ArabicNewSplits7_OSS_usingWellWrittenEssays_FineTuningAraBERT_run3_AugV5_k1_task1_organization

This model is a fine-tuned version of aubmindlab/bert-base-arabertv02 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.8396
  • Qwk: 0.6619
  • Mse: 0.8396
  • Rmse: 0.9163

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: 8
  • eval_batch_size: 8
  • 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 Qwk Mse Rmse
No log 0.25 2 7.9385 -0.0293 7.9385 2.8175
No log 0.5 4 5.1544 0.0484 5.1544 2.2703
No log 0.75 6 3.7354 0.1164 3.7354 1.9327
No log 1.0 8 2.3826 0.0789 2.3826 1.5436
No log 1.25 10 1.7186 0.2569 1.7186 1.3110
No log 1.5 12 1.7228 0.1887 1.7228 1.3126
No log 1.75 14 1.7162 0.1165 1.7162 1.3100
No log 2.0 16 1.6679 0.1165 1.6679 1.2915
No log 2.25 18 1.6137 0.1538 1.6137 1.2703
No log 2.5 20 1.6243 0.1905 1.6243 1.2745
No log 2.75 22 1.7201 0.2056 1.7201 1.3115
No log 3.0 24 1.6855 0.1887 1.6855 1.2983
No log 3.25 26 1.5561 0.2243 1.5561 1.2474
No log 3.5 28 1.4150 0.2593 1.4150 1.1896
No log 3.75 30 1.2890 0.3303 1.2890 1.1354
No log 4.0 32 1.2181 0.3604 1.2181 1.1037
No log 4.25 34 1.1879 0.4348 1.1879 1.0899
No log 4.5 36 1.1841 0.4615 1.1841 1.0882
No log 4.75 38 1.2040 0.4793 1.2040 1.0973
No log 5.0 40 1.2823 0.3902 1.2823 1.1324
No log 5.25 42 1.1835 0.4839 1.1835 1.0879
No log 5.5 44 1.1127 0.5203 1.1127 1.0548
No log 5.75 46 1.0054 0.5669 1.0054 1.0027
No log 6.0 48 0.9667 0.5827 0.9667 0.9832
No log 6.25 50 0.9029 0.6418 0.9029 0.9502
No log 6.5 52 0.8206 0.6765 0.8206 0.9059
No log 6.75 54 0.9222 0.5984 0.9222 0.9603
No log 7.0 56 1.2301 0.4844 1.2301 1.1091
No log 7.25 58 1.1927 0.5312 1.1927 1.0921
No log 7.5 60 1.0617 0.5692 1.0617 1.0304
No log 7.75 62 0.8915 0.6222 0.8915 0.9442
No log 8.0 64 0.8804 0.6765 0.8804 0.9383
No log 8.25 66 0.8770 0.5821 0.8770 0.9365
No log 8.5 68 1.0373 0.6667 1.0373 1.0185
No log 8.75 70 1.2794 0.5848 1.2794 1.1311
No log 9.0 72 1.1820 0.5844 1.1820 1.0872
No log 9.25 74 1.1164 0.6043 1.1164 1.0566
No log 9.5 76 1.1808 0.5755 1.1808 1.0867
No log 9.75 78 1.0817 0.6099 1.0817 1.0400
No log 10.0 80 0.9407 0.7051 0.9407 0.9699
No log 10.25 82 0.9551 0.6792 0.9551 0.9773
No log 10.5 84 1.0323 0.6667 1.0323 1.0160
No log 10.75 86 0.8741 0.675 0.8741 0.9350
No log 11.0 88 0.7790 0.7123 0.7790 0.8826
No log 11.25 90 0.8553 0.7034 0.8553 0.9249
No log 11.5 92 0.8644 0.6573 0.8644 0.9298
No log 11.75 94 0.8155 0.6993 0.8155 0.9031
