Data_lm.save() will crash the colab

Part 1 (2019)
1

Hi everyone,

I’m trying to follow Lesson 3 and used the same way to the Kaggle competion.
The problem is when I want to save the databunch,it will crash the colab even when there are half about the RAM left(6G).
Here is the code:
data_lm = (TextList
.from_csv(path, ‘train.csv’, cols=‘comment_text’)
.split_by_rand_pct(0.2,66)
.label_for_lm()
.databunch(bs=12))
data_lm.save()

I’d like to know why.
Is the train.csv too large?

1 Like