IRyS-NER-Recruitment

This model is a fine-tuned version of roberta-base on a resume dataset. It achieves the following results on the evaluation set:

• Loss: 0.0775
• Precision: 0.7828
• Recall: 0.8439
• F1: 0.8122
• Accuracy: 0.9778

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: 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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	17	0.4337	0.6471	0.0799	0.1422	0.9024
No log	2.0	34	0.2148	0.4065	0.3612	0.3825	0.9312
No log	3.0	51	0.1374	0.6051	0.7160	0.6559	0.9607
No log	4.0	68	0.0988	0.6835	0.7995	0.7369	0.9669
No log	5.0	85	0.0926	0.7103	0.8321	0.7664	0.9692
No log	6.0	102	0.0880	0.7364	0.8721	0.7985	0.9723
No log	7.0	119	0.0804	0.7542	0.8185	0.7850	0.9733
No log	8.0	136	0.0839	0.7490	0.8639	0.8024	0.9733
No log	9.0	153	0.0805	0.7720	0.8267	0.7984	0.9767
No log	10.0	170	0.0799	0.7786	0.8267	0.8019	0.9761
No log	11.0	187	0.0777	0.7841	0.8339	0.8083	0.9775
No log	12.0	204	0.0804	0.7644	0.8566	0.8079	0.9761
No log	13.0	221	0.0775	0.7828	0.8439	0.8122	0.9778
No log	14.0	238	0.0810	0.7674	0.8593	0.8108	0.9771
No log	15.0	255	0.0823	0.7717	0.8557	0.8115	0.9776

Framework versions

• Transformers 4.27.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2