Can't load GLUE/MNLI dataset from Hub due to schema issues

🤗Datasets
1

Hello, I’m trying to load the MNLI dataset from GLUE on Databricks using datasets 2.13.2 and facing the error depicted in the image below. Since it’s a standard and widely used dataset I was expecting it to work out-of-the-box. Any special config I should provide?

Thanks in advance, Felipe

Screenshot 2024-02-29 at 15.15.08
Screenshot 2024-02-29 at 15.15.081900×932 144 KB

2

Hi! You can avoid this error by updating datasets (pip install -U datasets) to a newer version and then downloading the dataset.

3

Hi Mario, thanks for the quick reply. I did but still getting the below:

Screenshot 2024-02-29 at 18.48.31
Screenshot 2024-02-29 at 18.48.311878×1148 316 KB

4

Found the problem. It was missing pyspark.

With pyspark 3.5.1 and datasets 2.17.1 everything is working now.

Thanks for the help.

5

This topic was automatically closed 12 hours after the last reply. New replies are no longer allowed.