RoBERTa-legal-de-cased_German_legal_SQuAD_part_augmented_17
This model was trained from scratch on the None dataset.
It achieves the following results on the evaluation set:
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: 2e-05
- train_batch_size: 128
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 17
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|
| No log | 1.0 | 4 | 6.3320 |
| No log | 2.0 | 8 | 6.1329 |
| No log | 3.0 | 12 | 5.8040 |
| No log | 4.0 | 16 | 5.4681 |
| No log | 5.0 | 20 | 5.3900 |
| No log | 6.0 | 24 | 5.1992 |
| No log | 7.0 | 28 | 4.9927 |
| No log | 8.0 | 32 | 4.8472 |
| No log | 9.0 | 36 | 4.7646 |
| No log | 10.0 | 40 | 4.6557 |
| No log | 11.0 | 44 | 4.5620 |
| No log | 12.0 | 48 | 4.5213 |
| No log | 13.0 | 52 | 4.4434 |
| No log | 14.0 | 56 | 4.3200 |
| No log | 15.0 | 60 | 4.2448 |
| No log | 16.0 | 64 | 4.2161 |
| No log | 17.0 | 68 | 4.2145 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.14.7
- Tokenizers 0.15.0