No log 12.0 96 0.7934 0.6806 0.7934 0.8907
No log 12.25 98 0.9552 0.6536 0.9552 0.9774
No log 12.5 100 1.1599 0.6257 1.1599 1.0770
No log 12.75 102 1.1224 0.6509 1.1224 1.0594
No log 13.0 104 0.9196 0.6494 0.9196 0.9590
No log 13.25 106 0.7888 0.6667 0.7888 0.8882
No log 13.5 108 0.8430 0.6714 0.8430 0.9181
No log 13.75 110 1.0141 0.6087 1.0141 1.0070
No log 14.0 112 1.1154 0.5455 1.1154 1.0561
No log 14.25 114 1.0894 0.4962 1.0894 1.0437
No log 14.5 116 1.0026 0.5672 1.0026 1.0013
No log 14.75 118 0.8806 0.6165 0.8806 0.9384
No log 15.0 120 0.8388 0.7101 0.8388 0.9159
No log 15.25 122 0.8445 0.6522 0.8445 0.9190
No log 15.5 124 0.9593 0.6923 0.9593 0.9794
No log 15.75 126 1.0445 0.6347 1.0445 1.0220
No log 16.0 128 1.0292 0.6545 1.0292 1.0145
No log 16.25 130 0.9376 0.6928 0.9376 0.9683
No log 16.5 132 0.8708 0.6993 0.8708 0.9332
No log 16.75 134 0.8123 0.6761 0.8123 0.9013
No log 17.0 136 0.8259 0.6621 0.8259 0.9088
No log 17.25 138 0.9630 0.6714 0.9630 0.9813
No log 17.5 140 1.0882 0.5985 1.0882 1.0432
No log 17.75 142 1.0453 0.5821 1.0453 1.0224
No log 18.0 144 0.9859 0.6389 0.9859 0.9929
No log 18.25 146 0.9407 0.6389 0.9407 0.9699
No log 18.5 148 0.9134 0.6277 0.9134 0.9557
No log 18.75 150 0.8856 0.6165 0.8856 0.9411
No log 19.0 152 0.8267 0.6418 0.8267 0.9092
No log 19.25 154 0.8717 0.6573 0.8717 0.9337
No log 19.5 156 0.9435 0.6579 0.9435 0.9713
No log 19.75 158 0.9744 0.6795 0.9744 0.9871
No log 20.0 160 0.9995 0.6795 0.9995 0.9997
No log 20.25 162 0.8358 0.7237 0.8358 0.9142
No log 20.5 164 0.7856 0.6857 0.7856 0.8863
No log 20.75 166 0.7375 0.6765 0.7375 0.8588
No log 21.0 168 0.7717 0.6950 0.7717 0.8785
No log 21.25 170 0.8712 0.6573 0.8712 0.9334
No log 21.5 172 0.8944 0.6573 0.8944 0.9457
No log 21.75 174 0.8328 0.6619 0.8328 0.9126
No log 22.0 176 0.7638 0.7007 0.7638 0.8739
No log 22.25 178 0.7465 0.7101 0.7465 0.8640
No log 22.5 180 0.8438 0.6710 0.8438 0.9186
No log 22.75 182 0.8801 0.6541 0.8801 0.9382
No log 23.0 184 0.8689 0.6497 0.8689 0.9322
No log 23.25 186 0.9716 0.6460 0.9716 0.9857
No log 23.5 188 0.9518 0.65 0.9518 0.9756
No log 23.75 190 0.9260 0.6582 0.9260 0.9623
No log 24.0 192 0.9295 0.6541 0.9295 0.9641
No log 24.25 194 1.0014 0.6424 1.0014 1.0007
No log 24.5 196 1.0334 0.6626 1.0334 1.0166
No log 24.75 198 0.8691 0.6853 0.8691 0.9323
No log 25.0 200 0.7628 0.6715 0.7628 0.8734
No log 25.25 202 0.7787 0.6993 0.7787 0.8824
