Easy training on custom dataset. Various backends (MobileNet and SqueezeNet) supported. A YOLO demo to detect raccoon run entirely in brower is accessible at https://git.io/vF7vI (not on Windows).
Hi, firstly thanks for your project. It is very helpful for me to learning YOLO.
I have a problem when i use your Blood_detection code, after 4-5 epoch, it will shows that "val_loss is not improved by 0.02185". And it let the file early stop.
rnirdhar
commented
5 years ago
Hi, firstly thanks for your project. It is very helpful for me to learning YOLO. I have a problem when i use your Blood_detection code, after 4-5 epoch, it will shows that "val_loss is not improved by 0.02185". And it let the file early stop.