Closed rusty-ai closed 11 months ago
Hi @rusty-ai ,
Currently, within the context of the CLI, we cannot, as we use a high-level abstraction on top of the transformers API called pipeline. However, you can get the same directly from transformers: https://github.com/huggingface/transformers/issues/25138#issuecomment-1836554598
Is there any way to just detect the language of the audio, as it was in the original Whisper? Eg.