sentiment-seq_bn-rf64-1

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: 0.3480
• Accuracy: 0.8496
• Precision: 0.8202
• Recall: 0.8136
• F1: 0.8167

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: 30
• eval_batch_size: 8
• 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	Accuracy	Precision	Recall	F1
0.5624	1.0	122	0.5217	0.7243	0.6565	0.6199	0.6280
0.5051	2.0	244	0.5208	0.7193	0.6836	0.7114	0.6888
0.4776	3.0	366	0.4669	0.7719	0.7253	0.7286	0.7269
0.4447	4.0	488	0.4394	0.7794	0.7353	0.7089	0.7191
0.4309	5.0	610	0.4312	0.7995	0.7598	0.7806	0.7680
0.395	6.0	732	0.4173	0.8020	0.7638	0.7899	0.7733
0.3841	7.0	854	0.4012	0.8246	0.7884	0.7884	0.7884
0.3621	8.0	976	0.3882	0.8346	0.8062	0.7830	0.7929
0.3562	9.0	1098	0.3912	0.8321	0.7977	0.7962	0.7969
0.3428	10.0	1220	0.3767	0.8496	0.8211	0.8111	0.8158
0.3282	11.0	1342	0.3736	0.8596	0.8389	0.8132	0.8243
0.3308	12.0	1464	0.3691	0.8571	0.8299	0.8214	0.8255
0.3143	13.0	1586	0.3631	0.8596	0.8424	0.8082	0.8223
0.3173	14.0	1708	0.3592	0.8546	0.8263	0.8196	0.8229
0.305	15.0	1830	0.3542	0.8496	0.8202	0.8136	0.8167
0.2968	16.0	1952	0.3541	0.8546	0.8254	0.8221	0.8238
0.3049	17.0	2074	0.3487	0.8546	0.8284	0.8146	0.8210
0.3001	18.0	2196	0.3514	0.8546	0.8239	0.8272	0.8255
0.2986	19.0	2318	0.3479	0.8622	0.8385	0.8225	0.8298
0.2894	20.0	2440	0.3480	0.8496	0.8202	0.8136	0.8167

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1