Datasets:
ZH-EN does not work
#1
by
Muennighoff
- opened
load_dataset("wmt19", "zh-en")
TypeError: Couldn't cast array of type
struct<c[hn]: string, en: string, zh: string>
to
struct<en: string, zh: string>
you can try
!git clone https://github.com/huggingface/datasets
data_dir is your clone datasets ,for example data_dir = '/tmp/datasets/datasets'
data = datasets.load_dataset(os.path.join(data_dir, 'wmt19'),name='zh-en', cache_dir="./sample_data")
Hi! This is an easier fix suggested by @lhoestq : https://github.com/huggingface/datasets/issues/4575#issuecomment-1223844270
albertvillanova
changed discussion status to
closed