metadata
language:
- id
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
- magic_data
- TITML
metrics:
- wer
model-index:
- name: Whisper Medium Indonesian
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_11_0 id
type: mozilla-foundation/common_voice_11_0
config: id
split: test
metrics:
- name: Wer
type: wer
value: 3.993359771281011
Whisper Medium Indonesian
This model is a fine-tuned version of openai/whisper-medium on the mozilla-foundation/common_voice_11_0, magic_data, titml id dataset. It achieves the following results on the evaluation set:
- Loss: 0.0698
- Wer: 3.9934
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-06
- train_batch_size: 16
- 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: 10000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0252 | 0.35 | 1000 | 0.0733 | 4.4822 |
| 0.0222 | 0.7 | 2000 | 0.0698 | 4.3392 |
| 0.0166 | 1.06 | 3000 | 0.0696 | 4.2378 |
| 0.0117 | 1.41 | 4000 | 0.0679 | 4.0810 |
| 0.0295 | 1.76 | 5000 | 0.0671 | 4.0856 |
| 0.0147 | 2.11 | 6000 | 0.0690 | 4.0302 |
| 0.0157 | 2.47 | 7000 | 0.0698 | 3.9934 |
| 0.0113 | 2.82 | 8000 | 0.0698 | 4.0302 |
| 0.0101 | 3.17 | 9000 | 0.0707 | 4.0579 |
| 0.0076 | 3.52 | 10000 | 0.0708 | 4.0625 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2