Edit model card

T5_summ_gen_v1

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

  • Loss: 2.0950
  • Rouge1: 0.1986
  • Rouge2: 0.1044
  • Rougel: 0.1726
  • Rougelsum: 0.1727
  • Gen Len: 19.0

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: 16
  • eval_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 20

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 62 2.2294 0.1988 0.1023 0.1715 0.1714 19.0
No log 2.0 124 2.2038 0.1998 0.1024 0.1727 0.1725 19.0
No log 3.0 186 2.1890 0.2011 0.1049 0.1744 0.1746 19.0
No log 4.0 248 2.1767 0.2002 0.1059 0.1736 0.1737 19.0
No log 5.0 310 2.1593 0.2015 0.1064 0.1739 0.1741 19.0
No log 6.0 372 2.1522 0.2022 0.1059 0.1747 0.175 19.0
No log 7.0 434 2.1404 0.2028 0.1078 0.1746 0.1748 19.0
No log 8.0 496 2.1369 0.2015 0.1061 0.1735 0.1737 19.0
2.382 9.0 558 2.1299 0.1999 0.1053 0.1723 0.1725 19.0
2.382 10.0 620 2.1205 0.2003 0.1058 0.173 0.1729 19.0
2.382 11.0 682 2.1170 0.1998 0.105 0.1727 0.1727 19.0
2.382 12.0 744 2.1122 0.2003 0.1057 0.1734 0.1734 19.0
2.382 13.0 806 2.1084 0.1993 0.1042 0.1725 0.1726 19.0
2.382 14.0 868 2.1046 0.1988 0.1037 0.1723 0.1725 19.0
2.382 15.0 930 2.1023 0.1992 0.1047 0.1727 0.1729 19.0
2.382 16.0 992 2.1006 0.1992 0.1047 0.1727 0.1729 19.0
2.2855 17.0 1054 2.0979 0.1983 0.1034 0.1722 0.1723 19.0
2.2855 18.0 1116 2.0961 0.1988 0.1046 0.1729 0.173 19.0
2.2855 19.0 1178 2.0953 0.1986 0.1044 0.1725 0.1726 19.0
2.2855 20.0 1240 2.0950 0.1986 0.1044 0.1726 0.1727 19.0

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.5
  • Tokenizers 0.14.1
Downloads last month
8

Finetuned from

Dataset used to train PavanPasidu/T5_summ_gen_v1

Evaluation results