UsmanAXAI's picture
Upload tokenizer
478c273 verified
metadata
language:
  - ar
license: apache-2.0
tags:
  - ar-asr-leaderboard
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_16_1
metrics:
  - wer
base_model: openai/whisper-medium
model-index:
  - name: Whisper Medium Ar - AxAI
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: Common Voice 16.1
          type: mozilla-foundation/common_voice_16_1
          config: default
          split: None
          args: 'config: ar, split: test[:10%]'
        metrics:
          - type: wer
            value: 100
            name: Wer

Whisper Medium Ar - AxAI

This model is a fine-tuned version of openai/whisper-medium on the Common Voice 16.1 dataset. It achieves the following results on the evaluation set:

  • Loss: 1.4298
  • Wer: 100.0

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1e-06
  • train_batch_size: 4
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • training_steps: 1000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.0045 35.71 500 1.3081 100.0
0.0012 71.43 1000 1.4298 100.0

Framework versions

  • Transformers 4.37.2
  • Pytorch 2.2.0+cu121
  • Datasets 2.17.1
  • Tokenizers 0.15.2