metadata
language:
- sr
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- espnet/yodas
metrics:
- wer
model-index:
- name: Whisper Small Yodas
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Yodas
type: espnet/yodas
config: sr
split: test
args: sr
metrics:
- name: Wer
type: wer
value: 0.11913993655269652
Whisper Small Yodas
This model is a fine-tuned version of openai/whisper-small on the Yodas dataset. It achieves the following results on the evaluation set:
- Loss: 0.1748
- Wer Ortho: 0.2143
- Wer: 0.1191
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: 50
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
|---|---|---|---|---|---|
| 0.7316 | 0.24 | 500 | 0.2554 | 0.2942 | 0.2184 |
| 0.6996 | 0.49 | 1000 | 0.2136 | 0.2535 | 0.1563 |
| 0.6073 | 0.73 | 1500 | 0.1979 | 0.2374 | 0.1452 |
| 0.6032 | 0.98 | 2000 | 0.1872 | 0.2228 | 0.1280 |
| 0.4603 | 1.22 | 2500 | 0.1811 | 0.2136 | 0.1218 |
| 0.4142 | 1.46 | 3000 | 0.1767 | 0.2152 | 0.1200 |
| 0.4457 | 1.71 | 3500 | 0.1759 | 0.2159 | 0.1234 |
| 0.4376 | 1.95 | 4000 | 0.1748 | 0.2143 | 0.1191 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.0.1+cu117
- Datasets 2.18.0
- Tokenizers 0.15.1