3_bert-base-NER-finetuned_20241018_v2

This model is a fine-tuned version of dslim/bert-base-NER on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0039
• Precision: 0.9993
• Recall: 0.9996
• F1: 0.9994
• Accuracy: 0.9995

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: 32
• eval_batch_size: 32
• 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	Precision	Recall	F1	Accuracy
0.0012	1.0	3556	0.0045	0.9968	0.9986	0.9977	0.9989
0.0034	2.0	7112	0.0034	0.9981	0.9991	0.9986	0.9991
0.0022	3.0	10668	0.0029	0.9984	0.9994	0.9989	0.9993
0.0038	4.0	14224	0.0024	0.9986	0.9993	0.9989	0.9994
0.0019	5.0	17780	0.0025	0.9986	0.9994	0.9990	0.9994
0.0013	6.0	21336	0.0030	0.9988	0.9994	0.9991	0.9994
0.0004	7.0	24892	0.0027	0.9986	0.9992	0.9989	0.9993
0.0	8.0	28448	0.0023	0.9986	0.9994	0.9990	0.9994
0.0001	9.0	32004	0.0027	0.9986	0.9993	0.9990	0.9993
0.0	10.0	35560	0.0029	0.9990	0.9995	0.9993	0.9994
0.0001	11.0	39116	0.0029	0.9990	0.9995	0.9992	0.9995
0.0	12.0	42672	0.0029	0.9991	0.9993	0.9992	0.9994
0.0005	13.0	46228	0.0035	0.9990	0.9994	0.9992	0.9994
0.0021	14.0	49784	0.0033	0.9991	0.9993	0.9992	0.9993
0.0012	15.0	53340	0.0038	0.9991	0.9994	0.9992	0.9994
0.0	16.0	56896	0.0034	0.9991	0.9995	0.9993	0.9994
0.0	17.0	60452	0.0039	0.9993	0.9996	0.9994	0.9995
0.0001	18.0	64008	0.0041	0.9992	0.9996	0.9994	0.9994
0.0	19.0	67564	0.0037	0.9991	0.9995	0.9993	0.9994
0.0	20.0	71120	0.0041	0.9991	0.9996	0.9993	0.9994
0.0	21.0	74676	0.0040	0.9991	0.9995	0.9993	0.9994
0.0	22.0	78232	0.0042	0.9991	0.9995	0.9993	0.9995
0.0	23.0	81788	0.0050	0.9992	0.9995	0.9994	0.9994
0.0	24.0	85344	0.0049	0.9992	0.9995	0.9993	0.9994
0.0	25.0	88900	0.0050	0.9992	0.9995	0.9994	0.9995

Framework versions

• Transformers 4.44.1
• Pytorch 2.2.0a0+81ea7a4
• Datasets 2.21.0
• Tokenizers 0.19.1