ArabicNewSplits7_FineTuningAraBERT_run1_AugV5_k2_task5_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: 0.5317
  • Qwk: 0.6147
  • Mse: 0.5317
  • Rmse: 0.7292

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 4.1135 0.0130 4.1135 2.0282
No log 0.5 4 2.2846 0.1047 2.2846 1.5115
No log 0.75 6 1.3250 -0.0030 1.3250 1.1511
No log 1.0 8 1.0522 0.1837 1.0522 1.0258
No log 1.25 10 1.0442 0.1360 1.0442 1.0219
No log 1.5 12 1.0736 0.1545 1.0736 1.0362
No log 1.75 14 1.1683 0.1454 1.1683 1.0809
No log 2.0 16 1.3520 -0.0320 1.3520 1.1628
No log 2.25 18 1.4032 0.0 1.4032 1.1846
No log 2.5 20 1.0958 0.2125 1.0958 1.0468
No log 2.75 22 0.8511 0.3476 0.8511 0.9226
No log 3.0 24 0.8826 0.3432 0.8826 0.9395
No log 3.25 26 0.8140 0.3922 0.8140 0.9022
No log 3.5 28 0.8160 0.3353 0.8160 0.9033
No log 3.75 30 0.7580 0.4409 0.7580 0.8706
No log 4.0 32 0.7589 0.4980 0.7589 0.8712
No log 4.25 34 0.7522 0.4613 0.7522 0.8673
No log 4.5 36 0.7229 0.5879 0.7229 0.8502
No log 4.75 38 0.7666 0.5674 0.7666 0.8755
No log 5.0 40 0.6977 0.5763 0.6977 0.8353
No log 5.25 42 0.6845 0.6187 0.6845 0.8273
No log 5.5 44 0.7104 0.6144 0.7104 0.8428
No log 5.75 46 0.7129 0.6052 0.7129 0.8443
No log 6.0 48 0.6838 0.5816 0.6838 0.8269
No log 6.25 50 0.6372 0.6186 0.6372 0.7983
No log 6.5 52 0.6234 0.6310 0.6234 0.7895
No log 6.75 54 0.6459 0.6215 0.6459 0.8037
No log 7.0 56 0.6598 0.6206 0.6598 0.8123
No log 7.25 58 0.6649 0.6365 0.6649 0.8154
No log 7.5 60 0.6388 0.6011 0.6388 0.7993
No log 7.75 62 0.6563 0.6564 0.6563 0.8102
No log 8.0 64 0.6015 0.6555 0.6015 0.7755
No log 8.25 66 0.6032 0.6593 0.6032 0.7766
No log 8.5 68 0.7643 0.5833 0.7643 0.8742
No log 8.75 70 0.8268 0.6312 0.8268 0.9093
No log 9.0 72 0.7127 0.6193 0.7127 0.8442
No log 9.25 74 0.6450 0.6272 0.6450 0.8031
No log 9.5 76 0.6335 0.6598 0.6335 0.7959
No log 9.75 78 0.6357 0.6570 0.6357 0.7973
No log 10.0 80 0.7572 0.6304 0.7572 0.8702
No log 10.25 82 0.9212 0.5504 0.9212 0.9598
No log 10.5 84 0.7210 0.6831 0.7210 0.8491
No log 10.75 86 0.6260 0.6988 0.6260 0.7912
No log 11.0 88 0.6695 0.6259 0.6695 0.8182
No log 11.25 90 0.6167 0.6681 0.6167 0.7853
No log 11.5 92 0.7750 0.6312 0.7750 0.8803
No log 11.75 94 0.8945 0.4995 0.8945 0.9458
No log 12.0 96 0.7172 0.6082 0.7172 0.8469
No log 12.25 98 0.6394 0.6424 0.6394 0.7996
No log 12.5 100 0.6456 0.6066 0.6456 0.8035
No log 12.75 102 0.6289 0.6649 0.6289 0.7930
No log 13.0 104 0.6887 0.6062 0.6887 0.8299
No log 13.25 106 0.6350 0.6426 0.6350 0.7969
No log 13.5 108 0.6074 0.6649 0.6074 0.7794
No log 13.75 110 0.6064 0.6349 0.6064 0.7787
No log 14.0 112 0.6658 0.6184 0.6658 0.8159
No log 14.25 114 0.6647 0.5653 0.6647 0.8153
No log 14.5 116 0.5837 0.7089 0.5837 0.7640
No log 14.75 118 0.5879 0.7239 0.5879 0.7667
No log 15.0 120 0.6079 0.6677 0.6079 0.7797
No log 15.25 122 0.5616 0.7303 0.5616 0.7494
No log 15.5 124 0.5601 0.6805 0.5601 0.7484
No log 15.75 126 0.5788 0.6460 0.5788 0.7608
No log 16.0 128 0.6200 0.6460 0.6200 0.7874
