metadata

tags:
  - generated_from_trainer
datasets:
  - i2b22014
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: electramed-small-deid2014-ner-v3
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: i2b22014
          type: i2b22014
          config: i2b22014-deid
          split: train
          args: i2b22014-deid
        metrics:
          - name: Precision
            type: precision
            value: 0.7776378519384726
          - name: Recall
            type: recall
            value: 0.7946502435885652
          - name: F1
            type: f1
            value: 0.7860520094562647
          - name: Accuracy
            type: accuracy
            value: 0.9908687950002661

electramed-small-deid2014-ner-v3

This model is a fine-tuned version of giacomomiolo/electramed_small_scivocab on the i2b22014 dataset. It achieves the following results on the evaluation set:

Loss: 0.0354
Precision: 0.7776
Recall: 0.7947
F1: 0.7861
Accuracy: 0.9909

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
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0125	1.0	1838	0.1338	0.3514	0.3812	0.3657	0.9715
0.0032	2.0	3676	0.0856	0.4444	0.5156	0.4774	0.9778
0.0012	3.0	5514	0.0678	0.5222	0.5994	0.5581	0.9819
0.0006	4.0	7352	0.0547	0.6900	0.7025	0.6962	0.9865
0.018	5.0	9190	0.0466	0.7227	0.7468	0.7345	0.9881
0.0002	6.0	11028	0.0419	0.7396	0.7664	0.7528	0.9891
0.0002	7.0	12866	0.0390	0.7730	0.7693	0.7712	0.9901
0.0002	8.0	14704	0.0368	0.7778	0.7822	0.7800	0.9906
0.0001	9.0	16542	0.0359	0.7765	0.7898	0.7831	0.9907
0.0001	10.0	18380	0.0354	0.7776	0.7947	0.7861	0.9909

Framework versions

Transformers 4.21.3
Pytorch 1.12.1+cu113
Datasets 2.4.0
Tokenizers 0.12.1