Hugging Face's logo

pki
/
t5-small-finetuned_xsum

Text2Text Generation
Transformers
PyTorch
TensorBoard
Safetensors
t5
Generated from Trainer
Eval Results
text-generation-inference
Inference Endpoints
Model card Files Metrics Community
2
t5-small-finetuned_xsum / README.md
pki's picture
pki
Update README.md
65e211b
preview code
|
raw
history blame
6.93 kB
metadata
tags:
  - generated_from_trainer
datasets:
  - xsum
metrics:
  - rouge
model-index:
  - name: t5-small-finetuned_xsum
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: xsum
          type: xsum
          args: default
        metrics:
          - name: Rouge1
            type: rouge
            value: 34.0559

t5-small-finetuned_xsum

This model is a fine-tuned version of pki/t5-small-finetuned_xsum on the xsum dataset. It achieves the following results on the evaluation set:

  • Loss: 2.0479
  • Rouge1: 34.0559
  • Rouge2: 12.7506
  • Rougel: 27.6762
  • Rougelsum: 27.68
  • Gen Len: 18.7924

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: 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: 50

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
2.1176 1.0 12753 2.0913 33.1548 11.8434 26.7805 26.7751 18.7805
2.1019 2.0 25506 2.0875 33.231 11.9329 26.8674 26.861 18.7992
2.1044 3.0 38259 2.0846 33.3643 11.9807 26.9817 26.9764 18.773
2.0874 4.0 51012 2.0832 33.3562 12.0681 27.0178 27.0189 18.7988
2.0791 5.0 63765 2.0803 33.38 12.081 27.0368 27.0344 18.7844
2.0894 6.0 76518 2.0787 33.2549 11.9662 26.8674 26.8669 18.7975
2.0802 7.0 89271 2.0777 33.3978 12.0828 27.0461 27.0443 18.7757
2.0719 8.0 102024 2.0743 33.4083 12.1141 27.0523 27.0457 18.7928
2.0782 9.0 114777 2.0748 33.3673 12.1637 27.0696 27.0663 18.7902
2.0736 10.0 127530 2.0713 33.5771 12.2219 27.1707 27.1706 18.7945
2.0816 11.0 140283 2.0703 33.5099 12.2069 27.1822 27.1835 18.8002
2.057 12.0 153036 2.0693 33.5853 12.2427 27.2096 27.2109 18.806
2.0584 13.0 165789 2.0676 33.4883 12.2674 27.1582 27.154 18.7857
2.0475 14.0 178542 2.0662 33.5529 12.2765 27.1897 27.1901 18.79
2.0426 15.0 191295 2.0643 33.6543 12.3545 27.2946 27.2928 18.8036
2.0373 16.0 204048 2.0648 33.6671 12.349 27.2649 27.2707 18.7905
2.0178 17.0 216801 2.0637 33.6794 12.4545 27.3015 27.3079 18.7948
2.0235 18.0 229554 2.0626 33.7635 12.423 27.3475 27.3446 18.7892
2.0296 19.0 242307 2.0622 33.7574 12.4651 27.3879 27.3882 18.8134
2.0319 20.0 255060 2.0595 33.9093 12.5389 27.5003 27.5001 18.7915
2.0208 21.0 267813 2.0583 33.7875 12.4912 27.4243 27.4332 18.7982
2.0151 22.0 280566 2.0581 33.8516 12.4805 27.46 27.4647 18.816
2.0188 23.0 293319 2.0575 33.7744 12.4548 27.381 27.382 18.802
2.0087 24.0 306072 2.0579 33.8953 12.4984 27.4675 27.4727 18.7819
2.0186 25.0 318825 2.0557 33.7766 12.4414 27.4025 27.4024 18.8005
2.0051 26.0 331578 2.0555 33.8973 12.5796 27.5338 27.5339 18.8153
2.0024 27.0 344331 2.0557 33.8709 12.5116 27.4684 27.4664 18.7911
1.9947 28.0 357084 2.0545 33.8499 12.5242 27.4677 27.4716 18.8025
1.9931 29.0 369837 2.0545 33.7957 12.5272 27.4129 27.4174 18.8
1.9826 30.0 382590 2.0548 33.9723 12.6665 27.5598 27.5662 18.7958
1.999 31.0 395343 2.0522 33.9702 12.6435 27.5788 27.579 18.795
1.9872 32.0 408096 2.0525 33.9546 12.638 27.5985 27.5949 18.7976
1.991 33.0 420849 2.0520 33.9792 12.6073 27.5686 27.5707 18.8056
2.0044 34.0 433602 2.0504 34.0736 12.6511 27.647 27.6472 18.8093
1.9972 35.0 446355 2.0513 34.0506 12.711 27.6533 27.6537 18.7984
1.9901 36.0 459108 2.0504 33.9991 12.638 27.626 27.6272 18.7996
1.9742 37.0 471861 2.0507 33.9357 12.6636 27.5673 27.5716 18.8064
1.984 38.0 484614 2.0502 33.9476 12.6589 27.58 27.5813 18.8037
1.9864 39.0 497367 2.0499 34.0733 12.7198 27.6926 27.6992 18.8061
1.9734 40.0 510120 2.0492 33.9483 12.6486 27.5571 27.5598 18.8033
1.9895 41.0 522873 2.0490 33.9753 12.684 27.6058 27.6086 18.8011
1.964 42.0 535626 2.0487 33.9528 12.6376 27.576 27.5824 18.7919
1.9849 43.0 548379 2.0487 33.9868 12.6936 27.6116 27.6158 18.7966
1.9798 44.0 561132 2.0491 34.0379 12.7161 27.6227 27.6315 18.7889
1.9837 45.0 573885 2.0473 34.0046 12.6559 27.5931 27.5988 18.7996
1.9556 46.0 586638 2.0483 34.0378 12.712 27.6346 27.6446 18.7942
1.9844 47.0 599391 2.0479 34.0301 12.7121 27.6492 27.6554 18.7999
1.9869 48.0 612144 2.0474 34.0463 12.7151 27.6542 27.6604 18.7919
1.9851 49.0 624897 2.0476 34.0549 12.7384 27.6542 27.6555 18.7924
1.9912 50.0 637650 2.0479 34.0559 12.7506 27.6762 27.68 18.7924

Framework versions

  • Transformers 4.12.0.dev0
  • Pytorch 1.10.1
  • Datasets 1.14.0
  • Tokenizers 0.10.3