ValueError: cannot reshape array of size 0 into shape

[rakhimovv]

Hello! thank you for an amazing project!

Could you please help to resolve the issue when I run:

python train.py --dataset_path data/shadow_hand --experiment_name shadow_hand_experiment

The output:

============ Training LBS model ============
[Train] LRS dataloader
[PointCloud] Found 5046 samples in data/shadow_hand
  0%|                                                                                                                                                                                                     | 0/1 [00:00<?, ?it/s[Cache] get_canonical_pose                                                                                                                                                                               | 0/630 [00:00<?, ?it/s]
  0%|                                                                                                                                                                                                     | 0/1 [00:24<?, ?it/s]
Traceback (most recent call last):ng Average: 0.0004029735955761829:  63%|█████████████████████████████████████████████████████████████████████████▎                                          | 398/630 [00:24<00:09, 24.96it/s]
  File "/home/rusrakhimov/projects/drrobot/train.py", line 266, in <module>
    three_stage_training(lp.extract(args), None, op.extract(args), pp.extract(args), args.test_every, args.save_every, args.checkpoint_every, args.start_checkpoint, args.debug_from, args.experiment_name)
  File "/home/rusrakhimov/projects/drrobot/train.py", line 228, in three_stage_training
    train_lrs(gaussians)
  File "/home/rusrakhimov/projects/drrobot/lbs/nn.py", line 79, in train_lrs
    for i, (pcd, pose) in enumerate(dataloader):
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 630, in __next__
    data = self._next_data()
           ^^^^^^^^^^^^^^^^^
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 1345, in _next_data
    return self._process_data(data)
           ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 1371, in _process_data
    data.reraise()
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/torch/_utils.py", line 694, in reraise
    raise exception
ValueError: Caught ValueError in DataLoader worker process 3.
Original Traceback (most recent call last):
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/torch/utils/data/_utils/worker.py", line 308, in _worker_loop
    data = fetcher.fetch(index)
           ^^^^^^^^^^^^^^^^^^^^
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/torch/utils/data/_utils/fetch.py", line 51, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/torch/utils/data/_utils/fetch.py", line 51, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
            ~~~~~~~~~~~~^^^^^
  File "/home/rusrakhimov/projects/drrobot/utils/lbs_utils.py", line 162, in __getitem__
    pose = np.load(os.path.join(sample_path, 'joint_positions.npy'))
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/numpy/lib/npyio.py", line 456, in load
    return format.read_array(fid, allow_pickle=allow_pickle,
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/rusrakhimov/micromamba/envs/dr3/lib/python3.11/site-packages/numpy/lib/format.py", line 839, in read_array
    array.shape = shape
    ^^^^^^^^^^^
ValueError: cannot reshape array of size 0 into shape (24,)

Loss: 0.0001564089034218341, Running Average: 0.0004029735955761829:  63%|█████████████████████████████████████████████████████████████████████████▍                                          | 399/630 [00:24<00:14, 16.31it/s]

[rakhimovv]

I have a gcc compilator. The problem is of another nature.

Made a small investigation in the class PointCloudDataset

self.dirs = []
for d in sorted(glob.glob(os.path.join(self.path, 'sample_*'))):
    try:
        o3d.io.read_point_cloud(os.path.join(d, 'pc.ply'))
        np.load(os.path.join(d, 'joint_positions.npy'))
        self.dirs.append(d)
    except Exception as e:
        print(f"{d} is damaged due to {e}")

Exception example:

[Open3D WARNING] Read PLY failed: unable to open file: data/shadow_hand/sample_4979.lock/pc.ply
data/shadow_hand/sample_4979.lock is damaged due to [Errno 20] Not a directory: 'data/shadow_hand/sample_4979.lock/joint_positions.npy'
RPly: Unable to open file

Found out that glob captures garbage folders. After filtering such folders the training works