Edit model card

arabert_cross_relevance_task4_fold6

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: 0.3394
  • Qwk: 0.1601
  • Mse: 0.3398

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: 64
  • eval_batch_size: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Qwk Mse
No log 0.1111 2 0.4458 0.0318 0.4461
No log 0.2222 4 0.3469 0.0319 0.3472
No log 0.3333 6 0.3536 0.0805 0.3537
No log 0.4444 8 0.3107 0.0717 0.3106
No log 0.5556 10 0.2972 0.1151 0.2971
No log 0.6667 12 0.2762 0.1226 0.2762
No log 0.7778 14 0.2644 0.1492 0.2647
No log 0.8889 16 0.2552 0.1559 0.2557
No log 1.0 18 0.2658 0.2351 0.2661
No log 1.1111 20 0.2702 0.2761 0.2706
No log 1.2222 22 0.2662 0.2796 0.2669
No log 1.3333 24 0.2499 0.2669 0.2508
No log 1.4444 26 0.2460 0.2651 0.2468
No log 1.5556 28 0.2523 0.2545 0.2531
No log 1.6667 30 0.2554 0.2137 0.2561
No log 1.7778 32 0.2471 0.2220 0.2478
No log 1.8889 34 0.2416 0.2899 0.2423
No log 2.0 36 0.2485 0.2753 0.2491
No log 2.1111 38 0.2427 0.2134 0.2433
No log 2.2222 40 0.2377 0.2055 0.2383
No log 2.3333 42 0.2333 0.2092 0.2338
No log 2.4444 44 0.2390 0.2332 0.2394
No log 2.5556 46 0.2482 0.2477 0.2487
No log 2.6667 48 0.2585 0.2256 0.2591
No log 2.7778 50 0.2626 0.2812 0.2633
No log 2.8889 52 0.2502 0.2684 0.2508
No log 3.0 54 0.2402 0.1962 0.2407
No log 3.1111 56 0.2434 0.1883 0.2439
No log 3.2222 58 0.2516 0.2024 0.2522
No log 3.3333 60 0.2671 0.3187 0.2678
No log 3.4444 62 0.2665 0.2324 0.2672
No log 3.5556 64 0.2719 0.1977 0.2725
No log 3.6667 66 0.2689 0.2130 0.2695
No log 3.7778 68 0.2612 0.2140 0.2618
No log 3.8889 70 0.2566 0.2395 0.2572
No log 4.0 72 0.2553 0.2583 0.2560
No log 4.1111 74 0.2604 0.2476 0.2610
No log 4.2222 76 0.2643 0.2402 0.2649
No log 4.3333 78 0.2719 0.2220 0.2725
No log 4.4444 80 0.2816 0.1895 0.2822
No log 4.5556 82 0.2776 0.2106 0.2783
No log 4.6667 84 0.2758 0.2275 0.2765
No log 4.7778 86 0.2876 0.2033 0.2882
No log 4.8889 88 0.3127 0.1941 0.3132
No log 5.0 90 0.3256 0.1643 0.3260
No log 5.1111 92 0.3061 0.1895 0.3065
No log 5.2222 94 0.2811 0.2186 0.2816
No log 5.3333 96 0.2823 0.2186 0.2827
No log 5.4444 98 0.2910 0.2137 0.2913
No log 5.5556 100 0.3094 0.2076 0.3096
No log 5.6667 102 0.3152 0.2118 0.3154
No log 5.7778 104 0.3477 0.1897 0.3478
No log 5.8889 106 0.3597 0.1780 0.3598
No log 6.0 108 0.3476 0.1780 0.3479
No log 6.1111 110 0.3267 0.1992 0.3272
No log 6.2222 112 0.3089 0.2125 0.3095
No log 6.3333 114 0.3128 0.1908 0.3134
No log 6.4444 116 0.3211 0.1742 0.3215
No log 6.5556 118 0.3339 0.1590 0.3343
No log 6.6667 120 0.3109 0.1741 0.3113
No log 6.7778 122 0.2827 0.2060 0.2832
No log 6.8889 124 0.2767 0.2060 0.2772
No log 7.0 126 0.2730 0.2130 0.2736
No log 7.1111 128 0.2806 0.2009 0.2812
No log 7.2222 130 0.2954 0.2098 0.2959
No log 7.3333 132 0.3020 0.2033 0.3025
No log 7.4444 134 0.3007 0.2098 0.3012
No log 7.5556 136 0.2958 0.2054 0.2964
No log 7.6667 138 0.3086 0.1781 0.3091
No log 7.7778 140 0.3217 0.1623 0.3222
No log 7.8889 142 0.3310 0.1677 0.3315
No log 8.0 144 0.3269 0.1668 0.3274
No log 8.1111 146 0.3173 0.1682 0.3179
No log 8.2222 148 0.3173 0.1682 0.3178
No log 8.3333 150 0.3261 0.1634 0.3266
No log 8.4444 152 0.3255 0.1634 0.3259
No log 8.5556 154 0.3229 0.1711 0.3234
No log 8.6667 156 0.3219 0.1668 0.3225
No log 8.7778 158 0.3264 0.1634 0.3269
No log 8.8889 160 0.3314 0.1558 0.3319
No log 9.0 162 0.3302 0.1558 0.3307
No log 9.1111 164 0.3292 0.1601 0.3297
No log 9.2222 166 0.3320 0.1601 0.3325
No log 9.3333 168 0.3371 0.1601 0.3376
No log 9.4444 170 0.3396 0.1601 0.3400
No log 9.5556 172 0.3414 0.1601 0.3418
No log 9.6667 174 0.3407 0.1601 0.3411
No log 9.7778 176 0.3400 0.1601 0.3404
No log 9.8889 178 0.3392 0.1601 0.3396
No log 10.0 180 0.3394 0.1601 0.3398

Framework versions

  • Transformers 4.44.0
  • Pytorch 2.4.0
  • Datasets 2.21.0
  • Tokenizers 0.19.1
Downloads last month
2
Safetensors
Model size
135M params
Tensor type
F32
·
Inference API
Unable to determine this model's library. Check the docs .

Model tree for salbatarni/arabert_cross_relevance_task4_fold6

Finetuned
(694)
this model