Fixes for YOLOv4 tiny training

[stellasphere]

Description

Fixes for the YOLOv4 training notebook, addressing issue #28

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• This change requires a documentation update

How has this change been tested, please provide a testcase or example of how you tested the change?

Tested on the BCCD dataset with Colab Free

Any specific deployment considerations

For example, documentation changes, usability, usage/costs, secrets, etc.

Docs

• Docs updated? What were the changes:

[NihadBadalov]

Is this going to get merged? #28 is open for already two years

[capjamesg]

This notebook is almost ready to go.

Can you replace the contents of the last cell with:

#/test has images that we can test our detector on
test_images = [f for f in os.listdir('./BCCD-4/test/') if f.endswith('.jpg')]
import random
img_path = "./BCCD-4/test/" + random.choice(test_images);

#test out our detector!
!./darknet detect /content/darknet/cfg/custom-yolov4-tiny-detector.cfg /content/darknet/backup/custom-yolov4-tiny-detector_last.weights {img_path} -dont-show
imShow('/content/darknet/predictions.jpg')

As is, the model weights, configuration, and image do not load properly.