README.md

metadata

library_name: transformers
license: mit
datasets:
  - Tarakeshwaran/Whisper-train-data
language:
  - en
metrics:
  - wer
base_model:
  - openai/whisper-small
pipeline_tag: automatic-speech-recognition
tags:
  - generated_from_trainer
model-index:
  - name: Whisper Small En - Tarakeshwaran
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: minimal_common_voice_en
          type: Tarakeshwaran/Whisper-train-data
          args: 'config: en, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 13.170732

Whisper Small En - Tarakeshwaran

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

• Loss: 1.041972
• Wer: 13.170732

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: 16
• eval_batch_size: 8
• seed: 42
• optimizer: adamw_bnb_8bit
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 25
• training_steps: 100
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.47.0.dev0
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3