--- tags: - generated_from_trainer model-index: - name: RoBERTa-legal-de-cased_German_legal_SQuAD_100 results: [] --- # RoBERTa-legal-de-cased_German_legal_SQuAD_100 This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.1554 ## 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: 100 ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:-----:|:----:|:---------------:| | No log | 1.0 | 3 | 6.3212 | | No log | 2.0 | 6 | 6.2478 | | No log | 3.0 | 9 | 6.0992 | | No log | 4.0 | 12 | 5.8910 | | No log | 5.0 | 15 | 5.3419 | | No log | 6.0 | 18 | 5.2077 | | No log | 7.0 | 21 | 5.0437 | | No log | 8.0 | 24 | 4.8946 | | No log | 9.0 | 27 | 4.7876 | | No log | 10.0 | 30 | 4.4466 | | No log | 11.0 | 33 | 4.2105 | | No log | 12.0 | 36 | 3.8482 | | No log | 13.0 | 39 | 3.6657 | | No log | 14.0 | 42 | 3.5262 | | No log | 15.0 | 45 | 3.4486 | | No log | 16.0 | 48 | 3.2999 | | No log | 17.0 | 51 | 3.2247 | | No log | 18.0 | 54 | 3.1381 | | No log | 19.0 | 57 | 2.9505 | | No log | 20.0 | 60 | 2.9027 | | No log | 21.0 | 63 | 2.8186 | | No log | 22.0 | 66 | 2.8264 | | No log | 23.0 | 69 | 2.6828 | | No log | 24.0 | 72 | 2.6195 | | No log | 25.0 | 75 | 2.4215 | | No log | 26.0 | 78 | 2.3506 | | No log | 27.0 | 81 | 2.4491 | | No log | 28.0 | 84 | 2.2961 | | No log | 29.0 | 87 | 2.2478 | | No log | 30.0 | 90 | 2.2342 | | No log | 31.0 | 93 | 2.1921 | | No log | 32.0 | 96 | 2.1144 | | No log | 33.0 | 99 | 1.8770 | | No log | 34.0 | 102 | 2.0223 | | No log | 35.0 | 105 | 1.9032 | | No log | 36.0 | 108 | 1.8026 | | No log | 37.0 | 111 | 1.8888 | | No log | 38.0 | 114 | 1.7613 | | No log | 39.0 | 117 | 1.7523 | | No log | 40.0 | 120 | 1.6841 | | No log | 41.0 | 123 | 1.6183 | | No log | 42.0 | 126 | 1.6989 | | No log | 43.0 | 129 | 1.5156 | | No log | 44.0 | 132 | 1.6684 | | No log | 45.0 | 135 | 1.6410 | | No log | 46.0 | 138 | 1.3884 | | No log | 47.0 | 141 | 1.4724 | | No log | 48.0 | 144 | 1.4343 | | No log | 49.0 | 147 | 1.3841 | | No log | 50.0 | 150 | 1.2559 | | No log | 51.0 | 153 | 1.3226 | | No log | 52.0 | 156 | 1.4074 | | No log | 53.0 | 159 | 1.3355 | | No log | 54.0 | 162 | 1.3603 | | No log | 55.0 | 165 | 1.3612 | | No log | 56.0 | 168 | 1.3112 | | No log | 57.0 | 171 | 1.3595 | | No log | 58.0 | 174 | 1.3395 | | No log | 59.0 | 177 | 1.2835 | | No log | 60.0 | 180 | 1.3380 | | No log | 61.0 | 183 | 1.3221 | | No log | 62.0 | 186 | 1.2596 | | No log | 63.0 | 189 | 1.2257 | | No log | 64.0 | 192 | 1.2678 | | No log | 65.0 | 195 | 1.2746 | | No log | 66.0 | 198 | 1.2055 | | No log | 67.0 | 201 | 1.1874 | | No log | 68.0 | 204 | 1.2531 | | No log | 69.0 | 207 | 1.2856 | | No log | 70.0 | 210 | 1.2025 | | No log | 71.0 | 213 | 1.1595 | | No log | 72.0 | 216 | 1.2038 | | No log | 73.0 | 219 | 1.2413 | | No log | 74.0 | 222 | 1.2031 | | No log | 75.0 | 225 | 1.1367 | | No log | 76.0 | 228 | 1.1172 | | No log | 77.0 | 231 | 1.1705 | | No log | 78.0 | 234 | 1.2407 | | No log | 79.0 | 237 | 1.2265 | | No log | 80.0 | 240 | 1.1681 | | No log | 81.0 | 243 | 1.1485 | | No log | 82.0 | 246 | 1.1626 | | No log | 83.0 | 249 | 1.1758 | | No log | 84.0 | 252 | 1.1984 | | No log | 85.0 | 255 | 1.2050 | | No log | 86.0 | 258 | 1.1858 | | No log | 87.0 | 261 | 1.1767 | | No log | 88.0 | 264 | 1.1605 | | No log | 89.0 | 267 | 1.1615 | | No log | 90.0 | 270 | 1.1741 | | No log | 91.0 | 273 | 1.1772 | | No log | 92.0 | 276 | 1.1721 | | No log | 93.0 | 279 | 1.1670 | | No log | 94.0 | 282 | 1.1610 | | No log | 95.0 | 285 | 1.1554 | | No log | 96.0 | 288 | 1.1532 | | No log | 97.0 | 291 | 1.1576 | | No log | 98.0 | 294 | 1.1576 | | No log | 99.0 | 297 | 1.1564 | | No log | 100.0 | 300 | 1.1554 | ### Framework versions - Transformers 4.36.2 - Pytorch 2.1.2+cu121 - Datasets 2.14.7 - Tokenizers 0.15.0