igorsterner/german-english-roberta-base-binary-ner

german-english-binary-ner-roberta-base-30-final

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

• Loss: 0.1571
• Precision: 0.7113
• Recall: 0.8042
• F1: 0.7549

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: 1e-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
No log	2.36	250	0.0813	0.6063	0.7778	0.6814
0.0606	4.72	500	0.0871	0.6745	0.8290	0.7438
0.0606	7.08	750	0.1051	0.7218	0.7746	0.7473
0.0099	9.43	1000	0.1103	0.7428	0.7628	0.7527
0.0099	11.79	1250	0.1156	0.7349	0.7843	0.7588
0.0037	14.15	1500	0.1200	0.7323	0.7881	0.7592
0.0037	16.51	1750	0.1415	0.7139	0.7977	0.7535
0.0018	18.87	2000	0.1339	0.7218	0.7880	0.7534
0.0018	21.23	2250	0.1423	0.7533	0.7820	0.7674
0.001	23.58	2500	0.1506	0.7192	0.7806	0.7486
0.001	25.94	2750	0.1521	0.7165	0.8077	0.7594
0.0006	28.3	3000	0.1571	0.7113	0.8042	0.7549

Framework versions

• Transformers 4.36.1
• Pytorch 2.1.2+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0