whisper-large-sv / README.md
birgermoell's picture
update model card README.md
b0275f0
metadata
language:
  - sv
license: apache-2.0
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_11_0
metrics:
  - wer
model-index:
  - name: whisper-large-sv
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 11.0
          type: mozilla-foundation/common_voice_11_0
          config: sv-SE
          split: train[:1%]+validation[:1%]
          args: sv-SE
        metrics:
          - name: Wer
            type: wer
            value: 30.935251798561154

whisper-large-sv

This model is a fine-tuned version of openai/whisper-large on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

  • Loss: 1.5259
  • Wer: 30.9353

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: 1
  • 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: 1
  • training_steps: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
4.5521 0.04 5 3.5048 48.2014
1.8009 0.08 10 1.5259 30.9353

Framework versions

  • Transformers 4.25.0.dev0
  • Pytorch 1.13.0+cu117
  • Datasets 2.6.1
  • Tokenizers 0.13.1