nerugm-lora-r4a2d0.1

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: 0.1302
• Precision: 0.7375
• Recall: 0.8605
• F1: 0.7943
• Accuracy: 0.9573

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
0.7665	1.0	528	0.4290	0.3803	0.1255	0.1887	0.8711
0.336	2.0	1056	0.2177	0.6187	0.7751	0.6882	0.9335
0.2067	3.0	1584	0.1743	0.6523	0.8187	0.7261	0.9410
0.1734	4.0	2112	0.1525	0.7026	0.8443	0.7670	0.9500
0.1557	5.0	2640	0.1442	0.7125	0.8512	0.7757	0.9524
0.146	6.0	3168	0.1445	0.7085	0.8629	0.7781	0.9520
0.1397	7.0	3696	0.1444	0.7145	0.8768	0.7874	0.9525
0.1338	8.0	4224	0.1386	0.7262	0.8675	0.7906	0.9545
0.1277	9.0	4752	0.1365	0.7395	0.8629	0.7965	0.9561
0.1255	10.0	5280	0.1332	0.7348	0.8629	0.7937	0.9563
0.1215	11.0	5808	0.1330	0.7242	0.8652	0.7885	0.9557
0.1189	12.0	6336	0.1340	0.7342	0.8652	0.7943	0.9561
0.1179	13.0	6864	0.1295	0.7445	0.8582	0.7973	0.9571
0.114	14.0	7392	0.1295	0.7446	0.8675	0.8014	0.9579
0.1128	15.0	7920	0.1317	0.7371	0.8652	0.7960	0.9571
0.1115	16.0	8448	0.1300	0.7376	0.8675	0.7973	0.9575
0.1109	17.0	8976	0.1307	0.7357	0.8652	0.7952	0.9577
0.1097	18.0	9504	0.1319	0.7386	0.8652	0.7969	0.9575
0.1086	19.0	10032	0.1296	0.7375	0.8605	0.7943	0.9573
0.1094	20.0	10560	0.1302	0.7375	0.8605	0.7943	0.9573

Framework versions

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