wheat_detection

This is my repository with a baseline model for Wheat Detection challenge on Kaggle

Main frameworks used:

• hydra
• pytorch-lightning

To use it for training, perform the following steps:

• download the data, unzip in and put in some folder;
• define that folder in config conf/data/data.yaml as a value of the key data.folder_path
• run run_hydra.py script

There is no script for prediction, because in this competition we have to make prediction in kernels.

Refer to my kernel for more information: https://www.kaggle.com/artgor/object-detection-with-pytorch-lightning