k2e_asr-scr_w2v2-base_003
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: -inf
- Per: 0.1504
- Pcc: 0.6412
- Ctc Loss: 0.4563
- Mse Loss: 0.8278
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: 3333
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2235
- training_steps: 22350
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Per | Pcc | Ctc Loss | Mse Loss |
|---|---|---|---|---|---|---|---|
| 19.0581 | 1.0 | 745 | 4.7610 | 0.9890 | 0.4254 | 3.8212 | 1.0115 |
| 4.6142 | 2.0 | 1490 | 4.3367 | 0.9890 | 0.5426 | 3.5146 | 1.0524 |
| 2.5298 | 3.0 | 2235 | 1.8275 | 0.2518 | 0.5558 | 0.9450 | 0.8809 |
| 1.4011 | 4.01 | 2980 | 1.4794 | 0.1752 | 0.6324 | 0.5673 | 0.8902 |
| 1.0517 | 5.01 | 3725 | 1.1440 | 0.1652 | 0.6501 | 0.5285 | 0.6974 |
| 0.6284 | 6.01 | 4470 | 1.3914 | 0.1602 | 0.6319 | 0.4873 | 0.9713 |
| 0.0986 | 7.01 | 5215 | 0.6875 | 0.1604 | 0.6455 | 0.4865 | 0.7234 |
| -1.2 | 8.01 | 5960 | -1.4751 | 0.1509 | 0.6417 | 0.4542 | 0.8851 |
| -9.3211 | 9.01 | 6705 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 10.01 | 7450 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 11.01 | 8195 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 12.02 | 8940 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 13.02 | 9685 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 14.02 | 10430 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 15.02 | 11175 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 16.02 | 11920 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 17.02 | 12665 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 18.02 | 13410 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 19.03 | 14155 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 20.03 | 14900 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 21.03 | 15645 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 22.03 | 16390 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 23.03 | 17135 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 24.03 | 17880 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 25.03 | 18625 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 26.03 | 19370 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 27.04 | 20115 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 28.04 | 20860 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 29.04 | 21605 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
| 0.0 | 30.04 | 22350 | -inf | 0.1504 | 0.6412 | 0.4563 | 0.8278 |
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/k2e_asr-scr_w2v2-base_003
Base model
facebook/wav2vec2-base