Text: BucketIterator goes into infinite loop

[DeepInEvil]

When I use the bucketIterator with a tabulardataset object, it is going into infinite loop in training.
Here is the example of the use:
train = data.TabularDataset(
path= review_dir + d_review_dir + 'train.tsv', format='tsv', skip_header=True,
fields=[('text', TEXT),
('labels', LABELS)])

train_iter, dev_iter, test_iter = data.Iterator.splits(
                            (train, val, test), device = -1,
                            batch_sizes=(args.batch_size, len(val), len(test)))

When I loop through the train_iter object with: for batch in train_iter:
it goes into an infinite loop, please fix the error.

[keitakurita]

If you do not want the iterator to go into an infinite loop, you need to pass the keyword argument repeat=False like

train_iter, dev_iter, test_iter = data.Iterator.splits(
                            (train, val, test), device = -1,
                            repeat=False,
                            batch_sizes=(args.batch_size, len(val), len(test)))