<!-- 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. -->
fine-tuned-DatasetQAS-TYDI-QA-ID-with-indobert-base-uncased-with-ITTL-without-freeze-LR-1e-05
This model is a fine-tuned version of indolem/indobert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1675
- Exact Match: 61.4311
- F1: 76.0013
- Precision: 77.2642
- Recall: 81.7278
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.06
- num_epochs: 16
Training results
| Training Loss | Epoch | Step | Validation Loss | Exact Match | F1 | Precision | Recall |
|---|---|---|---|---|---|---|---|
| 6.2602 | 0.5 | 38 | 4.9592 | 0.5236 | 11.0968 | 11.1980 | 26.6188 |
| 5.556 | 0.99 | 76 | 3.0406 | 12.7400 | 25.1945 | 25.6795 | 39.6592 |
| 3.3395 | 1.5 | 114 | 2.4880 | 21.1169 | 34.7076 | 32.9184 | 52.2869 |
| 2.4731 | 1.99 | 152 | 2.2257 | 27.3997 | 40.0066 | 39.3514 | 53.5252 |
| 2.4731 | 2.5 | 190 | 2.0431 | 32.6353 | 44.5789 | 44.7211 | 55.0020 |
| 2.162 | 2.99 | 228 | 1.8362 | 38.9180 | 50.4876 | 50.7087 | 59.7434 |
| 1.8755 | 3.5 | 266 | 1.6441 | 43.9791 | 56.8266 | 57.4538 | 65.5751 |
| 1.5888 | 3.99 | 304 | 1.4664 | 52.0070 | 63.9616 | 65.2046 | 70.0798 |
| 1.5888 | 4.5 | 342 | 1.3509 | 54.4503 | 68.6979 | 70.2140 | 76.0813 |
| 1.333 | 4.99 | 380 | 1.2571 | 54.7993 | 68.9857 | 70.8728 | 75.8745 |
| 1.2051 | 5.5 | 418 | 1.2440 | 56.5445 | 70.2921 | 72.4571 | 75.9313 |
| 1.0522 | 5.99 | 456 | 1.1808 | 57.5916 | 72.1230 | 73.6246 | 78.8092 |
| 1.0522 | 6.5 | 494 | 1.1575 | 58.9878 | 73.1594 | 74.9064 | 79.2545 |
| 0.9584 | 6.99 | 532 | 1.1553 | 58.9878 | 73.5139 | 75.2615 | 79.3901 |
| 0.9006 | 7.5 | 570 | 1.1112 | 60.0349 | 74.5273 | 75.8170 | 81.1555 |
| 0.8102 | 7.99 | 608 | 1.1164 | 59.8604 | 74.5013 | 76.2748 | 80.2875 |
| 0.8102 | 8.5 | 646 | 1.1371 | 60.0349 | 74.2469 | 75.9082 | 79.9186 |
| 0.773 | 8.99 | 684 | 1.1410 | 60.7330 | 74.9095 | 76.7045 | 80.9178 |
| 0.7482 | 9.5 | 722 | 1.1307 | 60.3839 | 74.7594 | 76.8954 | 80.6364 |
| 0.6878 | 9.99 | 760 | 1.1219 | 61.0820 | 74.9064 | 76.4266 | 81.4087 |
| 0.6878 | 10.5 | 798 | 1.1362 | 62.1291 | 76.5097 | 77.5924 | 82.8049 |
| 0.6401 | 10.99 | 836 | 1.1266 | 61.0820 | 75.8874 | 77.0263 | 81.7467 |
| 0.634 | 11.5 | 874 | 1.1570 | 61.7801 | 75.9638 | 77.5661 | 80.8536 |
| 0.5856 | 11.99 | 912 | 1.1675 | 61.4311 | 76.0013 | 77.2642 | 81.7278 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu117
- Datasets 2.2.0
- Tokenizers 0.13.2
