ivangtorre's picture
Upload dataset
965d5a9 verified
metadata
language:
  - qu
  - gn
license: unknown
task_categories:
  - automatic-speech-recognition
  - translation
dataset_info:
  - config_name: bribri
    features:
      - name: audio
        dtype: audio
      - name: subset
        dtype: string
      - name: source_processed
        dtype: string
      - name: source_raw
        dtype: string
      - name: target_raw
        dtype: string
      - name: split
        dtype: string
    splits:
      - name: train
        num_bytes: 55969483
        num_examples: 495
      - name: dev
        num_bytes: 22453830
        num_examples: 250
    download_size: 78332438
    dataset_size: 78423313
  - config_name: guarani
    features:
      - name: audio
        dtype: audio
      - name: subset
        dtype: string
      - name: source_processed
        dtype: string
      - name: source_raw
        dtype: string
      - name: target_raw
        dtype: string
      - name: split
        dtype: string
    splits:
      - name: train
        num_bytes: 37288028
        num_examples: 293
      - name: dev
        num_bytes: 14154952
        num_examples: 93
    download_size: 85618314
    dataset_size: 51442980
  - config_name: kotiria
    features:
      - name: audio
        dtype: audio
      - name: subset
        dtype: string
      - name: source_processed
        dtype: string
      - name: source_raw
        dtype: string
      - name: target_raw
        dtype: string
      - name: split
        dtype: string
    splits:
      - name: train
        num_bytes: 291865656.624
        num_examples: 1972
      - name: dev
        num_bytes: 34442493
        num_examples: 254
    download_size: 341252496
    dataset_size: 326308149.624
  - config_name: quechua
    features:
      - name: audio
        dtype: audio
      - name: subset
        dtype: string
      - name: source_processed
        dtype: string
      - name: source_raw
        dtype: string
      - name: target_raw
        dtype: string
      - name: split
        dtype: string
    splits:
      - name: train
        num_bytes: 192806183
        num_examples: 573
      - name: dev
        num_bytes: 239435023
        num_examples: 250
    download_size: 862887688
    dataset_size: 432241206
configs:
  - config_name: bribri
    data_files:
      - split: train
        path: bribri/train-*
      - split: dev
        path: bribri/dev-*
  - config_name: guarani
    data_files:
      - split: train
        path: guarani/train-*
      - split: dev
        path: guarani/dev-*
  - config_name: kotiria
    data_files:
      - split: train
        path: kotiria/train-*
      - split: dev
        path: kotiria/dev-*
  - config_name: quechua
    data_files:
      - split: train
        path: quechua/train-*
      - split: dev
        path: quechua/dev-*

Citation

@inproceedings{ebrahimi2022findings,
  title={Findings of the Second AmericasNLP Competition on Speech-to-Text Translation},
  author={Ebrahimi, Abteen and Mager, Manuel and Wiemerslage, Adam and Denisov, Pavel and Oncevay, Arturo and Liu, Danni and Koneru, Sai and Ugan, Enes Yavuz and Li, Zhaolin and Niehues, Jan and others},
  booktitle={NeurIPS 2022 Competition Track},
  pages={217--232},
  year={2022},
  organization={PMLR}
}