Edit model card

lora-roberta-large-no-anger-f4-0927

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

  • Loss: 0.7106
  • Accuracy: 0.7405
  • Prec: 0.7387
  • Recall: 0.7405
  • F1: 0.7387
  • B Acc: 0.5982
  • Micro F1: 0.7405
  • Prec Joy: 0.7558
  • Recall Joy: 0.7617
  • F1 Joy: 0.7587
  • Prec Anger: 0.6294
  • Recall Anger: 0.5631
  • F1 Anger: 0.5944
  • Prec Disgust: 0.4637
  • Recall Disgust: 0.3854
  • F1 Disgust: 0.4209
  • Prec Fear: 0.4892
  • Recall Fear: 0.5817
  • F1 Fear: 0.5315
  • Prec Neutral: 0.8292
  • Recall Neutral: 0.8481
  • F1 Neutral: 0.8385
  • Prec Sadness: 0.6600
  • Recall Sadness: 0.6140
  • F1 Sadness: 0.6362
  • Prec Surprise: 0.5320
  • Recall Surprise: 0.4331
  • F1 Surprise: 0.4775

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: 0.001
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.05
  • num_epochs: 25.0

Training results

Training Loss Epoch Step Validation Loss Accuracy Prec Recall F1 B Acc Micro F1 Prec Joy Recall Joy F1 Joy Prec Anger Recall Anger F1 Anger Prec Disgust Recall Disgust F1 Disgust Prec Fear Recall Fear F1 Fear Prec Neutral Recall Neutral F1 Neutral Prec Sadness Recall Sadness F1 Sadness Prec Surprise Recall Surprise F1 Surprise
0.8167 1.25 2049 0.7756 0.7130 0.7003 0.7130 0.7030 0.5272 0.7130 0.7252 0.7430 0.7340 0.6026 0.3749 0.4622 0.4187 0.3646 0.3898 0.5369 0.4170 0.4694 0.7763 0.8629 0.8173 0.6123 0.5784 0.5949 0.4797 0.3495 0.4044
0.7639 2.5 4098 0.7302 0.7293 0.7206 0.7293 0.7224 0.5662 0.7293 0.7361 0.7617 0.7487 0.6187 0.5198 0.5649 0.3881 0.4229 0.4048 0.5851 0.4247 0.4922 0.7961 0.8570 0.8254 0.6380 0.6185 0.6281 0.532 0.3585 0.4283
0.7395 3.75 6147 0.7348 0.7287 0.7328 0.7287 0.7271 0.5793 0.7287 0.6989 0.8136 0.7519 0.6786 0.4384 0.5327 0.4180 0.3875 0.4022 0.4632 0.5830 0.5162 0.8480 0.8134 0.8303 0.6648 0.5950 0.6280 0.5210 0.4241 0.4676
0.789 5.0 8196 0.7419 0.7275 0.7206 0.7275 0.7180 0.5511 0.7275 0.6888 0.8113 0.7450 0.6014 0.5183 0.5568 0.4038 0.4021 0.4029 0.5747 0.4305 0.4923 0.8063 0.8420 0.8238 0.6861 0.5838 0.6308 0.6224 0.2695 0.3762
0.7439 6.25 10245 0.7608 0.7207 0.7317 0.7207 0.7224 0.5858 0.7207 0.6882 0.8143 0.7459 0.6198 0.5004 0.5537 0.3944 0.3542 0.3732 0.4556 0.5843 0.5120 0.8599 0.7888 0.8228 0.7047 0.5590 0.6235 0.4535 0.4996 0.4754
0.712 7.5 12294 0.7240 0.7298 0.7270 0.7298 0.7263 0.5809 0.7298 0.7057 0.8043 0.7518 0.6313 0.4795 0.5450 0.4141 0.4271 0.4205 0.5707 0.4517 0.5043 0.8329 0.8214 0.8271 0.6126 0.6459 0.6288 0.5209 0.4367 0.4751
0.7032 8.75 14343 0.7095 0.7344 0.7328 0.7344 0.7317 0.5833 0.7344 0.7557 0.7479 0.7518 0.6391 0.5302 0.5796 0.4311 0.3521 0.3876 0.4724 0.6062 0.5310 0.8188 0.8498 0.8340 0.6472 0.6140 0.6301 0.5605 0.3827 0.4549
