Potatoasdasdasdasda's picture
End of training
2f54dca
metadata
language:
  - es
license: apache-2.0
base_model: openai/whisper-base
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_11_0
metrics:
  - wer
model-index:
  - name: Whisper base Spanish Improved
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 11.0
          type: mozilla-foundation/common_voice_11_0
          config: es
          split: test
          args: 'config: es, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 20.435869264920363

Whisper base Spanish Improved

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

  • Loss: 0.3238
  • Wer: 20.4359

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: 5e-06
  • train_batch_size: 16
  • 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: 8000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.3177 0.12 1000 0.3974 23.5886
0.294 0.25 2000 0.3681 22.2548
0.3409 0.38 3000 0.3512 21.6964
0.26 0.5 4000 0.3407 21.2621
0.3503 0.62 5000 0.3345 20.8259
0.3067 0.75 6000 0.3297 20.5207
0.2324 0.88 7000 0.3243 20.4956
0.3413 1.0 8000 0.3238 20.4359

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.1
  • Datasets 2.15.0
  • Tokenizers 0.15.0