Got very high recall but low precision in my own two-classes dataset

[lantudou]

Hi, I tried to train the original yolov4 model on my own datasets. Why I will got a model with very high recall but very low precision? , It seems like my model can locate the object but cannot classify the objects correctly.

I can promise the dataset is correct because I have tested it using darknet.

By the way, I observed that the cls loss is very low at the beginning of my training, it is only 0.0016 in the first epoch. Anyone can tell me it is normal?

[lantudou]

For testing the code, I use the only one dataset as the training dataset and the valid dataset