ntdxyg / deeppseudo Goto Github PK

was trying to run the file and met with this issue. Could not find any methods to solve it , python 3.7.10 , torch text 0.9.0 , torch 1.8.0

Traceback (most recent call last):
File "DeepPseudo.py", line 583, in
transformer.load_state_dict(torch.load(save_path))
File "/home/ec2-user/DeepPseudo/venv/lib64/python3.7/site-packages/torch/serialization.py", line 593, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "/home/ec2-user/DeepPseudo/venv/lib64/python3.7/site-packages/torch/serialization.py", line 762, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
EOFError: Ran out of input

In the DeepPseudo.py,what does SCALE mean? Why multiply by self.scale in 320?And why self.scale =np.sqrt(0.5)?