Datasets:

mesolitica
/

Zeroshot-Audio-Classification-Instructions

language:
  - en
dataset_info:
  features:
    - name: question
      dtype: string
    - name: answer
      dtype: string
    - name: audio_filename
      dtype: string
    - name: metadata
      dtype: string
  splits:
    - name: vggsound_train
      num_bytes: 1189877449
      num_examples: 182615
    - name: vggsound_test
      num_bytes: 99939976
      num_examples: 15341
    - name: nonspeech7k_train
      num_bytes: 2451589
      num_examples: 6289
    - name: nonspeech7k_test
      num_bytes: 285542
      num_examples: 725
    - name: vocalsound_train
      num_bytes: 3646349
      num_examples: 15531
    - name: vocalsound_test
      num_bytes: 842929
      num_examples: 3591
    - name: urbansound8k
      num_bytes: 3453567
      num_examples: 8732
    - name: emotion
      num_bytes: 16432408
      num_examples: 19974
    - name: age
      num_bytes: 62968387
      num_examples: 48767
    - name: gender
      num_bytes: 60029824
      num_examples: 48767
    - name: language
      num_bytes: 93887480
      num_examples: 48767
    - name: fsd50k
      num_bytes: 175199243
      num_examples: 51197
    - name: CochlScene_train
      num_bytes: 17713725
      num_examples: 60855
    - name: CochlScene_test
      num_bytes: 2230052
      num_examples: 7687
    - name: CochlScene_val
      num_bytes: 2189339
      num_examples: 7573
    - name: tau2022
      num_bytes: 100770546
      num_examples: 230350
    - name: esd_emotion
      num_bytes: 4977344
      num_examples: 35000
    - name: BirdCLEF_2021_scientific_name
      num_bytes: 265135140
      num_examples: 27740
    - name: BirdCLEF_2021_common_name
      num_bytes: 250890697
      num_examples: 27740
    - name: audioset
      num_bytes: 2616069868
      num_examples: 312883
    - name: emobox
      num_bytes: 38409103
      num_examples: 184821
  download_size: 99507577
  dataset_size: 5007400557
configs:
  - config_name: default
    data_files:
      - split: vggsound_train
        path: data/vggsound_train-*
      - split: vggsound_test
        path: data/vggsound_test-*
      - split: nonspeech7k_train
        path: data/nonspeech7k_train-*
      - split: nonspeech7k_test
        path: data/nonspeech7k_test-*
      - split: vocalsound_train
        path: data/vocalsound_train-*
      - split: vocalsound_test
        path: data/vocalsound_test-*
      - split: urbansound8k
        path: data/urbansound8k-*
      - split: emotion
        path: data/emotion-*
      - split: age
        path: data/age-*
      - split: gender
        path: data/gender-*
      - split: language
        path: data/language-*
      - split: fsd50k
        path: data/fsd50k-*
      - split: CochlScene_train
        path: data/CochlScene_train-*
      - split: CochlScene_test
        path: data/CochlScene_test-*
      - split: CochlScene_val
        path: data/CochlScene_val-*
      - split: tau2022
        path: data/tau2022-*
      - split: esd_emotion
        path: data/esd_emotion-*
      - split: BirdCLEF_2021_scientific_name
        path: data/BirdCLEF_2021_scientific_name-*
      - split: BirdCLEF_2021_common_name
        path: data/BirdCLEF_2021_common_name-*
      - split: audioset
        path: data/audioset-*
      - split: emobox
        path: data/emobox-*

Zeroshot-Audio-Classification-Instructions

Convert audio classification dataset into zero-shot format speech instructions, support both single label and multi-label,

We also converted huge WAV files into MP3 16k sample rate to reduce storage size.

To prevent leakage, please do not include test set in training session.

how to prepare the dataset

huggingface-cli download \
mesolitica/Zeroshot-Audio-Classification-Instructions \
--include "*.zip" \
--repo-type "dataset" \
--local-dir './'

huggingface-cli download \
mesolitica/Audio-Adversarial-Instructions \
--include "*.zip" \
--repo-type "dataset" \
--local-dir './'

huggingface-cli download \
mesolitica/Animal-Sound-Instructions \
--include "*.zip" \
--repo-type "dataset" \
--local-dir './'

huggingface-cli download \
mesolitica/EmoBox \
--include "*.zip" \
--repo-type "dataset" \
--local-dir './'

wget https://gist.githubusercontent.com/huseinzol05/2e26de4f3b29d99e993b349864ab6c10/raw/9b2251f3ff958770215d70c8d82d311f82791b78/unzip.py
python3 unzip.py

Acknowledgement

Special thanks to https://www.sns.com.my and Nvidia for 8x H100 node!