layoutlmv2-base-uncased_finetuned_docvqa_on_1200

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: 4.6669

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
5.26 0.2212 50 4.5357
4.3552 0.4425 100 4.0284
4.0237 0.6637 150 3.7961
3.7428 0.8850 200 3.5727
3.6213 1.1062 250 3.7866
3.2334 1.3274 300 3.1121
3.0382 1.5487 350 2.9537
2.8353 1.7699 400 2.8318
2.4759 1.9912 450 2.6736
1.9881 2.2124 500 3.0365
1.9279 2.4336 550 2.4144
1.9336 2.6549 600 2.1754
1.772 2.8761 650 2.1086
1.5504 3.0973 700 2.7056
1.4621 3.3186 750 2.8930
1.4227 3.5398 800 2.4620
1.3924 3.7611 850 2.1275
1.3063 3.9823 900 2.2443
1.0697 4.2035 950 2.6747
0.9476 4.4248 1000 2.7229
1.0868 4.6460 1050 2.9257
0.8726 4.8673 1100 2.7007
0.9436 5.0885 1150 2.8765
0.7219 5.3097 1200 2.5301
0.6919 5.5310 1250 2.9763
0.491 5.7522 1300 3.1198
0.5382 5.9735 1350 3.0883
0.462 6.1947 1400 3.2955
0.6533 6.4159 1450 3.3370
0.6477 6.6372 1500 3.3794
0.4849 6.8584 1550 3.3798
0.4881 7.0796 1600 3.2085
0.3952 7.3009 1650 3.2885
0.161 7.5221 1700 3.6201
0.6895 7.7434 1750 3.4253
0.4638 7.9646 1800 3.4787
0.2186 8.1858 1850 3.7668
0.2531 8.4071 1900 3.7723
0.3971 8.6283 1950 3.7131
0.5665 8.8496 2000 3.5627
0.3377 9.0708 2050 3.1885
0.208 9.2920 2100 3.3734
0.1775 9.5133 2150 4.0609
0.3295 9.7345 2200 3.7039
0.2627 9.9558 2250 3.6028
0.1988 10.1770 2300 3.6288
0.1772 10.3982 2350 3.5394
0.0719 10.6195 2400 4.2068
0.1629 10.8407 2450 4.2701
0.1921 11.0619 2500 4.0440
0.164 11.2832 2550 3.9099
0.1281 11.5044 2600 3.7753
0.0586 11.7257 2650 3.9491
0.1436 11.9469 2700 4.2734
0.0405 12.1681 2750 4.4347
0.0664 12.3894 2800 4.2338
0.0864 12.6106 2850 3.8694
0.103 12.8319 2900 3.9883
0.0456 13.0531 2950 4.5064
0.05 13.2743 3000 4.1434
0.0436 13.4956 3050 4.3928
0.0798 13.7168 3100 4.5576
0.0919 13.9381 3150 4.4114
0.0988 14.1593 3200 4.4998
0.0332 14.3805 3250 4.3948
0.0326 14.6018 3300 4.3823
0.0434 14.8230 3350 4.2468
0.0926 15.0442 3400 4.3909
0.027 15.2655 3450 4.5539
0.047 15.4867 3500 4.5799
0.0189 15.7080 3550 4.3943
0.0096 15.9292 3600 4.4218
0.0467 16.1504 3650 4.6181
0.0144 16.3717 3700 4.5609
0.0339 16.5929 3750 4.5994
0.074 16.8142 3800 4.5598
0.018 17.0354 3850 4.5528
0.0043 17.2566 3900 4.6133
0.0179 17.4779 3950 4.5414
0.039 17.6991 4000 4.4690
0.0134 17.9204 4050 4.4789
0.0094 18.1416 4100 4.5317
0.004 18.3628 4150 4.5711
0.0064 18.5841 4200 4.6237
0.0505 18.8053 4250 4.6148
0.0312 19.0265 4300 4.6302
0.0127 19.2478 4350 4.6577
0.0169 19.4690 4400 4.6685
0.0192 19.6903 4450 4.6626
0.0232 19.9115 4500 4.6669

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.1
  • Tokenizers 0.19.1
Downloads last month
34
Safetensors
Model size
200M 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 sahil-everlign/layoutlmv2-base-uncased_finetuned_docvqa_on_1200

Finetuned
(67)
this model