Whisper Base Noise Ko - Dearlie
This model is a fine-tuned version of openai/whisper-base on the Noise Data dataset. It achieves the following results on the evaluation set:
- Loss: 0.9216
- Cer: 37.5124
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: 0.0001
- 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 | Cer |
|---|---|---|---|---|
| 1.3536 | 0.8780 | 1000 | 1.3752 | 55.5864 |
| 0.944 | 1.7559 | 2000 | 1.0808 | 51.4185 |
| 0.5985 | 2.6339 | 3000 | 0.9612 | 40.2651 |
| 0.3207 | 3.5119 | 4000 | 0.9216 | 37.5124 |
Framework versions
- Transformers 4.41.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1
- Downloads last month
- 2
Model tree for Dearlie/whisper-noise4
Base model
openai/whisper-base