metadata
language:
- is
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- language-and-voice-lab/samromur_asr
metrics:
- wer
model-index:
- name: Whisper Small Icelandic
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: samromur
type: language-and-voice-lab/samromur_asr
config: samromur_asr
split: test
args: 'split: test'
metrics:
- name: Wer
type: wer
value: 23.040907733651835
Whisper Small Icelandic
This model is a fine-tuned version of openai/whisper-small on the samromur dataset. It achieves the following results on the evaluation set:
- Loss: 0.2613
- Wer: 23.0409
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.3551 | 0.18 | 1000 | 0.4322 | 35.0421 |
| 0.2541 | 0.36 | 2000 | 0.3249 | 27.4721 |
| 0.231 | 0.53 | 3000 | 0.2781 | 24.2234 |
| 0.2277 | 0.71 | 4000 | 0.2613 | 23.0409 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.7.1
- Tokenizers 0.13.2