ner_fine_tuned

This model is a fine-tuned version of cahya/bert-base-indonesian-NER on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0080
• Precision: 0.6970
• Recall: 0.5349
• F1: 0.6053
• Accuracy: 0.8900

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: 8
• eval_batch_size: 8
• 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	8	0.5649	0.625	0.4651	0.5333	0.8832
No log	2.0	16	0.6457	0.7857	0.5116	0.6197	0.9003
No log	3.0	24	0.7181	0.6471	0.5116	0.5714	0.8832
No log	4.0	32	0.8134	0.6970	0.5349	0.6053	0.8900
No log	5.0	40	0.8528	0.6667	0.5116	0.5789	0.8866
No log	6.0	48	0.8893	0.6667	0.5116	0.5789	0.8866
No log	7.0	56	0.9148	0.6667	0.5116	0.5789	0.8866
No log	8.0	64	0.9440	0.6667	0.5116	0.5789	0.8866
No log	9.0	72	0.9744	0.6970	0.5349	0.6053	0.8900
No log	10.0	80	0.9895	0.6765	0.5349	0.5974	0.8900
No log	11.0	88	0.9968	0.6970	0.5349	0.6053	0.8900
No log	12.0	96	1.0015	0.6970	0.5349	0.6053	0.8900
No log	13.0	104	1.0049	0.6970	0.5349	0.6053	0.8900
No log	14.0	112	1.0072	0.6970	0.5349	0.6053	0.8900
No log	15.0	120	1.0080	0.6970	0.5349	0.6053	0.8900

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1