<!-- 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. -->
20230830005651
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: 0.5971
- Accuracy: 0.6346
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: 0.009
- train_batch_size: 16
- eval_batch_size: 8
- seed: 44
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 80.0
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
No log | 1.0 | 35 | 0.7436 | 0.5673 |
No log | 2.0 | 70 | 0.5847 | 0.625 |
No log | 3.0 | 105 | 0.6059 | 0.6346 |
No log | 4.0 | 140 | 1.1626 | 0.375 |
No log | 5.0 | 175 | 0.6864 | 0.6346 |
No log | 6.0 | 210 | 0.7208 | 0.5192 |
No log | 7.0 | 245 | 0.9688 | 0.375 |
No log | 8.0 | 280 | 0.5961 | 0.6346 |
No log | 9.0 | 315 | 1.0239 | 0.375 |
No log | 10.0 | 350 | 0.8614 | 0.6346 |
No log | 11.0 | 385 | 0.6080 | 0.5865 |
No log | 12.0 | 420 | 0.9951 | 0.6346 |
No log | 13.0 | 455 | 1.2255 | 0.6346 |
No log | 14.0 | 490 | 0.8444 | 0.4615 |
1.0856 | 15.0 | 525 | 0.8650 | 0.4038 |
1.0856 | 16.0 | 560 | 0.6598 | 0.6346 |
1.0856 | 17.0 | 595 | 0.8892 | 0.4135 |
1.0856 | 18.0 | 630 | 0.7671 | 0.6346 |
1.0856 | 19.0 | 665 | 0.8710 | 0.6346 |
1.0856 | 20.0 | 700 | 0.6549 | 0.6346 |
1.0856 | 21.0 | 735 | 0.6067 | 0.6346 |
1.0856 | 22.0 | 770 | 0.5914 | 0.6442 |
1.0856 | 23.0 | 805 | 0.5947 | 0.6058 |
1.0856 | 24.0 | 840 | 1.2091 | 0.6346 |
1.0856 | 25.0 | 875 | 0.6322 | 0.6346 |
1.0856 | 26.0 | 910 | 0.9031 | 0.4038 |
1.0856 | 27.0 | 945 | 0.6210 | 0.5192 |
1.0856 | 28.0 | 980 | 0.8715 | 0.3846 |
0.9189 | 29.0 | 1015 | 0.5853 | 0.625 |
0.9189 | 30.0 | 1050 | 0.6031 | 0.6346 |
0.9189 | 31.0 | 1085 | 0.8324 | 0.3654 |
0.9189 | 32.0 | 1120 | 0.6193 | 0.4904 |
0.9189 | 33.0 | 1155 | 0.8076 | 0.4327 |
0.9189 | 34.0 | 1190 | 0.6063 | 0.6346 |
0.9189 | 35.0 | 1225 | 0.7284 | 0.6346 |
0.9189 | 36.0 | 1260 | 0.5846 | 0.6442 |
0.9189 | 37.0 | 1295 | 0.5876 | 0.6538 |
0.9189 | 38.0 | 1330 | 0.6024 | 0.6346 |
0.9189 | 39.0 | 1365 | 0.6396 | 0.6346 |
0.9189 | 40.0 | 1400 | 0.6092 | 0.6154 |
0.9189 | 41.0 | 1435 | 0.8573 | 0.3654 |
0.9189 | 42.0 | 1470 | 0.8101 | 0.3654 |
0.7966 | 43.0 | 1505 | 1.0529 | 0.3654 |
0.7966 | 44.0 | 1540 | 0.5920 | 0.6058 |
0.7966 | 45.0 | 1575 | 0.6194 | 0.4808 |
0.7966 | 46.0 | 1610 | 0.9256 | 0.6346 |
0.7966 | 47.0 | 1645 | 0.6016 | 0.6442 |
0.7966 | 48.0 | 1680 | 0.6049 | 0.6346 |
0.7966 | 49.0 | 1715 | 0.5900 | 0.6346 |
0.7966 | 50.0 | 1750 | 0.6643 | 0.4327 |
0.7966 | 51.0 | 1785 | 0.8735 | 0.3654 |
0.7966 | 52.0 | 1820 | 0.6986 | 0.3654 |
0.7966 | 53.0 | 1855 | 0.6106 | 0.6346 |
0.7966 | 54.0 | 1890 | 0.8216 | 0.3654 |
0.7966 | 55.0 | 1925 | 0.6384 | 0.6346 |
0.7966 | 56.0 | 1960 | 0.6011 | 0.6346 |
0.7966 | 57.0 | 1995 | 0.7289 | 0.3654 |
0.7473 | 58.0 | 2030 | 0.6678 | 0.4231 |
0.7473 | 59.0 | 2065 | 0.6058 | 0.6346 |
0.7473 | 60.0 | 2100 | 0.6821 | 0.6346 |
0.7473 | 61.0 | 2135 | 0.6128 | 0.6346 |
0.7473 | 62.0 | 2170 | 0.7182 | 0.3846 |
0.7473 | 63.0 | 2205 | 0.5843 | 0.6635 |
0.7473 | 64.0 | 2240 | 0.6061 | 0.6346 |
0.7473 | 65.0 | 2275 | 0.5895 | 0.6346 |
0.7473 | 66.0 | 2310 | 0.5848 | 0.6635 |
0.7473 | 67.0 | 2345 | 0.6607 | 0.6346 |
0.7473 | 68.0 | 2380 | 0.6923 | 0.4038 |
0.7473 | 69.0 | 2415 | 0.6541 | 0.6346 |
0.7473 | 70.0 | 2450 | 0.5853 | 0.6538 |
0.7473 | 71.0 | 2485 | 0.6062 | 0.6346 |
0.6913 | 72.0 | 2520 | 0.5920 | 0.625 |
0.6913 | 73.0 | 2555 | 0.7628 | 0.3846 |
0.6913 | 74.0 | 2590 | 0.6675 | 0.4615 |
0.6913 | 75.0 | 2625 | 0.6010 | 0.5962 |
0.6913 | 76.0 | 2660 | 0.5902 | 0.6442 |
0.6913 | 77.0 | 2695 | 0.7894 | 0.3462 |
0.6913 | 78.0 | 2730 | 0.6302 | 0.4808 |
0.6913 | 79.0 | 2765 | 0.5948 | 0.6442 |
0.6913 | 80.0 | 2800 | 0.5971 | 0.6346 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3