Problem loading the model

[AGFoersch]

I finally wanted to test the QuPath SAM extension, but I fail to get the SAM API to run.

I'm sorry if I'm making a dumb mistake, but when I run ...

uvicorn samapi.main:app

The application attempts to load the necessary models, but I receive warnings about overwriting model registrations. Eventually, the application fails with a PytorchStreamReader error, forcing the inference to run on CPU despite CUDA being detected. Below is the full error trace:

Full Traceback:

[path_to_env]\mobile_sam\modeling\tiny_vit_sam.py:656: UserWarning: Overwriting tiny_vit_5m_224 in registry with mobile_sam.modeling.tiny_vit_sam.tiny_vit_5m_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  return register_model(fn_wrapper)
[path_to_env]\mobile_sam\modeling\tiny_vit_sam.py:656: UserWarning: Overwriting tiny_vit_11m_224 in registry with mobile_sam.modeling.tiny_vit_sam.tiny_vit_11m_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  return register_model(fn_wrapper)
[path_to_env]\mobile_sam\modeling\tiny_vit_sam.py:656: UserWarning: Overwriting tiny_vit_21m_224 in registry with mobile_sam.modeling.tiny_vit_sam.tiny_vit_21m_224. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  return register_model(fn_wrapper)
[path_to_env]\mobile_sam\modeling\tiny_vit_sam.py:656: UserWarning: Overwriting tiny_vit_21m_384 in registry with mobile_sam.modeling.tiny_vit_sam.tiny_vit_21m_384. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  return register_model(fn_wrapper)
[path_to_env]\mobile_sam\modeling\tiny_vit_sam.py:656: UserWarning: Overwriting tiny_vit_21m_512 in registry with mobile_sam.modeling.tiny_vit_sam.tiny_vit_21m_512. This is because the name being registered conflicts with an existing name. Please check if this is not expected.
  return register_model(fn_wrapper)
[path_to_env]\samapi\main.py:238: UserWarning: cuda device found but got the error PytorchStreamReader failed reading zip archive: failed finding central directory - using CPU for inference
  warnings.warn(
Traceback (most recent call last):
  File "[path_to_env]\runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "[path_to_env]\runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "[path_to_env]\uvicorn.exe\__main__.py", line 7, in <module>
    sys.exit(main())
  File "[path_to_env]\click\core.py", line 1157, in __call__
    return self.main(*args, **kwargs)
  File "[path_to_env]\click\core.py", line 1078, in main
    rv = self.invoke(ctx)
  File "[path_to_env]\click\core.py", line 1434, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "[path_to_env]\click\core.py", line 783, in invoke
    return __callback(*args, **kwargs)
  File "[path_to_env]\uvicorn\main.py", line 412, in main
    run(
  File "[path_to_env]\uvicorn\main.py", line 579, in run
    server.run()
  File "[path_to_env]\uvicorn\server.py", line 65, in run
    return asyncio.run(self.serve(sockets=sockets))
  File "[path_to_env]\asyncio\runners.py", line 44, in run
    return loop.run_until_complete(main)
  File "[path_to_env]\asyncio\base_events.py", line 649, in run_until_complete
    return future.result()
  File "[path_to_env]\uvicorn\server.py", line 69, in serve
    await self._serve(sockets)
  File "[path_to_env]\uvicorn\server.py", line 76, in _serve
    config.load()
  File "[path_to_env]\uvicorn\config.py", line 434, in load
    self.loaded_app = import_from_string(self.app)
  File "[path_to_env]\uvicorn\importer.py", line 19, in import_from_string
    module = importlib.import_module(module_str)
  File "[path_to_env]\importlib\__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1050, in _gcd_import
  File "<frozen importlib._bootstrap>", line 1027, in _find_and_load
  File "<frozen importlib._bootstrap>", line 1006, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 688, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 883, in exec_module
  File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
  File "[path_to_env]\samapi\main.py", line 291, in <module>
    _register_default_weights()
  File "[path_to_env]\samapi\main.py", line 287, in _register_default_weights
    register_state_dict_from_url(model_type, checkpoint_url, f"default")
  File "[path_to_env]\samapi\main.py", line 272, in register_state_dict_from_url
    raise e
  File "[path_to_env]\samapi\main.py", line 261, in register_state_dict_from_url
    _, filepath = load_state_dict_from_url(
  File "[path_to_env]\samapi\hub_extension.py", line 244, in load_state_dict_from_url
    return torch.load(cached_file, map_location=map_location), cached_file
  File "[path_to_env]\torch\serialization.py", line 1004, in load
    with _open_zipfile_reader(opened_file) as opened_zipfile:
  File "[path_to_env]\torch\serialization.py", line 456, in __init__
    super().__init__(torch._C.PyTorchFileReader(name_or_buffer))
RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory

Any assistance or guidance on how to resolve this issue would be greatly appreciated!

Thank you so much and keep up the great work @ksugar !!!

P.S. A similar error returns when I set --workers 2 ...

[ksugar]

Hi @AGFoersch, thank you for testing samapi and reporting the issue.

The warning messages are always present and are not the root cause of the problem.
(I understand they are not preferable for users, so I will work on suppressing them soon.)

It seems that some of the model files were downloaded but corrupted.

Please try deleting the files and directories at the following paths, and then restart the server.

Windows

%homepath%\.cache\torch\hub\checkpoints
%homepath%\.samapi

Mac OS / Linux

~/.cache/torch/hub/checkpoints
~/.samapi

I hope this solves the issue, but if it doesn’t, I’d be happy to support further.

[AGFoersch]

It works! Thank you so much for your support.

[Elena983]

Hi, if i didn't have it before.
Do I need to create it?
And then to restart?

[ksugar]

Hi @Elena983, you don't need to create them manually. They will be automatically created. If they are missing for some reason, restarting should trigger their creation. Let me know if you run into any issues!

[TiTian25]

why i can’t find %homepath%.cache\torch\hub\checkpoints?