README.md

metadata

license: mit
tags:
  - generated_from_trainer
datasets:
  - pubmed-summarization
metrics:
  - rouge
model-index:
  - name: bart-finetuned-summarization-pubmed
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: pubmed-summarization
          type: pubmed-summarization
          config: section
          split: validation
          args: section
        metrics:
          - name: Rouge1
            type: rouge
            value: 43.1219

bart-finetuned-summarization-pubmed

This model is a fine-tuned version of facebook/bart-large-cnn on the pubmed-summarization dataset. It achieves the following results on the evaluation set:

• Loss: 1.7193
• Rouge1: 43.1219
• Rouge2: 18.7311
• Rougel: 28.1006
• Rougelsum: 38.0914
• Gen Len: 128.6263

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: 10
• eval_batch_size: 10
• seed: 42
• gradient_accumulation_steps: 5
• total_train_batch_size: 50
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.8564	1.0	2398	1.7437	43.2294	18.867	28.2156	38.1868	128.4766
1.75	2.0	4796	1.7193	43.1219	18.7311	28.1006	38.0914	128.6263

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3