<!-- 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. -->
retrain5_oneTimeTraining_MTL-1epoch
This model is a fine-tuned version of alexziweiwang/exp21-uaspeech-foundation on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 7.1861
- Acc: 0.285
- Wer: 1.1126
- Correct: 57
- Total: 200
- Strlen: 200
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: 9e-06
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1.0
Training results
Training Loss | Epoch | Step | Validation Loss | Acc | Wer | Correct | Total | Strlen |
---|---|---|---|---|---|---|---|---|
No log | 0.02 | 5 | 13.9337 | 0.01 | 1.2925 | 2 | 200 | 200 |
12.4373 | 0.04 | 10 | 13.7513 | 0.08 | 1.5296 | 16 | 200 | 200 |
12.4373 | 0.06 | 15 | 13.5517 | 0.125 | 2.1126 | 25 | 200 | 200 |
12.6667 | 0.08 | 20 | 13.3400 | 0.165 | 2.5791 | 33 | 200 | 200 |
12.6667 | 0.11 | 25 | 13.1141 | 0.205 | 3.6561 | 41 | 200 | 200 |
11.1856 | 0.13 | 30 | 12.8805 | 0.22 | 2.7451 | 44 | 200 | 200 |
11.1856 | 0.15 | 35 | 12.6423 | 0.245 | 2.5178 | 49 | 200 | 200 |
10.6635 | 0.17 | 40 | 12.4028 | 0.27 | 2.4308 | 54 | 200 | 200 |
10.6635 | 0.19 | 45 | 12.1660 | 0.3 | 2.1818 | 60 | 200 | 200 |
10.7952 | 0.21 | 50 | 11.9291 | 0.305 | 1.9348 | 61 | 200 | 200 |
10.7952 | 0.23 | 55 | 11.6945 | 0.31 | 1.6858 | 62 | 200 | 200 |
10.3867 | 0.25 | 60 | 11.4608 | 0.315 | 1.5237 | 63 | 200 | 200 |
10.3867 | 0.27 | 65 | 11.2313 | 0.315 | 1.3953 | 63 | 200 | 200 |
10.252 | 0.3 | 70 | 11.0102 | 0.315 | 1.3162 | 63 | 200 | 200 |
10.252 | 0.32 | 75 | 10.7918 | 0.315 | 1.2826 | 63 | 200 | 200 |
10.1788 | 0.34 | 80 | 10.5736 | 0.315 | 1.2628 | 63 | 200 | 200 |
10.1788 | 0.36 | 85 | 10.3607 | 0.32 | 1.2391 | 64 | 200 | 200 |
9.1361 | 0.38 | 90 | 10.1527 | 0.31 | 1.2253 | 62 | 200 | 200 |
9.1361 | 0.4 | 95 | 9.9507 | 0.31 | 1.2036 | 62 | 200 | 200 |
9.5447 | 0.42 | 100 | 9.7553 | 0.315 | 1.2095 | 63 | 200 | 200 |
9.5447 | 0.44 | 105 | 9.5599 | 0.31 | 1.2016 | 62 | 200 | 200 |
9.1579 | 0.46 | 110 | 9.3711 | 0.295 | 1.1996 | 59 | 200 | 200 |
9.1579 | 0.48 | 115 | 9.1892 | 0.295 | 1.1897 | 59 | 200 | 200 |
7.9217 | 0.51 | 120 | 9.0143 | 0.3 | 1.1858 | 60 | 200 | 200 |
7.9217 | 0.53 | 125 | 8.8493 | 0.305 | 1.1719 | 61 | 200 | 200 |
8.4439 | 0.55 | 130 | 8.6946 | 0.305 | 1.1739 | 61 | 200 | 200 |
8.4439 | 0.57 | 135 | 8.5492 | 0.31 | 1.1581 | 62 | 200 | 200 |
8.0639 | 0.59 | 140 | 8.4153 | 0.315 | 1.1502 | 63 | 200 | 200 |
8.0639 | 0.61 | 145 | 8.2872 | 0.32 | 1.1482 | 64 | 200 | 200 |
8.4173 | 0.63 | 150 | 8.1649 | 0.33 | 1.1443 | 66 | 200 | 200 |
8.4173 | 0.65 | 155 | 8.0500 | 0.325 | 1.1403 | 65 | 200 | 200 |
7.8991 | 0.67 | 160 | 7.9422 | 0.33 | 1.1364 | 66 | 200 | 200 |
7.8991 | 0.7 | 165 | 7.8410 | 0.32 | 1.1344 | 64 | 200 | 200 |
6.9206 | 0.72 | 170 | 7.7469 | 0.32 | 1.1304 | 64 | 200 | 200 |
6.9206 | 0.74 | 175 | 7.6601 | 0.325 | 1.1285 | 65 | 200 | 200 |
7.1911 | 0.76 | 180 | 7.5832 | 0.305 | 1.1206 | 61 | 200 | 200 |
7.1911 | 0.78 | 185 | 7.5163 | 0.305 | 1.1225 | 61 | 200 | 200 |
7.201 | 0.8 | 190 | 7.4565 | 0.305 | 1.1245 | 61 | 200 | 200 |
7.201 | 0.82 | 195 | 7.4049 | 0.295 | 1.1245 | 59 | 200 | 200 |
7.1507 | 0.84 | 200 | 7.3568 | 0.295 | 1.1225 | 59 | 200 | 200 |
7.1507 | 0.86 | 205 | 7.3139 | 0.3 | 1.1206 | 60 | 200 | 200 |
6.6223 | 0.89 | 210 | 7.2774 | 0.295 | 1.1186 | 59 | 200 | 200 |
6.6223 | 0.91 | 215 | 7.2469 | 0.295 | 1.1186 | 59 | 200 | 200 |
7.1645 | 0.93 | 220 | 7.2220 | 0.295 | 1.1166 | 59 | 200 | 200 |
7.1645 | 0.95 | 225 | 7.2041 | 0.29 | 1.1146 | 58 | 200 | 200 |
6.2562 | 0.97 | 230 | 7.1921 | 0.29 | 1.1146 | 58 | 200 | 200 |
6.2562 | 0.99 | 235 | 7.1861 | 0.285 | 1.1126 | 57 | 200 | 200 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 1.18.3
- Tokenizers 0.13.2