metadata

language:
  - sv-SE
license: apache-2.0
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - mozilla-foundation/common_voice_11_0
metrics:
  - wer
model-index:
  - name: Whisper Small Hi - Swedish
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice 11.0
          type: mozilla-foundation/common_voice_11_0
          config: null
          split: None
          args: 'config: sv-SE, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 51.31886746793579

Whisper Small Hi - Swedish

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

Loss: 1.0765
Wer: 51.3189

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: 0.0005
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-06
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
1.5005	1.29	1000	1.7517	84.8861
0.8752	2.59	2000	1.2958	68.8688
0.4382	3.88	3000	1.1835	60.4152
0.0694	5.17	4000	1.1659	55.8442
0.0091	6.47	5000	1.0765	51.3189

Framework versions

Transformers 4.26.0.dev0
Pytorch 1.9.0+cu102
Datasets 2.7.1
Tokenizers 0.13.2