metadata

license: apache-2.0
tags:
  - summarization
  - generated_from_trainer
datasets:
  - wiki_lingua
metrics:
  - rouge
model-index:
  - name: wiki_lingua-fr-8-3-5.6e-05-mt5-small-finetuned
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: wiki_lingua
          type: wiki_lingua
          config: fr
          split: test
          args: fr
        metrics:
          - name: Rouge1
            type: rouge
            value: 19.9596

wiki_lingua-fr-8-3-5.6e-05-mt5-small-finetuned

This model is a fine-tuned version of google/mt5-small on the wiki_lingua dataset. It achieves the following results on the evaluation set:

Loss: 1.9117
Rouge1: 19.9596
Rouge2: 7.5052
Rougel: 17.4363
Rougelsum: 19.5192

Baseline LEAD-64

Rouge1: 22.4
Rouge2: 5.92
Rougel: 14.44
Rougelsum: 14.44

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: 5.6e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
2.8962	1.0	5428	2.0026	18.8621	6.6127	16.0264	18.4354
2.313	2.0	10856	1.9260	19.7274	7.2791	17.0466	19.2904
2.2248	3.0	16284	1.9117	19.9596	7.5052	17.4363	19.5192

Framework versions

Transformers 4.27.4
Pytorch 1.13.0
Datasets 2.1.0
Tokenizers 0.13.2