No log 16.25 130 0.6486 0.5635 0.6486 0.8053
No log 16.5 132 0.6611 0.5794 0.6611 0.8131
No log 16.75 134 0.6112 0.6460 0.6112 0.7818
No log 17.0 136 0.5856 0.6788 0.5856 0.7653
No log 17.25 138 0.5840 0.6924 0.5840 0.7642
No log 17.5 140 0.5983 0.6491 0.5983 0.7735
No log 17.75 142 0.6557 0.6811 0.6557 0.8098
No log 18.0 144 0.6277 0.6254 0.6277 0.7923
No log 18.25 146 0.6655 0.6012 0.6655 0.8158
No log 18.5 148 0.6444 0.6177 0.6444 0.8028
No log 18.75 150 0.6370 0.5714 0.6370 0.7981
No log 19.0 152 0.6244 0.5274 0.6244 0.7902
No log 19.25 154 0.6407 0.6138 0.6407 0.8004
No log 19.5 156 0.6243 0.6256 0.6243 0.7901
No log 19.75 158 0.6335 0.6444 0.6335 0.7959
No log 20.0 160 0.7065 0.6745 0.7064 0.8405
No log 20.25 162 0.6679 0.6570 0.6679 0.8172
No log 20.5 164 0.5981 0.6452 0.5981 0.7733
No log 20.75 166 0.6486 0.6282 0.6486 0.8053
No log 21.0 168 0.6551 0.6464 0.6551 0.8094
No log 21.25 170 0.6059 0.6333 0.6059 0.7784
No log 21.5 172 0.6247 0.6526 0.6247 0.7904
No log 21.75 174 0.6673 0.6154 0.6673 0.8169
No log 22.0 176 0.6872 0.5826 0.6872 0.8290
No log 22.25 178 0.6612 0.6032 0.6612 0.8132
No log 22.5 180 0.6207 0.6509 0.6207 0.7878
No log 22.75 182 0.6530 0.6218 0.6530 0.8081
No log 23.0 184 0.6115 0.6177 0.6115 0.7820
No log 23.25 186 0.6004 0.6426 0.6004 0.7749
No log 23.5 188 0.6531 0.5873 0.6531 0.8081
No log 23.75 190 0.6373 0.6194 0.6373 0.7983
No log 24.0 192 0.5724 0.6636 0.5724 0.7566
No log 24.25 194 0.5672 0.6509 0.5672 0.7531
No log 24.5 196 0.6090 0.6664 0.6090 0.7804
No log 24.75 198 0.6033 0.6536 0.6033 0.7767
No log 25.0 200 0.6090 0.6452 0.6090 0.7804
No log 25.25 202 0.6431 0.5966 0.6431 0.8019
No log 25.5 204 0.6304 0.6269 0.6304 0.7940
No log 25.75 206 0.6107 0.6546 0.6107 0.7815
No log 26.0 208 0.6722 0.6199 0.6722 0.8199
No log 26.25 210 0.7143 0.6101 0.7143 0.8452
No log 26.5 212 0.6679 0.6255 0.6679 0.8173
No log 26.75 214 0.6080 0.6164 0.6080 0.7798
No log 27.0 216 0.6258 0.6174 0.6258 0.7911
No log 27.25 218 0.6103 0.6426 0.6103 0.7812
No log 27.5 220 0.5868 0.6729 0.5868 0.7660
No log 27.75 222 0.5818 0.6788 0.5818 0.7628
No log 28.0 224 0.5923 0.6602 0.5923 0.7696
No log 28.25 226 0.5878 0.6681 0.5878 0.7667
No log 28.5 228 0.5893 0.6858 0.5893 0.7677
No log 28.75 230 0.6067 0.5770 0.6067 0.7789
No log 29.0 232 0.6073 0.5326 0.6073 0.7793
No log 29.25 234 0.5736 0.7201 0.5736 0.7574
No log 29.5 236 0.5514 0.7041 0.5514 0.7426
No log 29.75 238 0.5635 0.6623 0.5635 0.7507
No log 30.0 240 0.5790 0.6575 0.5790 0.7609
No log 30.25 242 0.5537 0.6548 0.5537 0.7441
No log 30.5 244 0.5359 0.7176 0.5359 0.7320
No log 30.75 246 0.5394 0.6958 0.5394 0.7345
No log 31.0 248 0.5418 0.6958 0.5418 0.7361
No log 31.25 250 0.5430 0.6850 0.5430 0.7369
No log 31.5 252 0.5406 0.6978 0.5406 0.7352
No log 31.75 254 0.5605 0.6519 0.5605 0.7487
No log 32.0 256 0.5790 0.6157 0.5790 0.7609
No log 32.25 258 0.5576 0.6519 0.5576 0.7467
No log 32.5 260 0.5577 0.6846 0.5577 0.7468
