ptt5-cstnews-1024 / README.md
arthurmluz's picture
Model save
900abb1
metadata
license: mit
base_model: unicamp-dl/ptt5-base-portuguese-vocab
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: ptt5-cstnews-1024
    results: []

ptt5-cstnews-1024

This model is a fine-tuned version of unicamp-dl/ptt5-base-portuguese-vocab on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 2.4269
  • Rouge1: 0.086
  • Rouge2: 0.0557
  • Rougel: 0.0758
  • Rougelsum: 0.0833
  • Gen Len: 19.0

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: 2
  • eval_batch_size: 2
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 88 3.0796 0.0605 0.0381 0.0539 0.0582 19.0
No log 2.0 176 2.7409 0.0841 0.0507 0.0737 0.0817 19.0
3.8249 3.0 264 2.6518 0.0866 0.0508 0.0747 0.0833 19.0
3.8249 4.0 352 2.5961 0.0869 0.0526 0.0758 0.0839 19.0
2.7351 5.0 440 2.5584 0.0869 0.0539 0.0759 0.0841 19.0
2.7351 6.0 528 2.5364 0.0858 0.0521 0.0743 0.0826 19.0
2.5802 7.0 616 2.5092 0.0847 0.0516 0.0736 0.0815 19.0
2.5802 8.0 704 2.5026 0.0855 0.055 0.075 0.0827 19.0
2.5802 9.0 792 2.4862 0.0852 0.0551 0.0749 0.0825 19.0
2.4864 10.0 880 2.4744 0.0853 0.0553 0.0751 0.0826 19.0
2.4864 11.0 968 2.4676 0.0871 0.0561 0.0764 0.0843 19.0
2.4328 12.0 1056 2.4627 0.0865 0.0561 0.0763 0.0837 19.0
2.4328 13.0 1144 2.4566 0.0877 0.0562 0.0765 0.0846 19.0
2.3615 14.0 1232 2.4495 0.0869 0.0559 0.0761 0.0842 19.0
2.3615 15.0 1320 2.4439 0.0869 0.0559 0.0761 0.0842 19.0
2.2926 16.0 1408 2.4447 0.0869 0.0559 0.0761 0.0842 19.0
2.2926 17.0 1496 2.4437 0.0866 0.0555 0.0759 0.0839 19.0
2.2926 18.0 1584 2.4345 0.0862 0.0557 0.076 0.0834 19.0
2.2657 19.0 1672 2.4342 0.0871 0.056 0.0764 0.0843 19.0
2.2657 20.0 1760 2.4328 0.0871 0.056 0.0764 0.0843 19.0
2.2425 21.0 1848 2.4317 0.0863 0.0558 0.0761 0.0836 19.0
2.2425 22.0 1936 2.4311 0.0863 0.0558 0.0761 0.0836 19.0
2.2338 23.0 2024 2.4292 0.0863 0.0558 0.0761 0.0836 19.0
2.2338 24.0 2112 2.4268 0.0861 0.056 0.0759 0.0836 19.0
2.203 25.0 2200 2.4270 0.0857 0.0559 0.0756 0.0833 19.0
2.203 26.0 2288 2.4290 0.0857 0.0557 0.0756 0.0833 19.0
2.203 27.0 2376 2.4272 0.0857 0.0557 0.0756 0.0833 19.0
2.1676 28.0 2464 2.4265 0.0857 0.0557 0.0756 0.0833 19.0
2.1676 29.0 2552 2.4273 0.086 0.0557 0.0758 0.0833 19.0
2.1984 30.0 2640 2.4269 0.086 0.0557 0.0758 0.0833 19.0

Framework versions

  • Transformers 4.34.0
  • Pytorch 2.0.1+cu117
  • Datasets 2.14.5
  • Tokenizers 0.14.1