TypeError: unsupported operand type(s) for +: 'NoneType' and 'int'

[woodcore-an]

when i want to train the tless python -m surfemb.scripts.train tless:

Traceback (most recent call last):
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/home/zzz/github/surfemb/surfemb/scripts/train.py", line 121, in <module>
    main()
  File "/home/zzz/github/surfemb/surfemb/scripts/train.py", line 61, in main
    model = SurfaceEmbeddingModel(n_objs=len(obj_ids), **vars(args))
  File "/home/zzz/github/surfemb/surfemb/surface_embedding.py", line 48, in __init__
    n_class=(emb_dim + 1) if separate_decoders else n_objs * (emb_dim + 1),
TypeError: unsupported operand type(s) for +: 'NoneType' and 'int'

It Seems the emb_dim is None. But it has the default value emb_dim=12.

[rasmushaugaard]

I think it's because the argument parser has no default value for emb_dim. I'll add that at some point. Try python -m surfemb.scripts.train tless --emb-dim 12

[woodcore-an]

It works! By the way, in the train.py, the argument should be --gpus ,not the --gpu.

    parser.add_argument('--gpu', type=int, nargs='+', default=[0])

it should like this:

    parser.add_argument('--gpus', type=int, nargs='+', default=[0])

--gpu will cause the error:

Traceback (most recent call last):
  File "/home/zzz/miniconda3/envs/surf/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/home/zzz/miniconda3/envs/surf/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/home/zzz/woodcore/environments/surfemb/surfemb/scripts/train.py", line 121, in <module>
    main()
  File "/home/zzz/woodcore/environments/surfemb/surfemb/scripts/train.py", line 110, in main
    logger=logger, gpus=args.gpus, max_steps=args.max_steps,
AttributeError: 'Namespace' object has no attribute 'gpus'

And i want to know which version of pytorch are you using? I run conda env create -f environment.yml, it will install the pytorch=1.11.0. Ubuntu16.04, 2070super. But it will cause the following problem:

Traceback (most recent call last):
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/home/zzz/github/surfemb/surfemb/scripts/train.py", line 121, in <module>
    main()
  File "/home/zzz/github/surfemb/surfemb/scripts/train.py", line 61, in main
    model = SurfaceEmbeddingModel(n_objs=len(obj_ids), **vars(args))
  File "/home/zzz/github/surfemb/surfemb/surface_embedding.py", line 47, in __init__
    self.cnn = ResNetUNet(
  File "/home/zzz/github/surfemb/surfemb/dep/unet.py", line 21, in __init__
    self.base_model = torchvision.models.resnet18(pretrained=True)
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/site-packages/torchvision/models/resnet.py", line 309, in resnet18
    return _resnet("resnet18", BasicBlock, [2, 2, 2, 2], pretrained, progress, **kwargs)
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/site-packages/torchvision/models/resnet.py", line 296, in _resnet
    state_dict = load_state_dict_from_url(model_urls[arch], progress=progress)
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/site-packages/torch/hub.py", line 595, in load_state_dict_from_url
    return torch.load(cached_file, map_location=map_location)
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/site-packages/torch/serialization.py", line 705, in load
    with _open_zipfile_reader(opened_file) as opened_zipfile:
  File "/home/zzz/miniconda3/envs/surfemb/lib/python3.8/site-packages/torch/serialization.py", line 243, in __init__
    super(_open_zipfile_reader, self).__init__(torch._C.PyTorchFileReader(name_or_buffer))
RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory

It seems that the higher version cannot read the models.

[rasmushaugaard]

I've added a default value for emb_dim and changed the argument name from gpu to gpus.

The model loading error you're experiencing seems to be for the pretrained resnet18 backbone, not the pretrained surfemb models. I'm not sure what may have went wrong there.

I'm using pytorch 1.11.0 and torchvision 0.12.0.

[woodcore-an]

Thanks!