torchvision data downloader aborts if ipywidgets are not available

🐛 Bug

I am testing some PyTorch code in Deepnote and the following call to torchvision aborts since the environment doesn’t support ipywidgets

test_data = torchvision.datasets.MNIST('../data', train=False, download=True, transform=local_transforms)

To Reproduce

Check the following notebook - https://deepnote.com/project/0b12c588-a1de-4d66-bb0f-fa6b8f9b1b87

ImportError: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
<ipython-input-3-c62d245b04eb> in <module>
      5 local_transforms = torchvision.transforms.Compose([local_transform_1, local_transform_2])
      6 
----> 7 test_data = torchvision.datasets.MNIST('../data', train=False, download=True, transform=local_transforms)

/shared-libs/python3.7/py/lib/python3.7/site-packages/torchvision/datasets/mnist.py in __init__(self, root, train, transform, target_transform, download)
     77 
     78         if download:
---> 79             self.download()
     80 
     81         if not self._check_exists():

/shared-libs/python3.7/py/lib/python3.7/site-packages/torchvision/datasets/mnist.py in download(self)
    144         for url, md5 in self.resources:
    145             filename = url.rpartition('/')[2]
--> 146             download_and_extract_archive(url, download_root=self.raw_folder, filename=filename, md5=md5)
    147 
    148         # process and save as torch files

/shared-libs/python3.7/py/lib/python3.7/site-packages/torchvision/datasets/utils.py in download_and_extract_archive(url, download_root, extract_root, filename, md5, remove_finished)
    254         filename = os.path.basename(url)
    255 
--> 256     download_url(url, download_root, filename, md5)
    257 
    258     archive = os.path.join(download_root, filename)

/shared-libs/python3.7/py/lib/python3.7/site-packages/torchvision/datasets/utils.py in download_url(url, root, filename, md5)
     70             urllib.request.urlretrieve(
     71                 url, fpath,
---> 72                 reporthook=gen_bar_updater()
     73             )
     74         except (urllib.error.URLError, IOError) as e:  # type: ignore[attr-defined]

/shared-libs/python3.7/py/lib/python3.7/site-packages/torchvision/datasets/utils.py in gen_bar_updater()
     13 
     14 def gen_bar_updater() -> Callable[[int, int, int], None]:
---> 15     pbar = tqdm(total=None)
     16 
     17     def bar_update(count, block_size, total_size):

/shared-libs/python3.7/py/lib/python3.7/site-packages/tqdm/notebook.py in __init__(self, *args, **kwargs)
    246         unit_scale = 1 if self.unit_scale is True else self.unit_scale or 1
    247         total = self.total * unit_scale if self.total else self.total
--> 248         self.container = self.status_printer(self.fp, total, self.desc, self.ncols)
    249         self.container.pbar = self
    250         if display_here:

/shared-libs/python3.7/py/lib/python3.7/site-packages/tqdm/notebook.py in status_printer(_, total, desc, ncols)
    113         if IProgress is None:  # #187 #451 #558 #872
    114             raise ImportError(
--> 115                 "IProgress not found. Please update jupyter and ipywidgets."
    116                 " See https://ipywidgets.readthedocs.io/en/stable"
    117                 "/user_install.html")

ImportError: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html

Expected behavior

Maybe a check can be added that skips call to tqdm if it’s not available?

cc @pmeier

[jaintj95]

which version of tqdm deepnote provides ?

@vfdev-5 From my environment it seems tqdm version 4.55.1

[jaintj95]

@jaintj95 a solution like tqdm/tqdm#619 (comment) should help , as it disables the tqdm bar IIUC.

If this doesn’t work, it might be because PyTorch now defaults to importing tqdm from tqdm.auto, so you might need to patch it from there

Thanks. I’ve already tried the solution you linked. It doesn’t work. Let me see if I can figure out the patch.