README.md

metadata

language:
  - it
license: apache-2.0
datasets:
  - common_voice
  - mozilla-foundation/common_voice_8_0
metrics:
  - wer
  - cer
tags:
  - audio
  - automatic-speech-recognition
  - hf-asr-leaderboard
  - it
  - mozilla-foundation/common_voice_8_0
  - speech
  - wav2vec2
model-index:
  - name: XLS-R Wav2Vec2 Italian by radiogroup crits
    results:
      - task:
          name: Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice it
          type: mozilla-foundation/common_voice_8_0
          args: it
        metrics:
          - name: Test WER
            type: wer
            value: 9.04
          - name: Test CER
            type: cer
            value: 2.2
          - name: Test WER (+LM)
            type: wer
            value: 6.24
          - name: Test CER (+LM)
            type: cer
            value: 1.67

XLS-R-1B-ITALIAN-DOC4LM-5GRAM

Language model information

Our language model was generated using a dataset of Italian wikipedia articles and manual transcriptions of radio newspapers and television programs.

Download CommonVoice8.0 dataset for italian language

from datasets import load_dataset

dataset = load_dataset("mozilla-foundation/common_voice_8_0", "it", use_auth_token=True)

Evaluation Commands

To evaluate on mozilla-foundation/common_voice_8_0 with split test:

python eval.py --model_id radiogroup-crits/wav2vec2-xls-r-1b-italian-doc4lm-5gram --dataset mozilla-foundation/common_voice_8_0 --config it --split test --log_outputs --greedy
mv log_mozilla-foundation_common_voice_8_0_it_test_predictions.txt log_mozilla-foundation_common_voice_8_0_it_test_predictions_greedy.txt
mv mozilla-foundation_common_voice_8_0_it_test_eval_results.txt mozilla-foundation_common_voice_8_0_it_test_eval_results_greedy.txt

Citation

If you want to cite this model you can use this:

@misc{crits2022wav2vec2-xls-r-1b-italian-doc4lm-5gram,
  title={XLS-R Wav2Vec2 Italian by radiogroup crits},
  author={Teraoni Prioletti Raffaele, Casagranda Paolo and Russo Francesco},
  publisher={Hugging Face},
  journal={Hugging Face Hub},
  howpublished={\url{https://huggingface.co/radiogroup-crits/wav2vec2-xls-r-1b-italian-doc4lm-5gram}},
  year={2022}
}