How to properly train the Yolo network?

Question

Neural networks

Hello, how to properly train the Yolo network?
I have the following in my network config file
[net]
# Testing
# batch=1
# subdivisions=1
# Training
batch=64
subdivisions=16
width=608
height=608
channels=3
momentum=0.9
decay=0.0005
angle=0
saturation = 1.5
exposure = 1.5
hue=.1
Image size for training should be 608x608 as per config file, network will resize to 608x608 pix?
Or will it simply take objects from the image from the train.txt file with coordinates, and the size of the image is not important?

Answer 1

The size of images for training is not tied to the size specified in the configuration file. The main rule is training dataset image sizes = test image sizes.
Yolo automatically resizes images to different resolutions during training.
I recommend using YOLO v3 in 2019.
Chic - guide