No log 25.5 204 0.8950 0.7037 0.8950 0.9460
No log 25.75 206 1.0624 0.6786 1.0624 1.0307
No log 26.0 208 1.0469 0.6667 1.0469 1.0232
No log 26.25 210 0.9168 0.7013 0.9168 0.9575
No log 26.5 212 0.8216 0.7114 0.8216 0.9064
No log 26.75 214 0.8163 0.7034 0.8163 0.9035
No log 27.0 216 0.8886 0.6525 0.8886 0.9427
No log 27.25 218 0.9120 0.6483 0.9120 0.9550
No log 27.5 220 0.9052 0.6483 0.9052 0.9514
No log 27.75 222 0.8583 0.6471 0.8583 0.9264
No log 28.0 224 0.8477 0.6714 0.8477 0.9207
No log 28.25 226 0.8665 0.6569 0.8665 0.9309
No log 28.5 228 0.8860 0.6522 0.8860 0.9413
No log 28.75 230 0.8931 0.6377 0.8931 0.9450
No log 29.0 232 0.9202 0.6351 0.9202 0.9593
No log 29.25 234 0.8643 0.6711 0.8643 0.9297
No log 29.5 236 0.7376 0.7234 0.7376 0.8588
No log 29.75 238 0.6859 0.7324 0.6859 0.8282
No log 30.0 240 0.6908 0.7310 0.6908 0.8311
No log 30.25 242 0.6848 0.7483 0.6848 0.8275
No log 30.5 244 0.7587 0.7582 0.7587 0.8711
No log 30.75 246 0.8236 0.7097 0.8236 0.9075
No log 31.0 248 0.7762 0.7123 0.7762 0.8810
No log 31.25 250 0.7813 0.6806 0.7813 0.8839
No log 31.5 252 0.7838 0.6968 0.7838 0.8853
No log 31.75 254 0.8272 0.675 0.8272 0.9095
No log 32.0 256 0.9148 0.6667 0.9148 0.9564
No log 32.25 258 0.8767 0.6577 0.8767 0.9363
No log 32.5 260 0.7436 0.7050 0.7436 0.8623
No log 32.75 262 0.7408 0.6567 0.7408 0.8607
No log 33.0 264 0.7745 0.6906 0.7745 0.8800
No log 33.25 266 0.8395 0.6241 0.8395 0.9162
No log 33.5 268 0.9384 0.6483 0.9384 0.9687
No log 33.75 270 0.9325 0.625 0.9325 0.9657
No log 34.0 272 0.8306 0.6619 0.8306 0.9114
No log 34.25 274 0.7886 0.6901 0.7886 0.8880
No log 34.5 276 0.7890 0.6857 0.7890 0.8882
No log 34.75 278 0.8653 0.7030 0.8653 0.9302
No log 35.0 280 1.0008 0.6627 1.0008 1.0004
No log 35.25 282 1.0011 0.6824 1.0011 1.0006
No log 35.5 284 0.8275 0.7368 0.8275 0.9097
No log 35.75 286 0.6655 0.7297 0.6655 0.8158
No log 36.0 288 0.6579 0.7222 0.6579 0.8111
No log 36.25 290 0.6633 0.7092 0.6633 0.8144
No log 36.5 292 0.7094 0.7361 0.7094 0.8422
No log 36.75 294 0.8295 0.6667 0.8295 0.9108
No log 37.0 296 0.8943 0.6707 0.8943 0.9456
No log 37.25 298 0.8828 0.6623 0.8828 0.9396
No log 37.5 300 0.8250 0.7092 0.8250 0.9083
No log 37.75 302 0.7853 0.7050 0.7853 0.8862
No log 38.0 304 0.8093 0.7050 0.8093 0.8996
No log 38.25 306 0.8771 0.6667 0.8771 0.9365
No log 38.5 308 0.9167 0.6269 0.9167 0.9574
No log 38.75 310 0.8900 0.6370 0.8900 0.9434
No log 39.0 312 0.8143 0.6812 0.8143 0.9024
