When I tried to use the training example of Patch-based pipelines in the help document, the following error occurred

🐛Bug

To reproduce

import torch
import torchio as tio
from torch.utils.data import DataLoader

patch_size = 96
queue_length = 300
samples_per_volume = 10
sampler = tio.data.UniformSampler(patch_size)
subject = tio.datasets.Colin27()
subjects_dataset = tio.SubjectsDataset(10 * [subject])
patches_queue = tio.Queue(
    subjects_dataset,
    queue_length,
    samples_per_volume,
    sampler,
    num_workers=0,
)
patches_loader = DataLoader(patches_queue, batch_size=16)
num_epochs = 2
model = torch.nn.Identity()
for epoch_index in range(num_epochs):
    for patches_batch in patches_loader:
        inputs = patches_batch['t1'][tio.DATA]  # key 't1' is in subject
        targets = patches_batch['brain'][tio.DATA]  # key 'brain' is in subject
        logits = model(inputs)  # model being an instance of torch.nn.Module

Expected behavior

It should be an iterable object

Actual behavior

TypeError: randint(): argument 'high' (position 1) must be int, not numpy.int32

System info

Output of

TypeError: randint(): argument 'high' (position 1) must be int, not numpy.int32

[rayryeng]

This occurs because you have an older version of Numpy. Please update to the latest version. Also make sure you update PyTorch to at least version 1.6.0

[fepegar]

This occurs because you have an older version of Numpy. Please update to the latest version. Also make sure you update PyTorch to at least version 1.6.0

Thanks, @rayryeng. @Key-lei, feel free to reopen if that solution does not help.