t5 / README.md
palimisis's picture
End of training
4093988
metadata
license: apache-2.0
base_model: t5-small
tags:
  - generated_from_trainer
datasets:
  - billsum
metrics:
  - rouge
model-index:
  - name: t5
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: billsum
          type: billsum
          config: default
          split: ca_test
          args: default
        metrics:
          - name: Rouge1
            type: rouge
            value: 0.1248

t5

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

  • Loss: 2.0846
  • Rouge1: 0.1248
  • Rouge2: 0.0719
  • Rougel: 0.1096
  • Rougelsum: 0.1097
  • 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: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 8
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 248 2.3488 0.0936 0.0361 0.0801 0.0799 19.0
No log 2.0 496 2.2214 0.1098 0.0519 0.0953 0.0953 19.0
2.8859 3.0 744 2.1603 0.1243 0.0682 0.1082 0.1081 19.0
2.8859 4.0 992 2.1276 0.1255 0.0717 0.1098 0.1099 19.0
2.3699 5.0 1240 2.1063 0.1252 0.0715 0.1101 0.1102 19.0
2.3699 6.0 1488 2.0908 0.1254 0.0723 0.1099 0.1101 19.0
2.2841 7.0 1736 2.0862 0.1247 0.0718 0.1093 0.1094 19.0
2.2841 8.0 1984 2.0846 0.1248 0.0719 0.1096 0.1097 19.0

Framework versions

  • Transformers 4.36.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.0