Hugging Face's logo

yonigo
/
distilbert-base-multilingual-cased-pii

Token Classification
Transformers
Safetensors
distilbert
Generated from Trainer
Inference Endpoints
Model card Files Community
distilbert-base-multilingual-cased-pii / README.md
yonigo's picture
yonigo
Training in progress, step 1000
93a6a1f verified
preview code
|
raw
history blame
No virus
12 kB
metadata
license: apache-2.0
base_model: distilbert-base-cased
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: distilbert-base-cased-pii-en
    results: []

distilbert-base-cased-pii-en

This model is a fine-tuned version of distilbert-base-cased on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0635
  • Bod F1: 0.9622
  • Building F1: 0.9808
  • Cardissuer F1: 0.0
  • City F1: 0.9589
  • Country F1: 0.9741
  • Date F1: 0.9176
  • Driverlicense F1: 0.9355
  • Email F1: 0.9855
  • Geocoord F1: 0.9814
  • Givenname1 F1: 0.8496
  • Givenname2 F1: 0.7261
  • Idcard F1: 0.9306
  • Ip F1: 0.9850
  • Lastname1 F1: 0.7974
  • Lastname2 F1: 0.6672
  • Lastname3 F1: 0.6393
  • Pass F1: 0.9005
  • Passport F1: 0.9405
  • Postcode F1: 0.9794
  • Secaddress F1: 0.9745
  • Sex F1: 0.9738
  • Socialnumber F1: 0.9437
  • State F1: 0.9794
  • Street F1: 0.9651
  • Tel F1: 0.9674
  • Time F1: 0.9638
  • Title F1: 0.9587
  • Username F1: 0.9469
  • Precision: 0.9318
  • Recall: 0.9460
  • F1: 0.9388
  • Accuracy: 0.9913

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: 128
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.2
  • lr_scheduler_warmup_steps: 3000
  • training_steps: 20000

