---
dataset_info:
  features:
  - name: 'Unnamed: 0'
    dtype: int64
  - name: path
    dtype: audio
  - name: label
    dtype: int64
  splits:
  - name: train
    num_bytes: 10950822720
    num_examples: 6000
  - name: valid
    num_bytes: 3665953567
    num_examples: 2000
  - name: test
    num_bytes: 3779356929.                  0
    num_examples: 2000
  - name: train_low
    num_bytes: 2226351300.6
    num_examples: 1200
  download_size: 10721885223
  dataset_size: 20622484516.6
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
  - split: train_low
    path: data/train_low-*
pretty_name: BEANS Bats
---
# Dataset Card for "beans_bats"


## Dataset Description

**Paper:** https://doi.org/10.1038/sdata.2017.143               

## Dataset Summary

This dataset contains annotated recordings of Egyptian fruit bats with splits and preprocessing like described in BEANS. It is used for **classification** tasks.

## Data Splits

 | train | train_low | valid | test | 
 |------|-----| ------ | ------ | 
 | 6000 |  1200 | 2000 | 2000 |