andrewhsin / yolo-nas-pytorch Goto Github PK

💡 Your Question

I found this class in the original repository “YoloDarknetFormatDetectionDataset”，
I found that such a format is supported in its examples1,
data_dir
└── train/test/val
├── images
│ ├─ 0001.jpg
│ ├─ 0002.jpg
│ └─ ...
└── labels
├─ 0001.txt
├─ 0002.txt
└─ ...
but building a trainer does not require both train and val，
Can you teach me the correct way to use it to train this type of custom dataset?

Versions

No response

📚 Documentation Improvement

When I tried to access the segmentation on a custom dataset drive link, I got an access denied page. This was true of most of the links. Thanks!