|
--- |
|
license: mit |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: bart-cnn-pubmed-arxiv-pubmed-v3-e32 |
|
results: [] |
|
--- |
|
|
|
<!-- 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-pubmed-v3-e32 |
|
|
|
This model is a fine-tuned version of [theojolliffe/bart-cnn-pubmed-arxiv-pubmed](https://huggingface.co/theojolliffe/bart-cnn-pubmed-arxiv-pubmed) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.9707 |
|
- Rouge1: 58.6575 |
|
- Rouge2: 47.1055 |
|
- Rougel: 50.0715 |
|
- Rougelsum: 57.58 |
|
- 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: 2 |
|
- eval_batch_size: 2 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 32 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len | |
|
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:--------:| |
|
| No log | 1.0 | 398 | 0.9589 | 52.4374 | 32.0538 | 34.189 | 49.8178 | 142.0 | |
|
| 1.1222 | 2.0 | 796 | 0.8144 | 54.363 | 35.2782 | 37.5982 | 51.9121 | 142.0 | |
|
| 0.6686 | 3.0 | 1194 | 0.7747 | 53.3334 | 34.9112 | 38.1684 | 50.9676 | 142.0 | |
|
| 0.4394 | 4.0 | 1592 | 0.7660 | 53.2391 | 34.1677 | 38.4917 | 50.582 | 142.0 | |
|
| 0.4394 | 5.0 | 1990 | 0.7508 | 54.3922 | 36.631 | 39.6881 | 52.4238 | 142.0 | |
|
| 0.2962 | 6.0 | 2388 | 0.8112 | 53.9595 | 36.1326 | 38.937 | 51.8107 | 142.0 | |
|
| 0.201 | 7.0 | 2786 | 0.7842 | 55.3659 | 38.4021 | 41.1556 | 53.3145 | 142.0 | |
|
| 0.1414 | 8.0 | 3184 | 0.7557 | 54.8476 | 38.7707 | 41.8756 | 53.3081 | 142.0 | |
|
| 0.107 | 9.0 | 3582 | 0.8296 | 55.7594 | 39.3691 | 41.6456 | 53.9381 | 142.0 | |
|
| 0.107 | 10.0 | 3980 | 0.8298 | 54.8163 | 38.9233 | 42.4104 | 52.9344 | 142.0 | |
|
| 0.0838 | 11.0 | 4378 | 0.8492 | 56.3438 | 41.5532 | 44.6348 | 54.6106 | 141.8704 | |
|
| 0.0637 | 12.0 | 4776 | 0.8619 | 56.8559 | 41.2682 | 43.4566 | 54.7799 | 142.0 | |
|
| 0.051 | 13.0 | 5174 | 0.8733 | 57.4154 | 42.6009 | 44.401 | 56.0209 | 142.0 | |
|
| 0.04 | 14.0 | 5572 | 0.8777 | 58.3095 | 44.7657 | 47.8527 | 56.7276 | 142.0 | |
|
| 0.04 | 15.0 | 5970 | 0.8711 | 57.6542 | 43.1785 | 46.3796 | 56.0532 | 142.0 | |
|
| 0.0341 | 16.0 | 6368 | 0.9038 | 57.7274 | 43.5198 | 45.8797 | 56.1525 | 142.0 | |
|
| 0.0272 | 17.0 | 6766 | 0.8845 | 58.4461 | 44.9513 | 47.6616 | 57.0634 | 142.0 | |
|
| 0.0231 | 18.0 | 7164 | 0.9108 | 58.5774 | 46.2637 | 49.9201 | 57.1939 | 141.963 | |
|
| 0.018 | 19.0 | 7562 | 0.9059 | 58.7442 | 44.7141 | 47.6061 | 57.3604 | 142.0 | |
|
| 0.018 | 20.0 | 7960 | 0.9133 | 57.2809 | 43.7722 | 46.2016 | 55.4421 | 142.0 | |
|
| 0.0159 | 21.0 | 8358 | 0.9245 | 57.1685 | 44.5445 | 48.5015 | 55.9304 | 142.0 | |
|
| 0.012 | 22.0 | 8756 | 0.9149 | 57.4727 | 44.2417 | 48.0224 | 56.1341 | 141.9444 | |
|
| 0.0109 | 23.0 | 9154 | 0.9472 | 58.3537 | 45.2341 | 47.8222 | 56.8061 | 141.8148 | |
|
| 0.0082 | 24.0 | 9552 | 0.9426 | 58.1553 | 45.6645 | 49.019 | 56.7908 | 142.0 | |
|
| 0.0082 | 25.0 | 9950 | 0.9407 | 58.3571 | 46.0699 | 49.382 | 57.1456 | 142.0 | |
|
| 0.0071 | 26.0 | 10348 | 0.9654 | 59.5689 | 47.2126 | 50.5317 | 58.2492 | 142.0 | |
|
| 0.0057 | 27.0 | 10746 | 0.9651 | 58.2261 | 46.2797 | 49.8995 | 57.0725 | 142.0 | |
|
| 0.0049 | 28.0 | 11144 | 0.9555 | 57.3502 | 44.2364 | 47.6214 | 55.69 | 142.0 | |
|
| 0.0043 | 29.0 | 11542 | 0.9591 | 57.3909 | 44.5927 | 47.541 | 56.2071 | 142.0 | |
|
| 0.0043 | 30.0 | 11940 | 0.9637 | 58.3275 | 46.1513 | 49.4288 | 57.073 | 142.0 | |
|
| 0.0033 | 31.0 | 12338 | 0.9705 | 58.4669 | 46.613 | 49.5711 | 57.3531 | 142.0 | |
|
| 0.0031 | 32.0 | 12736 | 0.9707 | 58.6575 | 47.1055 | 50.0715 | 57.58 | 142.0 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.18.0 |
|
- Pytorch 1.11.0+cu113 |
|
- Datasets 2.1.0 |
|
- Tokenizers 0.12.1 |
|
|
