Library Deprecation while converting to EXE

[Alan S]

I was trying to pack my pytorch code using the command , some of the pytorch packages get deprecated while doing it. does anyone know any alternate ways or any changes in the command to properly pack those packages? im including the deprecation log in this.

(venv) PS D:\ai> pyinstaller --onefile --windowed --name AIQualityCheckerv2 --hidden-import=torchvision.models --hidden-import=ultralytics --hidden-import=watchdog.observers --hidden-import=watchdog.events --collect-all ultralytics applicationv11.py

32219 INFO: Processing standard module hook ‘hook-torch.py’ from 'D:\ai\venv\Lib\site-packages\pyinstaller_hooks_contrib\stdhooks’
1501 WARNING: Failed to collect submodules for ‘torch.utils.tensorboard’ because importing ‘torch.utils.tensorboard’ raised: ModuleNotFoundError: No module named ‘tensorboard’
W1031 20:12:52.144000 19208 Lib\site-packages\torch\distributed\elastic\multiprocessing\redirects.py:29] NOTE: Redirects are currently not supported in Windows or MacOs.
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init.py:54: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
ALLREDUCE = partial(ddp_comm_hook_wrapper, comm_hook=default.allreduce_hook)
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init.py:55: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
FP16_COMPRESS = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:58: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
BF16_COMPRESS = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:61: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
QUANTIZE_PER_TENSOR = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:64: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
QUANTIZE_PER_CHANNEL = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:67: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
POWER_SGD = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:74: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
POWER_SGD_RANK2 = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:80: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
BATCHED_POWER_SGD = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:85: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
BATCHED_POWER_SGD_RANK2 = partial(
D:\ai\venv\Lib\site-packages\torch\distributed\algorithms\ddp_comm_hooks_init_.py:90: FutureWarning: functools.partial will be a method descriptor in future Python versions; wrap it in enum.member() if you want to preserve the old behavior
NOOP = partial(
D:\ai\venv\Lib\site-packages\PyInstaller\utils\hooks_init_.py:665: DeprecationWarning: torch.distributed._sharding_spec will be deprecated, use torch.distributed._shard.sharding_spec instead
import(name)
D:\ai\venv\Lib\site-packages\PyInstaller\utils\hooks_init_.py:665: DeprecationWarning: torch.distributed._sharded_tensor will be deprecated, use torch.distributed._shard.sharded_tensor instead
import(name)
D:\ai\venv\Lib\site-packages\PyInstaller\utils\hooks_init_.py:665: DeprecationWarning: torch.distributed._shard.checkpoint will be deprecated, use torch.distributed.checkpoint instead
import(name)
52587 INFO: hook-torch: this torch build does not depend on MKL…
54695 INFO: Processing standard module hook ‘hook-matplotlib.py’ from ‘D:\ai\venv\Lib\site-packages\PyInstaller\hooks’
55239 INFO: Processing pre-safe-import-module hook ‘hook-packaging.py’ from ‘D:\ai\venv\Lib\site-packages\PyInstaller\hooks\pre_safe_import_module’
55427 INFO: Processing pre-safe-import-module hook ‘hook-gi.py’ from ‘D:\ai\venv\Lib\site-packages\PyInstaller\hooks\pre_safe_import_module’
55758 INFO: Processing standard module hook ‘hook-matplotlib.backend_bases.py’ from ‘D:\ai\venv\Lib\site-packages\PyInstaller\hooks’

[bwoodsend]

They should be safe to ignore. PyInstaller isn't specifically trying to import those deprecated modules – it's just loading everything that's there so that it can see what's in the package. 

[Alan S]

thanks for the response, but when i ignore that and continue to use the file, the part of the code which uses that packages runs into error which makes the whole application useless. can you suggest me some ways to rectify that?

[Brenainn Woodsend]

What exactly is the error you see at runtime? It's unlikely to be anything to do with those deprecation warnings.