Edit model card

bert-base-cased-finetuned-ner-cadec-no-iob

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

  • Loss: 0.4487
  • Precision: 0.6037
  • Recall: 0.6491
  • F1: 0.6256
  • Accuracy: 0.9313
  • Adr Precision: 0.5441
  • Adr Recall: 0.6103
  • Adr F1: 0.5753
  • Disease Precision: 0.5
  • Disease Recall: 0.375
  • Disease F1: 0.4286
  • Drug Precision: 0.8649
  • Drug Recall: 0.8889
  • Drug F1: 0.8767
  • Finding Precision: 0.2903
  • Finding Recall: 0.2812
  • Finding F1: 0.2857
  • Symptom Precision: 0.4839
  • Symptom Recall: 0.5172
  • Symptom F1: 0.5000
  • Macro Avg F1: 0.5333
  • Weighted Avg F1: 0.6256

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: 35

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy Adr Precision Adr Recall Adr F1 Disease Precision Disease Recall Disease F1 Drug Precision Drug Recall Drug F1 Finding Precision Finding Recall Finding F1 Symptom Precision Symptom Recall Symptom F1 Macro Avg F1 Weighted Avg F1
No log 1.0 125 0.2340 0.5044 0.6003 0.5482 0.9191 0.4397 0.5938 0.5053 0.3529 0.375 0.3636 0.7574 0.85 0.8010 0.1818 0.0625 0.0930 0.0 0.0 0.0 0.3526 0.5328
No log 2.0 250 0.2068 0.5546 0.6227 0.5867 0.9253 0.4770 0.6 0.5315 0.55 0.3438 0.4231 0.8256 0.8944 0.8587 0.3158 0.1875 0.2353 0.4286 0.1034 0.1667 0.4430 0.5782
No log 3.0 375 0.2031 0.5633 0.6161 0.5885 0.9281 0.5150 0.5670 0.5397 0.5 0.4062 0.4483 0.8093 0.8722 0.8396 0.2 0.2188 0.2090 0.375 0.5172 0.4348 0.4943 0.5891
0.209 4.0 500 0.2331 0.5483 0.6069 0.5761 0.9273 0.5009 0.5897 0.5417 0.0 0.0 0.0 0.8404 0.8778 0.8587 0.14 0.2188 0.1707 0.5 0.3103 0.3830 0.3908 0.5724
0.209 5.0 625 0.2376 0.5878 0.6491 0.6169 0.9324 0.5129 0.6165 0.5599 0.5312 0.5312 0.5312 0.8703 0.8944 0.8822 0.1429 0.0625 0.0870 0.5652 0.4483 0.5000 0.5121 0.6130
0.209 6.0 750 0.2523 0.5646 0.6346 0.5975 0.9258 0.5114 0.6021 0.5530 0.4 0.375 0.3871 0.8649 0.8889 0.8767 0.0857 0.0938 0.0896 0.4516 0.4828 0.4667 0.4746 0.6000
0.209 7.0 875 0.2753 0.5748 0.6438 0.6073 0.9249 0.5209 0.6165 0.5647 0.4762 0.3125 0.3774 0.8670 0.9056 0.8859 0.1458 0.2188 0.1750 0.5 0.3103 0.3830 0.4772 0.6096
0.0561 8.0 1000 0.2769 0.5868 0.6557 0.6193 0.9284 0.5288 0.6247 0.5728 0.6 0.375 0.4615 0.8703 0.8944 0.8822 0.2424 0.25 0.2462 0.3611 0.4483 0.4000 0.5125 0.6212
