Benedict-L's picture
End of training
bb069a8 verified
metadata
license: cc-by-nc-sa-4.0
base_model: microsoft/layoutlmv2-base-uncased
tags:
  - generated_from_trainer
model-index:
  - name: layoutlmv2-base-uncased_finetuned_docvqa
    results: []

layoutlmv2-base-uncased_finetuned_docvqa

This model is a fine-tuned version of microsoft/layoutlmv2-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 4.6126

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: 5e-05
  • train_batch_size: 4
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 20

Training results

Training Loss Epoch Step Validation Loss
5.2641 0.2212 50 4.8696
4.6026 0.4425 100 4.2722
4.3484 0.6637 150 4.0328
3.879 0.8850 200 3.6907
3.5541 1.1062 250 3.3437
3.3072 1.3274 300 3.1499
3.1514 1.5487 350 2.9365
2.9353 1.7699 400 2.7036
2.4954 1.9912 450 2.7155
1.9393 2.2124 500 2.8356
1.8631 2.4336 550 2.4434
1.9553 2.6549 600 2.5365
2.0108 2.8761 650 2.5717
1.8383 3.0973 700 2.5751
1.3356 3.3186 750 2.5472
1.3101 3.5398 800 2.6720
1.3699 3.7611 850 2.4359
1.421 3.9823 900 2.9012
1.1819 4.2035 950 2.9297
0.9407 4.4248 1000 2.7371
1.0575 4.6460 1050 2.3495
0.9061 4.8673 1100 2.5941
0.8149 5.0885 1150 2.7071
0.7002 5.3097 1200 3.2910
1.009 5.5310 1250 2.7820
0.6106 5.7522 1300 2.9551
0.7998 5.9735 1350 3.0283
0.5198 6.1947 1400 3.0532
0.5274 6.4159 1450 3.3331
0.4868 6.6372 1500 3.0930
0.4724 6.8584 1550 3.3668
0.6184 7.0796 1600 3.1645
0.4337 7.3009 1650 3.3045
0.4681 7.5221 1700 3.3785
0.3815 7.7434 1750 3.6287
0.4704 7.9646 1800 3.6386
0.2866 8.1858 1850 3.8093
0.4064 8.4071 1900 3.6475
0.4187 8.6283 1950 3.4646
0.4037 8.8496 2000 3.8256
0.3989 9.0708 2050 3.7898
0.1772 9.2920 2100 3.9931
0.2577 9.5133 2150 3.7201
0.3283 9.7345 2200 3.7783
0.416 9.9558 2250 3.7312
0.1935 10.1770 2300 3.8151
0.1934 10.3982 2350 3.6563
0.2502 10.6195 2400 3.9194
0.3274 10.8407 2450 3.6391
0.0669 11.0619 2500 3.9782
0.144 11.2832 2550 3.9159
0.1992 11.5044 2600 4.2785
0.1433 11.7257 2650 4.3765
0.204 11.9469 2700 4.1064
0.094 12.1681 2750 4.0756
0.0549 12.3894 2800 4.3475
0.1252 12.6106 2850 4.3339
0.2964 12.8319 2900 4.0766
0.0759 13.0531 2950 4.0707
0.019 13.2743 3000 4.2173
0.1115 13.4956 3050 4.2590
0.0624 13.7168 3100 4.1736
0.1996 13.9381 3150 4.2134
0.1371 14.1593 3200 4.3083
0.0826 14.3805 3250 4.3719
0.0729 14.6018 3300 4.3055
0.0893 14.8230 3350 4.2607
0.0209 15.0442 3400 4.3385
0.0463 15.2655 3450 4.5433
0.0498 15.4867 3500 4.4161
0.0544 15.7080 3550 4.5817
0.1237 15.9292 3600 4.3659
0.0696 16.1504 3650 4.1952
0.0654 16.3717 3700 4.2650
0.1063 16.5929 3750 4.1685
0.0564 16.8142 3800 4.2705
0.0212 17.0354 3850 4.3499
0.0131 17.2566 3900 4.3843
0.0044 17.4779 3950 4.4541
0.0719 17.6991 4000 4.4613
0.0271 17.9204 4050 4.5354
0.0073 18.1416 4100 4.6207
0.0037 18.3628 4150 4.6541
0.0171 18.5841 4200 4.6636
0.0345 18.8053 4250 4.6466
0.103 19.0265 4300 4.5768
0.0232 19.2478 4350 4.6006
0.0162 19.4690 4400 4.6079
0.0261 19.6903 4450 4.6057
0.0083 19.9115 4500 4.6126

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.3.1+cu121
  • Datasets 2.19.2
  • Tokenizers 0.19.1