run_3
This model is a fine-tuned version of bert-base-uncased on the wikitext dataset. It achieves the following results on the evaluation set:
- Loss: 7.1422
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: 0.005
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 7.8139 | 0.07 | 50 | 7.3922 |
| 7.3173 | 0.14 | 100 | 7.2946 |
| 7.2587 | 0.21 | 150 | 7.2339 |
| 7.2122 | 0.27 | 200 | 7.2167 |
| 7.1908 | 0.34 | 250 | 7.1945 |
| 7.171 | 0.41 | 300 | 7.1875 |
| 7.2054 | 0.48 | 350 | 7.1893 |
| 7.1899 | 0.55 | 400 | 7.1889 |
| 7.1839 | 0.62 | 450 | 7.1801 |
| 7.1571 | 0.69 | 500 | 7.1759 |
| 7.1577 | 0.75 | 550 | 7.1725 |
| 7.1799 | 0.82 | 600 | 7.1757 |
| 7.1698 | 0.89 | 650 | 7.1715 |
| 7.1705 | 0.96 | 700 | 7.1651 |
| 7.1712 | 1.03 | 750 | 7.1677 |
| 7.1418 | 1.1 | 800 | 7.1699 |
| 7.1692 | 1.17 | 850 | 7.1659 |
| 7.1376 | 1.24 | 900 | 7.1656 |
| 7.1703 | 1.3 | 950 | 7.1643 |
| 7.1534 | 1.37 | 1000 | 7.1676 |
| 7.1445 | 1.44 | 1050 | 7.1607 |
| 7.1552 | 1.51 | 1100 | 7.1596 |
| 7.1475 | 1.58 | 1150 | 7.1599 |
| 7.1401 | 1.65 | 1200 | 7.1593 |
| 7.161 | 1.72 | 1250 | 7.1606 |
| 7.1513 | 1.78 | 1300 | 7.1564 |
| 7.1465 | 1.85 | 1350 | 7.1548 |
| 7.1603 | 1.92 | 1400 | 7.1529 |
| 7.1203 | 1.99 | 1450 | 7.1533 |
| 7.1308 | 2.06 | 1500 | 7.1546 |
| 7.1244 | 2.13 | 1550 | 7.1546 |
| 7.1437 | 2.2 | 1600 | 7.1561 |
| 7.1618 | 2.26 | 1650 | 7.1517 |
| 7.1502 | 2.33 | 1700 | 7.1519 |
| 7.146 | 2.4 | 1750 | 7.1514 |
| 7.1088 | 2.47 | 1800 | 7.1520 |
| 7.1335 | 2.54 | 1850 | 7.1483 |
| 7.1388 | 2.61 | 1900 | 7.1472 |
| 7.1502 | 2.68 | 1950 | 7.1470 |
| 7.1511 | 2.75 | 2000 | 7.1479 |
| 7.1288 | 2.81 | 2050 | 7.1506 |
| 7.1416 | 2.88 | 2100 | 7.1488 |
| 7.1568 | 2.95 | 2150 | 7.1512 |
| 7.133 | 3.02 | 2200 | 7.1497 |
| 7.1178 | 3.09 | 2250 | 7.1501 |
| 7.1482 | 3.16 | 2300 | 7.1506 |
| 7.1242 | 3.23 | 2350 | 7.1504 |
| 7.1181 | 3.29 | 2400 | 7.1497 |
| 7.1133 | 3.36 | 2450 | 7.1495 |
| 7.1199 | 3.43 | 2500 | 7.1468 |
| 7.146 | 3.5 | 2550 | 7.1467 |
| 7.1284 | 3.57 | 2600 | 7.1455 |
| 7.1356 | 3.64 | 2650 | 7.1464 |
| 7.1372 | 3.71 | 2700 | 7.1445 |
| 7.1307 | 3.77 | 2750 | 7.1429 |
| 7.1407 | 3.84 | 2800 | 7.1427 |
| 7.126 | 3.91 | 2850 | 7.1426 |
| 7.1288 | 3.98 | 2900 | 7.1425 |
| 7.1223 | 4.05 | 2950 | 7.1428 |
| 7.1169 | 4.12 | 3000 | 7.1429 |
| 7.139 | 4.19 | 3050 | 7.1441 |
| 7.1231 | 4.26 | 3100 | 7.1433 |
| 7.1114 | 4.32 | 3150 | 7.1429 |
| 7.1204 | 4.39 | 3200 | 7.1429 |
| 7.0994 | 4.46 | 3250 | 7.1430 |
| 7.1039 | 4.53 | 3300 | 7.1434 |
| 7.1489 | 4.6 | 3350 | 7.1428 |
| 7.1315 | 4.67 | 3400 | 7.1426 |
| 7.1173 | 4.74 | 3450 | 7.1426 |
| 7.1241 | 4.8 | 3500 | 7.1428 |
| 7.1001 | 4.87 | 3550 | 7.1427 |
| 7.137 | 4.94 | 3600 | 7.1422 |
Framework versions
- Transformers 4.33.1
- Pytorch 1.12.1
- Datasets 2.14.6
- Tokenizers 0.13.3
- Downloads last month
- 9
Model tree for temporary0-0name/run_3
Base model
google-bert/bert-base-uncased