Closed tuvuumass closed 5 years ago
Hi Tu,
Yes, the DROP dataset is yet not fully converted (it also does not appear in the datasets available as described in the readme)
It will be fully added in the next few weeks. Thanks for understanding, Alon
Hello Alon,
I tried to build the dataset as in the README file.
python build_dataset.py --dataset_name DROP --split train --output_file DROP.jsonl.gz --n_processes 10
But I got this error: