metadata
license: mit
base_model: facebook/bart-large-cnn
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: BART1
results: []
BART1
This model is a fine-tuned version of facebook/bart-large-cnn on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 3.8706
- Rouge1: 57.2472
- Rouge2: 23.1787
- Rougel: 41.8726
- Rougelsum: 53.8183
- Gen Len: 234.4232
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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 1
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 5.8303 | 0.0835 | 100 | 5.6762 | 48.0404 | 16.526 | 33.0315 | 45.2714 | 234.4232 |
| 5.2419 | 0.1671 | 200 | 5.1330 | 49.5121 | 17.8978 | 34.5708 | 46.291 | 234.4232 |
| 5.0085 | 0.2506 | 300 | 4.8037 | 52.3507 | 19.2179 | 36.3445 | 48.7473 | 234.4232 |
| 4.676 | 0.3342 | 400 | 4.5745 | 51.4939 | 19.2534 | 37.2441 | 48.7288 | 234.4232 |
| 4.4521 | 0.4177 | 500 | 4.4154 | 52.9389 | 20.2028 | 38.4139 | 49.9981 | 234.4232 |
| 4.4572 | 0.5013 | 600 | 4.2389 | 54.6029 | 21.0796 | 39.2355 | 51.1397 | 234.4232 |
| 4.2836 | 0.5848 | 700 | 4.1267 | 55.5174 | 22.1184 | 40.2744 | 52.0886 | 234.4232 |
| 4.2862 | 0.6684 | 800 | 4.0549 | 56.305 | 22.433 | 40.8636 | 52.6987 | 234.4232 |
| 4.0806 | 0.7519 | 900 | 3.9673 | 57.3033 | 22.873 | 41.2543 | 53.5936 | 234.4232 |
| 4.0806 | 0.8355 | 1000 | 3.9154 | 56.3519 | 22.7588 | 41.4512 | 52.9385 | 234.4232 |
| 3.8885 | 0.9190 | 1100 | 3.8706 | 57.2472 | 23.1787 | 41.8726 | 53.8183 | 234.4232 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1