Edit model card

Whisper Base French

This model is a fine-tuned version of openai/whisper-base on the mozilla-foundation/common_voice_11_0 fr dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4968
  • Wer on mozilla-foundation/common_voice_11_0 fr: 24.0648
  • Wer on google/fleurs fr_fr: 24.20
  • Wer on facebook/voxpopuli fr: 23.66

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: 32
  • eval_batch_size: 16
  • 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: 5000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.534 0.2 1000 0.5710 27.4408
0.4409 1.2 2000 0.5279 25.1981
0.3095 2.2 3000 0.5117 25.0818
0.3285 3.2 4000 0.4995 24.0601
0.3032 4.2 5000 0.4968 24.0648

Framework versions

  • Transformers 4.26.0.dev0
  • Pytorch 1.11.0+cu102
  • Datasets 2.7.1.dev0
  • Tokenizers 0.13.2
Downloads last month
14

Dataset used to train qanastek/whisper-base-french-cased

Evaluation results