stbl_clinical_bert_ft_rs4

This model is a fine-tuned version of emilyalsentzer/Bio_ClinicalBERT on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1088
• F1: 0.9076

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss	F1
0.2994	1.0	101	0.0977	0.8416
0.0639	2.0	202	0.0846	0.8689
0.0318	3.0	303	0.0781	0.8879
0.0173	4.0	404	0.0770	0.8934
0.0099	5.0	505	0.0905	0.9021
0.005	6.0	606	0.0963	0.9020
0.0031	7.0	707	0.1024	0.9095
0.002	8.0	808	0.1063	0.9057
0.0017	9.0	909	0.1072	0.9076
0.0014	10.0	1010	0.1103	0.9089
0.0013	11.0	1111	0.1093	0.9087
0.0008	12.0	1212	0.1088	0.9076

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1+cu113
• Datasets 2.5.1
• Tokenizers 0.12.1