Edit model card

bart-large-cnn-finetuned-roundup-64

This model is a fine-tuned version of facebook/bart-large-cnn on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 2.4772
  • Rouge1: 46.5444
  • Rouge2: 27.4056
  • Rougel: 29.6779
  • Rougelsum: 44.0905
  • Gen Len: 142.0

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: 2
  • eval_batch_size: 2
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 64
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 132 1.3213 48.3389 28.6641 31.4086 45.6679 142.0
No log 2.0 264 1.2325 48.798 29.3068 31.4329 45.7945 142.0
No log 3.0 396 1.2791 47.1449 27.3965 30.56 44.4704 142.0
0.9574 4.0 528 1.3134 46.2319 25.6249 28.7673 43.7555 140.3
0.9574 5.0 660 1.3187 46.7313 25.3467 29.3873 43.9495 142.0
0.9574 6.0 792 1.4271 48.1638 27.8874 30.5334 45.9944 142.0
0.9574 7.0 924 1.4876 46.7481 25.7259 29.7214 43.7042 140.5
0.3303 8.0 1056 1.5259 46.7075 26.0716 29.5521 43.7312 142.0
0.3303 9.0 1188 1.6223 48.012 27.2795 30.4989 45.4644 142.0
0.3303 10.0 1320 1.6842 48.0074 26.8831 29.3396 45.1937 142.0
0.3303 11.0 1452 1.7317 46.52 26.5152 29.5124 43.8797 142.0
0.1478 12.0 1584 1.8087 47.5887 27.0488 29.8569 44.7318 140.8
0.1478 13.0 1716 1.8263 46.1251 25.8576 30.1698 42.7228 142.0
0.1478 14.0 1848 1.9459 46.4034 25.7039 28.2542 43.7254 142.0
0.1478 15.0 1980 1.9539 44.4666 24.5827 27.7147 41.9769 142.0
0.0779 16.0 2112 1.9654 47.2267 26.4562 29.7352 44.0823 142.0
0.0779 17.0 2244 1.9580 48.5086 28.0294 30.8311 45.6336 142.0
0.0779 18.0 2376 2.0065 48.293 28.5678 30.0243 45.1384 142.0
0.0499 19.0 2508 1.9313 49.0549 28.9695 32.0711 46.3834 142.0
0.0499 20.0 2640 2.0176 47.0121 25.1606 29.0108 44.1556 142.0
0.0499 21.0 2772 2.0711 48.3754 28.2221 30.772 45.8547 140.95
0.0499 22.0 2904 2.0848 45.7392 25.254 29.0833 43.0381 142.0
0.0335 23.0 3036 2.0711 47.2931 27.4573 30.718 44.5932 142.0
0.0335 24.0 3168 2.1200 50.515 30.4253 33.7045 47.6158 142.0
0.0335 25.0 3300 2.1097 46.4737 26.3055 29.0148 43.2135 142.0
0.0335 26.0 3432 2.1695 46.9099 26.5227 29.7757 44.0613 142.0
0.0249 27.0 3564 2.1494 47.8319 27.6364 31.3593 45.065 141.95
0.0249 28.0 3696 2.1510 47.504 26.8971 31.7196 45.0328 142.0
0.0249 29.0 3828 2.1612 46.8789 27.266 30.1009 43.8248 142.0
0.0249 30.0 3960 2.1579 47.7012 27.7761 30.935 44.3686 142.0
0.018 31.0 4092 2.1981 48.4703 29.167 31.9815 45.8005 142.0
0.018 32.0 4224 2.2332 45.9512 25.8111 29.2467 42.9234 142.0
0.018 33.0 4356 2.1944 47.7189 28.1413 30.9692 44.9361 142.0
0.018 34.0 4488 2.2589 50.9687 32.3987 36.5644 48.3938 142.0
0.0132 35.0 4620 2.2269 47.8241 28.0442 31.5535 44.9394 142.0
0.0132 36.0 4752 2.2865 47.4383 27.0825 30.4109 44.194 142.0
0.0132 37.0 4884 2.3267 49.1786 29.6416 32.875 46.8821 142.0
0.0095 38.0 5016 2.2872 48.2085 28.3304 32.1473 45.3571 142.0
0.0095 39.0 5148 2.3340 46.6762 26.1637 29.0149 43.5923 142.0
0.0095 40.0 5280 2.3425 46.7561 26.1645 29.6337 43.6188 142.0
0.0095 41.0 5412 2.3111 49.4118 29.9761 33.4765 46.601 142.0
0.0076 42.0 5544 2.3892 45.3335 25.0161 28.4124 41.9873 142.0
0.0076 43.0 5676 2.3808 46.2506 26.4283 29.3841 42.7488 142.0
0.0076 44.0 5808 2.3825 45.6823 26.0048 29.5501 42.6475 142.0
0.0076 45.0 5940 2.3592 47.9127 26.7924 30.2353 44.791 142.0
0.0051 46.0 6072 2.4206 46.0415 27.0681 29.9602 43.1225 142.0
0.0051 47.0 6204 2.4214 48.1229 29.0913 31.1828 45.0022 142.0
0.0051 48.0 6336 2.4176 47.3825 27.7622 30.4138 43.9047 142.0
0.0051 49.0 6468 2.4137 48.2544 28.277 31.5548 45.6053 142.0
0.0041 50.0 6600 2.4384 49.6459 30.186 33.0059 47.0483 142.0
0.0041 51.0 6732 2.4433 47.7279 27.7857 30.2982 45.0842 142.0
0.0041 52.0 6864 2.4068 48.6047 28.1758 31.2744 45.8336 142.0
0.0041 53.0 6996 2.4362 48.7095 29.3335 31.9509 46.4161 142.0
0.003 54.0 7128 2.4307 48.836 29.6069 32.4004 46.1986 142.0
0.003 55.0 7260 2.4292 47.2945 26.7577 28.9719 43.8988 142.0
0.003 56.0 7392 2.4425 45.2261 25.6879 28.8129 42.6474 142.0
0.0024 57.0 7524 2.4386 47.967 28.5415 32.2049 45.5111 142.0
0.0024 58.0 7656 2.4528 47.5552 27.6397 30.9151 44.2627 142.0
0.0024 59.0 7788 2.4574 46.7821 27.3368 30.6334 44.0533 142.0
0.0024 60.0 7920 2.4659 47.3507 26.8371 30.4566 44.4452 142.0
0.0018 61.0 8052 2.4766 47.9847 28.2678 30.0664 45.0071 142.0
0.0018 62.0 8184 2.4682 46.8392 27.1275 30.144 43.6379 142.0
0.0018 63.0 8316 2.4754 45.6338 26.2812 29.4831 42.8744 142.0
0.0018 64.0 8448 2.4772 46.5444 27.4056 29.6779 44.0905 142.0

Framework versions

  • Transformers 4.18.0
  • Pytorch 1.11.0+cu113
  • Datasets 2.1.0
  • Tokenizers 0.12.1
Downloads last month
9