0.0561 9.0 1125 0.3161 0.5719 0.6240 0.5968 0.9281 0.5091 0.5794 0.5419 0.5263 0.3125 0.3922 0.8757 0.9 0.8877 0.1739 0.25 0.2051 0.48 0.4138 0.4444 0.4943 0.5998
0.0561 10.0 1250 0.3101 0.5867 0.6385 0.6115 0.9297 0.5343 0.5938 0.5625 0.4839 0.4688 0.4762 0.8791 0.8889 0.8840 0.1818 0.25 0.2105 0.4483 0.4483 0.4483 0.5163 0.6160
0.0561 11.0 1375 0.3321 0.5862 0.6412 0.6125 0.9295 0.5245 0.5959 0.5579 0.6 0.4688 0.5263 0.8556 0.8889 0.8719 0.2286 0.25 0.2388 0.4516 0.4828 0.4667 0.5323 0.6142
0.0206 12.0 1500 0.3459 0.5923 0.6517 0.6206 0.9303 0.5323 0.6124 0.5695 0.5517 0.5 0.5246 0.875 0.8944 0.8846 0.2581 0.25 0.2540 0.375 0.4138 0.3934 0.5252 0.6224
0.0206 13.0 1625 0.3489 0.5866 0.6214 0.6035 0.9270 0.5327 0.5876 0.5588 0.4667 0.4375 0.4516 0.8370 0.8556 0.8462 0.24 0.1875 0.2105 0.4138 0.4138 0.4138 0.4962 0.6023
0.0206 14.0 1750 0.3762 0.5709 0.6214 0.5951 0.9270 0.5047 0.5588 0.5303 0.5 0.4375 0.4667 0.8811 0.9056 0.8932 0.2143 0.2812 0.2432 0.4242 0.4828 0.4516 0.5170 0.5987
0.0206 15.0 1875 0.3729 0.5806 0.6412 0.6094 0.9280 0.5149 0.6041 0.5560 0.5652 0.4062 0.4727 0.8503 0.8833 0.8665 0.3 0.2812 0.2903 0.4286 0.4138 0.4211 0.5213 0.6098
0.0093 16.0 2000 0.3980 0.5748 0.6385 0.6050 0.9265 0.5229 0.6124 0.5641 0.4762 0.3125 0.3774 0.8525 0.8667 0.8595 0.2326 0.3125 0.2667 0.4074 0.3793 0.3929 0.4921 0.6073
0.0093 17.0 2125 0.3885 0.5951 0.6359 0.6148 0.9285 0.5343 0.5938 0.5625 0.6087 0.4375 0.5091 0.8587 0.8778 0.8681 0.25 0.25 0.25 0.4375 0.4828 0.4590 0.5297 0.6157
0.0093 18.0 2250 0.4024 0.6015 0.6491 0.6244 0.9310 0.5368 0.6021 0.5675 0.5 0.4375 0.4667 0.8811 0.9056 0.8932 0.2857 0.25 0.2667 0.4545 0.5172 0.4839 0.5356 0.6247
0.0093 19.0 2375 0.4019 0.6025 0.6478 0.6243 0.9302 0.5399 0.6 0.5684 0.5714 0.5 0.5333 0.8703 0.8944 0.8822 0.2667 0.25 0.2581 0.4545 0.5172 0.4839 0.5452 0.6251
0.0053 20.0 2500 0.4061 0.5847 0.6332 0.6080 0.9291 0.5268 0.5876 0.5556 0.5652 0.4062 0.4727 0.8595 0.8833 0.8712 0.2286 0.25 0.2388 0.4054 0.5172 0.4545 0.5186 0.6098
0.0053 21.0 2625 0.4219 0.5903 0.6425 0.6153 0.9288 0.5213 0.6062 0.5605 0.55 0.3438 0.4231 0.8587 0.8778 0.8681 0.3103 0.2812 0.2951 0.5357 0.5172 0.5263 0.5346 0.6153
0.0053 22.0 2750 0.4190 0.6024 0.6557 0.6279 0.9309 0.5420 0.6247 0.5805 0.5185 0.4375 0.4746 0.8548 0.8833 0.8689 0.32 0.25 0.2807 0.4643 0.4483 0.4561 0.5321 0.6271
0.0053 23.0 2875 0.4272 0.5870 0.6412 0.6129 0.9287 0.5192 0.5856 0.5504 0.6154 0.5 0.5517 0.8610 0.8944 0.8774 0.2564 0.3125 0.2817 0.5172 0.5172 0.5172 0.5557 0.6155
