sentiment-seq_bn-rf64-2

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: 0.3398
• Accuracy: 0.8697
• Precision: 0.8520
• Recall: 0.8253
• F1: 0.8368

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
0.5604	1.0	122	0.5191	0.7168	0.6431	0.6021	0.6083
0.5023	2.0	244	0.5118	0.7293	0.6835	0.7010	0.6894
0.4676	3.0	366	0.4660	0.7544	0.7047	0.7087	0.7066
0.4415	4.0	488	0.4403	0.7845	0.7401	0.7300	0.7346
0.4208	5.0	610	0.4252	0.8120	0.7731	0.7745	0.7738
0.383	6.0	732	0.4142	0.8170	0.7790	0.7981	0.7869
0.3751	7.0	854	0.3982	0.8271	0.8007	0.7651	0.7791
0.3554	8.0	976	0.3822	0.8396	0.8187	0.7790	0.7944
0.3502	9.0	1098	0.3767	0.8521	0.8201	0.8279	0.8238
0.3326	10.0	1220	0.3653	0.8571	0.8322	0.8164	0.8236
0.3246	11.0	1342	0.3637	0.8571	0.8335	0.8139	0.8226
0.3255	12.0	1464	0.3571	0.8546	0.8284	0.8146	0.8210
0.3096	13.0	1586	0.3600	0.8471	0.8321	0.7843	0.8023
0.3123	14.0	1708	0.3455	0.8596	0.8347	0.8207	0.8272
0.2937	15.0	1830	0.3460	0.8622	0.8467	0.8100	0.8249
0.2941	16.0	1952	0.3415	0.8596	0.8336	0.8232	0.8281
0.3031	17.0	2074	0.3417	0.8647	0.8410	0.8267	0.8333
0.3003	18.0	2196	0.3399	0.8622	0.8361	0.8275	0.8316
0.2976	19.0	2318	0.3402	0.8672	0.8496	0.8210	0.8332
0.2956	20.0	2440	0.3398	0.8697	0.8520	0.8253	0.8368

Framework versions

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