whisper-small-hyper-tuned-v2
This model is a fine-tuned version of openai/whisper-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2138
- Wer: 0.3859
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: 5e-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: 1000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.2719 | 0.0330 | 100 | 0.3551 | 0.5211 |
| 0.2793 | 0.0660 | 200 | 0.3262 | 0.4921 |
| 0.2831 | 0.0990 | 300 | 0.3306 | 0.4927 |
| 0.2775 | 0.1320 | 400 | 0.3631 | 0.5363 |
| 0.2849 | 0.1650 | 500 | 0.3488 | 0.5040 |
| 0.2692 | 0.1980 | 600 | 0.3202 | 0.4967 |
| 0.2528 | 0.2309 | 700 | 0.2838 | 0.4400 |
| 0.2155 | 0.2639 | 800 | 0.2489 | 0.4116 |
| 0.1929 | 0.2969 | 900 | 0.2220 | 0.3912 |
| 0.1709 | 0.3299 | 1000 | 0.2138 | 0.3859 |
Framework versions
- Transformers 4.45.1
- Pytorch 2.1.0+cu118
- Datasets 3.0.1
- Tokenizers 0.20.0
- Downloads last month
- 0
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.