Datasets:

Teklia

/

NorHand-v2-line

like 0

Follow

Teklia 25

The number of samples in each split is incorrect in the README. This results in the following error when loading the dataset:

datasets.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=0, num_examples=145008, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='train', num_bytes=3567896931, num_examples=145009, shard_lengths=[30000, 30000, 30000, 20000, 30000, 5009], dataset_name='nor_hand-v2-line')}, {'expected': SplitInfo(name='validation', num_bytes=0, num_examples=14965, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='validation', num_bytes=364518531, num_examples=14966, shard_lengths=None, dataset_name='nor_hand-v2-line')}, {'expected': SplitInfo(name='test', num_bytes=0, num_examples=1792, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='test', num_bytes=31054952, num_examples=1562, shard_lengths=None, dataset_name='nor_hand-v2-line')}]

The large difference in the test set is expected, as the split was changed to match the NorHand v3 test set when it was released.