whisper_wermet_nosup_0015
This model is a fine-tuned version of openai/whisper-tiny on an unknown dataset. It achieves the following results on the evaluation set:
- Train Loss: 0.2198
- Train Accuracy: 0.0320
- Train Wermet: 3.8215
- Validation Loss: 0.4835
- Validation Accuracy: 0.0311
- Validation Wermet: 4.4217
- Epoch: 14
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:
- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': 1e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: float32
Training results
Train Loss | Train Accuracy | Train Wermet | Validation Loss | Validation Accuracy | Validation Wermet | Epoch |
---|---|---|---|---|---|---|
5.0729 | 0.0113 | 42.1824 | 4.4421 | 0.0120 | 27.3059 | 0 |
4.3249 | 0.0126 | 23.9224 | 4.0443 | 0.0141 | 18.2054 | 1 |
3.8845 | 0.0144 | 12.7780 | 3.4577 | 0.0169 | 10.3356 | 2 |
2.7411 | 0.0198 | 15.0018 | 1.8774 | 0.0244 | 14.5666 | 3 |
1.5621 | 0.0250 | 9.7248 | 1.2443 | 0.0273 | 5.4731 | 4 |
1.0745 | 0.0272 | 7.1512 | 0.9802 | 0.0285 | 4.4745 | 5 |
0.8261 | 0.0284 | 6.2358 | 0.8209 | 0.0293 | 5.6600 | 6 |
0.6673 | 0.0292 | 5.8338 | 0.7182 | 0.0298 | 4.0874 | 7 |
0.5548 | 0.0298 | 5.0555 | 0.6489 | 0.0301 | 4.4537 | 8 |
0.4694 | 0.0303 | 4.3895 | 0.6038 | 0.0304 | 2.8294 | 9 |
0.4011 | 0.0307 | 4.4178 | 0.5673 | 0.0306 | 3.5806 | 10 |
0.3446 | 0.0311 | 4.1189 | 0.5329 | 0.0308 | 4.7372 | 11 |
0.2968 | 0.0314 | 3.9189 | 0.5138 | 0.0309 | 2.2499 | 12 |
0.2553 | 0.0317 | 3.7758 | 0.4990 | 0.0310 | 3.3427 | 13 |
0.2198 | 0.0320 | 3.8215 | 0.4835 | 0.0311 | 4.4217 | 14 |
Framework versions
- Transformers 4.27.0.dev0
- TensorFlow 2.11.0
- Datasets 2.10.0
- Tokenizers 0.13.2
- Downloads last month
- 4
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.