<!-- 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. -->
Pegasus-x-base-govreport-12288-numbeam-2
This model is a fine-tuned version of google/pegasus-x-base on the govreport-summarization dataset. It achieves the following results on the evaluation set:
- Loss: 1.6327
'ROUGE': {'rouge1': 0.4843586873168665, 'rouge2': 0.20893237725290087, 'rougeL': 0.24808730732880646, 'rougeLsum': 0.248335395434983}
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: 0.0001
- train_batch_size: 1
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.2439 | 0.03 | 120 | 2.1257 |
| 2.1672 | 0.05 | 240 | 1.9610 |
| 2.1165 | 0.08 | 360 | 1.9005 |
| 2.1627 | 0.11 | 480 | 1.8735 |
| 1.8851 | 0.14 | 600 | 1.8362 |
| 1.9804 | 0.16 | 720 | 1.8153 |
| 1.9568 | 0.19 | 840 | 1.8008 |
| 1.8842 | 0.22 | 960 | 1.8016 |
| 2.0134 | 0.25 | 1080 | 1.7772 |
| 1.9086 | 0.27 | 1200 | 1.7640 |
| 1.9177 | 0.3 | 1320 | 1.7553 |
| 1.9547 | 0.33 | 1440 | 1.7377 |
| 1.793 | 0.36 | 1560 | 1.7411 |
| 1.8007 | 0.38 | 1680 | 1.7440 |
| 1.9048 | 0.41 | 1800 | 1.7244 |
| 1.9304 | 0.44 | 1920 | 1.7164 |
| 1.881 | 0.47 | 2040 | 1.7169 |
| 1.8999 | 0.49 | 2160 | 1.7107 |
| 1.9304 | 0.52 | 2280 | 1.7086 |
| 1.8168 | 0.55 | 2400 | 1.7031 |
| 2.0034 | 0.58 | 2520 | 1.7011 |
| 1.8952 | 0.6 | 2640 | 1.6996 |
| 1.8706 | 0.63 | 2760 | 1.6947 |
| 1.8435 | 0.66 | 2880 | 1.6919 |
| 1.7279 | 0.69 | 3000 | 1.6909 |
| 1.8097 | 0.71 | 3120 | 1.6789 |
| 1.9044 | 0.74 | 3240 | 1.6797 |
| 1.7076 | 0.77 | 3360 | 1.6813 |
| 1.9164 | 0.79 | 3480 | 1.6734 |
| 1.7003 | 0.82 | 3600 | 1.6724 |
| 1.8885 | 0.85 | 3720 | 1.6726 |
| 1.7595 | 0.88 | 3840 | 1.6704 |
| 1.9149 | 0.9 | 3960 | 1.6660 |
| 1.7987 | 0.93 | 4080 | 1.6616 |
| 1.8056 | 0.96 | 4200 | 1.6612 |
| 1.7138 | 0.99 | 4320 | 1.6714 |
| 1.6865 | 1.01 | 4440 | 1.6697 |
| 1.7406 | 1.04 | 4560 | 1.6592 |
| 1.7533 | 1.07 | 4680 | 1.6624 |
| 1.7599 | 1.1 | 4800 | 1.6620 |
| 1.7448 | 1.12 | 4920 | 1.6558 |
| 1.7208 | 1.15 | 5040 | 1.6574 |
| 1.7782 | 1.18 | 5160 | 1.6505 |
| 1.6579 | 1.21 | 5280 | 1.6559 |
| 1.8094 | 1.23 | 5400 | 1.6526 |
| 1.9198 | 1.26 | 5520 | 1.6450 |
| 1.6689 | 1.29 | 5640 | 1.6471 |
| 1.807 | 1.32 | 5760 | 1.6490 |
| 1.9385 | 1.34 | 5880 | 1.6423 |
| 1.6097 | 1.37 | 6000 | 1.6464 |
| 1.6278 | 1.4 | 6120 | 1.6489 |
| 1.7366 | 1.42 | 6240 | 1.6467 |
| 1.6839 | 1.45 | 6360 | 1.6418 |
| 1.8194 | 1.48 | 6480 | 1.6423 |
| 1.6548 | 1.51 | 6600 | 1.6412 |
| 1.7561 | 1.53 | 6720 | 1.6443 |
| 1.8182 | 1.56 | 6840 | 1.6410 |
| 1.7861 | 1.59 | 6960 | 1.6389 |
| 1.6587 | 1.62 | 7080 | 1.6361 |
| 1.7151 | 1.64 | 7200 | 1.6412 |
| 1.7458 | 1.67 | 7320 | 1.6430 |
| 1.677 | 1.7 | 7440 | 1.6354 |
| 1.7061 | 1.73 | 7560 | 1.6340 |
| 1.6471 | 1.75 | 7680 | 1.6364 |
| 1.6342 | 1.78 | 7800 | 1.6339 |
| 1.8216 | 1.81 | 7920 | 1.6362 |
| 1.7511 | 1.84 | 8040 | 1.6361 |
| 1.7565 | 1.86 | 8160 | 1.6317 |
| 1.766 | 1.89 | 8280 | 1.6332 |
| 1.6302 | 1.92 | 8400 | 1.6326 |
| 1.7421 | 1.95 | 8520 | 1.6328 |
| 1.7077 | 1.97 | 8640 | 1.6327 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3
