Bio_ClinicalBERT_fold_5_binary_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: 1.6689
• F1: 0.8148

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	288	0.4406	0.8072
0.4043	2.0	576	0.4952	0.8059
0.4043	3.0	864	0.4988	0.8222
0.2025	4.0	1152	0.8866	0.7948
0.2025	5.0	1440	0.9027	0.8176
0.0865	6.0	1728	1.1263	0.8003
0.035	7.0	2016	1.2498	0.7998
0.035	8.0	2304	1.3188	0.8093
0.0133	9.0	2592	1.4641	0.8021
0.0133	10.0	2880	1.4972	0.8042
0.0119	11.0	3168	1.5511	0.8057
0.0119	12.0	3456	1.5184	0.8108
0.0131	13.0	3744	1.5716	0.8017
0.0067	14.0	4032	1.5305	0.8176
0.0067	15.0	4320	1.4945	0.8227
0.0113	16.0	4608	1.5241	0.8216
0.0113	17.0	4896	1.5571	0.8182
0.0072	18.0	5184	1.6044	0.8107
0.0072	19.0	5472	1.6129	0.8156
0.002	20.0	5760	1.6990	0.8126
0.0036	21.0	6048	1.6867	0.8109
0.0036	22.0	6336	1.7301	0.8100
0.0021	23.0	6624	1.6595	0.8167
0.0021	24.0	6912	1.6577	0.8132
0.0029	25.0	7200	1.6689	0.8148

Framework versions

• Transformers 4.21.0
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1