Pytorch Unable to download Dataset

nlp
1

I am currently trying to use WikiTest103 dataset from the pytorch torchtext module.
When I try to run the below code, I get the error as:

from torchtext.datasets import WikiText103
X_train= WikiText103()
print(X_train.shape)

Error:

Traceback (most recent call last):
  File "D:/Freiburg/MasterProject/TestWikiText103.py", line 2, in <module>
    X_train= WikiText103()
TypeError: __init__() missing 2 required positional arguments: 'path' and 'text_field'

However,
When I run the same piece of code on Google Colab, it works perfectly.
My torchtext version is 0.8.1 and pytorch version is 1.7.1

2

Have you tried this way?
http://man.hubwiz.com/docset/torchtext.docset/Contents/Resources/Documents/experimental_datasets.html#wikitext103

3

I had the same issue, I solved it by upgrading torchtext to version 0.11.0 with:

 pip install torchtext --upgrade
4

It works. Thank you.