Navigate to the prepare_data directory and follow the instructions in the README.md file to prepare the data.

Install the training environment according to requirements.txt. Note that this may differ from prepare_data/requirements.txt, so please ensure you install the correct dependencies.

See run.sh

for seed in 42 43 44 45 46; do

  python main.py \
    --save_dir exp/resnet34+oversampling+use_mixup+seed_${seed} \
    --oversampling \
    --use_mixup \
    --model resnet34 \
    --seed $seed 
    
done

This will generate a scores.txt file in the exp/resnet34+oversampling+use_mixup+seed_${seed} directory, containing prediction scores for both validation and test sets.