bart-large-asqa-ob
This model is a fine-tuned version of facebook/bart-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.5919
- Rougelsum: 19.1048
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: 5e-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rougelsum |
|---|---|---|---|---|
| No log | 1.0 | 355 | 1.6256 | 18.3700 |
| 1.8462 | 2.0 | 710 | 1.5966 | 18.3464 |
| 1.6704 | 3.0 | 1065 | 1.5906 | 18.6009 |
| 1.6704 | 4.0 | 1420 | 1.5841 | 18.2794 |
| 1.6087 | 5.0 | 1775 | 1.5852 | 18.4272 |
| 1.5364 | 6.0 | 2130 | 1.5989 | 18.9977 |
| 1.5364 | 7.0 | 2485 | 1.5902 | 18.7631 |
| 1.4746 | 8.0 | 2840 | 1.5917 | 18.9565 |
| 1.4336 | 9.0 | 3195 | 1.5919 | 19.1048 |
Framework versions
- Transformers 4.23.0.dev0
- Pytorch 1.12.1+cu102
- Datasets 2.4.0
- Tokenizers 0.12.1