nerugm-lora-r2a1d0.1

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: 0.1332
• Precision: 0.7287
• Recall: 0.8536
• F1: 0.7862
• Accuracy: 0.9555

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
0.7886	1.0	528	0.4607	0.3243	0.0837	0.1330	0.8597
0.3911	2.0	1056	0.2542	0.6081	0.6915	0.6471	0.9293
0.2384	3.0	1584	0.1934	0.6527	0.7937	0.7163	0.9376
0.1934	4.0	2112	0.1678	0.6880	0.8187	0.7477	0.9446
0.172	5.0	2640	0.1589	0.6901	0.8373	0.7566	0.9468
0.1602	6.0	3168	0.1533	0.6931	0.8489	0.7631	0.9488
0.1532	7.0	3696	0.1505	0.6935	0.8559	0.7662	0.9498
0.1457	8.0	4224	0.1456	0.7103	0.8536	0.7754	0.9522
0.1401	9.0	4752	0.1418	0.7301	0.8536	0.7870	0.9543
0.1375	10.0	5280	0.1388	0.7308	0.8582	0.7894	0.9551
0.1331	11.0	5808	0.1360	0.7308	0.8582	0.7894	0.9555
0.1304	12.0	6336	0.1365	0.7258	0.8536	0.7845	0.9549
0.1285	13.0	6864	0.1343	0.7380	0.8512	0.7906	0.9559
0.1255	14.0	7392	0.1345	0.7401	0.8605	0.7958	0.9559
0.1249	15.0	7920	0.1346	0.7332	0.8605	0.7918	0.9549
0.1238	16.0	8448	0.1342	0.7307	0.8559	0.7883	0.9551
0.1232	17.0	8976	0.1342	0.7326	0.8582	0.7905	0.9557
0.1215	18.0	9504	0.1351	0.7317	0.8605	0.7909	0.9549
0.1209	19.0	10032	0.1337	0.7278	0.8559	0.7866	0.9547
0.1207	20.0	10560	0.1332	0.7287	0.8536	0.7862	0.9555

Framework versions

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