metadata
license: mit
base_model: facebook/bart-large-xsum
tags:
- generated_from_trainer
metrics:
- rouge
- bleu
model-index:
- name: bart_samsum
results: []
datasets:
- samsum
pipeline_tag: summarization
bart_samsum
This model is a fine-tuned version of facebook/bart-large-xsum on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.4947
- Rouge1: 53.3294
- Rouge2: 28.6009
- Rougel: 44.2008
- Rougelsum: 49.2031
- Bleu: 0.0
- Meteor: 0.4887
- Gen Len: 30.1209
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
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Bleu | Meteor | Gen Len |
|---|---|---|---|---|---|---|---|---|---|---|
| 1.3838 | 0.9997 | 1841 | 1.5631 | 52.3252 | 27.2646 | 42.5893 | 48.2397 | 0.0 | 0.4825 | 32.0415 |
| 1.0835 | 2.0 | 3683 | 1.4947 | 53.3294 | 28.6009 | 44.2008 | 49.2031 | 0.0 | 0.4887 | 30.1209 |
| 0.8345 | 2.9997 | 5524 | 1.5956 | 52.1812 | 27.1239 | 42.9864 | 47.6384 | 0.0 | 0.4774 | 30.5446 |
| 0.672 | 4.0 | 7366 | 1.6695 | 52.8148 | 27.4815 | 43.3732 | 48.4633 | 0.0 | 0.4836 | 31.0342 |
| 0.538 | 4.9986 | 9205 | 1.8055 | 52.0988 | 26.762 | 42.5505 | 47.3721 | 0.0 | 0.4738 | 29.8901 |
Framework versions
- Transformers 4.40.0
- Pytorch 2.2.1+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1