<!-- 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. -->
20230824084116
This model is a fine-tuned version of bert-large-cased on the super_glue dataset. It achieves the following results on the evaluation set:
- Loss: 0.6747
- Accuracy: 0.7329
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: 0.003
- train_batch_size: 4
- eval_batch_size: 8
- seed: 11
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 60.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.0144 | 1.0 | 623 | 1.2485 | 0.4729 |
| 0.8551 | 2.0 | 1246 | 0.7296 | 0.5415 |
| 0.9621 | 3.0 | 1869 | 1.3927 | 0.4729 |
| 0.8648 | 4.0 | 2492 | 0.6253 | 0.6173 |
| 0.8311 | 5.0 | 3115 | 0.6509 | 0.6606 |
| 0.8365 | 6.0 | 3738 | 0.6018 | 0.6895 |
| 0.772 | 7.0 | 4361 | 0.7314 | 0.6751 |
| 0.7306 | 8.0 | 4984 | 1.0930 | 0.5957 |
| 0.763 | 9.0 | 5607 | 0.7093 | 0.7076 |
| 0.6931 | 10.0 | 6230 | 0.6302 | 0.6968 |
| 0.6465 | 11.0 | 6853 | 1.1188 | 0.5776 |
| 0.6503 | 12.0 | 7476 | 0.6957 | 0.7112 |
| 0.6657 | 13.0 | 8099 | 0.6470 | 0.7112 |
| 0.6315 | 14.0 | 8722 | 0.7099 | 0.7112 |
| 0.5491 | 15.0 | 9345 | 0.5178 | 0.7184 |
| 0.4908 | 16.0 | 9968 | 0.6282 | 0.7365 |
| 0.4742 | 17.0 | 10591 | 0.6553 | 0.7256 |
| 0.4653 | 18.0 | 11214 | 0.5637 | 0.7112 |
| 0.492 | 19.0 | 11837 | 0.5870 | 0.7184 |
| 0.4519 | 20.0 | 12460 | 0.8201 | 0.7292 |
| 0.4198 | 21.0 | 13083 | 0.6294 | 0.7365 |
| 0.403 | 22.0 | 13706 | 0.6998 | 0.7220 |
| 0.4017 | 23.0 | 14329 | 0.8424 | 0.7220 |
| 0.368 | 24.0 | 14952 | 0.6179 | 0.7401 |
| 0.3514 | 25.0 | 15575 | 0.6303 | 0.7256 |
| 0.3458 | 26.0 | 16198 | 0.6241 | 0.7292 |
| 0.3488 | 27.0 | 16821 | 0.6348 | 0.7365 |
| 0.33 | 28.0 | 17444 | 0.6663 | 0.7292 |
| 0.3133 | 29.0 | 18067 | 0.6231 | 0.7437 |
| 0.3108 | 30.0 | 18690 | 0.6940 | 0.7220 |
| 0.3156 | 31.0 | 19313 | 0.7685 | 0.7256 |
| 0.2887 | 32.0 | 19936 | 0.5912 | 0.7365 |
| 0.2871 | 33.0 | 20559 | 0.6539 | 0.7401 |
| 0.2835 | 34.0 | 21182 | 0.7319 | 0.7292 |
| 0.2587 | 35.0 | 21805 | 0.6106 | 0.7365 |
| 0.2767 | 36.0 | 22428 | 0.6255 | 0.7329 |
| 0.2621 | 37.0 | 23051 | 0.7181 | 0.7329 |
| 0.2733 | 38.0 | 23674 | 0.6841 | 0.7365 |
| 0.2473 | 39.0 | 24297 | 0.7042 | 0.7329 |
| 0.2467 | 40.0 | 24920 | 0.6123 | 0.7329 |
| 0.2357 | 41.0 | 25543 | 0.6681 | 0.7365 |
| 0.2333 | 42.0 | 26166 | 0.7094 | 0.7292 |
| 0.2387 | 43.0 | 26789 | 0.6546 | 0.7365 |
| 0.2248 | 44.0 | 27412 | 0.7021 | 0.7329 |
| 0.2271 | 45.0 | 28035 | 0.6913 | 0.7545 |
| 0.2288 | 46.0 | 28658 | 0.6855 | 0.7365 |
| 0.2159 | 47.0 | 29281 | 0.6495 | 0.7401 |
| 0.2107 | 48.0 | 29904 | 0.6568 | 0.7292 |
| 0.2204 | 49.0 | 30527 | 0.7337 | 0.7329 |
| 0.2038 | 50.0 | 31150 | 0.6391 | 0.7365 |
| 0.2183 | 51.0 | 31773 | 0.6593 | 0.7437 |
| 0.2041 | 52.0 | 32396 | 0.6518 | 0.7220 |
| 0.2107 | 53.0 | 33019 | 0.6677 | 0.7256 |
| 0.2076 | 54.0 | 33642 | 0.6716 | 0.7292 |
| 0.1946 | 55.0 | 34265 | 0.6957 | 0.7256 |
| 0.1974 | 56.0 | 34888 | 0.6858 | 0.7256 |
| 0.2047 | 57.0 | 35511 | 0.6721 | 0.7329 |
| 0.2001 | 58.0 | 36134 | 0.6747 | 0.7365 |
| 0.1899 | 59.0 | 36757 | 0.6842 | 0.7329 |
| 0.1872 | 60.0 | 37380 | 0.6747 | 0.7329 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
