<!-- 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. -->
legal_bert_test2
This model is a fine-tuned version of nlpaueb/legal-bert-small-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.7119
- Accuracy: 0.4875
- Precision: 0.48
- Recall: 0.3
- F1: 0.3692
- D-index: 1.1458
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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | D-index |
|---|---|---|---|---|---|---|---|---|
| No log | 0.8 | 1 | 0.7095 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 1.8 | 2 | 0.7094 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 2.8 | 3 | 0.7092 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 3.8 | 4 | 0.7089 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 4.8 | 5 | 0.7085 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 5.8 | 6 | 0.7080 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 6.8 | 7 | 0.7074 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 7.8 | 8 | 0.7067 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 8.8 | 9 | 0.7060 | 0.5 | 0.0 | 0.0 | 0.0 | 1.1699 |
| No log | 9.8 | 10 | 0.7053 | 0.5125 | 1.0 | 0.025 | 0.0488 | 1.1939 |
| No log | 10.8 | 11 | 0.7045 | 0.5125 | 1.0 | 0.025 | 0.0488 | 1.1939 |
| No log | 11.8 | 12 | 0.7036 | 0.4875 | 0.3333 | 0.025 | 0.0465 | 1.1458 |
| No log | 12.8 | 13 | 0.7028 | 0.5 | 0.5 | 0.05 | 0.0909 | 1.1699 |
| No log | 13.8 | 14 | 0.7019 | 0.5 | 0.5 | 0.05 | 0.0909 | 1.1699 |
| No log | 14.8 | 15 | 0.7011 | 0.5 | 0.5 | 0.05 | 0.0909 | 1.1699 |
| No log | 15.8 | 16 | 0.7004 | 0.5 | 0.5 | 0.05 | 0.0909 | 1.1699 |
| No log | 16.8 | 17 | 0.6997 | 0.4875 | 0.4 | 0.05 | 0.0889 | 1.1458 |
| No log | 17.8 | 18 | 0.6991 | 0.5 | 0.5 | 0.1 | 0.1667 | 1.1699 |
| No log | 18.8 | 19 | 0.6985 | 0.4875 | 0.4444 | 0.1 | 0.1633 | 1.1458 |
| No log | 19.8 | 20 | 0.6981 | 0.5125 | 0.5333 | 0.2 | 0.2909 | 1.1939 |
| No log | 20.8 | 21 | 0.6978 | 0.475 | 0.4444 | 0.2 | 0.2759 | 1.1214 |
| No log | 21.8 | 22 | 0.6976 | 0.5125 | 0.5185 | 0.35 | 0.4179 | 1.1939 |
| No log | 22.8 | 23 | 0.6975 | 0.525 | 0.5263 | 0.5 | 0.5128 | 1.2176 |
| No log | 23.8 | 24 | 0.6977 | 0.5375 | 0.5366 | 0.55 | 0.5432 | 1.2412 |
| No log | 24.8 | 25 | 0.6980 | 0.525 | 0.5227 | 0.575 | 0.5476 | 1.2176 |
| No log | 25.8 | 26 | 0.6984 | 0.5 | 0.5 | 0.625 | 0.5556 | 1.1699 |
| No log | 26.8 | 27 | 0.6989 | 0.4625 | 0.4727 | 0.65 | 0.5474 | 1.0969 |
| No log | 27.8 | 28 | 0.6994 | 0.45 | 0.4667 | 0.7 | 0.56 | 1.0721 |
| No log | 28.8 | 29 | 0.7001 | 0.45 | 0.4688 | 0.75 | 0.5769 | 1.0721 |
| No log | 29.8 | 30 | 0.7006 | 0.4375 | 0.4615 | 0.75 | 0.5714 | 1.0471 |
| No log | 30.8 | 31 | 0.7011 | 0.45 | 0.4697 | 0.775 | 0.5849 | 1.0721 |
| No log | 31.8 | 32 | 0.7016 | 0.45 | 0.4697 | 0.775 | 0.5849 | 1.0721 |
| No log | 32.8 | 33 | 0.7019 | 0.4375 | 0.4615 | 0.75 | 0.5714 | 1.0471 |
| No log | 33.8 | 34 | 0.7022 | 0.4375 | 0.4615 | 0.75 | 0.5714 | 1.0471 |
| No log | 34.8 | 35 | 0.7024 | 0.4625 | 0.4762 | 0.75 | 0.5825 | 1.0969 |
| No log | 35.8 | 36 | 0.7028 | 0.4625 | 0.4762 | 0.75 | 0.5825 | 1.0969 |
| No log | 36.8 | 37 | 0.7031 | 0.4625 | 0.4754 | 0.725 | 0.5743 | 1.0969 |
| No log | 37.8 | 38 | 0.7036 | 0.45 | 0.4667 | 0.7 | 0.56 | 1.0721 |
| No log | 38.8 | 39 | 0.7041 | 0.45 | 0.4630 | 0.625 | 0.5319 | 1.0721 |
| No log | 39.8 | 40 | 0.7046 | 0.4375 | 0.4490 | 0.55 | 0.4944 | 1.0471 |
| No log | 40.8 | 41 | 0.7051 | 0.4375 | 0.4390 | 0.45 | 0.4444 | 1.0471 |
| No log | 41.8 | 42 | 0.7058 | 0.4625 | 0.4595 | 0.425 | 0.4416 | 1.0969 |
| No log | 42.8 | 43 | 0.7064 | 0.475 | 0.4706 | 0.4 | 0.4324 | 1.1214 |
| No log | 43.8 | 44 | 0.7072 | 0.4875 | 0.4839 | 0.375 | 0.4225 | 1.1458 |
| No log | 44.8 | 45 | 0.7080 | 0.475 | 0.4643 | 0.325 | 0.3824 | 1.1214 |
| No log | 45.8 | 46 | 0.7088 | 0.475 | 0.4615 | 0.3 | 0.3636 | 1.1214 |
| No log | 46.8 | 47 | 0.7097 | 0.475 | 0.4615 | 0.3 | 0.3636 | 1.1214 |
| No log | 47.8 | 48 | 0.7105 | 0.475 | 0.4615 | 0.3 | 0.3636 | 1.1214 |
| No log | 48.8 | 49 | 0.7112 | 0.475 | 0.4615 | 0.3 | 0.3636 | 1.1214 |
| No log | 49.8 | 50 | 0.7119 | 0.4875 | 0.48 | 0.3 | 0.3692 | 1.1458 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2
