Edit model card

qa-persian-bert-fa-zwnj-base

This model is a fine-tuned version of makhataei/qa-persian-bert-fa-zwnj-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 5.3843

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: 1.5625e-09
  • train_batch_size: 14
  • eval_batch_size: 14
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 100

Training results

Training Loss Epoch Step Validation Loss
5.125 1.0 9 5.3843
5.444 2.0 18 5.3843
5.3229 3.0 27 5.3843
5.2751 4.0 36 5.3843
5.375 5.0 45 5.3843
5.3294 6.0 54 5.3843
5.2187 7.0 63 5.3843
5.1924 8.0 72 5.3843
5.1873 9.0 81 5.3843
5.2205 10.0 90 5.3843
5.2005 11.0 99 5.3843
5.3034 12.0 108 5.3843
5.3776 13.0 117 5.3843
5.3787 14.0 126 5.3843
5.3281 15.0 135 5.3843
5.2438 16.0 144 5.3843
5.4077 17.0 153 5.3843
5.1741 18.0 162 5.3843
5.3613 19.0 171 5.3843
5.2547 20.0 180 5.3843
5.3764 21.0 189 5.3843
5.3364 22.0 198 5.3843
5.4365 23.0 207 5.3843
5.4776 24.0 216 5.3843
5.4576 25.0 225 5.3843
5.4154 26.0 234 5.3843
5.4173 27.0 243 5.3843
5.3745 28.0 252 5.3843
5.3702 29.0 261 5.3843
5.355 30.0 270 5.3843
5.4444 31.0 279 5.3843
5.3054 32.0 288 5.3843
5.3231 33.0 297 5.3843
5.4371 34.0 306 5.3843
5.3532 35.0 315 5.3843
5.4009 36.0 324 5.3843
5.3778 37.0 333 5.3843
5.4367 38.0 342 5.3843
5.4419 39.0 351 5.3843
5.3169 40.0 360 5.3843
5.4068 41.0 369 5.3843
5.4131 42.0 378 5.3843
5.438 43.0 387 5.3843
5.3087 44.0 396 5.3843
5.3824 45.0 405 5.3843
5.3873 46.0 414 5.3843
5.5421 47.0 423 5.3843
5.4856 48.0 432 5.3843
5.4418 49.0 441 5.3843
5.6124 50.0 450 5.3843
5.513 51.0 459 5.3843
5.4882 52.0 468 5.3843
5.4542 53.0 477 5.3843
5.5348 54.0 486 5.3843
5.4597 55.0 495 5.3843
5.498 56.0 504 5.3843
5.521 57.0 513 5.3843
5.4913 58.0 522 5.3843
5.4776 59.0 531 5.3843
5.4312 60.0 540 5.3843
5.4966 61.0 549 5.3843
5.6154 62.0 558 5.3843
5.5904 63.0 567 5.3843
5.5537 64.0 576 5.3843
5.5173 65.0 585 5.3843
5.7689 66.0 594 5.3843
5.6241 67.0 603 5.3843
5.4729 68.0 612 5.3843
5.6112 69.0 621 5.3843
5.6145 70.0 630 5.3843
5.6586 71.0 639 5.3843
5.5165 72.0 648 5.3843
5.626 73.0 657 5.3843
5.6916 74.0 666 5.3843
5.7136 75.0 675 5.3843
5.5844 76.0 684 5.3843
5.6844 77.0 693 5.3843
5.6695 78.0 702 5.3843
5.7414 79.0 711 5.3843
5.7771 80.0 720 5.3843
5.7905 81.0 729 5.3843
5.7723 82.0 738 5.3843
5.7076 83.0 747 5.3843
5.7264 84.0 756 5.3843
5.7092 85.0 765 5.3843
5.6928 86.0 774 5.3843
5.8473 87.0 783 5.3843
5.8604 88.0 792 5.3843
5.8269 89.0 801 5.3843
5.849 90.0 810 5.3843
5.9264 91.0 819 5.3843
5.827 92.0 828 5.3843
5.9179 93.0 837 5.3843
5.8771 94.0 846 5.3843
5.9364 95.0 855 5.3843
5.7973 96.0 864 5.3843
5.9082 97.0 873 5.3843
5.895 98.0 882 5.3843
5.8634 99.0 891 5.3843
5.957 100.0 900 5.3843

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.0.1+cu117
  • Datasets 2.15.0
  • Tokenizers 0.15.0
Downloads last month
6
Safetensors
Model size
118M params
Tensor type
F32

Finetuned from