Hugging Face's logo

ldos
/
text_shortening_model_v13

Text2Text Generation
Transformers
PyTorch
t5
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
1
text_shortening_model_v13 / README.md
ldos's picture
ldos
End of training
25da440
preview code
|
raw
history blame
No virus
16 kB
metadata
license: apache-2.0
base_model: t5-small
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: text_shortening_model_v13
    results: []

text_shortening_model_v13

This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.9666
  • Rouge1: 0.579
  • Rouge2: 0.3396
  • Rougel: 0.5272
  • Rougelsum: 0.5263
  • Bert precision: 0.8964
  • Bert recall: 0.8985
  • Average word count: 10.9714
  • Max word count: 16
  • Min word count: 6
  • Average token count: 16.1071
  • % shortened texts with length > 12: 22.8571

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: 0.0001
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 60

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Bert precision Bert recall Average word count Max word count Min word count Average token count % shortened texts with length > 12
1.4719 1.0 62 1.5878 0.5604 0.3377 0.508 0.5076 0.8858 0.8969 12.3214 17 4 17.0429 55.7143
1.2908 2.0 124 1.5191 0.5675 0.3397 0.5143 0.5145 0.8902 0.8974 11.8071 17 4 16.65 47.8571
1.1543 3.0 186 1.4723 0.5723 0.3432 0.5282 0.5278 0.8938 0.8988 11.45 17 5 16.4 39.2857
1.0903 4.0 248 1.4651 0.5785 0.3397 0.5266 0.5264 0.8888 0.9011 12.1071 17 7 17.2571 50.0
1.0048 5.0 310 1.4482 0.5768 0.334 0.5245 0.5249 0.8903 0.9007 12.1 17 6 17.1357 50.0
0.971 6.0 372 1.4231 0.5822 0.343 0.5346 0.5347 0.8954 0.9007 11.4071 16 6 16.4357 37.8571
0.9113 7.0 434 1.4383 0.585 0.3532 0.5375 0.5374 0.8937 0.9029 11.8143 17 6 16.8714 42.1429
0.8767 8.0 496 1.4240 0.5913 0.3592 0.5455 0.5449 0.8951 0.904 11.7071 16 6 16.8286 39.2857
0.8384 9.0 558 1.4367 0.5855 0.3567 0.5418 0.5413 0.8965 0.9016 11.4214 16 6 16.3857 34.2857
0.7953 10.0 620 1.4470 0.5878 0.3566 0.535 0.5346 0.8958 0.901 11.5143 16 6 16.4786 37.8571
0.7705 11.0 682 1.4686 0.585 0.3538 0.5325 0.5322 0.894 0.9002 11.6 16 7 16.75 37.8571
0.7354 12.0 744 1.4709 0.5875 0.3533 0.5415 0.5411 0.8947 0.9034 11.55 16 6 16.6643 35.7143
0.6957 13.0 806 1.5007 0.5912 0.3556 0.549 0.5481 0.8963 0.9027 11.5 16 7 16.6429 33.5714
0.6853 14.0 868 1.5181 0.5778 0.3461 0.5356 0.5348 0.8936 0.8993 11.4286 16 6 16.4571 28.5714
0.6519 15.0 930 1.5308 0.5817 0.3467 0.5355 0.5345 0.894 0.901 11.4714 16 7 16.5571 30.7143
0.656 16.0 992 1.5374 0.5821 0.3458 0.5331 0.5321 0.8961 0.8994 11.1429 16 6 16.1429 26.4286
0.6235 17.0 1054 1.5780 0.5913 0.352 0.5454 0.5452 0.8964 0.9045 11.4857 16 6 16.6929 29.2857
0.608 18.0 1116 1.6061 0.5871 0.3545 0.5436 0.5432 0.8953 0.9027 11.3786 16 6 16.6714 30.0
0.5764 19.0 1178 1.6206 0.5923 0.3592 0.5463 0.5458 0.8972 0.9026 11.2929 16 6 16.3857 30.7143
0.5657 20.0 1240 1.6333 0.5801 0.3407 0.5327 0.5322 0.894 0.9002 11.35 16 6 16.45 30.0
0.5405 21.0 1302 1.6460 0.5833 0.3433 0.5374 0.5366 0.8952 0.8978 11.0 16 6 16.1357 25.0
0.5335 22.0 1364 1.6747 0.5782 0.346 0.5376 0.5369 0.8977 0.8972 10.6857 16 6 15.6 22.1429
0.5398 23.0 1426 1.6644 0.5849 0.3528 0.542 0.5422 0.9003 0.8991 10.8429 16 6 15.7071 21.4286
0.5186 24.0 1488 1.6894 0.5741 0.3463 0.5341 0.5336 0.8961 0.8971 10.8714 16 6 15.8571 21.4286
0.4855 25.0 1550 1.6943 0.5849 0.3462 0.5368 0.5365 0.8955 0.8988 11.1071 16 6 16.0929 28.5714
