bart-base-finetuned-summarization-cnn-ver3

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: 2.9827
• Bertscore-mean-precision: 0.8811
• Bertscore-mean-recall: 0.8554
• Bertscore-mean-f1: 0.8679
• Bertscore-median-precision: 0.8809
• Bertscore-median-recall: 0.8545
• Bertscore-median-f1: 0.8669

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: 0.0003
• 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: 1

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
3.632	1.0	5742	2.9827	0.8811	0.8554	0.8679	0.8809	0.8545	0.8669

Framework versions

• Transformers 4.24.0
• Pytorch 1.12.1+cu113
• Datasets 2.7.1
• Tokenizers 0.13.2