TypeError: 'int' object is not callable during torch model weights load #8
Description
I followed the README.
- Setup the conda env based on the environment.yaml
- Downloaded the weights.tar and also the two npy stat files
- Compiled and installed wgrib2
- Used gdas script to download data for a 2023 summer day
- Trying to call inference. I checked that the file is in the right place. With
unzip -lit seems it's not corrupted but who knows.
fourcastnet) jupyter@instance-20250216-164054:~/FourCastNet/NCEP$ python inference.py 2023080400 -w model/ -i input_2023080400.npy -o output/ -l 336
model/weights.tar
Traceback (most recent call last):
File "/home/jupyter/FourCastNet/NCEP/inference.py", line 265, in <module>
fcn.run()
~~~~~~~^^
File "/home/jupyter/FourCastNet/NCEP/inference.py", line 196, in run
model = self.load_model(path)
File "/home/jupyter/FourCastNet/NCEP/inference.py", line 158, in load_model
checkpoint = torch.load(checkpoint_file, map_location=device)
File "/opt/conda/envs/fourcastnet/lib/python3.13/site-packages/torch/serialization.py", line 1462, in load
return _load(
opened_zipfile,
...<3 lines>...
**pickle_load_args,
)
File "/opt/conda/envs/fourcastnet/lib/python3.13/site-packages/torch/serialization.py", line 1964, in _load
result = unpickler.load()
File "/opt/conda/envs/fourcastnet/lib/python3.13/site-packages/torch/_weights_only_unpickler.py", line 512, in load
self.append(self.persistent_load(pid))
~~~~~~~~~~~~~~~~~~~~^^^^^
File "/opt/conda/envs/fourcastnet/lib/python3.13/site-packages/torch/serialization.py", line 1928, in persistent_load
typed_storage = load_tensor(
dtype, nbytes, key, _maybe_decode_ascii(location)
)
File "/opt/conda/envs/fourcastnet/lib/python3.13/site-packages/torch/serialization.py", line 1900, in load_tensor
wrap_storage=restore_location(storage, location),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^
File "/opt/conda/envs/fourcastnet/lib/python3.13/site-packages/torch/serialization.py", line 1811, in restore_location
result = map_location(storage, location)
TypeError: 'int' object is not callable
This is a blocker right now.