davidkern13
commented
11 months ago when run command with model name backgroundremover -i "image.jpeg" -m "u2net_human_seg" -o "output.png"
the model is download automaticaly for u2netp and u2net_human_seg and working whell, but for u2net still happen the error
I get this errror in console after run this command backgroundremover -i "./image.jpeg" -o "output.png"
Traceback (most recent call last): File "", line 198, in _run_module_as_main
File "", line 88, in _run_code
File "path\backgroundremover.exe__main__.py", line 7, in
File "path\backgroundremover\cmd\cli.py", line 239, in main
remove(
File "path\backgroundremover\bg.py", line 183, in remove
model = get_model(model_name)
File "path\backgroundremover\bg.py", line 171, in get_model
return detect.load_model(model_name="u2net")
File "path\backgroundremover\u2net\detect.py", line 82, in load_model
torch.load(
File "path\torch\serialization.py", line 815, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "path\torch\serialization.py", line 1051, in _legacy_load
typed_storage._untyped_storage._set_from_file(
RuntimeError: unexpected EOF, expected 4554729 more bytes. The file might be corrupted.
I think models its not downloaded and scripts dont find them and throw error. possible to redownload the train models automaticaly?
After put the error to chatgpt I get this response -> This error is coming from the torch library, which suggests that there might be an issue with the downloaded model file for the u2net model.
My environment:
Windows 10 FFmpeg version 5.1 Python version 3.11.4 Backgroundremover version 0.2.4 Torch version 2.0.1 Torchvision version 0.15.2