Hugging Face's logo

roequitz
/
bart-abs-1509-0313-lr-0.0003-bs-4-maxep-10

Safetensors
bart
Generated from Trainer
Model card Files Community
bart-abs-1509-0313-lr-0.0003-bs-4-maxep-10 / README.md
roequitz's picture
roequitz
End of training
a70d2c9 verified
preview code
|
raw
history blame
4.12 kB
metadata
license: apache-2.0
base_model: sshleifer/distilbart-xsum-12-6
tags:
  - generated_from_trainer
model-index:
  - name: bart-abs-1509-0313-lr-0.0003-bs-4-maxep-10
    results: []

bart-abs-1509-0313-lr-0.0003-bs-4-maxep-10

This model is a fine-tuned version of sshleifer/distilbart-xsum-12-6 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 4.7482
  • Rouge/rouge1: 0.4015
  • Rouge/rouge2: 0.1493
  • Rouge/rougel: 0.329
  • Rouge/rougelsum: 0.3294
  • Bertscore/bertscore-precision: 0.8894
  • Bertscore/bertscore-recall: 0.8807
  • Bertscore/bertscore-f1: 0.8849
  • Meteor: 0.3397
  • Gen Len: 33.2

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: 0.0003
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Rouge/rouge1 Rouge/rouge2 Rouge/rougel Rouge/rougelsum Bertscore/bertscore-precision Bertscore/bertscore-recall Bertscore/bertscore-f1 Meteor Gen Len
1.722 1.0 217 2.6756 0.4274 0.1866 0.3588 0.3592 0.8936 0.884 0.8886 0.3527 32.9545
1.7652 2.0 434 2.7321 0.416 0.1726 0.3511 0.3521 0.8944 0.8818 0.888 0.3352 31.5909
1.135 3.0 651 2.9372 0.3752 0.1441 0.3163 0.3158 0.8968 0.8736 0.8849 0.2976 26.4
0.9762 4.0 868 3.1311 0.3959 0.1535 0.3344 0.3353 0.8893 0.8777 0.8833 0.3296 33.1273
0.7207 5.0 1085 3.3741 0.4028 0.1562 0.3388 0.3389 0.8889 0.8818 0.8852 0.3324 34.3273
0.3986 6.0 1302 3.4504 0.4245 0.1689 0.3493 0.3501 0.892 0.8834 0.8876 0.351 34.4727
0.2471 7.0 1519 3.8316 0.4096 0.1536 0.3384 0.3389 0.8922 0.8814 0.8867 0.3376 32.7909
0.1613 8.0 1736 4.2439 0.4201 0.1621 0.346 0.347 0.8921 0.8815 0.8866 0.3503 33.3
0.0989 9.0 1953 4.4784 0.4115 0.1499 0.3394 0.3408 0.8904 0.8825 0.8863 0.3409 34.0
0.0644 10.0 2170 4.7482 0.4015 0.1493 0.329 0.3294 0.8894 0.8807 0.8849 0.3397 33.2

Framework versions

  • Transformers 4.44.0
  • Pytorch 2.4.0
  • Datasets 2.21.0
  • Tokenizers 0.19.1