tmvar_2e-05 / README.md
Brizape's picture
update model card README.md
4a46abb
metadata
license: mit
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: tmvar_2e-05
    results: []

tmvar_2e-05

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

  • Loss: 0.0136
  • Precision: 0.8308
  • Recall: 0.8757
  • F1: 0.8526
  • Accuracy: 0.9968

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: 16
  • eval_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • training_steps: 500

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
0.5077 1.47 25 0.1015 0.0 0.0 0.0 0.9843
0.0834 2.94 50 0.0463 0.3581 0.4162 0.3850 0.9877
0.0348 4.41 75 0.0315 0.3846 0.4324 0.4071 0.9896
0.0285 5.88 100 0.0234 0.5157 0.6216 0.5637 0.9927
0.0149 7.35 125 0.0174 0.7801 0.8054 0.7926 0.9957
0.0104 8.82 150 0.0156 0.78 0.8432 0.8104 0.9959
0.0059 10.29 175 0.0160 0.8360 0.8541 0.8449 0.9960
0.005 11.76 200 0.0139 0.8333 0.8649 0.8488 0.9964
0.003 13.24 225 0.0164 0.8263 0.8486 0.8373 0.9961
0.0024 14.71 250 0.0146 0.7980 0.8541 0.8251 0.9964
0.0023 16.18 275 0.0132 0.8267 0.9027 0.8630 0.9969
0.0016 17.65 300 0.0133 0.8274 0.8811 0.8534 0.9971
0.0015 19.12 325 0.0129 0.8235 0.9081 0.8638 0.9971
0.0014 20.59 350 0.0163 0.8703 0.8703 0.8703 0.9968
0.0013 22.06 375 0.0141 0.8402 0.8811 0.8602 0.9969
0.0013 23.53 400 0.0145 0.8438 0.8757 0.8594 0.9968
0.0011 25.0 425 0.0149 0.8482 0.8757 0.8617 0.9969
0.0011 26.47 450 0.0138 0.8351 0.8757 0.8549 0.9968
0.0011 27.94 475 0.0136 0.8308 0.8757 0.8526 0.9968

Framework versions

  • Transformers 4.27.4
  • Pytorch 2.0.0+cu118
  • Datasets 2.11.0
  • Tokenizers 0.13.2