SequentialFinetuningFromFolder
This model was trained from scratch on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.2211
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 100
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
No log | 1.0 | 21 | 1.0511 |
No log | 2.0 | 42 | 1.0333 |
No log | 3.0 | 63 | 1.0322 |
No log | 4.0 | 84 | 1.0051 |
No log | 5.0 | 105 | 0.9618 |
No log | 6.0 | 126 | 0.9609 |
No log | 7.0 | 147 | 0.9380 |
No log | 8.0 | 168 | 0.9073 |
No log | 9.0 | 189 | 0.9030 |
No log | 10.0 | 210 | 0.8811 |
No log | 11.0 | 231 | 0.8631 |
No log | 12.0 | 252 | 0.8400 |
No log | 13.0 | 273 | 0.8242 |
No log | 14.0 | 294 | 0.8133 |
No log | 15.0 | 315 | 0.7857 |
No log | 16.0 | 336 | 0.7744 |
No log | 17.0 | 357 | 0.7548 |
No log | 18.0 | 378 | 0.7549 |
No log | 19.0 | 399 | 0.7296 |
No log | 20.0 | 420 | 0.7169 |
No log | 21.0 | 441 | 0.7140 |
No log | 22.0 | 462 | 0.7026 |
No log | 23.0 | 483 | 0.7175 |
1.0179 | 24.0 | 504 | 0.6831 |
1.0179 | 25.0 | 525 | 0.6882 |
1.0179 | 26.0 | 546 | 0.6455 |
1.0179 | 27.0 | 567 | 0.6317 |
1.0179 | 28.0 | 588 | 0.6396 |
1.0179 | 29.0 | 609 | 0.6132 |
1.0179 | 30.0 | 630 | 0.5885 |
1.0179 | 31.0 | 651 | 0.5800 |
1.0179 | 32.0 | 672 | 0.5700 |
1.0179 | 33.0 | 693 | 0.5673 |
1.0179 | 34.0 | 714 | 0.5524 |
1.0179 | 35.0 | 735 | 0.5310 |
1.0179 | 36.0 | 756 | 0.5249 |
1.0179 | 37.0 | 777 | 0.5148 |
1.0179 | 38.0 | 798 | 0.5246 |
1.0179 | 39.0 | 819 | 0.4967 |
1.0179 | 40.0 | 840 | 0.4841 |
1.0179 | 41.0 | 861 | 0.4822 |
1.0179 | 42.0 | 882 | 0.4694 |
1.0179 | 43.0 | 903 | 0.4598 |
1.0179 | 44.0 | 924 | 0.4503 |
1.0179 | 45.0 | 945 | 0.4428 |
1.0179 | 46.0 | 966 | 0.4243 |
1.0179 | 47.0 | 987 | 0.4163 |
0.7797 | 48.0 | 1008 | 0.4187 |
0.7797 | 49.0 | 1029 | 0.4110 |
0.7797 | 50.0 | 1050 | 0.4013 |
0.7797 | 51.0 | 1071 | 0.4099 |
0.7797 | 52.0 | 1092 | 0.3870 |
0.7797 | 53.0 | 1113 | 0.3818 |
0.7797 | 54.0 | 1134 | 0.3783 |
0.7797 | 55.0 | 1155 | 0.3621 |
0.7797 | 56.0 | 1176 | 0.3591 |
0.7797 | 57.0 | 1197 | 0.3608 |
0.7797 | 58.0 | 1218 | 0.3447 |
0.7797 | 59.0 | 1239 | 0.3444 |
0.7797 | 60.0 | 1260 | 0.3390 |
0.7797 | 61.0 | 1281 | 0.3310 |
0.7797 | 62.0 | 1302 | 0.3201 |
0.7797 | 63.0 | 1323 | 0.3250 |
0.7797 | 64.0 | 1344 | 0.3115 |
0.7797 | 65.0 | 1365 | 0.3015 |
0.7797 | 66.0 | 1386 | 0.3014 |
0.7797 | 67.0 | 1407 | 0.3081 |
0.7797 | 68.0 | 1428 | 0.2892 |
0.7797 | 69.0 | 1449 | 0.3034 |
0.7797 | 70.0 | 1470 | 0.2828 |
0.7797 | 71.0 | 1491 | 0.2790 |
0.6123 | 72.0 | 1512 | 0.2727 |
0.6123 | 73.0 | 1533 | 0.2809 |
0.6123 | 74.0 | 1554 | 0.2694 |
0.6123 | 75.0 | 1575 | 0.2636 |
0.6123 | 76.0 | 1596 | 0.2613 |
0.6123 | 77.0 | 1617 | 0.2557 |
0.6123 | 78.0 | 1638 | 0.2529 |
0.6123 | 79.0 | 1659 | 0.2575 |
0.6123 | 80.0 | 1680 | 0.2539 |
0.6123 | 81.0 | 1701 | 0.2540 |
0.6123 | 82.0 | 1722 | 0.2423 |
0.6123 | 83.0 | 1743 | 0.2406 |
0.6123 | 84.0 | 1764 | 0.2383 |
0.6123 | 85.0 | 1785 | 0.2358 |
0.6123 | 86.0 | 1806 | 0.2371 |
0.6123 | 87.0 | 1827 | 0.2352 |
0.6123 | 88.0 | 1848 | 0.2335 |
0.6123 | 89.0 | 1869 | 0.2297 |
0.6123 | 90.0 | 1890 | 0.2305 |
0.6123 | 91.0 | 1911 | 0.2264 |
0.6123 | 92.0 | 1932 | 0.2255 |
0.6123 | 93.0 | 1953 | 0.2273 |
0.6123 | 94.0 | 1974 | 0.2220 |
0.6123 | 95.0 | 1995 | 0.2240 |
0.5063 | 96.0 | 2016 | 0.2214 |
0.5063 | 97.0 | 2037 | 0.2219 |
0.5063 | 98.0 | 2058 | 0.2202 |
0.5063 | 99.0 | 2079 | 0.2211 |
0.5063 | 100.0 | 2100 | 0.2211 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 5
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.