metadata
license: mit
tags:
- simplification
- generated_from_trainer
metrics:
- rouge
model-index:
- name: mbart-large-50-clara-med
results: []
mbart-large-50-clara-med
This model is a fine-tuned version of facebook/mbart-large-50 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.2175
- Rouge1: 48.3311
- Rouge2: 30.5638
- Rougel: 43.5214
- Rougelsum: 43.6488
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: 5.6e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 190 | 3.2394 | 16.8539 | 2.7013 | 12.425 | 12.5286 |
| No log | 2.0 | 380 | 1.7381 | 44.5316 | 27.8022 | 40.1591 | 40.3177 |
| 3.4249 | 3.0 | 570 | 1.7198 | 45.6463 | 28.6925 | 41.263 | 41.4703 |
| 3.4249 | 4.0 | 760 | 1.9450 | 43.0233 | 26.3397 | 38.7518 | 38.9154 |
| 0.8377 | 5.0 | 950 | 2.1068 | 46.5936 | 28.7218 | 41.7184 | 41.8448 |
| 0.8377 | 6.0 | 1140 | 2.2815 | 46.4517 | 28.5639 | 41.8107 | 41.9996 |
| 0.8377 | 7.0 | 1330 | 2.4726 | 46.0403 | 28.1887 | 40.9183 | 41.0318 |
| 0.3195 | 8.0 | 1520 | 2.5690 | 47.255 | 29.1482 | 42.4463 | 42.5728 |
| 0.3195 | 9.0 | 1710 | 2.6753 | 46.5967 | 28.5688 | 41.414 | 41.5889 |
| 0.1925 | 10.0 | 1900 | 2.7276 | 46.3251 | 28.4889 | 41.4556 | 41.581 |
| 0.1925 | 11.0 | 2090 | 2.7638 | 46.9325 | 29.2558 | 41.726 | 41.8413 |
| 0.1925 | 12.0 | 2280 | 2.8273 | 47.0344 | 29.1298 | 41.7291 | 41.9236 |
| 0.1313 | 13.0 | 2470 | 2.8633 | 47.5234 | 29.6376 | 42.3409 | 42.4372 |
| 0.1313 | 14.0 | 2660 | 2.8989 | 47.0396 | 29.117 | 41.9893 | 42.1846 |
| 0.1117 | 15.0 | 2850 | 2.9691 | 47.8406 | 29.889 | 42.5645 | 42.7676 |
| 0.1117 | 16.0 | 3040 | 2.9763 | 46.9489 | 28.9919 | 41.8404 | 42.0141 |
| 0.1117 | 17.0 | 3230 | 2.9985 | 47.6628 | 29.7341 | 42.6382 | 42.7649 |
| 0.0824 | 18.0 | 3420 | 3.0511 | 48.0627 | 30.4108 | 43.1693 | 43.3489 |
| 0.0824 | 19.0 | 3610 | 3.0102 | 48.05 | 29.9552 | 43.1462 | 43.3421 |
| 0.0467 | 20.0 | 3800 | 3.0520 | 47.5451 | 29.6129 | 42.6499 | 42.7968 |
| 0.0467 | 21.0 | 3990 | 3.0978 | 47.5042 | 29.6191 | 42.6093 | 42.7341 |
| 0.0467 | 22.0 | 4180 | 3.1270 | 47.8301 | 29.9484 | 42.6866 | 42.9179 |
| 0.0246 | 23.0 | 4370 | 3.1435 | 47.6683 | 30.1974 | 43.0456 | 43.1496 |
| 0.0246 | 24.0 | 4560 | 3.1599 | 47.8652 | 30.2751 | 43.0445 | 43.1898 |
| 0.013 | 25.0 | 4750 | 3.1750 | 48.1352 | 30.4185 | 43.0485 | 43.2456 |
| 0.013 | 26.0 | 4940 | 3.1939 | 47.9653 | 30.3968 | 43.1271 | 43.2522 |
| 0.013 | 27.0 | 5130 | 3.2054 | 48.2122 | 30.6 | 43.3461 | 43.4629 |
| 0.0071 | 28.0 | 5320 | 3.1964 | 47.924 | 30.3089 | 43.0402 | 43.2016 |
| 0.0071 | 29.0 | 5510 | 3.2123 | 48.2967 | 30.5088 | 43.431 | 43.5384 |
| 0.005 | 30.0 | 5700 | 3.2175 | 48.3311 | 30.5638 | 43.5214 | 43.6488 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0
- Datasets 2.8.0
- Tokenizers 0.12.1