Whisper Small Luganda
This model is a fine-tuned version of openai/whisper-small on the Common Voice 15.0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4482
- Wer: 43.0651
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: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- 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 |
---|---|---|---|---|
0.8094 | 0.0682 | 500 | 0.8624 | 73.2082 |
0.6104 | 0.1364 | 1000 | 0.6438 | 60.0304 |
0.5388 | 0.2045 | 1500 | 0.5586 | 51.9028 |
0.4618 | 0.2727 | 2000 | 0.5165 | 48.3825 |
0.4339 | 0.3409 | 2500 | 0.4886 | 46.0645 |
0.4229 | 0.4091 | 3000 | 0.4672 | 44.7809 |
0.3948 | 0.4772 | 3500 | 0.4544 | 43.5807 |
0.4017 | 0.5454 | 4000 | 0.4482 | 43.0651 |
Framework versions
- Transformers 4.40.0
- Pytorch 2.2.2+cu118
- Datasets 2.19.0
- Tokenizers 0.19.1
- Downloads last month
- 4
Finetuned from
Dataset used to train mn720/inctraining
Evaluation results
- Wer on Common Voice 15.0validation set self-reported43.065