Hugging Face's logo

joheras
/
NASES-clara-med

Text2Text Generation
Transformers
PyTorch
TensorBoard
bart
simplification
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
NASES-clara-med / README.md
joheras's picture
joheras
Update README.md
d5c4c27
preview code
|
raw
history blame
No virus
4.29 kB
metadata
tags:
  - simplification
  - generated_from_trainer
metrics:
  - rouge
  - sari
model-index:
  - name: NASES-clara-med
    results: []
license: cc-by-nc-sa-4.0
datasets:
  - lcampillos/CLARA-MeD
language:
  - es

NASES-clara-med

This model is a fine-tuned version of ELiRF/NASES on the CLARA-MeD dataset. It achieves the following results on the evaluation set:

  • Loss: 3.1754
  • Rouge1: 45.2398
  • Rouge2: 27.7502
  • Rougel: 39.4698
  • Rougelsum: 39.7208
  • SARI: 49.5333

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: 16
  • eval_batch_size: 16
  • 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
No log 1.0 190 2.1363 43.1891 26.0464 37.7101 37.8669
No log 2.0 380 2.0887 44.3709 26.66 38.7491 38.9616
1.8749 3.0 570 2.0998 45.3838 27.6296 39.6766 39.8786
1.8749 4.0 760 2.2080 45.3734 27.9361 39.8229 39.9957
0.6851 5.0 950 2.3240 44.8206 27.4094 39.0302 39.249
0.6851 6.0 1140 2.4336 45.2087 27.6721 39.6997 39.9306
0.6851 7.0 1330 2.5224 45.3472 28.0703 39.9099 40.1756
0.2487 8.0 1520 2.5796 45.215 27.7175 39.5083 39.7442
0.2487 9.0 1710 2.6675 45.3478 27.5316 39.7082 39.9943
0.1383 10.0 1900 2.7055 44.6361 27.3284 38.8978 39.1641
0.1383 11.0 2090 2.7401 45.537 27.9101 39.8044 40.0529
0.1383 12.0 2280 2.7837 45.3551 27.7135 39.6413 39.8563
0.0866 13.0 2470 2.8190 45.9865 28.3685 40.3313 40.626
0.0866 14.0 2660 2.8380 45.3839 27.9721 39.8318 40.0786
0.065 15.0 2850 2.9169 45.3779 27.8374 39.7026 39.9432
0.065 16.0 3040 2.9225 45.3323 27.6681 39.5425 39.8021
0.065 17.0 3230 2.9558 45.507 28.2007 40.0316 40.3505
0.0465 18.0 3420 3.0746 45.5661 27.6864 39.7771 40.042
0.0465 19.0 3610 3.0260 45.4173 28.1651 39.9385 40.265
0.0287 20.0 3800 2.9955 44.8573 27.7183 39.3235 39.6152
0.0287 21.0 3990 3.0956 44.9341 27.481 39.4431 39.6973
0.0287 22.0 4180 3.1569 44.8046 27.4202 38.9288 39.2948
0.0205 23.0 4370 3.1127 45.6665 27.9091 39.9312 40.1756
0.0205 24.0 4560 3.1214 45.2634 27.757 39.6646 39.9734
0.0149 25.0 4750 3.1522 45.4023 27.961 39.6511 39.9969
0.0149 26.0 4940 3.1694 45.3276 27.7616 39.5195 39.776
0.0149 27.0 5130 3.1682 45.4472 27.8223 39.6778 39.9427
0.0126 28.0 5320 3.1421 45.4602 27.9026 39.8116 40.1192
0.0126 29.0 5510 3.1576 45.4435 27.9545 39.7496 39.9925
0.01 30.0 5700 3.1754 45.2398 27.7502 39.4698 39.7208

Framework versions

  • Transformers 4.25.1
  • Pytorch 1.13.0
  • Datasets 2.8.0
  • Tokenizers 0.12.1