<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
output
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 4.1041
- Wer: 1.0
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: 0.1
- train_batch_size: 4
- 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: 10
- num_epochs: 4
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
9.9461 | 0.13 | 10 | 15.0608 | 1.0 |
14.0423 | 0.26 | 20 | 6.7504 | 1.0 |
8.3772 | 0.39 | 30 | 11.9379 | 0.9994 |
14.5978 | 0.52 | 40 | 11.9946 | 1.0 |
14.3409 | 0.65 | 50 | 27.0088 | 1.0 |
13.6096 | 0.78 | 60 | 11.5932 | 1.0 |
11.9665 | 0.91 | 70 | 11.2733 | 1.0 |
11.4185 | 1.04 | 80 | 11.3424 | 1.0 |
8.8849 | 1.17 | 90 | 9.9160 | 1.0 |
7.4571 | 1.3 | 100 | 8.4745 | 1.0 |
5.098 | 1.43 | 110 | 5.9112 | 1.0 |
4.3967 | 1.56 | 120 | 4.2660 | 1.0 |
4.4123 | 1.69 | 130 | 4.3839 | 1.0 |
4.1123 | 1.82 | 140 | 4.1882 | 1.0 |
3.8196 | 1.95 | 150 | 4.2341 | 1.0 |
3.9248 | 2.08 | 160 | 4.2814 | 1.0 |
3.7829 | 2.21 | 170 | 4.3074 | 1.0 |
3.8432 | 2.34 | 180 | 4.1633 | 1.0 |
3.7821 | 2.47 | 190 | 4.1557 | 1.0 |
3.8287 | 2.6 | 200 | 4.3340 | 1.0 |
3.7915 | 2.73 | 210 | 4.2436 | 1.0 |
3.7911 | 2.86 | 220 | 4.3358 | 1.0 |
3.8335 | 2.99 | 230 | 4.1469 | 1.0 |
3.7663 | 3.12 | 240 | 4.1145 | 1.0 |
3.7516 | 3.25 | 250 | 4.0924 | 1.0 |
3.7326 | 3.38 | 260 | 4.1800 | 1.0 |
3.7447 | 3.51 | 270 | 4.1080 | 1.0 |
3.7293 | 3.64 | 280 | 4.1557 | 1.0 |
3.7541 | 3.77 | 290 | 4.1289 | 1.0 |
3.7222 | 3.9 | 300 | 4.1041 | 1.0 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0+cu116
- Datasets 2.7.1
- Tokenizers 0.13.2