No log 39.25 314 0.7783 0.6812 0.7783 0.8822
No log 39.5 316 0.8228 0.6846 0.8228 0.9071
No log 39.75 318 0.8702 0.675 0.8702 0.9328
No log 40.0 320 0.8211 0.6839 0.8211 0.9061
No log 40.25 322 0.7526 0.7183 0.7526 0.8675
No log 40.5 324 0.7235 0.6950 0.7235 0.8506
No log 40.75 326 0.7397 0.7273 0.7397 0.8601
No log 41.0 328 0.8023 0.7059 0.8023 0.8957
No log 41.25 330 0.8483 0.6946 0.8483 0.9210
No log 41.5 332 0.9367 0.6784 0.9367 0.9678
No log 41.75 334 0.9445 0.6744 0.9445 0.9718
No log 42.0 336 0.8601 0.6977 0.8601 0.9274
No log 42.25 338 0.8004 0.7125 0.8004 0.8946
No log 42.5 340 0.8164 0.6797 0.8164 0.9035
No log 42.75 342 0.8745 0.6667 0.8745 0.9352
No log 43.0 344 0.9237 0.6014 0.9237 0.9611
No log 43.25 346 0.9099 0.6119 0.9099 0.9539
No log 43.5 348 0.8754 0.6119 0.8754 0.9356
No log 43.75 350 0.8705 0.6119 0.8705 0.9330
No log 44.0 352 0.8758 0.6269 0.8758 0.9358
No log 44.25 354 0.8529 0.6222 0.8529 0.9235
No log 44.5 356 0.8269 0.6806 0.8269 0.9093
No log 44.75 358 0.8254 0.6939 0.8254 0.9085
No log 45.0 360 0.8846 0.6538 0.8846 0.9405
No log 45.25 362 0.9899 0.6786 0.9899 0.9949
No log 45.5 364 1.0318 0.6587 1.0318 1.0158
No log 45.75 366 0.9882 0.6040 0.9882 0.9941
No log 46.0 368 0.9402 0.6759 0.9402 0.9696
No log 46.25 370 0.9150 0.6377 0.9150 0.9565
No log 46.5 372 0.9085 0.6377 0.9085 0.9532
No log 46.75 374 0.8812 0.7092 0.8812 0.9387
No log 47.0 376 0.8818 0.6277 0.8818 0.9391
No log 47.25 378 0.9066 0.6667 0.9066 0.9521
No log 47.5 380 0.9041 0.6667 0.9041 0.9508
No log 47.75 382 0.8372 0.6761 0.8372 0.9150
No log 48.0 384 0.7824 0.6471 0.7824 0.8845
No log 48.25 386 0.7885 0.6619 0.7885 0.8880
No log 48.5 388 0.8313 0.6759 0.8313 0.9117
No log 48.75 390 0.8372 0.6755 0.8372 0.9150
No log 49.0 392 0.8354 0.6792 0.8354 0.9140
No log 49.25 394 0.8954 0.6988 0.8954 0.9463
No log 49.5 396 0.8932 0.6909 0.8932 0.9451
No log 49.75 398 0.8389 0.6623 0.8389 0.9159
No log 50.0 400 0.8227 0.6806 0.8227 0.9070
No log 50.25 402 0.8370 0.6571 0.8370 0.9149
No log 50.5 404 0.8589 0.6809 0.8589 0.9268
No log 50.75 406 0.8835 0.6377 0.8835 0.9400
No log 51.0 408 0.9106 0.6241 0.9106 0.9542
No log 51.25 410 0.9342 0.6143 0.9342 0.9665
No log 51.5 412 0.9240 0.6377 0.9240 0.9612
No log 51.75 414 0.9043 0.6370 0.9043 0.9510
No log 52.0 416 0.8928 0.6370 0.8928 0.9449
No log 52.25 418 0.8829 0.6377 0.8829 0.9396
No log 52.5 420 0.9106 0.6497 0.9106 0.9543
