metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: headline_geneartor
results: []
headline_geneartor
This model is a fine-tuned version of facebook/bart-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.8033
- Rouge1: 0.4798
- Rouge2: 0.2432
- Rougel: 0.4377
- Rougelsum: 0.4375
- Gen Len: 17.7487
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.1366 | 1.0 | 3557 | 1.8742 | 0.4718 | 0.2344 | 0.4309 | 0.4308 | 17.7863 |
| 1.8889 | 2.0 | 7114 | 1.8210 | 0.4741 | 0.2392 | 0.433 | 0.4329 | 17.7028 |
| 1.7281 | 3.0 | 10671 | 1.8087 | 0.4797 | 0.2417 | 0.4374 | 0.4372 | 17.7501 |
| 1.6333 | 4.0 | 14228 | 1.8033 | 0.4798 | 0.2432 | 0.4377 | 0.4375 | 17.7487 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.13.0
- Tokenizers 0.13.3