| from faster_whisper import WhisperModel | |
| jfk_path = "jfk.flac" | |
| model = WhisperModel("tiny", device="cuda") | |
| segments, info = model.transcribe(jfk_path, word_timestamps=True) | |
| for segment in segments: | |
| print("[%.2fs -> %.2fs] %s" % (segment.start, segment.end, segment.text)) | |