<!-- 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_long_legal_test_sm
This model is a fine-tuned version of saibo/legal-longformer-base-4096 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.7030
- Accuracy: 0.4788
- Precision: 0.5067
- Recall: 0.5045
- F1: 0.5056
- D-index: 1.1273
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: 8
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | D-index |
|---|---|---|---|---|---|---|---|---|
| No log | 0.98 | 26 | 0.6940 | 0.4858 | 0.5366 | 0.1964 | 0.2876 | 1.1593 |
| No log | 2.0 | 53 | 0.6931 | 0.5165 | 0.5674 | 0.3571 | 0.4384 | 1.2106 |
| No log | 2.98 | 79 | 0.6912 | 0.5236 | 0.5294 | 0.8839 | 0.6622 | 1.1943 |
| No log | 4.0 | 106 | 0.6932 | 0.4953 | 0.5174 | 0.6652 | 0.5820 | 1.1510 |
| No log | 4.98 | 132 | 0.7097 | 0.5024 | 0.5389 | 0.4018 | 0.4604 | 1.1802 |
| No log | 6.0 | 159 | 0.7527 | 0.5212 | 0.5349 | 0.7188 | 0.6133 | 1.1992 |
| No log | 6.98 | 185 | 0.8177 | 0.5189 | 0.5510 | 0.4821 | 0.5143 | 1.2081 |
| No log | 8.0 | 212 | 1.0477 | 0.4858 | 0.5254 | 0.2768 | 0.3626 | 1.1547 |
| No log | 8.98 | 238 | 1.2084 | 0.4858 | 0.5192 | 0.3616 | 0.4263 | 1.1498 |
| No log | 10.0 | 265 | 1.3307 | 0.5118 | 0.5594 | 0.3571 | 0.4360 | 1.2013 |
| No log | 10.98 | 291 | 1.7030 | 0.4906 | 0.5213 | 0.4375 | 0.4757 | 1.1548 |
| No log | 12.0 | 318 | 1.8288 | 0.5142 | 0.5341 | 0.6295 | 0.5779 | 1.1904 |
| No log | 12.98 | 344 | 2.1054 | 0.4882 | 0.5198 | 0.4107 | 0.4589 | 1.1517 |
| No log | 14.0 | 371 | 2.2934 | 0.4811 | 0.5076 | 0.5982 | 0.5492 | 1.1265 |
| No log | 14.98 | 397 | 2.4791 | 0.5 | 0.5429 | 0.3393 | 0.4176 | 1.1792 |
| No log | 16.0 | 424 | 2.4843 | 0.4858 | 0.5138 | 0.5 | 0.5068 | 1.1418 |
| No log | 16.98 | 450 | 2.6494 | 0.4835 | 0.5141 | 0.4062 | 0.4539 | 1.1425 |
| No log | 18.0 | 477 | 2.6552 | 0.4788 | 0.5072 | 0.4732 | 0.4896 | 1.1291 |
| 0.3174 | 18.98 | 503 | 2.6959 | 0.4764 | 0.5045 | 0.5 | 0.5022 | 1.1228 |
| 0.3174 | 19.62 | 520 | 2.7030 | 0.4788 | 0.5067 | 0.5045 | 0.5056 | 1.1273 |
Framework versions
- Transformers 4.27.4
- Pytorch 1.13.1+cu116
- Datasets 2.11.0
- Tokenizers 0.13.2
