metadata
language:
- es
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Large Es - Javier Alonso
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: es
metrics:
- name: Wer
type: wer
value: 5.520113299724547
Whisper Large Es - Javier Alonso
This model is a fine-tuned version of openai/whisper-large on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1571
- Wer: 5.5201
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: 8
- eval_batch_size: 2
- 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: 10000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.211 | 0.1 | 1000 | 0.2293 | 8.3896 |
| 0.2227 | 0.2 | 2000 | 0.2215 | 8.2552 |
| 0.1496 | 0.3 | 3000 | 0.2121 | 8.0362 |
| 0.1851 | 0.4 | 4000 | 0.2018 | 7.5197 |
| 0.1917 | 0.5 | 5000 | 0.1916 | 7.1098 |
| 0.1857 | 0.6 | 6000 | 0.1817 | 6.5537 |
| 0.1294 | 0.7 | 7000 | 0.1752 | 6.4062 |
| 0.1358 | 0.8 | 8000 | 0.1670 | 5.9950 |
| 0.1542 | 0.9 | 9000 | 0.1604 | 5.7858 |
| 0.1554 | 1.0 | 10000 | 0.1571 | 5.5201 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.10.0+cu111
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2