README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: transcription
      dtype: string
    - name: language
      dtype: string
    - name: dataset_source
      dtype: string
    - name: duration
      dtype: float64
    - name: BounharAbdelaziz/Moroccan-Darija-STT-large-turbo-v1.5
      dtype: string
  splits:
    - name: validation
      num_bytes: 7693903
      num_examples: 105
  download_size: 7683733
  dataset_size: 7693903
configs:
  - config_name: default
    data_files:
      - split: validation
        path: data/validation-*

Overview

This dataset contains transcriptions produced by various Speech-to-Text (STT) models.

The dataset is built on the Atlasia Moroccan-Darija-Youtube-Commons-Eval, which itself is a subset of the evaluation dataset. This subset has been carefully filtered to include only audio samples with durations of less than 4 seconds. These shorter samples are typically associated with high-quality transcriptions, making them ideal for evaluating STT model performance.

Currently evaluated model

• "BounharAbdelaziz/Morocco-Darija-STT-tiny"
• "BounharAbdelaziz/Morocco-Darija-STT-small"
• "BounharAbdelaziz/Morocco-Darija-STT-large-v1.2"
• "openai/whisper-large-v3-turbo"
• "openai/whisper-large-v3"
• "boumehdi/wav2vec2-large-xlsr-moroccan-darija"
• "abdelkader12/whisper-small-ar"
• "ychafiqui/whisper-medium-darija"
• "ychafiqui/whisper-small-darija"
• ...please add yours after eval...