metadata
license: apache-2.0
tags:
- summarization
- generated_from_trainer
datasets:
- cnn_dailymail
model-index:
- name: bart-base-finetuned-summarization-cnn-ver1.1
results: []
bart-base-finetuned-summarization-cnn-ver1.1
This model is a fine-tuned version of facebook/bart-base on the cnn_dailymail dataset. It achieves the following results on the evaluation set:
- Loss: 2.3824
- Bertscore-mean-precision: 0.8904
- Bertscore-mean-recall: 0.8610
- Bertscore-mean-f1: 0.8753
- Bertscore-median-precision: 0.8893
- Bertscore-median-recall: 0.8606
- Bertscore-median-f1: 0.8744
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
- 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 | Bertscore-mean-precision | Bertscore-mean-recall | Bertscore-mean-f1 | Bertscore-median-precision | Bertscore-median-recall | Bertscore-median-f1 |
---|---|---|---|---|---|---|---|---|---|
2.4217 | 1.0 | 5742 | 2.3095 | 0.8824 | 0.8582 | 0.8700 | 0.8822 | 0.8559 | 0.8696 |
1.7335 | 2.0 | 11484 | 2.2855 | 0.8907 | 0.8610 | 0.8754 | 0.8907 | 0.8600 | 0.8746 |
1.3013 | 3.0 | 17226 | 2.3824 | 0.8904 | 0.8610 | 0.8753 | 0.8893 | 0.8606 | 0.8744 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2