Edit model card

electramed-small-NCBI-ner

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

  • Loss: 0.0664
  • Precision: 0.8083
  • Recall: 0.8875
  • F1: 0.8461
  • Accuracy: 0.9821

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
0.4787 1.0 340 0.5090 0.6090 0.5062 0.5529 0.9608
0.2029 2.0 680 0.1890 0.7643 0.8208 0.7916 0.9774
0.1402 3.0 1020 0.1106 0.7839 0.8802 0.8292 0.9807
0.075 4.0 1360 0.0876 0.8162 0.8698 0.8422 0.9817
0.0408 5.0 1700 0.0776 0.8090 0.8781 0.8422 0.9818
0.0308 6.0 2040 0.0697 0.8044 0.8823 0.8415 0.9825
0.0405 7.0 2380 0.0680 0.8118 0.8854 0.8470 0.9830
0.0138 8.0 2720 0.0665 0.8111 0.8854 0.8466 0.9826
0.0223 9.0 3060 0.0675 0.8064 0.8896 0.8460 0.9821
0.0395 10.0 3400 0.0664 0.8083 0.8875 0.8461 0.9821

Framework versions

  • Transformers 4.21.1
  • Pytorch 1.12.1+cu113
  • Datasets 2.4.0
  • Tokenizers 0.12.1
Downloads last month
1

Dataset used to train chintagunta85/electramed-small-NCBI-ner

Evaluation results