Variome_5e-05_29_03

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: 0.1054
• Precision: 0.5304
• Recall: 0.4586
• F1: 0.4919
• Accuracy: 0.9843

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: 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
0.5733	5.0	25	0.1581	0.0	0.0	0.0	0.9794
0.1685	10.0	50	0.1553	0.0	0.0	0.0	0.9794
0.1691	15.0	75	0.1547	0.0	0.0	0.0	0.9794
0.1606	20.0	100	0.1335	0.0	0.0	0.0	0.9794
0.1144	25.0	125	0.1044	0.2151	0.1504	0.1770	0.9802
0.0795	30.0	150	0.1023	0.2	0.1504	0.1717	0.9804
0.0612	35.0	175	0.1102	0.4118	0.2105	0.2786	0.9831
0.0465	40.0	200	0.0991	0.4158	0.3158	0.3590	0.9840
0.0352	45.0	225	0.0995	0.4653	0.3534	0.4017	0.9838
0.0281	50.0	250	0.0969	0.4685	0.3910	0.4262	0.9838
0.0223	55.0	275	0.0976	0.5684	0.4060	0.4737	0.9853
0.0183	60.0	300	0.0992	0.5093	0.4135	0.4564	0.9848
0.0154	65.0	325	0.0996	0.5816	0.4286	0.4935	0.9858
0.0131	70.0	350	0.1007	0.5221	0.4436	0.4797	0.9842
0.0109	75.0	375	0.1023	0.5130	0.4436	0.4758	0.9842
0.0094	80.0	400	0.1037	0.5566	0.4436	0.4937	0.9851
0.0085	85.0	425	0.1054	0.5304	0.4586	0.4919	0.9843

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2