nerugm-ia3

This model is a fine-tuned version of indolem/indobert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1662
• Precision: 0.6732
• Recall: 0.7937
• F1: 0.7285
• Accuracy: 0.9450

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: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20.0

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.8173	1.0	528	0.5014	0.2653	0.0302	0.0543	0.8480
0.4808	2.0	1056	0.3565	0.5217	0.3068	0.3864	0.8941
0.3767	3.0	1584	0.2893	0.5761	0.5520	0.5638	0.9194
0.3159	4.0	2112	0.2479	0.6181	0.6612	0.6390	0.9279
0.2785	5.0	2640	0.2236	0.6304	0.7008	0.6637	0.9331
0.2527	6.0	3168	0.2097	0.6449	0.7333	0.6862	0.9361
0.2365	7.0	3696	0.1997	0.6415	0.7519	0.6923	0.9376
0.2243	8.0	4224	0.1905	0.6534	0.7612	0.7032	0.9394
0.2134	9.0	4752	0.1857	0.6522	0.7705	0.7064	0.9398
0.2072	10.0	5280	0.1814	0.6562	0.7798	0.7127	0.9418
0.2009	11.0	5808	0.1756	0.6601	0.7798	0.7150	0.9426
0.1962	12.0	6336	0.1738	0.6589	0.7844	0.7162	0.9436
0.1921	13.0	6864	0.1720	0.6621	0.7914	0.7210	0.9442
0.1887	14.0	7392	0.1705	0.6621	0.7914	0.7210	0.9442
0.1857	15.0	7920	0.1688	0.6680	0.7937	0.7254	0.9448
0.1846	16.0	8448	0.1684	0.6712	0.7960	0.7283	0.9450
0.1833	17.0	8976	0.1676	0.6706	0.7937	0.7270	0.9446
0.1804	18.0	9504	0.1667	0.6719	0.7937	0.7278	0.9446
0.1816	19.0	10032	0.1664	0.6719	0.7937	0.7278	0.9448
0.1801	20.0	10560	0.1662	0.6732	0.7937	0.7285	0.9450

Framework versions

• Transformers 4.39.3
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.15.2