Datasets:
Tasks:
Audio Classification
Multilinguality:
multilingual
Size Categories:
10M<n<100M
Language Creators:
other
Annotations Creators:
machine-generated
Source Datasets:
extended|common_voice
License:
Streaming mode broken?
#3
by
jnemecek
- opened
Streaming mode was working a few weeks ago, but seems broken now. I'm not sure if the recent PR made the change, but attempts to load from previous commits did not solve the problem.
244 n_files_path = dl_manager.download(n_files_url)
245
--> 246 with open(n_files_path, "r", encoding="utf-8") as file:
247 n_files = int(file.read().strip()) # the file contains a number of archives
248
ValueError: I/O operation on closed file.
With some troubleshooting, I traced this back to fsspec. Full details on the huggingface github. Bottom line, fsspec needs to be < 2022.08.02.