Training results

Training Loss Epoch Step Validation Loss Bod F1 Building F1 Cardissuer F1 City F1 Country F1 Date F1 Driverlicense F1 Email F1 Geocoord F1 Givenname1 F1 Givenname2 F1 Idcard F1 Ip F1 Lastname1 F1 Lastname2 F1 Lastname3 F1 Pass F1 Passport F1 Postcode F1 Secaddress F1 Sex F1 Socialnumber F1 State F1 Street F1 Tel F1 Time F1 Title F1 Username F1 Precision Recall F1 Accuracy
0.23 2.1368 1000 0.1111 0.8804 0.9218 0.0 0.5854 0.8890 0.8009 0.5896 0.9469 0.7002 0.4149 0.0 0.5740 0.9056 0.5230 0.0 0.0 0.6551 0.6301 0.7819 0.5871 0.8396 0.7137 0.8771 0.6466 0.8567 0.9185 0.7411 0.8202 0.7127 0.7640 0.7374 0.9723
0.0687 4.2735 2000 0.0541 0.9365 0.9688 0.0 0.9115 0.9434 0.8841 0.8530 0.9768 0.9698 0.7276 0.3073 0.8720 0.9701 0.6430 0.1854 0.0 0.8169 0.8857 0.9638 0.9617 0.9312 0.8927 0.9418 0.9231 0.9438 0.9461 0.9232 0.9138 0.8698 0.9020 0.8856 0.9862
0.0452 6.4103 3000 0.0469 0.9550 0.9753 0.0 0.9404 0.9620 0.9014 0.9004 0.9831 0.9721 0.7664 0.4606 0.9099 0.97 0.6838 0.3818 0.0 0.8611 0.9079 0.9757 0.9715 0.9661 0.9197 0.9674 0.9435 0.9448 0.96 0.9354 0.9357 0.8975 0.9212 0.9092 0.9883
0.0311 8.5470 4000 0.0434 0.9581 0.9753 0.0 0.9481 0.9641 0.8973 0.9217 0.9833 0.9745 0.8085 0.6085 0.9232 0.9781 0.7519 0.4891 0.0849 0.8685 0.9280 0.9771 0.9704 0.9659 0.9278 0.9729 0.9549 0.9563 0.9594 0.9564 0.9407 0.9117 0.9344 0.9229 0.9897
0.023 10.6838 5000 0.0416 0.9578 0.9778 0.0 0.9503 0.9639 0.9043 0.9245 0.9804 0.9767 0.8277 0.6607 0.9031 0.9836 0.7790 0.5779 0.25 0.8791 0.9276 0.9767 0.9660 0.9708 0.9353 0.9749 0.9610 0.9681 0.9611 0.9545 0.9382 0.9185 0.9363 0.9273 0.9903
0.016 12.8205 6000 0.0436 0.9607 0.9778 0.0 0.9516 0.9698 0.9027 0.9274 0.9781 0.9677 0.8472 0.6913 0.9259 0.9877 0.7857 0.5925 0.3819 0.8853 0.9377 0.9731 0.9688 0.9752 0.9461 0.9756 0.9594 0.9613 0.9568 0.9545 0.9447 0.9247 0.9396 0.9321 0.9907
0.012 14.9573 7000 0.0447 0.9598 0.9808 0.0 0.9558 0.9737 0.9070 0.9339 0.9719 0.9654 0.8482 0.7009 0.9280 0.9859 0.7879 0.6027 0.4729 0.8909 0.9413 0.9781 0.9750 0.9740 0.9431 0.9778 0.9624 0.9623 0.9665 0.9555 0.9403 0.9275 0.9413 0.9344 0.9911
0.0086 17.0940 8000 0.0487 0.9584 0.9814 0.0 0.9529 0.9733 0.9114 0.9350 0.9857 0.9767 0.8493 0.7059 0.9306 0.9829 0.7918 0.6216 0.5292 0.8934 0.9330 0.9790 0.9766 0.9750 0.9395 0.9747 0.9624 0.9605 0.9631 0.9580 0.9399 0.9282 0.9423 0.9352 0.9908
0.0062 19.2308 9000 0.0509 0.9596 0.9795 0.0 0.9594 0.9708 0.9128 0.9299 0.9872 0.9837 0.8491 0.7188 0.9270 0.9859 0.7923 0.6468 0.5371 0.8919 0.9369 0.9783 0.9756 0.9749 0.9382 0.9778 0.9650 0.9647 0.9653 0.9559 0.9461 0.9337 0.9394 0.9365 0.9911
0.0045 21.3675 10000 0.0548 0.9559 0.9774 0.0 0.9524 0.9720 0.9080 0.9348 0.9827 0.9814 0.8446 0.7117 0.9271 0.9800 0.7977 0.6428 0.5266 0.8964 0.9351 0.9754 0.9716 0.9728 0.9478 0.9757 0.9584 0.9698 0.9587 0.9548 0.9423 0.9242 0.9457 0.9348 0.9907
0.0036 23.5043 11000 0.0560 0.9594 0.9781 0.0 0.9575 0.9720 0.9121 0.9367 0.9814 0.9814 0.8504 0.7209 0.9317 0.9807 0.7922 0.6507 0.5918 0.8864 0.9380 0.9769 0.9722 0.9745 0.9399 0.9771 0.9628 0.9675 0.9618 0.9581 0.9446 0.9288 0.9434 0.9361 0.9910
0.0026 25.6410 12000 0.0576 0.9596 0.9798 0.0 0.9575 0.9732 0.9130 0.9308 0.9831 0.9791 0.8471 0.7104 0.9268 0.9836 0.7967 0.6563 0.6222 0.8982 0.9345 0.9771 0.9739 0.9733 0.9402 0.9771 0.9673 0.9656 0.9642 0.9576 0.9480 0.9288 0.9446 0.9366 0.9910
0.002 27.7778 13000 0.0608 0.9555 0.9796 0.0 0.9561 0.9717 0.9045 0.9319 0.9848 0.9791 0.8488 0.7157 0.9268 0.9852 0.7909 0.6580 0.6039 0.8900 0.9360 0.9802 0.9717 0.9750 0.9361 0.9778 0.9646 0.9683 0.9615 0.9565 0.9465 0.9279 0.9433 0.9355 0.9909
0.0016 29.9145 14000 0.0601 0.9573 0.9801 0.0 0.9589 0.9722 0.9135 0.9353 0.9848 0.9837 0.8499 0.7202 0.9316 0.9871 0.7942 0.6677 0.6432 0.9017 0.9402 0.9799 0.9744 0.9740 0.9455 0.9783 0.9631 0.9719 0.9645 0.9600 0.9482 0.9331 0.9443 0.9387 0.9913
0.0013 32.0513 15000 0.0613 0.9606 0.9798 0.0 0.9571 0.9739 0.9155 0.9365 0.9840 0.9791 0.8466 0.7221 0.9321 0.9861 0.7948 0.6622 0.6281 0.9017 0.9407 0.9787 0.9739 0.9740 0.9475 0.9774 0.9632 0.9693 0.9644 0.9594 0.9469 0.9313 0.9457 0.9384 0.9912
0.001 34.1880 16000 0.0639 0.9611 0.9808 0.0 0.9601 0.9729 0.9149 0.9337 0.9867 0.9814 0.8483 0.7269 0.9307 0.9838 0.7956 0.6627 0.6154 0.9025 0.9397 0.9797 0.9733 0.9738 0.9383 0.9776 0.9636 0.9674 0.9637 0.9576 0.9480 0.9304 0.9457 0.9380 0.9911
0.0009 36.3248 17000 0.0621 0.9622 0.9811 0.0 0.9604 0.9741 0.9156 0.9359 0.9855 0.9814 0.8510 0.7273 0.9319 0.9859 0.7991 0.6646 0.6413 0.8999 0.9393 0.9789 0.9739 0.9740 0.9427 0.9789 0.9653 0.9687 0.9637 0.9597 0.9460 0.9324 0.9456 0.9390 0.9913
0.0008 38.4615 18000 0.0631 0.9620 0.9801 0.0 0.9582 0.9744 0.9190 0.9350 0.9853 0.9814 0.8514 0.7253 0.9298 0.9848 0.7992 0.6677 0.6434 0.8992 0.9401 0.9797 0.9739 0.9731 0.9421 0.9789 0.9653 0.9681 0.9643 0.9592 0.9462 0.9319 0.9457 0.9388 0.9913
0.0007 40.5983 19000 0.0633 0.9615 0.9806 0.0 0.9589 0.9741 0.9170 0.9358 0.9861 0.9814 0.8501 0.7268 0.9316 0.9857 0.7973 0.6662 0.6393 0.9002 0.9405 0.9797 0.9745 0.9738 0.9442 0.9796 0.9658 0.9679 0.9644 0.9586 0.9469 0.9323 0.9459 0.9391 0.9913
0.0008 42.7350 20000 0.0635 0.9622 0.9808 0.0 0.9589 0.9741 0.9176 0.9355 0.9855 0.9814 0.8496 0.7261 0.9306 0.9850 0.7974 0.6672 0.6393 0.9005 0.9405 0.9794 0.9745 0.9738 0.9437 0.9794 0.9651 0.9674 0.9638 0.9587 0.9469 0.9318 0.9460 0.9388 0.9913

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.3.1+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1