Edit model card

predict-perception-xlmr-blame-victim

This model is a fine-tuned version of xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 1.1098
  • Rmse: 0.6801
  • Rmse Blame::a La vittima: 0.6801
  • Mae: 0.5617
  • Mae Blame::a La vittima: 0.5617
  • R2: -1.5910
  • R2 Blame::a La vittima: -1.5910
  • Cos: -0.1304
  • Pair: 0.0
  • Rank: 0.5
  • Neighbors: 0.3333
  • Rsa: nan

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: 1e-05
  • train_batch_size: 20
  • eval_batch_size: 8
  • seed: 1996
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Rmse Rmse Blame::a La vittima Mae Mae Blame::a La vittima R2 R2 Blame::a La vittima Cos Pair Rank Neighbors Rsa
1.0422 1.0 15 0.4952 0.4542 0.4542 0.4095 0.4095 -0.1560 -0.1560 -0.1304 0.0 0.5 0.2971 nan
1.0434 2.0 30 0.4851 0.4496 0.4496 0.4054 0.4054 -0.1324 -0.1324 -0.1304 0.0 0.5 0.2971 nan
1.038 3.0 45 0.4513 0.4337 0.4337 0.3885 0.3885 -0.0536 -0.0536 -0.1304 0.0 0.5 0.2971 nan
1.0151 4.0 60 0.4395 0.4280 0.4280 0.3840 0.3840 -0.0262 -0.0262 -0.1304 0.0 0.5 0.2715 nan
0.9727 5.0 75 0.4490 0.4325 0.4325 0.3811 0.3811 -0.0482 -0.0482 0.2174 0.0 0.5 0.3338 nan
0.9733 6.0 90 0.4540 0.4349 0.4349 0.3860 0.3860 -0.0598 -0.0598 -0.2174 0.0 0.5 0.3248 nan
0.9396 7.0 105 0.4501 0.4331 0.4331 0.3849 0.3849 -0.0508 -0.0508 0.0435 0.0 0.5 0.2609 nan
0.8759 8.0 120 0.4597 0.4377 0.4377 0.3849 0.3849 -0.0731 -0.0731 0.3043 0.0 0.5 0.3898 nan
0.8768 9.0 135 0.4575 0.4366 0.4366 0.3784 0.3784 -0.0680 -0.0680 0.4783 0.0 0.5 0.4615 nan
0.8312 10.0 150 0.5363 0.4727 0.4727 0.4071 0.4071 -0.2520 -0.2520 -0.0435 0.0 0.5 0.2733 nan
0.7296 11.0 165 0.5291 0.4696 0.4696 0.4057 0.4057 -0.2353 -0.2353 0.3043 0.0 0.5 0.3898 nan
0.7941 12.0 180 0.5319 0.4708 0.4708 0.4047 0.4047 -0.2417 -0.2417 0.1304 0.0 0.5 0.3381 nan
0.6486 13.0 195 0.6787 0.5318 0.5318 0.4516 0.4516 -0.5846 -0.5846 0.1304 0.0 0.5 0.3381 nan
0.6241 14.0 210 1.0146 0.6502 0.6502 0.5580 0.5580 -1.3687 -1.3687 -0.1304 0.0 0.5 0.3509 nan
0.5868 15.0 225 0.7164 0.5464 0.5464 0.4682 0.4682 -0.6725 -0.6725 -0.0435 0.0 0.5 0.3333 nan
0.5305 16.0 240 0.9064 0.6146 0.6146 0.5173 0.5173 -1.1161 -1.1161 -0.0435 0.0 0.5 0.3333 nan
0.495 17.0 255 1.3860 0.7600 0.7600 0.6433 0.6433 -2.2358 -2.2358 -0.0435 0.0 0.5 0.2935 nan
0.566 18.0 270 0.7618 0.5634 0.5634 0.4730 0.4730 -0.7785 -0.7785 0.0435 0.0 0.5 0.3225 nan
0.4305 19.0 285 0.8849 0.6072 0.6072 0.5048 0.5048 -1.0659 -1.0659 -0.0435 0.0 0.5 0.3333 nan
0.5108 20.0 300 0.7376 0.5544 0.5544 0.4716 0.4716 -0.7220 -0.7220 0.0435 0.0 0.5 0.3225 nan
0.44 21.0 315 1.1611 0.6956 0.6956 0.5921 0.5921 -1.7108 -1.7108 -0.1304 0.0 0.5 0.3333 nan
0.395 22.0 330 1.3004 0.7361 0.7361 0.6078 0.6078 -2.0360 -2.0360 -0.2174 0.0 0.5 0.3587 nan
0.3945 23.0 345 0.9376 0.6251 0.6251 0.5272 0.5272 -1.1890 -1.1890 -0.2174 0.0 0.5 0.3188 nan
0.3093 24.0 360 1.3586 0.7524 0.7524 0.6219 0.6219 -2.1719 -2.1719 -0.2174 0.0 0.5 0.3587 nan
0.2676 25.0 375 1.2200 0.7130 0.7130 0.5994 0.5994 -1.8484 -1.8484 -0.2174 0.0 0.5 0.3587 nan
0.3257 26.0 390 1.2235 0.7140 0.7140 0.5900 0.5900 -1.8564 -1.8564 -0.2174 0.0 0.5 0.3587 nan
0.4004 27.0 405 1.0978 0.6763 0.6763 0.5624 0.5624 -1.5629 -1.5629 -0.2174 0.0 0.5 0.3587 nan
0.283 28.0 420 1.1454 0.6909 0.6909 0.5697 0.5697 -1.6742 -1.6742 -0.2174 0.0 0.5 0.3587 nan
0.3326 29.0 435 1.1214 0.6836 0.6836 0.5646 0.5646 -1.6181 -1.6181 -0.1304 0.0 0.5 0.3333 nan
0.2632 30.0 450 1.1098 0.6801 0.6801 0.5617 0.5617 -1.5910 -1.5910 -0.1304 0.0 0.5 0.3333 nan

Framework versions

  • Transformers 4.16.2
  • Pytorch 1.10.2+cu113
  • Datasets 1.18.3
  • Tokenizers 0.11.0
Downloads last month
11