# Frame-Level Speech Recognition — Data

来源：https://www.kaggle.com/competitions/11785-hw1p2-f23/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description


    - train-clean-100: training set
    - dev-clean folder: dev/validation set
    - test-clean folder: test set, sample_submission.csv