metadata

license: apache-2.0
base_model: openai/whisper-small
tags:
  - generated_from_trainer
datasets:
  - common_voice_8_0
metrics:
  - wer
model-index:
  - name: whisper-small-uz
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: common_voice_8_0
          type: common_voice_8_0
          config: uz
          split: validation[:20%]
          args: uz
        metrics:
          - name: Wer
            type: wer
            value: 29.159322033898306

whisper-small-uz

This model is a fine-tuned version of openai/whisper-small on the common_voice_8_0 dataset. It achieves the following results on the evaluation set:

Loss: 0.3743
Wer: 29.1593

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
0.3146	1.3514	1000	0.4122	35.3898
0.1332	2.7027	2000	0.3529	29.7356
0.0256	4.0541	3000	0.3658	29.2881
0.0134	5.4054	4000	0.3743	29.1593

Framework versions

Transformers 4.40.2
Pytorch 2.1.2
Datasets 2.19.1
Tokenizers 0.19.1