<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
bart-cnn-pubmed-arxiv-v3-e16
This model is a fine-tuned version of theojolliffe/bart-cnn-pubmed-arxiv on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.9340
- Rouge1: 57.6388
- Rouge2: 44.834
- Rougel: 47.5043
- Rougelsum: 56.1122
- Gen Len: 142.0
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: 2e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 16
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
---|---|---|---|---|---|---|---|---|
1.2407 | 1.0 | 795 | 0.9270 | 53.3842 | 33.8559 | 35.7393 | 50.6907 | 142.0 |
0.704 | 2.0 | 1590 | 0.8092 | 53.2159 | 35.0209 | 37.8641 | 50.9514 | 141.963 |
0.5277 | 3.0 | 2385 | 0.7588 | 52.7709 | 34.2453 | 36.6319 | 50.1137 | 142.0 |
0.3449 | 4.0 | 3180 | 0.7617 | 52.0249 | 34.5679 | 37.3669 | 49.7643 | 142.0 |
0.2668 | 5.0 | 3975 | 0.7575 | 54.3131 | 35.3985 | 38.9242 | 51.5667 | 142.0 |
0.1756 | 6.0 | 4770 | 0.8161 | 53.6214 | 36.4376 | 39.1745 | 51.3685 | 142.0 |
0.1326 | 7.0 | 5565 | 0.7848 | 55.7549 | 38.8517 | 42.0106 | 53.4243 | 142.0 |
0.1051 | 8.0 | 6360 | 0.7912 | 55.2709 | 39.952 | 42.7398 | 53.6479 | 142.0 |
0.0781 | 9.0 | 7155 | 0.8491 | 55.5698 | 40.0599 | 42.9521 | 53.6734 | 142.0 |
0.0685 | 10.0 | 7950 | 0.8684 | 55.1142 | 40.3136 | 43.699 | 53.5463 | 142.0 |
0.0494 | 11.0 | 8745 | 0.8886 | 57.7988 | 43.6659 | 46.0913 | 56.3383 | 142.0 |
0.0338 | 12.0 | 9540 | 0.8827 | 57.0166 | 42.7553 | 46.2344 | 55.2893 | 142.0 |
0.0296 | 13.0 | 10335 | 0.9111 | 56.7741 | 42.6116 | 45.1692 | 55.2065 | 142.0 |
0.0228 | 14.0 | 11130 | 0.9209 | 56.635 | 43.2461 | 46.314 | 55.049 | 142.0 |
0.0189 | 15.0 | 11925 | 0.9193 | 56.4404 | 43.4216 | 46.279 | 55.1403 | 142.0 |
0.0152 | 16.0 | 12720 | 0.9340 | 57.6388 | 44.834 | 47.5043 | 56.1122 | 142.0 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0+cu113
- Datasets 2.1.0
- Tokenizers 0.12.1