indobert-base-p1-multilabel-indonesian-hate-speech-modified

This model is a fine-tuned version of indobenchmark/indobert-base-p1 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3655
• F1: 0.8170
• Roc Auc: 0.8883
• Accuracy: 0.7608

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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.2112	1.0	1317	0.1797	0.7887	0.8552	0.7039
0.1389	2.0	2634	0.1804	0.7946	0.8742	0.6872
0.0838	3.0	3951	0.1981	0.8152	0.8825	0.7403
0.0529	4.0	5268	0.2520	0.8036	0.8671	0.7434
0.0349	5.0	6585	0.2577	0.8065	0.8908	0.7244
0.0203	6.0	7902	0.2994	0.7986	0.8776	0.7198
0.0154	7.0	9219	0.3155	0.8086	0.8927	0.7434
0.0112	8.0	10536	0.3284	0.8070	0.8789	0.7540
0.0061	9.0	11853	0.3507	0.8065	0.8782	0.7525
0.0079	10.0	13170	0.3514	0.8091	0.8956	0.7396
0.0052	11.0	14487	0.3420	0.8135	0.8854	0.7608
0.0047	12.0	15804	0.3503	0.8164	0.8842	0.7608
0.0046	13.0	17121	0.3558	0.8165	0.8871	0.7616
0.0014	14.0	18438	0.3790	0.8096	0.8926	0.7472
0.0021	15.0	19755	0.3655	0.8170	0.8883	0.7608

Framework versions

• Transformers 4.51.3
• Pytorch 2.7.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.1