<!-- 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. -->
scenario-kd-from-post-finetune-gold-silver-div-2-8000-data-smsa-model-haryoaw-sc
This model is a fine-tuned version of haryoaw/scenario-normal-finetune-clf-data-smsa-model-xlm-roberta-base on the smsa dataset. It achieves the following results on the evaluation set:
- Loss: 0.8265
- Accuracy: 0.9198
- F1: 0.8871
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: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6969
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
---|---|---|---|---|---|
No log | 0.4 | 100 | 1.5313 | 0.8960 | 0.8525 |
No log | 0.8 | 200 | 1.5166 | 0.8984 | 0.8531 |
No log | 1.2 | 300 | 1.4584 | 0.8968 | 0.8390 |
No log | 1.6 | 400 | 1.1168 | 0.9175 | 0.8830 |
1.3988 | 2.0 | 500 | 1.5231 | 0.8952 | 0.8554 |
1.3988 | 2.4 | 600 | 1.1988 | 0.9111 | 0.8775 |
1.3988 | 2.8 | 700 | 1.1218 | 0.9103 | 0.8743 |
1.3988 | 3.2 | 800 | 1.0986 | 0.9143 | 0.8766 |
1.3988 | 3.6 | 900 | 1.1321 | 0.9167 | 0.8737 |
0.6891 | 4.0 | 1000 | 1.0806 | 0.9103 | 0.8752 |
0.6891 | 4.4 | 1100 | 0.9855 | 0.9190 | 0.8878 |
0.6891 | 4.8 | 1200 | 1.1468 | 0.9127 | 0.8784 |
0.6891 | 5.2 | 1300 | 1.1624 | 0.9071 | 0.8699 |
0.6891 | 5.6 | 1400 | 0.8455 | 0.9183 | 0.8836 |
0.5436 | 6.0 | 1500 | 0.9643 | 0.9183 | 0.8825 |
0.5436 | 6.4 | 1600 | 0.9413 | 0.9087 | 0.8734 |
0.5436 | 6.8 | 1700 | 0.9458 | 0.9190 | 0.8890 |
0.5436 | 7.2 | 1800 | 0.9271 | 0.9127 | 0.8769 |
0.5436 | 7.6 | 1900 | 0.9143 | 0.9111 | 0.8754 |
0.4723 | 8.0 | 2000 | 0.9558 | 0.9151 | 0.8741 |
0.4723 | 8.4 | 2100 | 1.0649 | 0.9198 | 0.8807 |
0.4723 | 8.8 | 2200 | 0.8562 | 0.9183 | 0.8829 |
0.4723 | 9.2 | 2300 | 0.8350 | 0.9159 | 0.8828 |
0.4723 | 9.6 | 2400 | 0.8073 | 0.9159 | 0.8823 |
0.4263 | 10.0 | 2500 | 0.9557 | 0.9159 | 0.8791 |
0.4263 | 10.4 | 2600 | 0.9259 | 0.9159 | 0.8838 |
0.4263 | 10.8 | 2700 | 0.8317 | 0.9230 | 0.8917 |
0.4263 | 11.2 | 2800 | 0.8210 | 0.9183 | 0.8833 |
0.4263 | 11.6 | 2900 | 0.9217 | 0.9230 | 0.8881 |
0.3875 | 12.0 | 3000 | 0.7284 | 0.9183 | 0.8858 |
0.3875 | 12.4 | 3100 | 0.9787 | 0.9190 | 0.8848 |
0.3875 | 12.8 | 3200 | 0.8598 | 0.9159 | 0.8732 |
0.3875 | 13.2 | 3300 | 0.8222 | 0.9175 | 0.8805 |
0.3875 | 13.6 | 3400 | 0.8813 | 0.9222 | 0.8919 |
0.351 | 14.0 | 3500 | 0.9317 | 0.9230 | 0.8848 |
0.351 | 14.4 | 3600 | 0.9120 | 0.9159 | 0.8823 |
0.351 | 14.8 | 3700 | 0.7626 | 0.9270 | 0.8922 |
0.351 | 15.2 | 3800 | 0.8003 | 0.9206 | 0.8829 |
0.351 | 15.6 | 3900 | 0.8165 | 0.9238 | 0.8922 |
0.3285 | 16.0 | 4000 | 0.8977 | 0.9214 | 0.8863 |
0.3285 | 16.4 | 4100 | 0.8437 | 0.9151 | 0.8724 |
0.3285 | 16.8 | 4200 | 0.8861 | 0.9190 | 0.8884 |
0.3285 | 17.2 | 4300 | 0.7916 | 0.9254 | 0.8953 |
0.3285 | 17.6 | 4400 | 0.8829 | 0.9230 | 0.8915 |
0.3151 | 18.0 | 4500 | 0.8952 | 0.9183 | 0.8866 |
0.3151 | 18.4 | 4600 | 0.8962 | 0.9103 | 0.8746 |
0.3151 | 18.8 | 4700 | 0.8192 | 0.9190 | 0.8831 |
0.3151 | 19.2 | 4800 | 0.8661 | 0.9151 | 0.8860 |
0.3151 | 19.6 | 4900 | 0.8064 | 0.9198 | 0.8853 |
0.2938 | 20.0 | 5000 | 0.7614 | 0.9214 | 0.8891 |
0.2938 | 20.4 | 5100 | 0.8073 | 0.9190 | 0.8840 |
0.2938 | 20.8 | 5200 | 0.9247 | 0.9063 | 0.8653 |
0.2938 | 21.2 | 5300 | 0.7755 | 0.9159 | 0.8778 |
0.2938 | 21.6 | 5400 | 0.7942 | 0.9190 | 0.8786 |
0.2859 | 22.0 | 5500 | 0.7626 | 0.9175 | 0.8839 |
0.2859 | 22.4 | 5600 | 0.8849 | 0.9190 | 0.8804 |
0.2859 | 22.8 | 5700 | 0.7944 | 0.9183 | 0.8829 |
0.2859 | 23.2 | 5800 | 0.8265 | 0.9198 | 0.8871 |
Framework versions
- Transformers 4.33.3
- Pytorch 2.0.1
- Datasets 2.14.5
- Tokenizers 0.13.3