bogdancazan's picture
update model card README.md
e70b614
|
raw
history blame
2.16 kB
metadata
tags:
  - generated_from_trainer
datasets:
  - billsum
metrics:
  - rouge
model-index:
  - name: prophetnet_summarization_pretrained
    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: 0.4982

prophetnet_summarization_pretrained

This model is a fine-tuned version of microsoft/prophetnet-large-uncased on the billsum dataset. It achieves the following results on the evaluation set:

  • Loss: 2.3683
  • Rouge1: 0.4982
  • Rouge2: 0.2267
  • Rougel: 0.2983
  • Rougelsum: 0.2985
  • Gen Len: 139.3831

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: 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: 4

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 124 2.5178 0.4894 0.2223 0.2903 0.2903 139.8105
No log 2.0 248 2.4170 0.4973 0.2279 0.2975 0.297 140.6492
No log 3.0 372 2.3895 0.4964 0.2282 0.2984 0.2981 138.5323
No log 4.0 496 2.3683 0.4982 0.2267 0.2983 0.2985 139.3831

Framework versions

  • Transformers 4.29.2
  • Pytorch 2.0.1+cu118
  • Datasets 2.12.0
  • Tokenizers 0.13.3