roberta_ner_personal_info_v2

This model is a fine-tuned version of tner/roberta-large-ontonotes5 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0202
• Precision: 0.9744
• Recall: 0.9803
• F1: 0.9774
• Accuracy: 0.9956

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: 20
• eval_batch_size: 20
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.1952	1.0	621	0.0387	0.9334	0.9414	0.9374	0.9886
0.0407	2.0	1242	0.0267	0.9587	0.9606	0.9596	0.9923
0.0246	3.0	1863	0.0230	0.9616	0.9703	0.9660	0.9936
0.0163	4.0	2484	0.0206	0.9705	0.9730	0.9717	0.9946
0.0097	5.0	3105	0.0203	0.9702	0.9747	0.9724	0.9946
0.0076	6.0	3726	0.0194	0.9717	0.9774	0.9746	0.9950
0.0066	7.0	4347	0.0193	0.9751	0.9780	0.9766	0.9954
0.0058	8.0	4968	0.0201	0.9720	0.9781	0.9750	0.9952
0.0044	9.0	5589	0.0222	0.9717	0.9751	0.9734	0.9950
0.0037	10.0	6210	0.0204	0.9736	0.9786	0.9761	0.9954
0.0034	11.0	6831	0.0192	0.9771	0.9808	0.9789	0.9958
0.0028	12.0	7452	0.0210	0.9723	0.9789	0.9756	0.9953
0.0025	13.0	8073	0.0222	0.9747	0.9792	0.9769	0.9955
0.0023	14.0	8694	0.0198	0.9771	0.9806	0.9789	0.9958
0.002	15.0	9315	0.0202	0.9744	0.9803	0.9774	0.9956

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3