search

computervisioneng / train-yolov8-custom-dataset-step-by-step-guide

GNU Affero General Public License v3.0
257 stars 112 forks source link

Yolov8 training custom dataset #5

Open reemalsubaiee opened 1 year ago

reemalsubaiee commented 1 year ago

Good day Mr. Felipe, i hope all is well.

i really enjoyed your vidoe and it was very internsting and straight to the point especially as i beginner into computer vision myself.

i justed wanted to ask you, during the training procces i had a problem when no images is showing. i have create the foleders with the same name where images > train > contains the images. while labels > train > containts the labels . txt from CVAT.

i have a bunch of photos i have collect from the web and when applying the code this shows up:

YOLOv8n summary: 225 layers, 3011043 parameters, 3011027 gradients

TensorBoard: Start with 'tensorboard --logdir runs/detect/train', view at http://localhost:6006/ train: Scanning /content/drive/MyDrive/Object detection/Images/train.cache... 0 images, 20 backgrounds, 0 corrupt: 100%| 20/20 [00:00<?, ?it/s]

FileNotFoundError: train: No labels found in /content/drive/MyDrive/Object detection/Images/train.cache, can not start training.

gadgetmusician commented 1 year ago

I resolved this by personalising the file pathway to the minimum. Keep going because once it is running its so good to use; coming off the back of a fully custom architecture on my last project.