Bio_ClinicalBERT_fold_3_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: 1.8860
• F1: 0.8051

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	0.4493	0.7916
0.3975	2.0	578	0.4608	0.7909
0.3975	3.0	867	0.8364	0.7726
0.1885	4.0	1156	1.0380	0.7902
0.1885	5.0	1445	1.1612	0.7921
0.0692	6.0	1734	1.3894	0.7761
0.0295	7.0	2023	1.3730	0.7864
0.0295	8.0	2312	1.4131	0.7939
0.0161	9.0	2601	1.5538	0.7929
0.0161	10.0	2890	1.6417	0.7931
0.006	11.0	3179	1.5745	0.7974
0.006	12.0	3468	1.7212	0.7908
0.0132	13.0	3757	1.7349	0.7945
0.0062	14.0	4046	1.7593	0.7908
0.0062	15.0	4335	1.7420	0.8035
0.0073	16.0	4624	1.7620	0.8007
0.0073	17.0	4913	1.8286	0.7908
0.0033	18.0	5202	1.7863	0.7977
0.0033	19.0	5491	1.9275	0.7919
0.0035	20.0	5780	1.8481	0.8042
0.0035	21.0	6069	1.9465	0.8012
0.0035	22.0	6358	1.8177	0.8044
0.005	23.0	6647	1.8615	0.8030
0.005	24.0	6936	1.8427	0.8054
0.0011	25.0	7225	1.8860	0.8051

Framework versions

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