README.md

metadata

base_model: google/pegasus-x-large
tags:
  - summarization
  - summary
  - booksum
  - long-document
  - long-form
datasets:
  - ubaada/booksum-complete-cleaned
language:
  - en
pipeline_tag: summarization
metrics:
  - rouge
model-index:
  - name: ubaada/pegasus-x-large-booksum-16k
    results:
      - task:
          type: summarization
          name: Summarization
        dataset:
          name: ubaada/booksum-complete-cleaned
          type: BookSum
          config: ubaada--booksum
          split: test
        metrics:
          - type: rouge
            value: 30.947853
            name: ROUGE-1
            verified: false
          - type: rouge
            value: 5.568146
            name: ROUGE-2
            verified: false

pegasus-x-large-booksum-16k

This model is a fine-tuned version of google/pegasus-x-large on ubaada/booksum-complete-cleaned. It was trained on the 'train' split of chapters sub-dataset. It achieves the following results on the evaluation set:

• Loss: 1.9677
• Rouge1: 0.3504
• Rouge2: 0.0525
• Rougel: 0.1398

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: 8e-05
• train_batch_size: 8
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel
1.417	0.9996	628	1.9677	0.3504	0.0525	0.1398

Framework versions

• Transformers 4.40.2
• Pytorch 2.2.0
• Datasets 2.19.1
• Tokenizers 0.19.1