Hugging Face's logo

ldos
/
text_shortening_model_v8

Text2Text Generation
Transformers
PyTorch
t5
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
text_shortening_model_v8 / README.md
ldos's picture
ldos
End of training
18fae5a
preview code
|
raw
history blame
13.6 kB
metadata
license: apache-2.0
base_model: t5-small
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: text_shortening_model_v8
    results: []

text_shortening_model_v8

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

  • Loss: 2.3248
  • Rouge1: 0.43
  • Rouge2: 0.2172
  • Rougel: 0.3684
  • Rougelsum: 0.3674
  • Bert precision: 0.8551
  • Bert recall: 0.8369
  • Average word count: 9.8214
  • Max word count: 17
  • Min word count: 5
  • Average token count: 15.5857
  • % shortened texts with length > 12: 17.1429

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: 64
  • eval_batch_size: 64
  • 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 Bert precision Bert recall Average word count Max word count Min word count Average token count % shortened texts with length > 12
0.2688 1.0 8 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.284 2.0 16 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.264 3.0 24 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2564 4.0 32 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2727 5.0 40 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2924 6.0 48 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2666 7.0 56 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2662 8.0 64 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2631 9.0 72 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2844 10.0 80 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2653 11.0 88 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2649 12.0 96 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2972 13.0 104 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2553 14.0 112 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.261 15.0 120 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2832 16.0 128 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2635 17.0 136 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2484 18.0 144 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2612 19.0 152 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2996 20.0 160 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2562 21.0 168 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2503 22.0 176 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2763 23.0 184 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2692 24.0 192 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.284 25.0 200 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2838 26.0 208 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2729 27.0 216 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2685 28.0 224 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2599 29.0 232 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2829 30.0 240 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2275 31.0 248 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2605 32.0 256 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2855 33.0 264 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.251 34.0 272 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2629 35.0 280 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2679 36.0 288 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2453 37.0 296 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2853 38.0 304 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2542 39.0 312 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2627 40.0 320 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2668 41.0 328 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2742 42.0 336 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2746 43.0 344 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2768 44.0 352 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2729 45.0 360 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2729 46.0 368 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2788 47.0 376 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.286 48.0 384 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2484 49.0 392 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429
0.2679 50.0 400 2.3248 0.43 0.2172 0.3684 0.3674 0.8551 0.8369 9.8214 17 5 15.5857 17.1429

Framework versions

  • Transformers 4.32.1
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.4
  • Tokenizers 0.13.3