README.md

metadata

language:
  - ar
license: apache-2.0
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - MohammadJamalaldeen/Sudanese_Dialect
metrics:
  - wer
model-index:
  - name: Sudanese Whisper - MohammadJamalaldeen
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Sudanese Dialect
          type: MohammadJamalaldeen/Sudanese_Dialect
          args: 'config: ar, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 109.00443880786304

Sudanese Whisper - MohammadJamalaldeen

This model is a fine-tuned version of openai/whisper-small on the Sudanese Dialect dataset. It achieves the following results on the evaluation set:

• Loss: 0.9342
• Wer: 109.0044

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-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3694	1.66	1000	0.8267	83.6398
0.136	3.31	2000	0.8340	107.5460
0.0695	4.97	3000	0.8763	80.6806
0.0198	6.62	4000	0.9342	109.0044

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.0
• Datasets 2.1.0
• Tokenizers 0.13.2