Tflite
My predicted class/class when I test is 221529 in output data but my class is only 4.
I wonder where did that high number come from?
Input shape: [ 1 640 640 3] Output shape: [ 1 25200 9]
so when I deploy it in my android text it generate 178k, 203564, etc.
I tried to use detect my tflite model and it works fine, they can detect the class by adding my customdata.yaml
!cd yolov5 & python export.py --weights runs/train/exp/weights/best.pt data=yolov5/data/customdata.yaml --include saved_model pb
But when I tried to run in variable model my "cardboard" turns to "class 1"
