Data source: https://huggingface.co/datasets/gpt omni/VoiceAssistant 400K 1. Question and answer audios are extracted, which results in 940,108 audio files. 2. Raw conversations are formed as multi round chat format in data.jsonl , which has in total of 251,223 samples.
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy