---
dataset_info:
  features:
  - name: audio
    dtype: audio
  - name: class
    dtype: string
  - name: class_id
    dtype: int64
  - name: noise_measurement
    dtype: float64
  - name: latitude
    dtype: float64
  - name: longitude
    dtype: float64
  - name: altitude
    dtype: float64
  - name: accuracy
    dtype: float64
  - name: submitter_id
    dtype: int64
  - name: region
    dtype: string
  - name: timestamp
    dtype: string
  - name: sentence
    dtype: string
  splits:
  - name: train
    num_bytes: 48611940.0
    num_examples: 1000
  download_size: 48496430
  dataset_size: 48611940.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---


All right reserved by and credit to Sunbird/urban-noise-uganda-61k <br>
Original URL: https://huggingface.co/datasets/Sunbird/urban-noise-uganda-61k
<br>
Change:<br>
1. add a 'sentence' column for ASR training, value as '%nz' for all rows