Processed Fitzpatrick-17k train.csv and val.csv not found

[AnshulPundhir]

Dear Authors,
Thank you for providing the source code but I am not able to find script to do required pre-processing on the datasets.

# paths
ann_train = pd.read_csv('./dataset/Fitzpatrick-17k/processed/rand_seed={}/split/train.csv'.format(args.rand_seed), index_col=0)
ann_valid = pd.read_csv('./dataset/Fitzpatrick-17k/processed/rand_seed={}/split/val.csv'.format(args.rand_seed), index_col=0)
ann_train.reset_index(inplace=True)
ann_valid.reset_index(inplace=True)

train_pkl = './dataset/Fitzpatrick-17k/processed/rand_seed={}/pkls/train_images.pkl'.format(args.rand_seed)
valid_pkl = './dataset/Fitzpatrick-17k/processed/rand_seed={}/pkls/val_images.pkl'.format(args.rand_seed)

In above code (inside main.py)..... how to obtain csv files since official dataset does not contain these splits.
If there is any missing file, please share.

Thank you

[XuZikang]

The train.csv, val.csv and test.csv are generated by randomly splitting the whole dataset with a ratio of 8:1:1.

You can refer to the code here: MEDFAIR