Edit model card

lmv2-g-rai_1-995-doc-10-18

This model is a fine-tuned version of microsoft/layoutlmv2-base-uncased on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0163
  • Dob Key Precision: 0.7638
  • Dob Key Recall: 0.7638
  • Dob Key F1: 0.7638
  • Dob Key Number: 127
  • Dob Value Precision: 0.9767
  • Dob Value Recall: 0.9767
  • Dob Value F1: 0.9767
  • Dob Value Number: 129
  • Doctor Name Key Precision: 0.6970
  • Doctor Name Key Recall: 0.6866
  • Doctor Name Key F1: 0.6917
  • Doctor Name Key Number: 67
  • Doctor Name Value Precision: 0.9275
  • Doctor Name Value Recall: 0.9143
  • Doctor Name Value F1: 0.9209
  • Doctor Name Value Number: 70
  • Patient Name Key Precision: 0.7055
  • Patient Name Key Recall: 0.7357
  • Patient Name Key F1: 0.7203
  • Patient Name Key Number: 140
  • Patient Name Value Precision: 0.9724
  • Patient Name Value Recall: 0.9792
  • Patient Name Value F1: 0.9758
  • Patient Name Value Number: 144
  • Overall Precision: 0.8460
  • Overall Recall: 0.8523
  • Overall F1: 0.8492
  • Overall Accuracy: 0.9958

