sentiment-base-4

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: 0.8053
• Accuracy: 0.8922
• Precision: 0.8694
• Recall: 0.8712
• F1: 0.8703

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
0.371	1.0	122	0.2789	0.8847	0.8717	0.8434	0.8557
0.214	2.0	244	0.2703	0.8947	0.8682	0.8855	0.8760
0.141	3.0	366	0.4446	0.8747	0.8586	0.8313	0.8431
0.093	4.0	488	0.5896	0.8697	0.8520	0.8253	0.8368
0.0469	5.0	610	0.6099	0.8797	0.8530	0.8599	0.8563
0.0498	6.0	732	0.6610	0.8997	0.9016	0.8516	0.8715
0.0257	7.0	854	0.6781	0.8872	0.8917	0.8302	0.8532
0.0267	8.0	976	0.8200	0.8872	0.8951	0.8277	0.8523
0.016	9.0	1098	0.5966	0.8997	0.8740	0.8916	0.8819
0.0132	10.0	1220	0.6437	0.9023	0.8792	0.8883	0.8835
0.0161	11.0	1342	0.6797	0.9073	0.8920	0.8819	0.8867
0.0091	12.0	1464	0.6954	0.9098	0.8999	0.8787	0.8883
0.0101	13.0	1586	0.6751	0.9123	0.8910	0.9004	0.8955
0.0025	14.0	1708	0.7317	0.9023	0.8934	0.8658	0.8780
0.0088	15.0	1830	0.6789	0.8897	0.8670	0.8670	0.8670
0.0017	16.0	1952	0.7505	0.8897	0.8659	0.8695	0.8676
0.0017	17.0	2074	0.7756	0.8897	0.8659	0.8695	0.8676
0.0011	18.0	2196	0.8041	0.8922	0.8673	0.8763	0.8716
0.0017	19.0	2318	0.8064	0.8922	0.8694	0.8712	0.8703
0.0008	20.0	2440	0.8053	0.8922	0.8694	0.8712	0.8703

Framework versions

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