I am using YOLOv8 for segmentation on google colab ( train-yolov8-instance-segmentation-on-custom-dataset.ipynb - Colaboratory (google.com).
I have trained the model for 100 epochs and am able to test it on my own dataset however, the precision is high and the recall is low. I would like to change these parameters so that while testing even if the precision is low, the trained model segments most of the objects.
Any ideas on how I can tweak these parameters would be very helpful!