Hello i’m doing transfer learning on yolov8n-cls, with this script:
from ultralytics import YOLO
model = YOLO('yolov8n-cls.pt')
print(model.names)
results = model.train(
data='./dataset',
epochs=25,
imgsz=224,
device='cpu',
project='progetto_gatti',
name='run_',
patience=15,
freeze="9"
)
and this dataset:
dataset/
├── test/
│ ├── gatto/
│ └── non_gatto/
├── train/
│ ├── gatto/
│ └── non_gatto/
└── val/
├── gatto/
└── non_gatto/
and the testing of the model:
from ultralytics import YOLO
best_model = YOLO('./runs/classify/progetto_gatti/run_4/weights/best.pt')
metrics = best_model.val(data='./dataset', split='val', batch=1)
print(best_model.names)
however the confusion matrix looks like this:
and i can’t get why there is a background class even if the print(best_model.names) prints out:
{0: ‘gatto’, 1: ‘non_gatto’}