Bio_ClinicalBERT_fold_1_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: 1.9306
• F1: 0.7943

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	290	0.5314	0.7931
0.556	2.0	580	0.5572	0.8042
0.556	3.0	870	0.6381	0.7893
0.2427	4.0	1160	0.9330	0.7807
0.2427	5.0	1450	1.1033	0.7918
0.0987	6.0	1740	1.2582	0.7856
0.0381	7.0	2030	1.3816	0.7807
0.0381	8.0	2320	1.5874	0.7807
0.0199	9.0	2610	1.5712	0.7856
0.0199	10.0	2900	1.5860	0.7856
0.0139	11.0	3190	1.7511	0.7757
0.0139	12.0	3480	1.7635	0.7807
0.0056	13.0	3770	1.7726	0.7893
0.0077	14.0	4060	1.7565	0.7881
0.0077	15.0	4350	1.8406	0.7844
0.005	16.0	4640	1.8595	0.7893
0.005	17.0	4930	1.8831	0.7770
0.0036	18.0	5220	1.9561	0.7881
0.0017	19.0	5510	2.0062	0.7856
0.0017	20.0	5800	1.9275	0.7980
0.0004	21.0	6090	1.8752	0.7993
0.0004	22.0	6380	1.9238	0.7918
0.0022	23.0	6670	1.9552	0.7881
0.0022	24.0	6960	1.9274	0.7955
0.0009	25.0	7250	1.9306	0.7943

Framework versions

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