metadata

license: apache-2.0
base_model: facebook/bart-base
tags:
  - generated_from_trainer
datasets:
  - clupubhealth
metrics:
  - rouge
model-index:
  - name: pubhealth-expanded-1
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: clupubhealth
          type: clupubhealth
          config: expanded
          split: test
          args: expanded
        metrics:
          - name: Rouge1
            type: rouge
            value: 28.6755

pubhealth-expanded-1

This model is a fine-tuned version of facebook/bart-base on the clupubhealth dataset. It achieves the following results on the evaluation set:

Loss: 2.3198
Rouge1: 28.6755
Rouge2: 9.2869
Rougel: 21.9675
Rougelsum: 22.2946
Gen Len: 19.85

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: 12
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 10
total_train_batch_size: 120
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	Rougelsum	Gen Len
3.6788	0.08	40	2.3758	29.5273	9.3588	22.4799	22.6212	19.835
3.4222	0.15	80	2.3484	29.0821	9.1988	22.3907	22.5996	19.88
3.3605	0.23	120	2.3500	29.2893	9.296	22.1247	22.4075	19.94
3.3138	0.31	160	2.3504	29.039	8.907	21.9631	22.2506	19.91
3.2678	0.39	200	2.3461	29.678	9.4429	22.3439	22.6962	19.92
3.2371	0.46	240	2.3267	28.535	9.1858	21.3721	21.6634	19.915
3.204	0.54	280	2.3330	29.0796	9.4283	21.8953	22.1867	19.885
3.1881	0.62	320	2.3164	29.1456	9.1919	21.9529	22.235	19.945
3.1711	0.69	360	2.3208	29.3212	9.4823	22.1643	22.4159	19.895
3.1752	0.77	400	2.3239	29.0408	9.3615	21.8007	22.0795	19.945
3.1591	0.85	440	2.3218	28.6336	9.2799	21.5843	21.9422	19.845
3.1663	0.93	480	2.3198	28.6755	9.2869	21.9675	22.2946	19.85

Framework versions

Transformers 4.31.0
Pytorch 2.0.1+cu117
Datasets 2.7.1
Tokenizers 0.13.2