0.6972 10.0 16392 0.7108 0.7343 0.7325 0.7343 0.7317 0.5923 0.7343 0.7158 0.8038 0.7572 0.5785 0.5474 0.5625 0.3615 0.4729 0.4097 0.5714 0.4865 0.5255 0.8322 0.8288 0.8305 0.6797 0.5973 0.6358 0.5403 0.4097 0.4660
0.6859 11.25 18441 0.7211 0.7376 0.7321 0.7376 0.7322 0.5792 0.7376 0.7067 0.8093 0.7545 0.6216 0.5325 0.5736 0.4119 0.4188 0.4153 0.5720 0.4755 0.5193 0.8264 0.8407 0.8335 0.6677 0.6099 0.6375 0.5876 0.3675 0.4522
0.6542 12.5 20490 0.7143 0.7347 0.7294 0.7347 0.7307 0.5817 0.7347 0.7358 0.7824 0.7584 0.6263 0.5407 0.5804 0.3931 0.3792 0.3860 0.5700 0.4665 0.5131 0.8203 0.8364 0.8283 0.6158 0.6658 0.6398 0.5400 0.4007 0.4600
0.6463 13.75 22539 0.7022 0.7369 0.7366 0.7369 0.7354 0.5947 0.7369 0.7371 0.7864 0.7610 0.5452 0.6393 0.5885 0.5170 0.3167 0.3928 0.5519 0.4858 0.5168 0.8455 0.8218 0.8335 0.6062 0.6649 0.6342 0.5320 0.4483 0.4866
0.6333 15.0 24588 0.7106 0.7405 0.7387 0.7405 0.7387 0.5982 0.7405 0.7558 0.7617 0.7587 0.6294 0.5631 0.5944 0.4637 0.3854 0.4209 0.4892 0.5817 0.5315 0.8292 0.8481 0.8385 0.6600 0.6140 0.6362 0.5320 0.4331 0.4775
0.6184 16.25 26637 0.7199 0.7338 0.7389 0.7338 0.7348 0.6077 0.7338 0.7207 0.8008 0.7586 0.6140 0.5571 0.5842 0.3692 0.4292 0.3969 0.5024 0.5972 0.5457 0.8534 0.8079 0.8301 0.6714 0.6 0.6337 0.5109 0.4618 0.4851
0.5916 17.5 28686 0.7220 0.7368 0.7376 0.7368 0.7363 0.6003 0.7368 0.7426 0.7859 0.7636 0.5858 0.5713 0.5784 0.3743 0.4125 0.3925 0.5766 0.4653 0.5150 0.8479 0.8258 0.8367 0.5879 0.6676 0.6252 0.5146 0.4735 0.4932
0.5823 18.75 30735 0.7228 0.7376 0.7374 0.7376 0.7364 0.5960 0.7376 0.7210 0.8058 0.7610 0.6206 0.5534 0.5851 0.4056 0.3625 0.3828 0.5199 0.5631 0.5406 0.8460 0.8200 0.8328 0.6599 0.6126 0.6354 0.5254 0.4546 0.4875
0.5728 20.0 32784 0.7313 0.7344 0.7365 0.7344 0.7349 0.6090 0.7344 0.7295 0.7934 0.7601 0.5795 0.5907 0.5851 0.3927 0.4271 0.4092 0.5434 0.5161 0.5294 0.8462 0.8115 0.8285 0.6541 0.6311 0.6424 0.4928 0.4933 0.4930
0.5562 21.25 34833 0.7414 0.7376 0.7372 0.7376 0.7366 0.5995 0.7376 0.7372 0.7934 0.7643 0.6308 0.5258 0.5735 0.3946 0.425 0.4092 0.5324 0.5341 0.5332 0.8433 0.8267 0.8349 0.6139 0.6374 0.6254 0.5249 0.4537 0.4867
0.5348 22.5 36882 0.7398 0.7370 0.7374 0.7370 0.7365 0.6017 0.7370 0.7268 0.8039 0.7634 0.5844 0.5892 0.5868 0.4013 0.3937 0.3975 0.5331 0.5238 0.5284 0.8488 0.8163 0.8322 0.6473 0.6275 0.6372 0.5194 0.4573 0.4864
0.5202 23.75 38931 0.7423 0.7389 0.7379 0.7389 0.7381 0.6013 0.7389 0.7415 0.7893 0.7646 0.6020 0.5728 0.5871 0.4013 0.3896 0.3953 0.5341 0.5296 0.5318 0.8416 0.8279 0.8347 0.6410 0.6338 0.6374 0.5093 0.4663 0.4869

Framework versions

  • Transformers 4.33.1
  • Pytorch 2.0.1
  • Datasets 2.12.0
  • Tokenizers 0.13.3
Downloads last month
1
Unable to determine this model’s pipeline type. Check the docs .

Adapter for