Bio_ClinicalBERT_fold_4_ternary_v1

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: 1.7349
• F1: 0.8052

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: 25

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	1.0	289	0.5378	0.7818
0.5561	2.0	578	0.4835	0.8002
0.5561	3.0	867	0.6401	0.7978
0.2473	4.0	1156	0.8665	0.7842
0.2473	5.0	1445	0.9942	0.7965
0.1002	6.0	1734	1.1535	0.8015
0.0428	7.0	2023	1.2619	0.8027
0.0428	8.0	2312	1.4386	0.7990
0.017	9.0	2601	1.4864	0.8039
0.017	10.0	2890	1.4817	0.8015
0.0145	11.0	3179	1.5205	0.8052
0.0145	12.0	3468	1.6825	0.7842
0.0115	13.0	3757	1.6670	0.7990
0.0083	14.0	4046	1.7283	0.7904
0.0083	15.0	4335	1.6552	0.8039
0.0071	16.0	4624	1.6760	0.8076
0.0071	17.0	4913	1.6973	0.7891
0.0109	18.0	5202	1.6050	0.8027
0.0109	19.0	5491	1.6379	0.8126
0.0037	20.0	5780	1.6936	0.8039
0.0013	21.0	6069	1.7187	0.8027
0.0013	22.0	6358	1.7839	0.7965
0.0015	23.0	6647	1.7551	0.8015
0.0015	24.0	6936	1.7312	0.8064
0.001	25.0	7225	1.7349	0.8052

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1