<!-- 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_paraphraser
This model is a fine-tuned version of tuner007/pegasus_paraphrase on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.3852
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: 1
- 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: 7
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.8224 | 0.19 | 10 | 1.2976 |
| 1.3213 | 0.37 | 20 | 1.2302 |
| 1.2333 | 0.56 | 30 | 1.2068 |
| 1.1097 | 0.74 | 40 | 1.1991 |
| 1.3741 | 0.93 | 50 | 1.2260 |
| 1.0593 | 1.12 | 60 | 1.2005 |
| 1.1234 | 1.3 | 70 | 1.2103 |
| 1.0116 | 1.49 | 80 | 1.2098 |
| 0.8591 | 1.67 | 90 | 1.1709 |
| 0.9176 | 1.86 | 100 | 1.1830 |
| 0.7524 | 2.05 | 110 | 1.2122 |
| 0.7762 | 2.23 | 120 | 1.2398 |
| 0.677 | 2.42 | 130 | 1.2440 |
| 0.8364 | 2.6 | 140 | 1.2356 |
| 0.7489 | 2.79 | 150 | 1.2542 |
| 0.7113 | 2.98 | 160 | 1.2678 |
| 0.5462 | 3.16 | 170 | 1.3100 |
| 0.6775 | 3.35 | 180 | 1.3193 |
| 0.6417 | 3.53 | 190 | 1.3157 |
| 0.547 | 3.72 | 200 | 1.3172 |
| 0.5357 | 3.91 | 210 | 1.3311 |
| 0.6796 | 4.09 | 220 | 1.3236 |
| 0.4884 | 4.28 | 230 | 1.3288 |
| 0.483 | 4.47 | 240 | 1.3423 |
| 0.667 | 4.65 | 250 | 1.3702 |
| 0.5785 | 4.84 | 260 | 1.3817 |
| 0.6123 | 5.02 | 270 | 1.3728 |
| 0.4735 | 5.21 | 280 | 1.3731 |
| 0.5278 | 5.4 | 290 | 1.3783 |
| 0.5393 | 5.58 | 300 | 1.3904 |
| 0.4631 | 5.77 | 310 | 1.3884 |
| 0.4538 | 5.95 | 320 | 1.3800 |
| 0.5137 | 6.14 | 330 | 1.3766 |
| 0.5514 | 6.33 | 340 | 1.3815 |
| 0.4629 | 6.51 | 350 | 1.3849 |
| 0.5013 | 6.7 | 360 | 1.3855 |
| 0.4566 | 6.88 | 370 | 1.3852 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.1+cu117
- Datasets 2.12.0
- Tokenizers 0.13.3
