README.md

metadata

tags:
  - generated_from_trainer
datasets:
  - afrispeech-200
metrics:
  - wer
model-index:
  - name: afrispeech_large_A100
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: afrispeech-200
          type: afrispeech-200
          config: all
          split: train
          args: all
        metrics:
          - name: Wer
            type: wer
            value: 14.81

afrispeech_large_A100

This model is a fine-tuned version of openai/whisper-large-v2 on the afrispeech-200 dataset.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

https://huggingface.co/Seyfelislem/afrispeech_large_A100/tensorboard

Framework versions

• Transformers 4.29.1
• Pytorch 1.13.1
• Datasets 2.12.0
• Tokenizers 0.13.3