nerugm-lora-r2a2d0.05

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: 0.1346
• Precision: 0.7366
• Recall: 0.8629
• F1: 0.7948
• 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
0.7885	1.0	528	0.4616	0.3182	0.0813	0.1296	0.8599
0.3921	2.0	1056	0.2524	0.6053	0.6798	0.6404	0.9273
0.2392	3.0	1584	0.1932	0.6500	0.7844	0.7109	0.9382
0.1931	4.0	2112	0.1676	0.6905	0.8234	0.7511	0.9444
0.1719	5.0	2640	0.1583	0.7056	0.8396	0.7668	0.9478
0.1602	6.0	3168	0.1539	0.7115	0.8582	0.7780	0.9502
0.1533	7.0	3696	0.1520	0.7031	0.8629	0.7748	0.9506
0.1455	8.0	4224	0.1456	0.7263	0.8559	0.7858	0.9525
0.1398	9.0	4752	0.1425	0.7301	0.8536	0.7870	0.9537
0.1368	10.0	5280	0.1395	0.7229	0.8536	0.7828	0.9533
0.1331	11.0	5808	0.1365	0.7360	0.8536	0.7904	0.9551
0.1305	12.0	6336	0.1377	0.7332	0.8605	0.7918	0.9549
0.1279	13.0	6864	0.1357	0.7415	0.8582	0.7956	0.9565
0.1251	14.0	7392	0.1355	0.7371	0.8652	0.7960	0.9555
0.1239	15.0	7920	0.1359	0.7366	0.8629	0.7948	0.9549
0.1231	16.0	8448	0.1347	0.7351	0.8629	0.7939	0.9551
0.122	17.0	8976	0.1353	0.7351	0.8629	0.7939	0.9555
0.1205	18.0	9504	0.1356	0.7317	0.8605	0.7909	0.9549
0.1202	19.0	10032	0.1347	0.7351	0.8629	0.7939	0.9551
0.1204	20.0	10560	0.1346	0.7366	0.8629	0.7948	0.9555

Framework versions

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