TextDataBunch.from_csv throwing `BrokenProcessPool' Error on Google Colab

fastai
1

Hi, while I was trying to use 5-nn-imdb.ipynb notebook, part of Natural Language Processing course in Fast.ai, I am facing one issue. There is a step to load data and tokenize using TextDataBunch. The code snippet is

while error:

try:

# Preprocessing steps

data_lm = TextDataBunch.from_csv(path, ‘texts.csv’)

error = False

print(f’failure count is {count}\n’)

This is throwing error continuously on Google Colab. While searching for this error on stack overflow and other forums there was a suggestion to set number of cpu cores to 1 using the command
defaults.cpus=1 Though I tried this, it is not working. What is the fix for this issue?

Thanks and Regards,
-Hari