metadata
language:
- es
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- arturoapio/MadeUpWords
metrics:
- wer
base_model: openai/whisper-small
model-index:
- name: Whisper Small es - Galilei
results:
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: Common Voice Made up words
type: arturoapio/MadeUpWords
args: 'config: es, split: test'
metrics:
- type: wer
value: 7.2727272727272725
name: Wer
Whisper Small es - Galilei
This model is a fine-tuned version of openai/whisper-small on the Common Voice Made up words dataset. It achieves the following results on the evaluation set:
- Loss: 0.0000
- Wer: 7.2727
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 |
|---|---|---|---|---|
| 0.0 | 71.43 | 1000 | 0.0000 | 5.4545 |
| 0.0 | 142.86 | 2000 | 0.0000 | 9.0909 |
| 0.0 | 214.29 | 3000 | 0.0000 | 7.2727 |
| 0.0 | 285.71 | 4000 | 0.0000 | 7.2727 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.1.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.1