0.4889 26.0 1612 1.7254 0.575 0.3487 0.5346 0.5347 0.8987 0.8978 10.7214 16 6 15.7357 21.4286
0.4822 27.0 1674 1.7385 0.5782 0.3446 0.5338 0.5335 0.8989 0.8984 10.7143 16 5 15.6714 22.8571
0.4725 28.0 1736 1.7633 0.5795 0.3447 0.5368 0.5362 0.8983 0.8982 10.7 16 5 15.7429 22.1429
0.4507 29.0 1798 1.7773 0.5714 0.3382 0.5293 0.5297 0.8954 0.898 10.9286 16 6 16.0429 25.0
0.4637 30.0 1860 1.7915 0.5787 0.3438 0.536 0.5359 0.8958 0.8997 11.1714 16 6 16.2429 27.1429
0.4589 31.0 1922 1.8094 0.5755 0.3387 0.5278 0.5279 0.8934 0.8982 11.2143 16 6 16.4286 27.8571
0.4234 32.0 1984 1.8093 0.5783 0.3429 0.5315 0.5319 0.8957 0.8983 11.0643 16 6 16.1786 25.7143
0.4398 33.0 2046 1.8084 0.5842 0.3471 0.5355 0.5347 0.8964 0.9004 11.1786 16 6 16.3143 25.0
0.4168 34.0 2108 1.8310 0.5839 0.3467 0.5348 0.5351 0.8972 0.8994 11.0429 16 6 16.1786 24.2857
0.4174 35.0 2170 1.8377 0.5802 0.3436 0.5306 0.5302 0.8964 0.8992 11.1214 16 6 16.2571 26.4286
0.4149 36.0 2232 1.8449 0.5791 0.3461 0.5306 0.5297 0.8951 0.9002 11.2929 16 6 16.5 27.1429
0.4029 37.0 2294 1.8459 0.5812 0.3451 0.533 0.5322 0.898 0.8986 10.8857 16 6 15.9286 22.1429
0.386 38.0 2356 1.8505 0.5861 0.3513 0.5373 0.5364 0.8969 0.9014 11.2429 16 6 16.3357 27.1429
0.3946 39.0 2418 1.8668 0.5877 0.3551 0.539 0.5377 0.8977 0.9014 11.1857 16 6 16.3071 25.7143
0.3889 40.0 2480 1.8692 0.585 0.3463 0.5347 0.5335 0.8985 0.9007 11.0357 16 6 16.1214 22.8571
0.3769 41.0 2542 1.8718 0.5795 0.3461 0.5336 0.5323 0.897 0.899 10.9286 16 6 16.0214 21.4286
0.3667 42.0 2604 1.9021 0.5803 0.3494 0.5313 0.5306 0.8965 0.8996 11.1071 16 6 16.2714 23.5714
0.3603 43.0 2666 1.9108 0.584 0.3486 0.5363 0.5353 0.8964 0.8987 11.0571 16 6 16.2714 22.8571
0.3732 44.0 2728 1.8997 0.5807 0.3458 0.533 0.5319 0.8973 0.899 10.9286 16 6 16.15 21.4286
0.3731 45.0 2790 1.9185 0.5816 0.3465 0.5319 0.5316 0.8984 0.899 10.8571 16 6 15.9429 19.2857
0.3663 46.0 2852 1.9283 0.5799 0.3443 0.5323 0.5312 0.8962 0.9002 11.2 16 7 16.3071 26.4286
0.3643 47.0 2914 1.9332 0.5769 0.3474 0.5287 0.5278 0.8962 0.8998 11.0643 16 6 16.3429 22.8571
0.3638 48.0 2976 1.9375 0.5766 0.3465 0.5274 0.5271 0.8956 0.9001 11.2929 16 7 16.4786 27.8571
0.3555 49.0 3038 1.9419 0.5682 0.3353 0.5215 0.5212 0.8947 0.8968 10.9143 16 6 16.0571 21.4286
0.3678 50.0 3100 1.9431 0.5815 0.3461 0.5313 0.531 0.8977 0.9003 10.9714 16 6 16.0286 22.1429
0.3439 51.0 3162 1.9477 0.5771 0.3414 0.5277 0.5271 0.8962 0.8993 11.0857 16 6 16.1857 25.0
0.3538 52.0 3224 1.9492 0.5764 0.3386 0.5269 0.5262 0.8959 0.899 11.05 16 6 16.1286 23.5714
0.3556 53.0 3286 1.9604 0.5762 0.3343 0.5258 0.5254 0.8955 0.8989 11.1643 16 6 16.2429 27.8571
0.3385 54.0 3348 1.9604 0.5784 0.3389 0.5282 0.5271 0.8966 0.8982 10.95 16 6 16.05 23.5714
0.3353 55.0 3410 1.9585 0.5796 0.3449 0.5313 0.531 0.8969 0.8996 11.0786 16 6 16.1071 25.7143
0.3472 56.0 3472 1.9639 0.5778 0.3379 0.5287 0.5282 0.8964 0.899 11.0857 16 6 16.1214 25.7143
0.3352 57.0 3534 1.9661 0.5758 0.335 0.5254 0.5246 0.8958 0.8984 11.0571 16 6 16.1429 25.0
0.3401 58.0 3596 1.9666 0.5754 0.3353 0.5255 0.5249 0.8952 0.898 11.0643 16 6 16.2 24.2857
0.3446 59.0 3658 1.9664 0.5829 0.3413 0.5312 0.5309 0.8969 0.8994 10.9857 16 6 16.1357 23.5714
0.3376 60.0 3720 1.9666 0.579 0.3396 0.5272 0.5263 0.8964 0.8985 10.9714 16 6 16.1071 22.8571

Framework versions

  • Transformers 4.33.0
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.4
  • Tokenizers 0.13.3