metadata

license: apache-2.0
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: bert-medical-ner
    results: []

bert-medical-ner

This model is a fine-tuned version of bert-base-cased on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.4386
Precision: 0.7582
Recall: 0.7089
F1: 0.7327
Accuracy: 0.9434

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: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	63	0.4577	0.8113	0.6343	0.7119	0.9397
No log	2.0	126	0.4098	0.7941	0.6563	0.7187	0.9419
No log	3.0	189	0.3993	0.7369	0.6744	0.7043	0.9389
No log	4.0	252	0.4033	0.7312	0.7089	0.7199	0.9418
No log	5.0	315	0.4329	0.7509	0.6919	0.7202	0.9416
No log	6.0	378	0.4343	0.7545	0.6829	0.7169	0.9420
No log	7.0	441	0.4348	0.7168	0.7140	0.7154	0.9402
0.0142	8.0	504	0.4362	0.7285	0.7055	0.7168	0.9399
0.0142	9.0	567	0.4420	0.7573	0.7072	0.7314	0.9436
0.0142	10.0	630	0.4371	0.7452	0.7027	0.7233	0.9423
0.0142	11.0	693	0.4400	0.7648	0.6947	0.7281	0.9429
0.0142	12.0	756	0.4346	0.7556	0.7027	0.7282	0.9422
0.0142	13.0	819	0.4382	0.7504	0.7089	0.7291	0.9428
0.0142	14.0	882	0.4368	0.7536	0.7123	0.7323	0.9434
0.0142	15.0	945	0.4386	0.7582	0.7089	0.7327	0.9434

Framework versions

Transformers 4.29.2
Pytorch 2.0.1+cu117
Datasets 2.12.0
Tokenizers 0.13.3