|
--- |
|
license: mit |
|
base_model: facebook/bart-large-cnn |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: bart-large-xsumfinetuned-samsum |
|
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-large-xsumfinetuned-samsum |
|
|
|
This model is a fine-tuned version of [facebook/bart-large-cnn](https://huggingface.co./facebook/bart-large-cnn) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 2.9976 |
|
- Rouge1: 0.4246 |
|
- Rouge2: 0.2069 |
|
- Rougel: 0.3253 |
|
- Rougelsum: 0.3907 |
|
|
|
## 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: 4 |
|
- eval_batch_size: 4 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 6 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | |
|
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:---------:| |
|
| 0.2858 | 1.0 | 3683 | 2.3127 | 0.4051 | 0.1926 | 0.3072 | 0.3720 | |
|
| 0.3128 | 2.0 | 7366 | 2.3467 | 0.4007 | 0.1911 | 0.3037 | 0.3687 | |
|
| 0.2544 | 3.0 | 11049 | 2.3126 | 0.4145 | 0.2019 | 0.3159 | 0.3801 | |
|
| 0.1846 | 4.0 | 14732 | 2.6484 | 0.4088 | 0.1977 | 0.3096 | 0.3774 | |
|
| 0.1143 | 5.0 | 18415 | 2.7793 | 0.4173 | 0.1997 | 0.3155 | 0.3843 | |
|
| 0.0687 | 6.0 | 22098 | 2.9976 | 0.4246 | 0.2069 | 0.3253 | 0.3907 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.40.1 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.19.0 |
|
- Tokenizers 0.19.1 |
|
|
