Datasets:

Modalities:
Audio
Text
ArXiv:
Libraries:
Datasets
License:

How to make a sample from a dataset for two languages?

#1
by artyomboyko - opened

Good afternoon. Can you tell me how to make a selection from a dataset for two languages?

For example I use the following code for Whisper fine-tuning:

common_voice = DatasetDict()

common_voice["train"] = load_dataset("mozilla-foundation/common_voice_13_0", "ru", split="train+validation", use_auth_token=True)
common_voice["test"] = load_dataset("mozilla-foundation/common_voice_13_0", "ru", split="test", use_auth_token=True)

How can I take data from a dataset for two languages (for example, Russian and English)?

Sign up or log in to comment