metadata
license: apache-2.0
tags:
- summarization
- generated_from_trainer
datasets:
- multi_news
metrics:
- rouge
model-index:
- name: multi-news-diff-weight
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: multi_news
type: multi_news
config: default
split: train[:95%]
args: default
metrics:
- name: Rouge1
type: rouge
value: 9.815
multi-news-diff-weight
This model is a fine-tuned version of facebook/bart-base on the multi_news dataset. It achieves the following results on the evaluation set:
- Loss: 2.3427
- Rouge1: 9.815
- Rouge2: 3.8774
- Rougel: 7.6169
- Rougelsum: 8.9863
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: 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: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| 2.75 | 1.0 | 19225 | 2.4494 | 9.5021 | 3.5429 | 7.3531 | 8.6912 |
| 2.456 | 2.0 | 38450 | 2.3665 | 9.8103 | 3.8494 | 7.6256 | 8.9991 |
| 2.285 | 3.0 | 57675 | 2.3427 | 9.815 | 3.8774 | 7.6169 | 8.9863 |
Framework versions
- Transformers 4.29.1
- Pytorch 2.0.0
- Datasets 2.12.0
- Tokenizers 0.13.3