metadata
license: mit
tags:
- generated_from_trainer
model-index:
- name: bart-pt-asqa-ob
results: []
bart-pt-asqa-ob
This model is a fine-tuned version of vblagoje/bart_lfqa on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.4268
- Rougelsum: 24.2407
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: 1e-05
- 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 | 1.6269 | 19.0683 |
| 1.6035 | 2.0 | 710 | 1.6400 | 19.8336 |
| 1.3505 | 3.0 | 1065 | 1.6525 | 20.9906 |
| 1.3505 | 4.0 | 1420 | 1.7070 | 21.5381 |
| 1.1756 | 5.0 | 1775 | 1.7348 | 22.6130 |
| 1.0148 | 6.0 | 2130 | 1.8440 | 22.8553 |
| 1.0148 | 7.0 | 2485 | 1.8460 | 23.1281 |
| 0.8886 | 8.0 | 2840 | 1.9321 | 23.4357 |
| 0.7687 | 9.0 | 3195 | 2.0124 | 23.3538 |
| 0.6779 | 10.0 | 3550 | 2.0809 | 23.7958 |
| 0.6779 | 11.0 | 3905 | 2.1312 | 23.5703 |
| 0.5933 | 12.0 | 4260 | 2.2144 | 24.0672 |
| 0.5283 | 13.0 | 4615 | 2.2463 | 23.9667 |
| 0.5283 | 14.0 | 4970 | 2.3022 | 24.0211 |
| 0.4885 | 15.0 | 5325 | 2.3010 | 24.2634 |
| 0.4379 | 16.0 | 5680 | 2.3311 | 24.2333 |
| 0.4085 | 17.0 | 6035 | 2.4048 | 24.2417 |
| 0.4085 | 18.0 | 6390 | 2.4118 | 24.2201 |
| 0.3821 | 19.0 | 6745 | 2.4237 | 24.2905 |
| 0.3699 | 20.0 | 7100 | 2.4268 | 24.2407 |
Framework versions
- Transformers 4.23.0.dev0
- Pytorch 1.12.1+cu102
- Datasets 2.4.0
- Tokenizers 0.12.1