<!-- 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. -->
20230829194726
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: 0.3334
- 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: 0.003
- 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 | 0.4523 | 0.6154 |
No log | 2.0 | 70 | 0.3765 | 0.4808 |
No log | 3.0 | 105 | 0.3393 | 0.6346 |
No log | 4.0 | 140 | 0.3913 | 0.6346 |
No log | 5.0 | 175 | 0.3666 | 0.4808 |
No log | 6.0 | 210 | 0.3610 | 0.4808 |
No log | 7.0 | 245 | 0.3798 | 0.4135 |
No log | 8.0 | 280 | 0.3573 | 0.6346 |
No log | 9.0 | 315 | 0.3413 | 0.5192 |
No log | 10.0 | 350 | 0.7085 | 0.4038 |
No log | 11.0 | 385 | 0.3432 | 0.5865 |
No log | 12.0 | 420 | 0.4053 | 0.6058 |
No log | 13.0 | 455 | 0.4454 | 0.625 |
No log | 14.0 | 490 | 0.3583 | 0.6346 |
0.4577 | 15.0 | 525 | 0.3559 | 0.6346 |
0.4577 | 16.0 | 560 | 0.4828 | 0.4038 |
0.4577 | 17.0 | 595 | 0.3425 | 0.6538 |
0.4577 | 18.0 | 630 | 0.4212 | 0.4135 |
0.4577 | 19.0 | 665 | 0.6152 | 0.4038 |
0.4577 | 20.0 | 700 | 0.4090 | 0.6346 |
0.4577 | 21.0 | 735 | 0.4476 | 0.4038 |
0.4577 | 22.0 | 770 | 0.3392 | 0.6346 |
0.4577 | 23.0 | 805 | 0.5753 | 0.4038 |
0.4577 | 24.0 | 840 | 0.3580 | 0.4904 |
0.4577 | 25.0 | 875 | 0.4477 | 0.3846 |
0.4577 | 26.0 | 910 | 0.3906 | 0.4519 |
0.4577 | 27.0 | 945 | 0.3437 | 0.5192 |
0.4577 | 28.0 | 980 | 0.3392 | 0.6154 |
0.4318 | 29.0 | 1015 | 0.3824 | 0.5192 |
0.4318 | 30.0 | 1050 | 0.3307 | 0.6346 |
0.4318 | 31.0 | 1085 | 0.4567 | 0.4327 |
0.4318 | 32.0 | 1120 | 0.3840 | 0.4519 |
0.4318 | 33.0 | 1155 | 0.4854 | 0.6346 |
0.4318 | 34.0 | 1190 | 0.3319 | 0.625 |
0.4318 | 35.0 | 1225 | 0.3613 | 0.4327 |
0.4318 | 36.0 | 1260 | 0.3426 | 0.5769 |
0.4318 | 37.0 | 1295 | 0.5563 | 0.6346 |
0.4318 | 38.0 | 1330 | 0.3468 | 0.6346 |
0.4318 | 39.0 | 1365 | 0.3961 | 0.6346 |
0.4318 | 40.0 | 1400 | 0.3660 | 0.5096 |
0.4318 | 41.0 | 1435 | 0.4189 | 0.4135 |
0.4318 | 42.0 | 1470 | 0.4040 | 0.3942 |
0.4014 | 43.0 | 1505 | 0.3298 | 0.6442 |
0.4014 | 44.0 | 1540 | 0.3314 | 0.6346 |
0.4014 | 45.0 | 1575 | 0.3638 | 0.4712 |
0.4014 | 46.0 | 1610 | 0.3499 | 0.5673 |
0.4014 | 47.0 | 1645 | 0.3331 | 0.6538 |
0.4014 | 48.0 | 1680 | 0.3303 | 0.6346 |
0.4014 | 49.0 | 1715 | 0.3490 | 0.6346 |
0.4014 | 50.0 | 1750 | 0.3402 | 0.5673 |
0.4014 | 51.0 | 1785 | 0.3679 | 0.5192 |
0.4014 | 52.0 | 1820 | 0.3438 | 0.6442 |
0.4014 | 53.0 | 1855 | 0.3319 | 0.6538 |
0.4014 | 54.0 | 1890 | 0.3667 | 0.4231 |
0.4014 | 55.0 | 1925 | 0.3415 | 0.5385 |
0.4014 | 56.0 | 1960 | 0.3373 | 0.6346 |
0.4014 | 57.0 | 1995 | 0.3637 | 0.4904 |
0.3745 | 58.0 | 2030 | 0.3341 | 0.6346 |
0.3745 | 59.0 | 2065 | 0.4288 | 0.3846 |
0.3745 | 60.0 | 2100 | 0.3569 | 0.625 |
0.3745 | 61.0 | 2135 | 0.3497 | 0.6346 |
0.3745 | 62.0 | 2170 | 0.3318 | 0.6538 |
0.3745 | 63.0 | 2205 | 0.3324 | 0.6538 |
0.3745 | 64.0 | 2240 | 0.3794 | 0.6346 |
0.3745 | 65.0 | 2275 | 0.3343 | 0.6731 |
0.3745 | 66.0 | 2310 | 0.3367 | 0.6346 |
0.3745 | 67.0 | 2345 | 0.3404 | 0.6346 |
0.3745 | 68.0 | 2380 | 0.3412 | 0.5769 |
0.3745 | 69.0 | 2415 | 0.3373 | 0.6346 |
0.3745 | 70.0 | 2450 | 0.3498 | 0.6346 |
0.3745 | 71.0 | 2485 | 0.3315 | 0.6346 |
0.3659 | 72.0 | 2520 | 0.3408 | 0.5481 |
0.3659 | 73.0 | 2555 | 0.3337 | 0.6346 |
0.3659 | 74.0 | 2590 | 0.3481 | 0.4615 |
0.3659 | 75.0 | 2625 | 0.3390 | 0.625 |
0.3659 | 76.0 | 2660 | 0.3339 | 0.6538 |
0.3659 | 77.0 | 2695 | 0.3492 | 0.4904 |
0.3659 | 78.0 | 2730 | 0.3397 | 0.5769 |
0.3659 | 79.0 | 2765 | 0.3342 | 0.6442 |
0.3659 | 80.0 | 2800 | 0.3334 | 0.6346 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3