metadata
license: apache-2.0
base_model: facebook/bart-base
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: bart-base
results: []
bart-base
This model is a fine-tuned version of facebook/bart-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4156
- Rouge1: 41.7881
- Rouge2: 19.9952
- Rougel: 36.4308
- Rougelsum: 38.1089
- Gen Len: 18.0
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-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 0.3288 | 1.0 | 1557 | 0.2425 | 41.3515 | 19.3193 | 35.4878 | 38.2323 | 18.0 |
| 0.2455 | 2.0 | 3115 | 0.2323 | 41.1865 | 19.5825 | 35.7587 | 37.8016 | 18.0 |
| 0.2097 | 3.0 | 4672 | 0.2333 | 41.4261 | 20.2577 | 36.0437 | 38.1275 | 18.0 |
| 0.1818 | 4.0 | 6230 | 0.2400 | 42.7857 | 21.7524 | 37.4117 | 39.3948 | 18.0 |
| 0.1591 | 5.0 | 7787 | 0.2489 | 41.9402 | 21.3931 | 36.7999 | 38.7229 | 18.0 |
| 0.1392 | 6.0 | 9345 | 0.2530 | 42.1993 | 21.3725 | 36.6614 | 38.5415 | 18.0 |
| 0.1218 | 7.0 | 10902 | 0.2616 | 42.0991 | 20.7834 | 36.6721 | 38.6425 | 18.0 |
| 0.1061 | 8.0 | 12460 | 0.2794 | 41.4682 | 20.2185 | 36.0528 | 37.7626 | 18.0 |
| 0.0929 | 9.0 | 14017 | 0.2858 | 41.5178 | 20.0354 | 36.027 | 37.9562 | 18.0 |
| 0.0813 | 10.0 | 15575 | 0.3001 | 42.1686 | 20.7936 | 36.7589 | 38.6885 | 18.0 |
| 0.0715 | 11.0 | 17132 | 0.3113 | 41.5616 | 20.6733 | 36.2947 | 38.1556 | 18.0 |
| 0.0622 | 12.0 | 18690 | 0.3228 | 41.3672 | 20.0432 | 36.1746 | 38.0949 | 18.0 |
| 0.0544 | 13.0 | 20247 | 0.3296 | 41.4662 | 19.8484 | 35.9521 | 37.7284 | 18.0 |
| 0.0478 | 14.0 | 21805 | 0.3373 | 41.1417 | 20.1208 | 36.1864 | 37.9314 | 18.0 |
| 0.0423 | 15.0 | 23362 | 0.3440 | 41.1174 | 19.551 | 35.7777 | 37.5518 | 18.0 |
| 0.0373 | 16.0 | 24920 | 0.3581 | 40.7365 | 19.5894 | 35.5672 | 37.4447 | 18.0 |
| 0.0327 | 17.0 | 26477 | 0.3654 | 41.0895 | 19.4995 | 35.7195 | 37.3265 | 18.0 |
| 0.0294 | 18.0 | 28035 | 0.3750 | 40.8447 | 19.4098 | 35.557 | 37.3456 | 18.0 |
| 0.0262 | 19.0 | 29592 | 0.3790 | 41.0388 | 19.8022 | 35.946 | 37.6522 | 18.0 |
| 0.0237 | 20.0 | 31150 | 0.3841 | 41.6747 | 19.6307 | 35.9938 | 37.6853 | 18.0 |
| 0.0212 | 21.0 | 32707 | 0.3874 | 40.7796 | 19.2156 | 35.3642 | 37.1609 | 18.0 |
| 0.0192 | 22.0 | 34265 | 0.3942 | 41.2411 | 19.5756 | 35.8442 | 37.5498 | 18.0 |
| 0.0173 | 23.0 | 35822 | 0.3974 | 41.112 | 19.7216 | 35.8072 | 37.5629 | 18.0 |
| 0.0159 | 24.0 | 37380 | 0.4042 | 40.6911 | 19.1988 | 35.5312 | 37.3276 | 18.0 |
| 0.0144 | 25.0 | 38937 | 0.4090 | 41.0017 | 19.3834 | 35.7806 | 37.6217 | 18.0 |
| 0.0132 | 26.0 | 40495 | 0.4101 | 41.6159 | 19.4447 | 36.1746 | 37.9271 | 18.0 |
| 0.012 | 27.0 | 42052 | 0.4117 | 41.4618 | 19.3824 | 36.0425 | 37.8597 | 18.0 |
| 0.0112 | 28.0 | 43610 | 0.4137 | 41.5302 | 19.565 | 36.1323 | 37.8484 | 18.0 |
| 0.0105 | 29.0 | 45167 | 0.4147 | 41.5432 | 19.9581 | 36.2526 | 38.0642 | 18.0 |
| 0.0099 | 29.99 | 46710 | 0.4156 | 41.7881 | 19.9952 | 36.4308 | 38.1089 | 18.0 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.19.1
- Tokenizers 0.15.2