my_awesome_billsum_model

This model is a fine-tuned version of t5-small on the billsum dataset. It achieves the following results on the evaluation set:

• Loss: 2.1303
• Rouge1: 19.4885
• Rouge2: 9.7756
• Rougel: 16.7539
• Rougelsum: 18.153

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: 5.6e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
2.9287	1.0	124	2.3690	18.7685	8.721	15.7134	17.2109
2.5051	2.0	248	2.2540	19.5651	9.5886	16.5619	18.1252
2.4042	3.0	372	2.2140	19.4716	9.7429	16.6675	18.0006
2.3442	4.0	496	2.1800	19.5841	9.7078	16.7923	18.1682
2.3075	5.0	620	2.1562	19.4162	9.6647	16.5106	17.9637
2.2693	6.0	744	2.1394	19.5064	9.8462	16.6515	18.0461
2.2714	7.0	868	2.1321	19.475	9.7216	16.6698	18.1103
2.2413	8.0	992	2.1303	19.4885	9.7756	16.7539	18.153

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu116
• Datasets 2.7.1
• Tokenizers 0.13.2