t5-small-finetuned-multinews
This model is a fine-tuned version of t5-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.7276
- Rouge1: 14.7073
- Rouge2: 4.8849
- Rougel: 11.336
- Rougelsum: 13.1015
- Gen Len: 18.98
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: 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: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 3.2539 | 1.0 | 506 | 2.8142 | 14.3316 | 4.7443 | 11.1018 | 12.8337 | 18.98 |
| 3.0164 | 2.0 | 1012 | 2.7613 | 14.749 | 4.9494 | 11.3621 | 13.1838 | 18.98 |
| 2.9764 | 3.0 | 1518 | 2.7402 | 14.7452 | 4.8903 | 11.367 | 13.1816 | 18.98 |
| 2.9514 | 4.0 | 2024 | 2.7307 | 14.7309 | 4.8615 | 11.3391 | 13.1464 | 18.98 |
| 2.9446 | 5.0 | 2530 | 2.7276 | 14.7073 | 4.8849 | 11.336 | 13.1015 | 18.98 |
Framework versions
- Transformers 4.40.1
- Pytorch 1.13.1+cu117
- Datasets 2.19.0
- Tokenizers 0.19.1