Datasets:

openslr

/

librispeech_asr

like 115

Hello,
Thanks for contributing your dataset!
However, an error came to me when I try to load your dataset through the following code:

from datasets import load_dataset
import pandas as pd
from multiprocessing import Pool
import os
data_dir = '/disk/scratch2/s1905792/librispeech'
#dataset = load_dataset("librispeech_asr", cache_dir=data_dir)
test = load_dataset("librispeech_asr", "all", split="test.clean", cache_dir=data_dir)

The error returned to me is

/disk/scratch2/s1905792/anaconda3/envs/py3.7-gpu/lib/python3.7/site-packages/datasets/utils/info_utils.py in verify_splits(expected_splits, recorded_splits)
     89         return
     90     if len(set(expected_splits) - set(recorded_splits)) > 0:
---> 91         raise ExpectedMoreSplits(str(set(expected_splits) - set(recorded_splits)))
     92     if len(set(recorded_splits) - set(expected_splits)) > 0:
     93         raise UnexpectedSplits(str(set(recorded_splits) - set(expected_splits)))

ExpectedMoreSplits: {'train.clean.360', 'train.clean.100', 'test.clean', 'validation.other', 'validation.clean', 'test.other', 'train.other.500'}

Could you point out where I was wrong and how to fix this bug? Is it possibly caused by missing any file?
Thanks!