openai/whisper-base
This model is a fine-tuned version of openai/whisper-base on the pphuc25/EngMed dataset. It achieves the following results on the evaluation set:
- Loss: 1.2118
- Wer: 21.1498
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: 8
- 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: 100
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.5428 | 1.0 | 323 | 0.6628 | 36.9726 |
| 0.3049 | 2.0 | 646 | 0.7340 | 25.6329 |
| 0.1478 | 3.0 | 969 | 0.8008 | 32.5422 |
| 0.0905 | 4.0 | 1292 | 0.8517 | 21.2553 |
| 0.0556 | 5.0 | 1615 | 0.9244 | 26.4241 |
| 0.0474 | 6.0 | 1938 | 0.9692 | 25.3692 |
| 0.0338 | 7.0 | 2261 | 1.0099 | 25.7384 |
| 0.0196 | 8.0 | 2584 | 1.0844 | 27.6371 |
| 0.0152 | 9.0 | 2907 | 1.1063 | 22.7848 |
| 0.0062 | 10.0 | 3230 | 1.1242 | 22.6793 |
| 0.0064 | 11.0 | 3553 | 1.1909 | 26.1076 |
| 0.0046 | 12.0 | 3876 | 1.1556 | 21.7300 |
| 0.0021 | 13.0 | 4199 | 1.1804 | 20.8861 |
| 0.0023 | 14.0 | 4522 | 1.1757 | 21.2553 |
| 0.0003 | 15.0 | 4845 | 1.2014 | 22.9430 |
| 0.0003 | 16.0 | 5168 | 1.1849 | 21.7300 |
| 0.0004 | 17.0 | 5491 | 1.1936 | 21.6245 |
| 0.0002 | 18.0 | 5814 | 1.2106 | 20.9916 |
| 0.0002 | 19.0 | 6137 | 1.2111 | 20.9388 |
| 0.0001 | 20.0 | 6460 | 1.2118 | 21.1498 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 3