metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - summarize_from_feedback
metrics:
  - rouge
pipeline_tag: summarization
model-index:
  - name: flan-t5-large-finetuned-openai-summarize_from_feedback
    results:
      - task:
          type: text2text-generation
          name: Sequence-to-sequence Language Modeling
        dataset:
          name: summarize_from_feedback
          type: summarize_from_feedback
          config: comparisons
          split: train
          args: comparisons
        metrics:
          - type: rouge
            value: 30.2401
            name: Rouge1
          - type: rouge
            value: 11.4916
            name: Rouge2
          - type: rouge
            value: 24.6485
            name: RougeL
          - type: rouge
            value: 26.1801
            name: RougeLSum
      - task:
          type: summarization
          name: Summarization
        dataset:
          name: cnn_dailymail
          type: cnn_dailymail
          config: 3.0.0
          split: test
        metrics:
          - type: rouge
            value: 23.0407
            name: ROUGE-1
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZmViMTI3Mzg3ZDlkMDJlMjBiODU1NTIyNmVmOWFjMzIzZDVkNTc4NWI0MGIzYTJmYmUyMDM4ZTk2Y2Q5YzVjNSIsInZlcnNpb24iOjF9.oyqMHGGnGCE1f3JUNBg9c2ThycvlecuoZVWcGXvOcm0SbenpBobLEnczlFb4qx3ySwDUsL7uVtFW7F46Lz_CAg
          - type: rouge
            value: 8.5384
            name: ROUGE-2
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiOWZmOGYzODgyN2VjNjlmMWZiYzY3ZDIwNjAyZGExNDc3ZmZmMzIwMzk2NGVlNTZiMWIwZDUxM2IwNWU4OWQ5ZSIsInZlcnNpb24iOjF9.dCSVXAMFQQASe6fpPEOJu3Cfsd6Adm1L53xF0Job6W2Qd78M91wfl0715sUiFpsEKWKN9Z9bnhGA7d-SScVSAA
          - type: rouge
            value: 17.6719
            name: ROUGE-L
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZjkwOTA5YzNmYTc2YmRhNTFmYzc4MmE1ZjI5YmE1NDJiNjliNWRmNWU3MTg3MDgxY2RjZTZkMGZkZDY0MTIzNyIsInZlcnNpb24iOjF9.8-0VXD5ZGKIAGhjvuiBAchDZxyVWKczwiBxWDIQEItT3egSjYefGN8eOo9Z7R7sToX_li7IPeajVl3PbrQgPCg
          - type: rouge
            value: 20.9526
            name: ROUGE-LSUM
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZGNkYjdiMzE4MWQxZmIyMjYzM2Y0MDU3YmFlZDg1ZmE5NTBiZTVhM2NkYTRjMmU1ZWE5MmUwMmI2NmM5ODZkMyIsInZlcnNpb24iOjF9.qjFjlfIVHew80u5t_U44n9J6_PufNyv3faHaqML_pgo3VYBrbZWHX75jnBHThueWSK2hQhhmwaxSmR4ZYndRBA
          - type: loss
            value: 2.6858959197998047
            name: loss
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNWEwMzJiODEzY2Y0OTE3NzAyMzM1MjAxYzJjN2Y2ZGU3MGQ0MjFmZDg5ZGFlNGQ1YjJlN2I5MTFhNDE3NjZkYyIsInZlcnNpb24iOjF9.cPkbHIU3UQYMF7gUZx9Iqu-265jgv7pcgedRdVEsvxq2gfgdlyFDROQK9KI2cfk4GbQogsXEca91NdKohWFtCA
          - type: gen_len
            value: 18.9249
            name: gen_len
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiOTM3M2M0YjIzMTNmNTg2OTVmNDJhZjEwODEzNTBkODk0N2E0ZTZjNjg4MWY1OTk1ZGMzZTRmNzVkN2Y2ZDE4NyIsInZlcnNpb24iOjF9.Qo6HhXLr-j-aPKRL3ZVdMJMwTCQAUhAPUcLlN-2lGSS9tAoxVJEYr0O8SttMWbBDw3owivQdduxVre9SGUKuBQ

flan-t5-large-finetuned-openai-summarize_from_feedback

This model is a fine-tuned version of google/flan-t5-large on the summarize_from_feedback dataset. It achieves the following results on the evaluation set:

Loss: 2.3118
Rouge1: 30.2401
Rouge2: 11.4916
Rougel: 24.6485
Rougelsum: 26.1801
Gen Len: 18.8428

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: 5e-05
train_batch_size: 16
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 6

Training results

See Tensorboard

Citation

@misc {manuel_romero_2023,
    author       = { {Manuel Romero} },
    title        = { flan-t5-large-finetuned-openai-summarize_from_feedback (Revision 51666f9) },
    year         = 2023,
    url          = { https://huggingface.co/mrm8488/flan-t5-large-finetuned-openai-summarize_from_feedback },
    doi          = { 10.57967/hf/0266 },
    publisher    = { Hugging Face }
}

Framework versions

Transformers 4.25.1
Pytorch 1.13.0+cu116
Datasets 2.8.0
Tokenizers 0.13.2