scenario-TCR-NER_data-univner_en

This model is a fine-tuned version of xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2275
• Precision: 0.7534
• Recall: 0.6859
• F1: 0.7181
• Accuracy: 0.9722

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: 3e-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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0841	1.28	500	0.1016	0.7187	0.6674	0.6921	0.9716
0.025	2.55	1000	0.1149	0.7384	0.6683	0.7016	0.9705
0.0161	3.83	1500	0.1161	0.7189	0.6953	0.7069	0.9726
0.0105	5.1	2000	0.1658	0.7504	0.6520	0.6978	0.9694
0.0059	6.38	2500	0.1509	0.7280	0.6902	0.7086	0.9726
0.0045	7.65	3000	0.1774	0.7381	0.6686	0.7016	0.9714
0.0033	8.93	3500	0.1494	0.7248	0.7279	0.7263	0.9746
0.0027	10.2	4000	0.1712	0.7501	0.6899	0.7188	0.9725
0.002	11.48	4500	0.2144	0.7732	0.6472	0.7046	0.9700
0.0016	12.76	5000	0.1804	0.7570	0.6871	0.7203	0.9732
0.0023	14.03	5500	0.1733	0.7599	0.7060	0.7319	0.9739
0.0013	15.31	6000	0.1777	0.7421	0.7145	0.7280	0.9735
0.0011	16.58	6500	0.2035	0.7550	0.6835	0.7175	0.9723
0.0012	17.86	7000	0.2195	0.7617	0.6605	0.7075	0.9714
0.0008	19.13	7500	0.2064	0.7358	0.6891	0.7117	0.9722
0.0005	20.41	8000	0.2327	0.7599	0.6780	0.7166	0.9719
0.0006	21.68	8500	0.2303	0.7646	0.6611	0.7091	0.9712
0.0005	22.96	9000	0.2317	0.7535	0.6721	0.7104	0.9715
0.0004	24.23	9500	0.2436	0.7574	0.6670	0.7093	0.9708
0.0003	25.51	10000	0.2548	0.7661	0.6583	0.7082	0.9707
0.0004	26.79	10500	0.2275	0.7534	0.6859	0.7181	0.9722

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3