<!-- 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-IDK-MRC-with-indobert-large-p2-with-ITTL-with-freeze-LR-1e-05
This model is a fine-tuned version of indobenchmark/indobert-large-p2 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.2708
- Exact Match: 52.7487
- F1: 60.8071
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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 32
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Exact Match | F1 |
---|---|---|---|---|---|
6.4745 | 0.49 | 36 | 2.5724 | 35.6021 | 37.8405 |
3.5197 | 0.98 | 72 | 1.9912 | 28.0105 | 35.4278 |
2.1756 | 1.48 | 108 | 1.6669 | 35.7330 | 43.0612 |
2.1756 | 1.97 | 144 | 1.5047 | 39.3979 | 46.1664 |
1.6725 | 2.46 | 180 | 1.3222 | 45.9424 | 52.9355 |
1.336 | 2.95 | 216 | 1.3205 | 44.1099 | 51.6851 |
1.176 | 3.45 | 252 | 1.2526 | 47.5131 | 55.3298 |
1.176 | 3.94 | 288 | 1.2778 | 47.3822 | 54.7110 |
1.1089 | 4.44 | 324 | 1.2291 | 49.8691 | 57.2303 |
0.967 | 4.93 | 360 | 1.1944 | 52.4869 | 60.2202 |
0.967 | 5.42 | 396 | 1.2122 | 53.7958 | 61.3033 |
0.9202 | 5.91 | 432 | 1.2348 | 54.0576 | 61.6263 |
0.8719 | 6.41 | 468 | 1.2206 | 55.2356 | 62.9267 |
0.8205 | 6.9 | 504 | 1.2472 | 53.9267 | 61.6359 |
0.8205 | 7.4 | 540 | 1.2764 | 52.3560 | 60.2681 |
0.7907 | 7.89 | 576 | 1.2382 | 55.3665 | 63.0145 |
0.7533 | 8.38 | 612 | 1.2812 | 52.4869 | 60.4214 |
0.7533 | 8.87 | 648 | 1.2474 | 53.1414 | 60.6338 |
0.7345 | 9.37 | 684 | 1.2708 | 52.7487 | 60.8071 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu117
- Datasets 2.2.0
- Tokenizers 0.13.2