Explanation of pretrained weights

[astrobdr]

I would like some more information on the yolov3.weights pretrained weights.

Were they learned from trianing on 100% of the 2017 coco training data? Or was some data held back for validation?

Did you use the same script as is found in the file to create a custom model with 80 classes specified?

And did you use the same training procedure as in train.py? Same learning rates, batch sizes, etc?

Thanks!