emilstabil's picture
update model card README.md
a12ff5e
metadata
license: apache-2.0
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: DanSumT5-smallV_55565
    results: []

DanSumT5-smallV_55565

This model is a fine-tuned version of Danish-summarisation/DanSumT5-small on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 2.6152
  • Rouge1: 33.2076
  • Rouge2: 9.7687
  • Rougel: 19.2885
  • Rougelsum: 30.5358
  • Gen Len: 125.4515

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: 80
  • eval_batch_size: 80
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 320
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 11

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 6 2.7915 28.1925 6.0798 16.0171 25.4406 117.8734
No log 2.0 12 2.7309 30.7001 7.8441 17.6277 28.1658 123.7384
No log 3.0 18 2.6932 31.9139 8.8623 18.4491 29.2043 125.2321
No log 4.0 24 2.6673 32.1541 9.2757 18.7349 29.3827 125.1941
No log 5.0 30 2.6506 32.6317 9.6369 18.9798 30.0012 125.6034
No log 6.0 36 2.6391 32.7076 9.7264 18.9488 29.9797 125.3376
No log 7.0 42 2.6307 32.9958 9.8324 19.0395 30.2766 125.0
No log 8.0 48 2.6241 33.2035 9.9866 19.1625 30.5136 125.2321
No log 9.0 54 2.6190 33.4626 10.076 19.2999 30.6955 125.4515
No log 10.0 60 2.6161 33.3145 9.9106 19.3186 30.6521 125.4515
No log 11.0 66 2.6152 33.2076 9.7687 19.2885 30.5358 125.4515

Framework versions

  • Transformers 4.30.2
  • Pytorch 1.12.1+git7548e2f
  • Datasets 2.13.2
  • Tokenizers 0.13.3