zip file with the second zip file the resultant best.pt is poor as soon as quality of recognition, why ?
The command in Google Colab to retrain is the following:
!python train.py --hyp hyp.scratch-low.yaml --img 368 --batch 16 --epochs 150 --data /content/datasets/lechugaslampax.yaml --weights best.pt --cache
Very Thanks
Beta Was this translation helpful? Give feedback.
All reactions