Hugging Face's logo

vineetsharma
/
BioMedical_NER-maccrobat-bert

Token Classification
Transformers
PyTorch
Safetensors
bert
Generated from Trainer
Inference Endpoints
Model card Files Community
1
BioMedical_NER-maccrobat-bert / README.md
vineetsharma's picture
vineetsharma
Added Examples
a736c2d
preview code
|
raw
history blame
7.3 kB
metadata
license: apache-2.0
base_model: bert-base-uncased
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: BioMedical_NER-maccrobat-bert
    results: []
widget:
  - text: >-
      CASE: A 28-year-old previously healthy man presented with a 6-week history
      of palpitations. The symptoms occurred during rest, 2–3 times per week,
      lasted up to 30 minutes at a time and were associated with dyspnea. Except
      for a grade 2/6 holosystolic tricuspid regurgitation murmur (best heard at
      the left sternal border with inspiratory accentuation), physical
      examination yielded unremarkable findings.
    example_title: example 1
  - text: >-
      A 63-year-old woman with no known cardiac history presented with a sudden
      onset of dyspnea requiring intubation and ventilatory support out of
      hospital. She denied preceding symptoms of chest discomfort, palpitations,
      syncope or infection. The patient was afebrile and normotensive, with a
      sinus tachycardia of 140 beats/min.
    example_title: example 2
  - text: >-
      A 48 year-old female presented with vaginal bleeding and abnormal Pap
      smears. Upon diagnosis of invasive non-keratinizing SCC of the cervix, she
      underwent a radical hysterectomy with salpingo-oophorectomy which
      demonstrated positive spread to the pelvic lymph nodes and the
      parametrium. Pathological examination revealed that the tumour also
      extensively involved the lower uterine segment.
    example_title: example 3

BioMedical_NER-maccrobat-bert

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

  • Loss: 0.3418
  • Precision: 0.8668
  • Recall: 0.9491
  • F1: 0.9061
  • Accuracy: 0.9501

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 45 1.7363 0.4262 0.0055 0.0108 0.6274
No log 2.0 90 1.3805 0.3534 0.2073 0.2613 0.6565
No log 3.0 135 1.1713 0.4026 0.3673 0.3841 0.6908
No log 4.0 180 1.0551 0.4392 0.5309 0.4807 0.7149
No log 5.0 225 0.9591 0.4893 0.6012 0.5395 0.7496
No log 6.0 270 0.8656 0.5156 0.6483 0.5744 0.7722
No log 7.0 315 0.8613 0.5124 0.6871 0.5870 0.7716
No log 8.0 360 0.7524 0.5699 0.7114 0.6329 0.8110
No log 9.0 405 0.6966 0.5884 0.7374 0.6545 0.8265
No log 10.0 450 0.6564 0.6147 0.7678 0.6827 0.8373
No log 11.0 495 0.5950 0.6484 0.7826 0.7092 0.8563
0.9321 12.0 540 0.6083 0.6578 0.8001 0.7220 0.8587
0.9321 13.0 585 0.5821 0.6682 0.8206 0.7366 0.8688
0.9321 14.0 630 0.5578 0.6787 0.8324 0.7477 0.8744
0.9321 15.0 675 0.4819 0.7338 0.8484 0.7870 0.8974
0.9321 16.0 720 0.4775 0.7461 0.8573 0.7978 0.9020
0.9321 17.0 765 0.4786 0.7395 0.8600 0.7952 0.9020
0.9321 18.0 810 0.4481 0.7647 0.8740 0.8157 0.9102
0.9321 19.0 855 0.4597 0.7638 0.8799 0.8177 0.9108
0.9321 20.0 900 0.4551 0.7617 0.8835 0.8181 0.9096
0.9321 21.0 945 0.4365 0.7698 0.8873 0.8244 0.9142
0.9321 22.0 990 0.3993 0.7986 0.8957 0.8444 0.9247
0.2115 23.0 1035 0.4162 0.7950 0.9014 0.8449 0.9234
0.2115 24.0 1080 0.4188 0.8007 0.9042 0.8493 0.9248
0.2115 25.0 1125 0.3996 0.8105 0.9103 0.8575 0.9291
0.2115 26.0 1170 0.3775 0.8226 0.9134 0.8657 0.9333
0.2115 27.0 1215 0.3656 0.8297 0.9187 0.8720 0.9364
0.2115 28.0 1260 0.3744 0.8323 0.9217 0.8747 0.9371
0.2115 29.0 1305 0.3763 0.8296 0.9229 0.8738 0.9364
0.2115 30.0 1350 0.3506 0.8454 0.9272 0.8844 0.9414
0.2115 31.0 1395 0.3602 0.8441 0.9301 0.8850 0.9413
0.2115 32.0 1440 0.3617 0.8359 0.9303 0.8806 0.9400
0.2115 33.0 1485 0.3737 0.8352 0.9310 0.8805 0.9388
0.0818 34.0 1530 0.3541 0.8477 0.9352 0.8893 0.9438
0.0818 35.0 1575 0.3553 0.8487 0.9377 0.8910 0.9439
0.0818 36.0 1620 0.3583 0.8476 0.9367 0.8899 0.9438
0.0818 37.0 1665 0.3318 0.8642 0.9400 0.9005 0.9484
0.0818 38.0 1710 0.3449 0.8598 0.9409 0.8985 0.9471
0.0818 39.0 1755 0.3466 0.8591 0.9419 0.8986 0.9468
0.0818 40.0 1800 0.3494 0.8591 0.9426 0.8989 0.9473
0.0818 41.0 1845 0.3494 0.8591 0.9451 0.9001 0.9475
0.0818 42.0 1890 0.3545 0.8588 0.9462 0.9004 0.9477
0.0818 43.0 1935 0.3569 0.8599 0.9460 0.9009 0.9470
0.0818 44.0 1980 0.3465 0.8645 0.9468 0.9038 0.9492
0.0469 45.0 2025 0.3424 0.8663 0.9489 0.9057 0.9498
0.0469 46.0 2070 0.3460 0.8643 0.9481 0.9043 0.9490
0.0469 47.0 2115 0.3445 0.8658 0.9483 0.9052 0.9496
0.0469 48.0 2160 0.3387 0.8701 0.9500 0.9083 0.9508
0.0469 49.0 2205 0.3432 0.8671 0.9491 0.9063 0.9501
0.0469 50.0 2250 0.3418 0.8668 0.9491 0.9061 0.9501

Framework versions

  • Transformers 4.32.1
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.4
  • Tokenizers 0.13.3