metadata
library_name: transformers
license: apache-2.0
base_model: facebook/bart-large
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: bart-large-aeslc-100-cnt-supervised-basic
results: []
bart-large-aeslc-100-cnt-supervised-basic
This model is a fine-tuned version of facebook/bart-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 5.9904
- Rouge1: 0.2974
- Rouge2: 0.1434
- Rougel: 0.2907
- Rougelsum: 0.2902
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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.06
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| 5.1764 | 0.4 | 10 | 5.6551 | 0.1489 | 0.0635 | 0.1375 | 0.1374 |
| 3.9597 | 0.8 | 20 | 5.1629 | 0.2459 | 0.1214 | 0.2437 | 0.2436 |
| 3.4396 | 1.2 | 30 | 4.9109 | 0.293 | 0.146 | 0.2814 | 0.2812 |
| 3.0038 | 1.6 | 40 | 4.9415 | 0.2924 | 0.1489 | 0.2824 | 0.2822 |
| 2.8195 | 2.0 | 50 | 5.0323 | 0.2804 | 0.1449 | 0.2771 | 0.277 |
| 2.6185 | 2.4 | 60 | 4.9905 | 0.3058 | 0.1559 | 0.2963 | 0.2963 |
| 2.2683 | 2.8 | 70 | 5.0726 | 0.2994 | 0.1529 | 0.2923 | 0.292 |
| 2.1857 | 3.2 | 80 | 5.1591 | 0.2903 | 0.1441 | 0.2862 | 0.286 |
| 2.0587 | 3.6 | 90 | 5.1690 | 0.2985 | 0.1485 | 0.2912 | 0.2912 |
| 2.28 | 4.0 | 100 | 5.2317 | 0.3034 | 0.1508 | 0.2961 | 0.2961 |
| 1.7853 | 4.4 | 110 | 5.4102 | 0.2944 | 0.1462 | 0.2878 | 0.2877 |
| 1.9434 | 4.8 | 120 | 5.4227 | 0.3065 | 0.1529 | 0.3004 | 0.3 |
| 1.4693 | 5.2 | 130 | 5.4791 | 0.3053 | 0.1489 | 0.2984 | 0.2978 |
| 1.6331 | 5.6 | 140 | 5.5514 | 0.2964 | 0.1423 | 0.2906 | 0.2904 |
| 1.585 | 6.0 | 150 | 5.5008 | 0.2967 | 0.1436 | 0.2912 | 0.291 |
| 1.3414 | 6.4 | 160 | 5.5670 | 0.2981 | 0.147 | 0.2919 | 0.2917 |
| 1.2337 | 6.8 | 170 | 5.6861 | 0.2973 | 0.1471 | 0.2916 | 0.2911 |
| 1.1571 | 7.2 | 180 | 5.8000 | 0.2914 | 0.1401 | 0.2856 | 0.2854 |
| 1.6853 | 7.6 | 190 | 5.9097 | 0.2922 | 0.1406 | 0.2862 | 0.2861 |
| 1.0714 | 8.0 | 200 | 5.9439 | 0.2954 | 0.143 | 0.2884 | 0.2883 |
| 0.9668 | 8.4 | 210 | 5.9412 | 0.2975 | 0.1449 | 0.2905 | 0.2904 |
| 1.0741 | 8.8 | 220 | 5.9350 | 0.2989 | 0.1436 | 0.292 | 0.2918 |
| 1.0771 | 9.2 | 230 | 5.9855 | 0.2954 | 0.1425 | 0.2887 | 0.2883 |
| 0.96 | 9.6 | 240 | 5.9881 | 0.297 | 0.1435 | 0.2903 | 0.29 |
| 1.0673 | 10.0 | 250 | 5.9904 | 0.2974 | 0.1434 | 0.2907 | 0.2902 |
Framework versions
- Transformers 4.57.3
- Pytorch 2.9.1+cu128
- Datasets 3.6.0
- Tokenizers 0.22.1