In this repository, we have implemented our simplifed network of the YOLOv3 architecture based on Ultralytics' implementation, which can be found in https://github.com/ultralytics/yolov3.

In order to run the experiment, we simply run YOLOv3-experiment.ipynb where we can also make changes to the type of data we are using and how many samples and batches we want to run for how many epochs.

However, we also have to create a ./models/ folder and a ./repData/ folder before training in order to store our results.

To make changes to the hyperparameters, we can change the .yaml files to change the hyperparameters.