ShuffleNetV1

Running on AiQu

After starting the job, clone the repo to the root (same directory where project and data directories exist) directory. Once the repo is cloned, run the following commands

python3 divide_boat_no_boat.py --src-dir ../data/train/images --csv ../data/train_ship_segmentations_v2.csv --dest-dir ./data

python3 train.py --save-dir ../project/shuffleNet-results --img-height 768 --img-width 768 --num-classes 2 --groups 3 --batch-size 100 --epochs 10 --data-dir ../data/classifier_data

python3 train_alt.py --save-dir ../project/base_model_results --img-height 768 --img-width 768 --num-classes 2 --batch-size 100 --epochs 1