<!-- 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. -->
pegasuscnn_billsum_model
This model is a fine-tuned version of google/pegasus-cnn_dailymail on the billsum dataset. It achieves the following results on the evaluation set:
- Loss: 1.6372
- Rouge1: 0.4732
- Rouge2: 0.2262
- Rougel: 0.3173
- Rougelsum: 0.3176
- Gen Len: 124.3831
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: 5
- eval_batch_size: 5
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.595 | 1.0 | 198 | 1.8846 | 0.4347 | 0.1901 | 0.2927 | 0.2923 | 123.2056 |
| 2.0147 | 2.0 | 396 | 1.7779 | 0.4483 | 0.2039 | 0.301 | 0.3007 | 122.9718 |
| 1.8752 | 3.0 | 594 | 1.7352 | 0.4563 | 0.2111 | 0.3014 | 0.3017 | 121.9435 |
| 1.8744 | 4.0 | 792 | 1.7021 | 0.4718 | 0.2235 | 0.3153 | 0.3155 | 123.5242 |
| 1.7731 | 5.0 | 990 | 1.6803 | 0.4667 | 0.2209 | 0.3172 | 0.3171 | 123.1411 |
| 1.7142 | 6.0 | 1188 | 1.6607 | 0.4702 | 0.221 | 0.3143 | 0.3141 | 123.8871 |
| 1.7003 | 7.0 | 1386 | 1.6503 | 0.4723 | 0.2261 | 0.3179 | 0.3175 | 124.5444 |
| 1.6977 | 8.0 | 1584 | 1.6426 | 0.4697 | 0.2255 | 0.3175 | 0.3172 | 123.8548 |
| 1.672 | 9.0 | 1782 | 1.6390 | 0.4698 | 0.2241 | 0.3163 | 0.3163 | 123.6371 |
| 1.6622 | 10.0 | 1980 | 1.6372 | 0.4732 | 0.2262 | 0.3173 | 0.3176 | 124.3831 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1
