pegasus-arxiv-finetuned-pubmed

This model is a fine-tuned version of google/pegasus-arxiv on the pub_med_summarization_dataset dataset. It achieves the following results on the evaluation set:

• Loss: 1.8118
• Rouge1: 44.286
• Rouge2: 19.0477
• Rougel: 27.1122
• Rougelsum: 40.2609
• Gen Len: 230.586

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: 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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.65	1.0	1000	1.9848	40.6984	16.387	25.0097	36.4831	215.294
2.1317	2.0	2000	1.8524	43.6431	18.6794	26.7571	39.6642	224.646
2.0591	3.0	3000	1.8254	43.6707	18.5176	26.6015	39.6325	225.894
2.0109	4.0	4000	1.8138	44.1244	18.8866	26.8313	40.0913	229.656
1.9894	5.0	5000	1.8118	44.286	19.0477	27.1122	40.2609	230.586

Framework versions

• Transformers 4.17.0
• Pytorch 1.9.1
• Datasets 1.18.4
• Tokenizers 0.11.6