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: 4e-05
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Dob Key Precision Dob Key Recall Dob Key F1 Dob Key Number Dob Value Precision Dob Value Recall Dob Value F1 Dob Value Number Doctor Name Key Precision Doctor Name Key Recall Doctor Name Key F1 Doctor Name Key Number Doctor Name Value Precision Doctor Name Value Recall Doctor Name Value F1 Doctor Name Value Number Patient Name Key Precision Patient Name Key Recall Patient Name Key F1 Patient Name Key Number Patient Name Value Precision Patient Name Value Recall Patient Name Value F1 Patient Name Value Number Overall Precision Overall Recall Overall F1 Overall Accuracy
0.5034 1.0 796 0.0841 0.7143 0.7480 0.7308 127 0.7881 0.9225 0.85 129 0.0 0.0 0.0 67 0.0 0.0 0.0 70 0.5988 0.7143 0.6515 140 0.4908 0.9236 0.6410 144 0.5944 0.6603 0.6256 0.9887
0.0579 2.0 1592 0.0365 0.7231 0.7402 0.7315 127 0.9766 0.9690 0.9728 129 0.6462 0.6269 0.6364 67 0.9296 0.9429 0.9362 70 0.7103 0.7357 0.7228 140 0.9392 0.9653 0.9521 144 0.8282 0.8405 0.8343 0.9954
0.0317 3.0 2388 0.0297 0.7578 0.7638 0.7608 127 0.9767 0.9767 0.9767 129 0.7077 0.6866 0.6970 67 0.8676 0.8429 0.8551 70 0.6474 0.7214 0.6824 140 0.8993 0.9306 0.9147 144 0.8101 0.8316 0.8207 0.9943
0.0233 4.0 3184 0.0195 0.7638 0.7638 0.7638 127 0.9403 0.9767 0.9582 129 0.7015 0.7015 0.7015 67 0.9718 0.9857 0.9787 70 0.6164 0.7 0.6555 140 0.9724 0.9792 0.9758 144 0.8222 0.8538 0.8377 0.9958
0.0189 5.0 3980 0.0188 0.7462 0.7638 0.7549 127 0.9545 0.9767 0.9655 129 0.5606 0.5522 0.5564 67 0.9565 0.9429 0.9496 70 0.6228 0.7429 0.6775 140 0.9724 0.9792 0.9758 144 0.8054 0.8434 0.8240 0.9955
0.0174 6.0 4776 0.0167 0.7638 0.7638 0.7638 127 0.9767 0.9767 0.9767 129 0.5970 0.5970 0.5970 67 0.9714 0.9714 0.9714 70 0.6478 0.7357 0.6890 140 0.9724 0.9792 0.9758 144 0.8250 0.8493 0.8370 0.9956
0.0162 7.0 5572 0.0185 0.7578 0.7638 0.7608 127 0.9767 0.9767 0.9767 129 0.4272 0.6567 0.5176 67 0.9677 0.8571 0.9091 70 0.7007 0.7357 0.7178 140 0.9724 0.9792 0.9758 144 0.7997 0.8434 0.8210 0.9954
0.0153 8.0 6368 0.0170 0.7638 0.7638 0.7638 127 0.9767 0.9767 0.9767 129 0.5758 0.5672 0.5714 67 0.9571 0.9571 0.9571 70 0.7305 0.7357 0.7331 140 0.9724 0.9792 0.9758 144 0.8437 0.8449 0.8443 0.9957
0.0142 9.0 7164 0.0163 0.7638 0.7638 0.7638 127 0.9767 0.9767 0.9767 129 0.6970 0.6866 0.6917 67 0.9275 0.9143 0.9209 70 0.7055 0.7357 0.7203 140 0.9724 0.9792 0.9758 144 0.8460 0.8523 0.8492 0.9958
0.0136 10.0 7960 0.0177 0.7405 0.7638 0.7519 127 0.9767 0.9767 0.9767 129 0.6094 0.5821 0.5954 67 0.8358 0.8 0.8175 70 0.6541 0.7429 0.6957 140 0.9589 0.9722 0.9655 144 0.8075 0.8301 0.8186 0.9953
0.0131 11.0 8756 0.0202 0.7402 0.7402 0.7402 127 0.9767 0.9767 0.9767 129 0.5968 0.5522 0.5736 67 0.9403 0.9 0.9197 70 0.7305 0.7357 0.7331 140 0.9655 0.9722 0.9689 144 0.8390 0.8316 0.8353 0.9954
0.0134 12.0 9552 0.0195 0.7239 0.7638 0.7433 127 0.9237 0.8450 0.8826 129 0.5846 0.5672 0.5758 67 0.9041 0.9429 0.9231 70 0.7305 0.7357 0.7331 140 0.9722 0.9722 0.9722 144 0.8193 0.8168 0.8180 0.9949
0.0127 13.0 10348 0.0169 0.7638 0.7638 0.7638 127 0.9767 0.9767 0.9767 129 0.7077 0.6866 0.6970 67 0.9403 0.9 0.9197 70 0.6211 0.7143 0.6645 140 0.9724 0.9792 0.9758 144 0.8256 0.8464 0.8359 0.9957
0.0119 14.0 11144 0.0174 0.7638 0.7638 0.7638 127 0.9767 0.9767 0.9767 129 0.5821 0.5821 0.5821 67 0.9437 0.9571 0.9504 70 0.6897 0.7143 0.7018 140 0.9338 0.9792 0.9559 144 0.8261 0.8419 0.8339 0.9955