No log 52.75 422 0.9066 0.6541 0.9066 0.9521
No log 53.0 424 0.8735 0.6625 0.8735 0.9346
No log 53.25 426 0.8244 0.7125 0.8244 0.9080
No log 53.5 428 0.8313 0.7125 0.8313 0.9118
No log 53.75 430 0.8625 0.6918 0.8625 0.9287
No log 54.0 432 0.8686 0.6795 0.8686 0.9320
No log 54.25 434 0.8380 0.6525 0.8380 0.9154
No log 54.5 436 0.8240 0.6812 0.8240 0.9077
No log 54.75 438 0.8262 0.6765 0.8262 0.9089
No log 55.0 440 0.8516 0.6812 0.8516 0.9228
No log 55.25 442 0.9069 0.6619 0.9069 0.9523
No log 55.5 444 0.9401 0.6383 0.9401 0.9696
No log 55.75 446 0.9196 0.6619 0.9196 0.9590
No log 56.0 448 0.8690 0.6618 0.8690 0.9322
No log 56.25 450 0.8100 0.6906 0.8100 0.9000
No log 56.5 452 0.7858 0.7 0.7858 0.8865
No log 56.75 454 0.7827 0.6993 0.7827 0.8847
No log 57.0 456 0.8038 0.7034 0.8038 0.8965
No log 57.25 458 0.8494 0.6624 0.8494 0.9216
No log 57.5 460 0.8557 0.6625 0.8557 0.9250
No log 57.75 462 0.8148 0.6962 0.8148 0.9027
No log 58.0 464 0.7878 0.7013 0.7878 0.8876
No log 58.25 466 0.7784 0.7059 0.7784 0.8823
No log 58.5 468 0.7566 0.7083 0.7566 0.8698
No log 58.75 470 0.7536 0.7083 0.7536 0.8681
No log 59.0 472 0.7646 0.7083 0.7646 0.8744
No log 59.25 474 0.7964 0.6846 0.7964 0.8924
No log 59.5 476 0.8370 0.6667 0.8370 0.9149
No log 59.75 478 0.8531 0.6623 0.8531 0.9237
No log 60.0 480 0.8390 0.6525 0.8390 0.9160
No log 60.25 482 0.8144 0.6765 0.8144 0.9024
No log 60.5 484 0.8191 0.6765 0.8191 0.9050
No log 60.75 486 0.8457 0.6667 0.8457 0.9196
No log 61.0 488 0.9106 0.6497 0.9106 0.9543
No log 61.25 490 0.9652 0.6748 0.9652 0.9825
No log 61.5 492 0.9729 0.6748 0.9729 0.9864
No log 61.75 494 0.9705 0.6748 0.9705 0.9851
No log 62.0 496 0.9191 0.6541 0.9191 0.9587
No log 62.25 498 0.8433 0.6667 0.8433 0.9183
0.3138 62.5 500 0.7819 0.6861 0.7819 0.8843
0.3138 62.75 502 0.7713 0.7050 0.7713 0.8782
0.3138 63.0 504 0.7653 0.7050 0.7653 0.8748
0.3138 63.25 506 0.7631 0.6861 0.7631 0.8735
0.3138 63.5 508 0.7885 0.6763 0.7885 0.8880
0.3138 63.75 510 0.8295 0.6338 0.8295 0.9108
0.3138 64.0 512 0.8539 0.6447 0.8539 0.9241
0.3138 64.25 514 0.8396 0.6619 0.8396 0.9163

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.0+cu118
  • Datasets 2.21.0
  • Tokenizers 0.19.1
Downloads last month
-
Safetensors
Model size
0.1B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for MayBashendy/ArabicNewSplits7_OSS_usingWellWrittenEssays_FineTuningAraBERT_run3_AugV5_k1_task1_organization

Finetuned
(4023)
this model