20230826161130
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: 0.1582
- Accuracy: 0.39
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: 0.02
- train_batch_size: 16
- eval_batch_size: 8
- seed: 11
- 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 | 25 | 0.6392 | 0.43 |
No log | 2.0 | 50 | 0.1729 | 0.41 |
No log | 3.0 | 75 | 0.1658 | 0.61 |
No log | 4.0 | 100 | 0.1579 | 0.57 |
No log | 5.0 | 125 | 0.1678 | 0.4 |
No log | 6.0 | 150 | 0.1583 | 0.55 |
No log | 7.0 | 175 | 0.1650 | 0.6 |
No log | 8.0 | 200 | 0.1643 | 0.62 |
No log | 9.0 | 225 | 0.1594 | 0.48 |
No log | 10.0 | 250 | 0.1572 | 0.61 |
No log | 11.0 | 275 | 0.1660 | 0.4 |
No log | 12.0 | 300 | 0.1570 | 0.63 |
No log | 13.0 | 325 | 0.1589 | 0.51 |
No log | 14.0 | 350 | 0.1581 | 0.42 |
No log | 15.0 | 375 | 0.1582 | 0.5 |
No log | 16.0 | 400 | 0.1576 | 0.53 |
No log | 17.0 | 425 | 0.1580 | 0.52 |
No log | 18.0 | 450 | 0.1581 | 0.55 |
No log | 19.0 | 475 | 0.1583 | 0.45 |
0.621 | 20.0 | 500 | 0.1606 | 0.52 |
0.621 | 21.0 | 525 | 0.1583 | 0.52 |
0.621 | 22.0 | 550 | 0.1573 | 0.49 |
0.621 | 23.0 | 575 | 0.1582 | 0.43 |
0.621 | 24.0 | 600 | 0.1581 | 0.53 |
0.621 | 25.0 | 625 | 0.1582 | 0.49 |
0.621 | 26.0 | 650 | 0.1582 | 0.5 |
0.621 | 27.0 | 675 | 0.1583 | 0.53 |
0.621 | 28.0 | 700 | 0.1586 | 0.47 |
0.621 | 29.0 | 725 | 0.1585 | 0.48 |
0.621 | 30.0 | 750 | 0.1584 | 0.46 |
0.621 | 31.0 | 775 | 0.1582 | 0.55 |
0.621 | 32.0 | 800 | 0.1582 | 0.53 |
0.621 | 33.0 | 825 | 0.1583 | 0.51 |
0.621 | 34.0 | 850 | 0.1585 | 0.39 |
0.621 | 35.0 | 875 | 0.1582 | 0.69 |
0.621 | 36.0 | 900 | 0.1583 | 0.48 |
0.621 | 37.0 | 925 | 0.1582 | 0.61 |
0.621 | 38.0 | 950 | 0.1580 | 0.63 |
0.621 | 39.0 | 975 | 0.1581 | 0.47 |
0.4969 | 40.0 | 1000 | 0.1582 | 0.49 |
0.4969 | 41.0 | 1025 | 0.1583 | 0.49 |
0.4969 | 42.0 | 1050 | 0.1583 | 0.47 |
0.4969 | 43.0 | 1075 | 0.1581 | 0.52 |
0.4969 | 44.0 | 1100 | 0.1584 | 0.47 |
0.4969 | 45.0 | 1125 | 0.1584 | 0.35 |
0.4969 | 46.0 | 1150 | 0.1582 | 0.56 |
0.4969 | 47.0 | 1175 | 0.1582 | 0.54 |
0.4969 | 48.0 | 1200 | 0.1582 | 0.53 |
0.4969 | 49.0 | 1225 | 0.1582 | 0.56 |
0.4969 | 50.0 | 1250 | 0.1582 | 0.54 |
0.4969 | 51.0 | 1275 | 0.1582 | 0.57 |
0.4969 | 52.0 | 1300 | 0.1582 | 0.52 |
0.4969 | 53.0 | 1325 | 0.1581 | 0.59 |
0.4969 | 54.0 | 1350 | 0.1582 | 0.55 |
0.4969 | 55.0 | 1375 | 0.1585 | 0.41 |
0.4969 | 56.0 | 1400 | 0.1584 | 0.45 |
0.4969 | 57.0 | 1425 | 0.1583 | 0.54 |
0.4969 | 58.0 | 1450 | 0.1583 | 0.41 |
0.4969 | 59.0 | 1475 | 0.1583 | 0.42 |
0.4428 | 60.0 | 1500 | 0.1583 | 0.4 |
0.4428 | 61.0 | 1525 | 0.1583 | 0.59 |
0.4428 | 62.0 | 1550 | 0.1582 | 0.65 |
0.4428 | 63.0 | 1575 | 0.1581 | 0.64 |
0.4428 | 64.0 | 1600 | 0.1581 | 0.59 |
0.4428 | 65.0 | 1625 | 0.1583 | 0.42 |
0.4428 | 66.0 | 1650 | 0.1582 | 0.5 |
0.4428 | 67.0 | 1675 | 0.1583 | 0.43 |
0.4428 | 68.0 | 1700 | 0.1584 | 0.39 |
0.4428 | 69.0 | 1725 | 0.1583 | 0.5 |
0.4428 | 70.0 | 1750 | 0.1583 | 0.49 |
0.4428 | 71.0 | 1775 | 0.1583 | 0.48 |
0.4428 | 72.0 | 1800 | 0.1584 | 0.29 |
0.4428 | 73.0 | 1825 | 0.1583 | 0.4 |
0.4428 | 74.0 | 1850 | 0.1582 | 0.59 |
0.4428 | 75.0 | 1875 | 0.1582 | 0.59 |
0.4428 | 76.0 | 1900 | 0.1582 | 0.53 |
0.4428 | 77.0 | 1925 | 0.1583 | 0.33 |
0.4428 | 78.0 | 1950 | 0.1583 | 0.35 |
0.4428 | 79.0 | 1975 | 0.1583 | 0.36 |
0.4082 | 80.0 | 2000 | 0.1582 | 0.39 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
- Downloads last month
- 0