No log 32.75 262 0.5565 0.6748 0.5565 0.7460
No log 33.0 264 0.5525 0.7103 0.5525 0.7433
No log 33.25 266 0.5601 0.6602 0.5601 0.7484
No log 33.5 268 0.5645 0.6602 0.5645 0.7513
No log 33.75 270 0.5592 0.6927 0.5592 0.7478
No log 34.0 272 0.5666 0.6927 0.5666 0.7527
No log 34.25 274 0.5898 0.6215 0.5898 0.7680
No log 34.5 276 0.5792 0.6402 0.5792 0.7610
No log 34.75 278 0.5682 0.6725 0.5682 0.7538
No log 35.0 280 0.5221 0.7041 0.5221 0.7226
No log 35.25 282 0.5134 0.7049 0.5134 0.7165
No log 35.5 284 0.5208 0.6890 0.5208 0.7217
No log 35.75 286 0.5115 0.6778 0.5115 0.7152
No log 36.0 288 0.5052 0.7151 0.5052 0.7108
No log 36.25 290 0.5098 0.6846 0.5098 0.7140
No log 36.5 292 0.5100 0.6888 0.5100 0.7141
No log 36.75 294 0.5124 0.6890 0.5124 0.7158
No log 37.0 296 0.5300 0.6890 0.5300 0.7280
No log 37.25 298 0.5347 0.6890 0.5347 0.7312
No log 37.5 300 0.5183 0.6890 0.5183 0.7199
No log 37.75 302 0.5096 0.6995 0.5096 0.7139
No log 38.0 304 0.5142 0.6909 0.5142 0.7171
No log 38.25 306 0.5141 0.6995 0.5141 0.7170
No log 38.5 308 0.5188 0.7253 0.5188 0.7203
No log 38.75 310 0.5237 0.6778 0.5237 0.7237
No log 39.0 312 0.5306 0.6778 0.5306 0.7284
No log 39.25 314 0.5303 0.6890 0.5303 0.7282
No log 39.5 316 0.5255 0.6778 0.5255 0.7249
No log 39.75 318 0.5207 0.6737 0.5207 0.7216
No log 40.0 320 0.5222 0.7253 0.5222 0.7226
No log 40.25 322 0.5317 0.7285 0.5317 0.7292
No log 40.5 324 0.5282 0.7285 0.5282 0.7268
No log 40.75 326 0.5220 0.7305 0.5220 0.7225
No log 41.0 328 0.5317 0.7001 0.5317 0.7292
No log 41.25 330 0.5409 0.6854 0.5409 0.7355
No log 41.5 332 0.5497 0.6570 0.5497 0.7414
No log 41.75 334 0.5504 0.6570 0.5504 0.7419
No log 42.0 336 0.5425 0.6553 0.5425 0.7365
No log 42.25 338 0.5370 0.6709 0.5370 0.7328
No log 42.5 340 0.5302 0.6581 0.5302 0.7282
No log 42.75 342 0.5239 0.6627 0.5239 0.7238
No log 43.0 344 0.5251 0.6882 0.5251 0.7247
No log 43.25 346 0.5372 0.6716 0.5372 0.7329
No log 43.5 348 0.5602 0.7222 0.5602 0.7484
No log 43.75 350 0.5829 0.7056 0.5829 0.7635
No log 44.0 352 0.5963 0.7056 0.5963 0.7722
No log 44.25 354 0.6157 0.7009 0.6157 0.7847
No log 44.5 356 0.6063 0.7009 0.6063 0.7787
No log 44.75 358 0.5930 0.7009 0.5930 0.7701
No log 45.0 360 0.5849 0.7009 0.5849 0.7648
No log 45.25 362 0.6042 0.6883 0.6042 0.7773
No log 45.5 364 0.5784 0.7056 0.5784 0.7605
No log 45.75 366 0.5755 0.7056 0.5755 0.7586
No log 46.0 368 0.5578 0.7222 0.5578 0.7468
No log 46.25 370 0.5287 0.7326 0.5287 0.7271
No log 46.5 372 0.5301 0.7001 0.5301 0.7281
No log 46.75 374 0.5325 0.7001 0.5325 0.7297
No log 47.0 376 0.5141 0.6882 0.5141 0.7170
No log 47.25 378 0.5072 0.6770 0.5072 0.7122
No log 47.5 380 0.5083 0.6888 0.5083 0.7130
No log 47.75 382 0.5033 0.7041 0.5033 0.7094
No log 48.0 384 0.5033 0.7201 0.5033 0.7094
No log 48.25 386 0.5027 0.7201 0.5027 0.7090
No log 48.5 388 0.5036 0.7201 0.5036 0.7097
No log 48.75 390 0.5034 0.7201 0.5034 0.7095
No log 49.0 392 0.5051 0.6940 0.5051 0.7107
