nerugm-lora-r4a2d0.05

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: 0.1305
• Precision: 0.7407
• Recall: 0.8698
• F1: 0.8001
• Accuracy: 0.9579

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
0.7682	1.0	528	0.4394	0.4048	0.1185	0.1834	0.8663
0.3466	2.0	1056	0.2217	0.6022	0.7379	0.6632	0.9327
0.2131	3.0	1584	0.1728	0.6765	0.8396	0.7493	0.9428
0.1759	4.0	2112	0.1509	0.7221	0.8559	0.7833	0.9516
0.1563	5.0	2640	0.1422	0.7303	0.8605	0.7901	0.9533
0.1464	6.0	3168	0.1429	0.7202	0.8722	0.7890	0.9541
0.1394	7.0	3696	0.1440	0.7153	0.8745	0.7869	0.9525
0.1325	8.0	4224	0.1398	0.7274	0.8791	0.7961	0.9553
0.1269	9.0	4752	0.1341	0.7420	0.8675	0.7999	0.9579
0.124	10.0	5280	0.1331	0.7379	0.8768	0.8014	0.9565
0.1194	11.0	5808	0.1329	0.7389	0.8815	0.8039	0.9569
0.1171	12.0	6336	0.1337	0.7384	0.8791	0.8027	0.9567
0.1153	13.0	6864	0.1294	0.7447	0.8745	0.8044	0.9587
0.1119	14.0	7392	0.1310	0.7472	0.8791	0.8078	0.9573
0.1109	15.0	7920	0.1312	0.7457	0.8722	0.8040	0.9579
0.1102	16.0	8448	0.1309	0.7442	0.8791	0.8061	0.9581
0.1095	17.0	8976	0.1314	0.7447	0.8815	0.8073	0.9587
0.1073	18.0	9504	0.1323	0.7403	0.8745	0.8018	0.9577
0.107	19.0	10032	0.1300	0.7407	0.8698	0.8001	0.9581
0.1073	20.0	10560	0.1305	0.7407	0.8698	0.8001	0.9579

Framework versions

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