nerugm-lora-r8a2d0.05

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: 0.1266
• Precision: 0.7622
• Recall: 0.8698
• F1: 0.8125
• Accuracy: 0.9591

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
0.7039	1.0	528	0.3293	0.5553	0.4962	0.5241	0.9123
0.2536	2.0	1056	0.1835	0.6530	0.8210	0.7274	0.9424
0.1831	3.0	1584	0.1832	0.6678	0.8210	0.7365	0.9440
0.1623	4.0	2112	0.1463	0.7213	0.8466	0.7789	0.9535
0.1439	5.0	2640	0.1387	0.7173	0.8420	0.7747	0.9541
0.1348	6.0	3168	0.1383	0.7256	0.8652	0.7893	0.9553
0.1293	7.0	3696	0.1394	0.7242	0.8652	0.7885	0.9545
0.124	8.0	4224	0.1351	0.7353	0.8698	0.7969	0.9569
0.1176	9.0	4752	0.1304	0.7404	0.8536	0.7930	0.9561
0.1153	10.0	5280	0.1278	0.7582	0.8582	0.8051	0.9585
0.111	11.0	5808	0.1304	0.7386	0.8652	0.7969	0.9579
0.109	12.0	6336	0.1323	0.7415	0.8652	0.7986	0.9565
0.1077	13.0	6864	0.1253	0.7649	0.8675	0.8130	0.9597
0.1032	14.0	7392	0.1243	0.7639	0.8629	0.8104	0.9593
0.1035	15.0	7920	0.1261	0.7664	0.8675	0.8138	0.9597
0.1017	16.0	8448	0.1258	0.7470	0.8559	0.7977	0.9577
0.1004	17.0	8976	0.1278	0.7576	0.8698	0.8098	0.9589
0.099	18.0	9504	0.1284	0.7510	0.8675	0.8051	0.9585
0.0991	19.0	10032	0.1256	0.7572	0.8605	0.8055	0.9581
0.0984	20.0	10560	0.1266	0.7622	0.8698	0.8125	0.9591

Framework versions

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