---
dataset_info:
  features:
  - name: input_features
    sequence:
      sequence: float32
  - name: labels
    sequence: int64
  splits:
  - name: train
    num_bytes: 2427438472
    num_examples: 2526
  - name: test
    num_bytes: 566988584
    num_examples: 590
  download_size: 1269442169
  dataset_size: 2994427056
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---

Cleaned (noise reduced, normalized, silence trimmed and computed log-Mel input features) 3000-row dataset(about 10hours) from Fleurs Tajik dataset
!!!Only for whisper!!!