Anyone encountering this Cast error with NeighborLoader?

[agapiR]

Hi community,

I'm encountering an issue when running the NeighborLoader. Specifically, I am running the following code snippet copied from the pyg docs:

from torch_geometric.datasets import Planetoid
from torch_geometric.loader import NeighborLoader

data = Planetoid(path, name='Cora')[0]

loader = NeighborLoader(
    data,
    # Sample 30 neighbors for each node for 2 iterations
    num_neighbors=[30] * 2,
    # Use a batch size of 128 for sampling training nodes
    batch_size=128,
    input_nodes=data.train_mask,
)

sampled_data = next(iter(loader))

I get the following error:

Traceback (most recent call last):
  File "/app/test_neighbor_loader.py", line 54, in <module>
    main()
  File "/app/test_neighbor_loader.py", line 50, in main
    sampled_data = next(iter(loader))
  File "/home/user/micromamba/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 628, in __next__
    data = self._next_data()
  File "/home/user/micromamba/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 671, in _next_data
    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
  File "/home/user/micromamba/lib/python3.9/site-packages/torch/utils/data/_utils/fetch.py", line 61, in fetch
    return self.collate_fn(data)
  File "/home/user/micromamba/lib/python3.9/site-packages/torch_geometric/loader/node_loader.py", line 134, in collate_fn
    out = self.node_sampler.sample_from_nodes(input_data)
  File "/home/user/micromamba/lib/python3.9/site-packages/torch_geometric/sampler/neighbor_sampler.py", line 246, in sample_from_nodes
    out = node_sample(inputs, self._sample)
  File "/home/user/micromamba/lib/python3.9/site-packages/torch_geometric/sampler/neighbor_sampler.py", line 504, in node_sample
    out = sample_fn(seed, seed_time)
  File "/home/user/micromamba/lib/python3.9/site-packages/torch_geometric/sampler/neighbor_sampler.py", line 391, in _sample
    out = torch.ops.pyg.neighbor_sample(*args)
  File "/home/user/micromamba/lib/python3.9/site-packages/torch/_ops.py", line 442, in __call__
    return self._op(*args, **kwargs or {})
RuntimeError: pyg::neighbor_sample() Expected a value of type 'Optional[Tensor]' for argument 'edge_weight' but instead found type 'bool'.
Position: 7
Value: True
Declaration: pyg::neighbor_sample(Tensor rowptr, Tensor col, Tensor seed, int[] num_neighbors, Tensor? node_time=None, Tensor? edge_time=None, Tensor? seed_time=None, Tensor? edge_weight=None, bool csc=False, bool replace=False, bool directed=True, bool disjoint=False, str temporal_strategy="uniform", bool return_edge_id=True) -> (Tensor, Tensor, Tensor, Tensor?, int[], int[])
Cast error details: Unable to cast True to Tensor

Has anyone encountered a similar issue or have any insights into what might be going wrong? I have tested with multiple configurations and encountered the same problem.

Current environment details:

• torch_geometric 2.4.0,
• pyg-lib 0.4.0,
• 1.13.1-cuda11.7,
• Ubuntu 22.04

Thanks in advance for any assistance or insights!

[rusty1s]

Please see pyg-team/pyg-lib#299. Sorry for any inconvenience.