0.013 15.0 11940 0.0174 0.6953 0.7008 0.6980 127 0.9767 0.9767 0.9767 129 0.6164 0.6716 0.6429 67 0.9706 0.9429 0.9565 70 0.6667 0.7143 0.6897 140 0.9583 0.9583 0.9583 144 0.8150 0.8331 0.8240 0.9950
0.0133 16.0 12736 0.0195 0.7008 0.7008 0.7008 127 0.9767 0.9767 0.9767 129 0.5823 0.6866 0.6301 67 0.9054 0.9571 0.9306 70 0.6174 0.6571 0.6367 140 0.9161 0.9097 0.9129 144 0.7860 0.8139 0.7997 0.9946
0.0154 17.0 13532 0.0239 0.6885 0.6614 0.6747 127 0.8623 0.9225 0.8914 129 0.5057 0.6567 0.5714 67 0.9403 0.9 0.9197 70 0.3727 0.5857 0.4556 140 0.9655 0.9722 0.9689 144 0.6829 0.7858 0.7308 0.9935
0.0163 18.0 14328 0.0437 0.6607 0.5827 0.6192 127 0.5736 0.8760 0.6933 129 0.4177 0.4925 0.4521 67 0.8243 0.8714 0.8472 70 0.4845 0.5571 0.5183 140 0.5990 0.7986 0.6845 144 0.5816 0.7001 0.6354 0.9887
0.0109 19.0 15124 0.0220 0.7578 0.7638 0.7608 127 0.9767 0.9767 0.9767 129 0.7097 0.6567 0.6822 67 0.9403 0.9 0.9197 70 0.6776 0.7357 0.7055 140 0.9724 0.9792 0.9758 144 0.8404 0.8479 0.8441 0.9955
0.0104 20.0 15920 0.0184 0.6093 0.7244 0.6619 127 0.976 0.9457 0.9606 129 0.6133 0.6866 0.6479 67 0.9437 0.9571 0.9504 70 0.6013 0.6571 0.6280 140 0.9724 0.9792 0.9758 144 0.7778 0.8272 0.8017 0.9950
0.0086 21.0 16716 0.0232 0.3889 0.4409 0.4133 127 0.9767 0.9767 0.9767 129 0.5270 0.5821 0.5532 67 0.9444 0.9714 0.9577 70 0.5245 0.5357 0.5300 140 0.9724 0.9792 0.9758 144 0.7143 0.7459 0.7298 0.9930
0.0085 22.0 17512 0.0197 0.7480 0.7480 0.7480 127 0.9767 0.9767 0.9767 129 0.6471 0.6567 0.6519 67 0.9189 0.9714 0.9444 70 0.6149 0.65 0.6319 140 0.9658 0.9792 0.9724 144 0.8165 0.8346 0.8254 0.9951
0.0083 23.0 18308 0.0220 0.7328 0.7559 0.7442 127 0.9692 0.9767 0.9730 129 0.6081 0.6716 0.6383 67 0.9571 0.9571 0.9571 70 0.6479 0.6571 0.6525 140 0.9592 0.9792 0.9691 144 0.8170 0.8375 0.8271 0.9952
0.0084 24.0 19104 0.0226 0.6418 0.6772 0.6590 127 0.9767 0.9767 0.9767 129 0.5 0.7164 0.5890 67 0.8919 0.9429 0.9167 70 0.5034 0.5286 0.5157 140 0.9724 0.9792 0.9758 144 0.7462 0.7991 0.7718 0.9942
0.0067 25.0 19900 0.0257 0.6691 0.7165 0.6920 127 0.9692 0.9767 0.9730 129 0.6267 0.7015 0.6620 67 0.9143 0.9143 0.9143 70 0.6828 0.7071 0.6947 140 0.94 0.9792 0.9592 144 0.8045 0.8390 0.8214 0.9949
0.0071 26.0 20696 0.0241 0.5828 0.6929 0.6331 127 0.9692 0.9767 0.9730 129 0.6029 0.6119 0.6074 67 0.8889 0.9143 0.9014 70 0.5563 0.5643 0.5603 140 0.9658 0.9792 0.9724 144 0.7602 0.7962 0.7778 0.9943
0.0072 27.0 21492 0.0222 0.6850 0.6850 0.6850 127 0.9767 0.9767 0.9767 129 0.5714 0.6567 0.6111 67 0.9178 0.9571 0.9371 70 0.6370 0.6643 0.6503 140 0.9592 0.9792 0.9691 144 0.7983 0.8242 0.8110 0.9948
0.0057 28.0 22288 0.0259 0.5909 0.6142 0.6023 127 0.9767 0.9767 0.9767 129 0.6714 0.7015 0.6861 67 0.9275 0.9143 0.9209 70 0.5734 0.5857 0.5795 140 0.9724 0.9792 0.9758 144 0.7820 0.7947 0.7883 0.9943
0.0054 29.0 23084 0.0299 0.6418 0.6772 0.6590 127 0.9618 0.9767 0.9692 129 0.6216 0.6866 0.6525 67 0.8873 0.9 0.8936 70 0.5306 0.5571 0.5436 140 0.9655 0.9722 0.9689 144 0.7678 0.7962 0.7817 0.9937
0.0066 30.0 23880 0.0254 0.5532 0.6142 0.5821 127 0.9259 0.9690 0.9470 129 0.5938 0.5672 0.5802 67 0.9130 0.9 0.9065 70 0.6738 0.6786 0.6762 140 0.9592 0.9792 0.9691 144 0.7747 0.7976 0.7860 0.9943

Framework versions

  • Transformers 4.24.0.dev0
  • Pytorch 1.12.1+cu113
  • Datasets 2.2.2
  • Tokenizers 0.13.1
Downloads last month
3
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.