datasets: Cannot download all datasets except mnist using command tfds.load().

**when I try to download the cifar10 using the below commands: ’‘’input:‘’‘ import tensorflow as tf import tensorflow_datasets as tfds data1=tfds.load(‘cifar10’)

’‘’result:‘’‘ E0724 13:57:10.078672 1204 registered.py:171] Failed to construct dataset cifar10 Traceback (most recent call last):

File “<ipython-input-15-157aebdffff9>”, line 1, in <module> data1=tfds.load(‘cifar10’)

File “C:\ProgramData\Anaconda3\lib\site-packages\tensorflow_datasets\core\api_utils.py”, line 52, in disallow_positional_args_dec return fn(*args, **kwargs)

File “C:\ProgramData\Anaconda3\lib\site-packages\tensorflow_datasets\core\registered.py”, line 296, in load dbuilder = builder(name, data_dir=data_dir, **builder_kwargs)

File “C:\ProgramData\Anaconda3\lib\site-packages\tensorflow_datasets\core\registered.py”, line 169, in builder return _DATASET_REGISTRYname

File “C:\ProgramData\Anaconda3\lib\site-packages\tensorflow_datasets\core\api_utils.py”, line 52, in disallow_positional_args_dec return fn(*args, **kwargs)

File “C:\ProgramData\Anaconda3\lib\site-packages\tensorflow_datasets\core\dataset_builder.py”, line 188, in init self.info.initialize_from_bucket()

File “C:\ProgramData\Anaconda3\lib\site-packages\tensorflow_datasets\core\dataset_info.py”, line 411, in initialize_from_bucket gcs_utils.download_gcs_file(fname, out_fname)

File “C:\ProgramData\Anaconda3\lib\site-packages\tensorflow_datasets\core\utils\gcs_utils.py”, line 40, in download_gcs_file resp = requests.get(url, stream=stream)

File “C:\ProgramData\Anaconda3\lib\site-packages\requests\api.py”, line 75, in get return request(‘get’, url, params=params, **kwargs)

File “C:\ProgramData\Anaconda3\lib\site-packages\requests\api.py”, line 60, in request return session.request(method=method, url=url, **kwargs)

File “C:\ProgramData\Anaconda3\lib\site-packages\requests\sessions.py”, line 533, in request resp = self.send(prep, **send_kwargs)

File “C:\ProgramData\Anaconda3\lib\site-packages\requests\sessions.py”, line 646, in send r = adapter.send(request, **kwargs)

File “C:\ProgramData\Anaconda3\lib\site-packages\requests\adapters.py”, line 516, in send raise ConnectionError(e, request=request)

ConnectionError: HTTPConnectionPool(host=‘storage.googleapis.com’, port=80): Max retries exceeded with url: /tfds-data/dataset_info/cifar10/1.0.2/image.image.json (Caused by NewConnectionError(‘<urllib3.connection.HTTPConnection object at 0x000001F638AEDBE0>: Failed to establish a new connection: [WinError 10060] 。’))

‘’‘supplement’‘’：sometimes, What I’ve amended source code, but I failed And when i use model tfds， i had closed the firewall.

Environment information (if applicable)

Operating System: Windows 10
Python version: 3.7.3
tensorflow-datasets version: 1.1.0
tensorflow version: 1.13.1
Anaconda version: 4.7.10

About this issue

Original URL
State: closed
Created 5 years ago
Comments: 18 (9 by maintainers)

Most upvoted comments

SAD. I have the same problem with you because we both live in CHINA!

xlcbingo1999 on Oct 6, 2020

If you have issue with GCS, could you try set _is_gcs_disabled = True and report if this works ? https://github.com/tensorflow/datasets/blob/5dd79ad752cf87126eb70d8d267ce13526007829/tensorflow_datasets/core/utils/gcs_utils.py#L34

If so, could you try updating this function to catch the error which is raised, and skip GCS:

https://github.com/tensorflow/datasets/blob/5dd79ad752cf87126eb70d8d267ce13526007829/tensorflow_datasets/core/utils/gcs_utils.py#L43

Conchylicultor on Oct 6, 2020

Yes, the GCS seems to be the issue. You can try to hack the code to disable the GCS

from tensorflow_datasets.core.utils import gcs_utils
gcs_utils.gcs_dataset_info_files = lambda *args, **kwargs: None
gcs_utils.is_dataset_on_gcs = lambda *args, **kwargs: False

tfds.load('cifar10')

Hopefully this should works. We should probably add an option so user can choose to disable GCS

Conchylicultor on Jul 25, 2019

@Conchylicultor maybe, however I’m in China, the special issue feel unwell. I will try use better vpn to build connection.

Yisiheng on Jul 26, 2019