def get_dataset()

in summarize_from_feedback/datasets/__init__.py [0:0]

• 12 lines of code
• 4 McCabe index (conditional complexity)

def get_dataset(name, split, layout, repeat=True, seed=None):
    if seed is None:
        seed = torch.initial_seed()

    data = list(_DATASETS[name](split))

    def shuffled():
        my_random = random.Random(seed)
        while True:
            my_random.shuffle(data)
            yield from data
            if not repeat:
                return

    return even_more_itertools.distribute(shuffled(), layout=layout)