No log 49.25 394 0.5041 0.6853 0.5041 0.7100
No log 49.5 396 0.5214 0.6501 0.5214 0.7221
No log 49.75 398 0.5286 0.6562 0.5286 0.7270
No log 50.0 400 0.5117 0.6536 0.5117 0.7154
No log 50.25 402 0.5030 0.7201 0.5030 0.7092
No log 50.5 404 0.5055 0.7033 0.5055 0.7110
No log 50.75 406 0.5073 0.7033 0.5073 0.7123
No log 51.0 408 0.5108 0.7033 0.5108 0.7147
No log 51.25 410 0.5133 0.7139 0.5133 0.7165
No log 51.5 412 0.5129 0.7352 0.5129 0.7162
No log 51.75 414 0.5283 0.6886 0.5283 0.7268
No log 52.0 416 0.5759 0.6272 0.5759 0.7589
No log 52.25 418 0.6138 0.6592 0.6138 0.7835
No log 52.5 420 0.5981 0.6420 0.5981 0.7734
No log 52.75 422 0.5469 0.6758 0.5469 0.7395
No log 53.0 424 0.5278 0.6553 0.5278 0.7265
No log 53.25 426 0.5438 0.6598 0.5438 0.7374
No log 53.5 428 0.5809 0.7179 0.5809 0.7622
No log 53.75 430 0.6145 0.6878 0.6145 0.7839
No log 54.0 432 0.5943 0.7056 0.5943 0.7709
No log 54.25 434 0.5704 0.6830 0.5704 0.7553
No log 54.5 436 0.5655 0.6822 0.5655 0.7520
No log 54.75 438 0.5528 0.6882 0.5528 0.7435
No log 55.0 440 0.5456 0.6911 0.5456 0.7386
No log 55.25 442 0.5454 0.6681 0.5454 0.7385
No log 55.5 444 0.5402 0.6681 0.5402 0.7350
No log 55.75 446 0.5333 0.7151 0.5333 0.7303
No log 56.0 448 0.5300 0.6838 0.5300 0.7280
No log 56.25 450 0.5254 0.6838 0.5254 0.7249
No log 56.5 452 0.5218 0.6896 0.5218 0.7224
No log 56.75 454 0.5201 0.6896 0.5201 0.7212
No log 57.0 456 0.5192 0.6896 0.5192 0.7206
No log 57.25 458 0.5188 0.6896 0.5188 0.7203
No log 57.5 460 0.5220 0.6896 0.5220 0.7225
No log 57.75 462 0.5194 0.6896 0.5194 0.7207
No log 58.0 464 0.5205 0.6940 0.5205 0.7214
No log 58.25 466 0.5208 0.6940 0.5208 0.7216
No log 58.5 468 0.5246 0.6433 0.5246 0.7243
No log 58.75 470 0.5319 0.6398 0.5319 0.7293
No log 59.0 472 0.5265 0.6333 0.5265 0.7256
No log 59.25 474 0.5263 0.6510 0.5263 0.7255
No log 59.5 476 0.5161 0.6510 0.5161 0.7184
No log 59.75 478 0.5054 0.6813 0.5054 0.7109
No log 60.0 480 0.5045 0.6813 0.5045 0.7103
No log 60.25 482 0.5075 0.6813 0.5075 0.7124
No log 60.5 484 0.5164 0.6969 0.5164 0.7186
No log 60.75 486 0.5160 0.6969 0.5160 0.7183
No log 61.0 488 0.5136 0.6969 0.5136 0.7166
No log 61.25 490 0.5125 0.6969 0.5125 0.7159
No log 61.5 492 0.5148 0.6491 0.5148 0.7175
No log 61.75 494 0.5109 0.6737 0.5109 0.7148
No log 62.0 496 0.5056 0.6882 0.5056 0.7110
No log 62.25 498 0.5056 0.7201 0.5056 0.7110
0.2005 62.5 500 0.5092 0.7201 0.5092 0.7136
0.2005 62.75 502 0.5162 0.7089 0.5162 0.7185
0.2005 63.0 504 0.5177 0.6911 0.5177 0.7195
0.2005 63.25 506 0.5297 0.6256 0.5297 0.7278
0.2005 63.5 508 0.5340 0.6147 0.5340 0.7308
0.2005 63.75 510 0.5317 0.6147 0.5317 0.7292

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.0+cu118
  • Datasets 2.21.0
  • Tokenizers 0.19.1
Downloads last month
7
Safetensors
Model size
135M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for MayBashendy/ArabicNewSplits7_FineTuningAraBERT_run1_AugV5_k2_task5_organization

Finetuned
(4205)
this model