Noctuaru/topik_syariah_p2_v6

results

This model is a fine-tuned version of indobenchmark/indobert-base-p2 on the None dataset. It achieves the following results on the evaluation set:

• Accuracy: 0.8841
• F1 Weighted: 0.8830
• Loss: 0.4070

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: 5.055195102732862e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Accuracy	F1 Weighted	Validation Loss
No log	1.0	8	0.3913	0.2201	1.0694
1.1334	2.0	16	0.6812	0.5872	0.6892
0.7706	3.0	24	0.7681	0.7472	0.6135
0.4775	4.0	32	0.8261	0.8161	0.4961
0.3038	5.0	40	0.8406	0.8433	0.3955
0.3038	6.0	48	0.8551	0.8509	0.4645
0.2063	7.0	56	0.8406	0.8422	0.3706
0.1441	8.0	64	0.8841	0.8830	0.4070
0.1155	9.0	72	0.8551	0.8534	0.4334
0.0703	10.0	80	0.8696	0.8703	0.4704
0.0703	11.0	88	0.8551	0.8477	0.5174
0.0535	12.0	96	0.8551	0.8531	0.4605
0.0377	13.0	104	0.8696	0.8667	0.4920

Framework versions

• Transformers 4.56.2
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1