README.md

metadata

tags:
  - summarization
  - en
  - seq2seq
  - mbart
  - Abstractive Summarization
  - generated_from_trainer
datasets:
  - cnn_dailymail
model-index:
  - name: mbert-finetune-en-cnn
    results:
      - task:
          type: summarization
          name: Summarization
        dataset:
          name: cnn_dailymail
          type: cnn_dailymail
          config: 3.0.0
          split: test
        metrics:
          - name: ROUGE-1
            type: rouge
            value: 40.5167
            verified: true
          - name: ROUGE-2
            type: rouge
            value: 17.9329
            verified: true
          - name: ROUGE-L
            type: rouge
            value: 27.6258
            verified: true
          - name: ROUGE-LSUM
            type: rouge
            value: 37.389
            verified: true
          - name: loss
            type: loss
            value: 2.2861809730529785
            verified: true
          - name: gen_len
            type: gen_len
            value: 87.9999
            verified: true

mbert-finetune-en-cnn

This model is a fine-tuned version of facebook/mbart-large-50 on the cnn_dailymail dataset. It achieves the following results on the evaluation set:

• Loss: 3.5577
• Rouge-1: 37.69
• Rouge-2: 16.47
• Rouge-l: 35.53
• Gen Len: 79.93
• Bertscore: 74.92

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: 0.0005
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 250
• num_epochs: 5
• label_smoothing_factor: 0.1

Training results

Framework versions

• Transformers 4.20.0
• Pytorch 1.11.0+cu113
• Datasets 2.3.2
• Tokenizers 0.12.1