OverflowError: Python int too large to convert to C long tensorflow datsets

Question

import tensorflow_datasets as tfds 

and then:

OverflowError                             Traceback (most recent call last)
<ipython-input-1-7342ce159b6a> in <module>
      1 import tensorflow as tf
----> 2 import tensorflow_datasets as tfds

~\anaconda3\lib\site-packages\tensorflow_datasets\__init__.py in <module>
     63 from tensorflow_datasets import translate
     64 from tensorflow_datasets import video
---> 65 from tensorflow_datasets import vision_language
     66 
     67 

~\anaconda3\lib\site-packages\tensorflow_datasets\vision_language\__init__.py in <module>
     19 from tensorflow_datasets.vision_language.grounded_scan import GroundedScan
     20 from tensorflow_datasets.vision_language.refcoco import RefCoco
---> 21 from tensorflow_datasets.vision_language.wit import Wit
     22 from tensorflow_datasets.vision_language.wit_kaggle import WitKaggle

~\anaconda3\lib\site-packages\tensorflow_datasets\vision_language\wit\__init__.py in <module>
     16 """wit dataset."""
     17 
---> 18 from tensorflow_datasets.vision_language.wit.wit import Wit

~\anaconda3\lib\site-packages\tensorflow_datasets\vision_language\wit\wit.py in <module>
     23 import tensorflow_datasets.public_api as tfds
     24 
---> 25 csv.field_size_limit(sys.maxsize)
     26 
     27 _DESCRIPTION = """

OverflowError: Python int too large to convert to C long

Answer 1

Issue is fixed in tensorflow-datasets 4.5.2

You can import tensorflow_datasets without any error.

import tensorflow_datasets as tfds