metadata
language:
- uz
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
- google/fleurs
metrics:
- wer
model-index:
- name: Whisper Small Uzbek
results:
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: mozilla-foundation/common_voice_11_0
type: mozilla-foundation/common_voice_11_0
config: uz
split: test
args: da
metrics:
- type: wer
value: 23.650914047642605
name: Wer
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: google/fleurs
type: google/fleurs
config: uz_uz
split: test
metrics:
- type: wer
value: 47.15
name: WER
Whisper Small Uzbek
This model is a fine-tuned version of openai/whisper-small trained on the mozilla-foundation/common_voice_11_0 uz and google/fleurs uz_uz datasets, and evaluated on the mozilla-foundation/common_voice_11_0 uz dataset. It achieves the following results on the evaluation set:
- Loss: 0.3872
- Wer: 23.6509
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: 64
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 400
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1542 | 0.2 | 1000 | 0.4711 | 30.8413 |
| 0.0976 | 0.4 | 2000 | 0.4040 | 26.6464 |
| 0.1088 | 1.0 | 3000 | 0.3765 | 24.4952 |
| 0.0527 | 1.21 | 4000 | 0.3872 | 23.6509 |
| 0.0534 | 1.41 | 5000 | 0.3843 | 23.6817 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2