0.0034 24.0 3000 0.4206 0.5887 0.6438 0.6150 0.9308 0.5160 0.6 0.5548 0.5769 0.4688 0.5172 0.8602 0.8889 0.8743 0.2963 0.25 0.2712 0.5385 0.4828 0.5091 0.5453 0.6154
0.0034 25.0 3125 0.4260 0.6037 0.6491 0.6256 0.9309 0.5365 0.6062 0.5692 0.52 0.4062 0.4561 0.8859 0.9056 0.8956 0.2692 0.2188 0.2414 0.4688 0.5172 0.4918 0.5308 0.6251
0.0034 26.0 3250 0.4341 0.5995 0.6478 0.6227 0.9310 0.5307 0.6062 0.5659 0.5417 0.4062 0.4643 0.8710 0.9 0.8852 0.2857 0.25 0.2667 0.5185 0.4828 0.5 0.5364 0.6223
0.0034 27.0 3375 0.4476 0.6010 0.6438 0.6217 0.9300 0.5314 0.5938 0.5609 0.56 0.4375 0.4912 0.8710 0.9 0.8852 0.3 0.2812 0.2903 0.5172 0.5172 0.5172 0.5490 0.6219
0.0025 28.0 3500 0.4281 0.6010 0.6478 0.6235 0.9299 0.5328 0.6021 0.5653 0.56 0.4375 0.4912 0.8663 0.9 0.8828 0.2667 0.25 0.2581 0.5556 0.5172 0.5357 0.5466 0.6235
0.0025 29.0 3625 0.4339 0.5988 0.6438 0.6205 0.9299 0.5378 0.6021 0.5681 0.52 0.4062 0.4561 0.8595 0.8833 0.8712 0.2903 0.2812 0.2857 0.4839 0.5172 0.5000 0.5362 0.6208
0.0025 30.0 3750 0.4408 0.6105 0.6596 0.6341 0.9311 0.5404 0.6206 0.5777 0.5909 0.4062 0.4815 0.8663 0.9 0.8828 0.36 0.2812 0.3158 0.5357 0.5172 0.5263 0.5568 0.6331
0.0025 31.0 3875 0.4450 0.6079 0.6504 0.6284 0.9309 0.5410 0.6124 0.5745 0.5417 0.4062 0.4643 0.8656 0.8944 0.8798 0.2917 0.2188 0.25 0.5357 0.5172 0.5263 0.5390 0.6268
0.0016 32.0 4000 0.4435 0.5988 0.6359 0.6168 0.9305 0.5345 0.5918 0.5616 0.52 0.4062 0.4561 0.8641 0.8833 0.8736 0.2857 0.25 0.2667 0.4839 0.5172 0.5000 0.5316 0.6165
0.0016 33.0 4125 0.4448 0.6017 0.6438 0.6221 0.9308 0.5369 0.6 0.5667 0.5417 0.4062 0.4643 0.8696 0.8889 0.8791 0.3103 0.2812 0.2951 0.4688 0.5172 0.4918 0.5394 0.6222
0.0016 34.0 4250 0.4459 0.6030 0.6451 0.6233 0.9304 0.5436 0.6041 0.5723 0.5 0.375 0.4286 0.8649 0.8889 0.8767 0.2812 0.2812 0.2812 0.4839 0.5172 0.5000 0.5318 0.6234
0.0016 35.0 4375 0.4487 0.6037 0.6491 0.6256 0.9313 0.5441 0.6103 0.5753 0.5 0.375 0.4286 0.8649 0.8889 0.8767 0.2903 0.2812 0.2857 0.4839 0.5172 0.5000 0.5333 0.6256

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.15.0
  • Tokenizers 0.15.0
Downloads last month
6
Safetensors
Model size
108M 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 mireiaplalis/bert-base-cased-finetuned-ner-cadec-no-iob

Finetuned
(1935)
this model