saq_asr-scr_w2v2-base_002
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: 1.5579
- Per: 0.1375
- Pcc: 0.6524
- Ctc Loss: 0.4848
- Mse Loss: 1.1172
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: 1
- seed: 2222
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 742
- training_steps: 7420
Training results
Training Loss | Epoch | Step | Validation Loss | Per | Pcc | Ctc Loss | Mse Loss |
---|---|---|---|---|---|---|---|
12.3448 | 1.0 | 742 | 4.7671 | 0.9994 | 0.5952 | 3.7706 | 1.2000 |
3.9935 | 2.0 | 1484 | 2.2612 | 0.4024 | 0.6003 | 1.4156 | 0.9141 |
1.7152 | 3.0 | 2226 | 1.5426 | 0.1756 | 0.6508 | 0.6567 | 0.8426 |
1.1701 | 4.0 | 2968 | 1.7292 | 0.1634 | 0.6518 | 0.5777 | 1.0791 |
0.7988 | 5.0 | 3710 | 1.6447 | 0.1547 | 0.6439 | 0.5390 | 1.0478 |
0.4267 | 6.0 | 4452 | 1.7815 | 0.1491 | 0.6515 | 0.5098 | 1.1760 |
0.0983 | 7.0 | 5194 | 1.4809 | 0.1432 | 0.6506 | 0.5025 | 1.0110 |
-0.1884 | 8.0 | 5936 | 1.5095 | 0.1400 | 0.6534 | 0.4844 | 1.0659 |
-0.3958 | 9.0 | 6678 | 1.5737 | 0.1375 | 0.6527 | 0.4815 | 1.1214 |
-0.5383 | 10.0 | 7420 | 1.5579 | 0.1375 | 0.6524 | 0.4848 | 1.1172 |
Framework versions
- Transformers 4.38.1
- Pytorch 2.0.1
- Datasets 2.16.1
- Tokenizers 0.15.2
- Downloads last month
- 0
Model tree for excalibur12/saq_asr-scr_w2v2-base_002
Base model
facebook/wav2vec2-base