Datasets:

Modalities:
Audio
Text
ArXiv:
Libraries:
Datasets
License:

🚩 Report

#15
by Hanbufei - opened

There is an error of data zh-CN in this dataset, which is downloaded repeatedly for many times. Each time, all the caches are cleared and downloaded from the beginning, but it cannot be expected correctly.


ReadError Traceback (most recent call last)
Cell In[2], line 5
1 from datasets import load_dataset, DatasetDict
3 common_voice = DatasetDict()
----> 5 common_voice["train"] = load_dataset(dataset_name, language_abbr, split="train")
6 common_voice["validation"] = load_dataset(dataset_name, language_abbr, split="validation")
8 common_voice["train"][0]

ReadError: unexpected end of data


model_name_or_path = "openai/whisper-large-v2"
model_dir = "models/whisper-large-v2-asr-int8"

language = "Chinese (China)"
language_abbr = "zh-CN"
task = "transcribe"
dataset_name = "mozilla-foundation/common_voice_11_0"

from datasets import load_dataset, DatasetDict

common_voice = DatasetDict()

common_voice["train"] = load_dataset(dataset_name, language_abbr, split="train")
common_voice["validation"] = load_dataset(dataset_name, language_abbr, split="validation")

common_voice["train"][0]

Sign up or log in to comment