Edit model card

Public100_1L_BERT_5epoch_again

This model is a fine-tuned version of Youssef320/Public100_1L_BERT_5epoch on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 3.4088
  • Top 1 Macro F1 Score: 0.1019
  • Top 1 Weighted F1score: 0.1606
  • Top 3 Macro F1 Score: 0.2198
  • Top3 3 Weighted F1 Score : 0.3153

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: 0.0002
  • train_batch_size: 64
  • eval_batch_size: 64
  • seed: 42
  • gradient_accumulation_steps: 32
  • total_train_batch_size: 2048
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant
  • num_epochs: 4.0

Training results

Training Loss Epoch Step Validation Loss Top 1 Macro F1 Score Top 1 Weighted F1score Top 3 Macro F1 Score Top3 3 Weighted F1 Score
3.5163 0.12 64 3.5267 0.0688 0.1259 0.1655 0.2710
3.4858 0.25 128 3.5224 0.0692 0.1260 0.1700 0.2728
3.444 0.38 192 3.5111 0.0751 0.1334 0.1730 0.2787
3.4475 0.5 256 3.5042 0.0746 0.1317 0.1738 0.2790
3.4461 0.62 320 3.4986 0.0750 0.1320 0.1731 0.2777
3.4652 0.75 384 3.4898 0.0781 0.1373 0.1796 0.2849
3.4443 0.88 448 3.4867 0.0802 0.1380 0.1811 0.2852
3.4828 1.0 512 3.4726 0.0797 0.1392 0.1836 0.2893
3.4113 1.12 576 3.4760 0.0819 0.1409 0.1863 0.2909
3.4054 1.25 640 3.4737 0.0822 0.1408 0.1827 0.2881
3.4218 1.38 704 3.4678 0.0826 0.1418 0.1861 0.2897
3.4095 1.5 768 3.4580 0.0847 0.1436 0.1890 0.2934
3.4153 1.62 832 3.4534 0.0858 0.1459 0.1904 0.2959
3.4154 1.75 896 3.4468 0.0855 0.1450 0.1921 0.2961
3.3818 1.88 960 3.4436 0.0836 0.1430 0.1905 0.2939
3.4033 2.0 1024 3.4368 0.0878 0.1481 0.1960 0.2996
3.3245 2.12 1088 3.4500 0.0894 0.1509 0.1972 0.3019
3.2943 2.25 1152 3.4536 0.0887 0.1485 0.1995 0.3015
3.3332 2.38 1216 3.4468 0.0900 0.1488 0.2005 0.3004
3.3483 2.5 1280 3.4377 0.0924 0.1523 0.2044 0.3035
3.3408 2.62 1344 3.4341 0.0923 0.1519 0.2066 0.3050
3.343 2.75 1408 3.4293 0.0928 0.1527 0.2052 0.3054
3.3487 2.88 1472 3.4235 0.0921 0.1525 0.2040 0.3038
3.348 3.0 1536 3.4169 0.0956 0.1563 0.2112 0.3108
3.2211 3.12 1600 3.4374 0.0966 0.1569 0.2123 0.3091
3.2275 3.25 1664 3.4398 0.0953 0.1549 0.2068 0.3073
3.2523 3.38 1728 3.4343 0.0967 0.1556 0.2109 0.3076
3.2741 3.5 1792 3.4355 0.0980 0.1561 0.2147 0.3081
3.2815 3.62 1856 3.4259 0.0994 0.1589 0.2160 0.3112
3.2517 3.75 1920 3.4184 0.0984 0.1577 0.2179 0.3116
3.2801 3.88 1984 3.4134 0.0996 0.1592 0.2173 0.3131
3.2925 4.0 2048 3.4088 0.1019 0.1606 0.2198 0.3153

Framework versions

  • Transformers 4.20.1
  • Pytorch 1.12.1+cu102
  • Datasets 2.0.0
  • Tokenizers 0.11.0
Downloads last month
4
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.