<!-- 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. -->
Finetune_Pegasus_1
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: 1.0942
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 |
|---|---|---|---|
| 1.7293 | 0.21 | 10 | 1.2156 |
| 1.3661 | 0.41 | 20 | 1.1203 |
| 1.3897 | 0.62 | 30 | 1.0665 |
| 1.3356 | 0.82 | 40 | 1.0304 |
| 1.171 | 1.03 | 50 | 1.0098 |
| 0.8665 | 1.23 | 60 | 1.0062 |
| 0.7864 | 1.44 | 70 | 1.0266 |
| 0.8785 | 1.64 | 80 | 1.0190 |
| 1.0596 | 1.85 | 90 | 1.0218 |
| 1.0386 | 2.05 | 100 | 1.0213 |
| 0.7452 | 2.26 | 110 | 1.0639 |
| 0.6807 | 2.46 | 120 | 1.0619 |
| 0.5764 | 2.67 | 130 | 1.0530 |
| 0.87 | 2.87 | 140 | 1.0571 |
| 0.7724 | 3.08 | 150 | 1.0563 |
| 0.5847 | 3.28 | 160 | 1.0692 |
| 0.6053 | 3.49 | 170 | 1.0652 |
| 0.6416 | 3.69 | 180 | 1.0531 |
| 0.6392 | 3.9 | 190 | 1.0416 |
| 0.6138 | 4.1 | 200 | 1.0489 |
| 0.6093 | 4.31 | 210 | 1.0668 |
| 0.5484 | 4.51 | 220 | 1.0843 |
| 0.6082 | 4.72 | 230 | 1.0771 |
| 0.56 | 4.92 | 240 | 1.0745 |
| 0.5796 | 5.13 | 250 | 1.0770 |
| 0.6597 | 5.33 | 260 | 1.0722 |
| 0.4834 | 5.54 | 270 | 1.0726 |
| 0.4232 | 5.74 | 280 | 1.0682 |
| 0.5432 | 5.95 | 290 | 1.0769 |
| 0.5944 | 6.15 | 300 | 1.0851 |
| 0.4663 | 6.36 | 310 | 1.0884 |
| 0.4568 | 6.56 | 320 | 1.0915 |
| 0.4565 | 6.77 | 330 | 1.0942 |
Framework versions
- Transformers 4.29.2
- Pytorch 2.0.1+cu117
- Datasets 2.12.0
- Tokenizers 0.13.3
