This dataset was created using the Voxtral ASR Fine-tuning Interface.

audio_path: Relative path to the audio file (stored in audio/ directory) text: Transcription of the audio

Number of examples: 10 Audio files uploaded: 10 Total dataset size: 10,015,583 bytes

from datasets import load_dataset, Audio

dataset = load_dataset("Tonic/voxtral-dataset-20250913_175651")

#… See the full description on the dataset page: https://huggingface.co/datasets/Tonic/voxtral-dataset-20250913_175651.