nerugm-lora-r8a2d0.15

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: 0.1281
• Precision: 0.7470
• Recall: 0.8629
• F1: 0.8008
• 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
0.7018	1.0	528	0.3353	0.5529	0.4800	0.5138	0.9115
0.2639	2.0	1056	0.1912	0.6494	0.8210	0.7252	0.9412
0.1862	3.0	1584	0.1672	0.6739	0.8536	0.7531	0.9466
0.1612	4.0	2112	0.1446	0.7238	0.8512	0.7824	0.9539
0.1439	5.0	2640	0.1390	0.7254	0.8582	0.7863	0.9545
0.1358	6.0	3168	0.1392	0.7256	0.8652	0.7893	0.9551
0.129	7.0	3696	0.1384	0.7267	0.8698	0.7919	0.9561
0.1228	8.0	4224	0.1339	0.7353	0.8698	0.7969	0.9575
0.1168	9.0	4752	0.1321	0.7439	0.8559	0.7960	0.9577
0.1146	10.0	5280	0.1300	0.7445	0.8582	0.7973	0.9581
0.1105	11.0	5808	0.1327	0.7333	0.8675	0.7948	0.9571
0.1083	12.0	6336	0.1333	0.7342	0.8652	0.7943	0.9569
0.106	13.0	6864	0.1265	0.7490	0.8582	0.7999	0.9591
0.1032	14.0	7392	0.1269	0.7445	0.8582	0.7973	0.9589
0.1023	15.0	7920	0.1291	0.7455	0.8629	0.7999	0.9585
0.1014	16.0	8448	0.1271	0.7400	0.8582	0.7947	0.9575
0.1002	17.0	8976	0.1281	0.7460	0.8722	0.8042	0.9589
0.0986	18.0	9504	0.1304	0.7416	0.8722	0.8016	0.9573
0.0978	19.0	10032	0.1271	0.7520	0.8652	0.8046	0.9589
0.0984	20.0	10560	0.1281	0.7470	0.8629	0.8008	0.9579

Framework versions

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