sentiment-lora-r4a2d0.05-1

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: 0.3356
• Accuracy: 0.8622
• Precision: 0.8399
• Recall: 0.8200
• F1: 0.8289

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
0.5657	1.0	122	0.5182	0.7243	0.6604	0.6424	0.6488
0.5109	2.0	244	0.5051	0.7243	0.6748	0.6874	0.6796
0.48	3.0	366	0.4643	0.7569	0.7047	0.6880	0.6948
0.434	4.0	488	0.4281	0.7920	0.7497	0.7378	0.7431
0.4106	5.0	610	0.4194	0.7920	0.7528	0.7778	0.7618
0.3812	6.0	732	0.3936	0.8296	0.8008	0.7744	0.7854
0.3689	7.0	854	0.3700	0.8521	0.8220	0.8204	0.8212
0.3489	8.0	976	0.3656	0.8346	0.8088	0.7780	0.7905
0.3502	9.0	1098	0.3640	0.8371	0.8101	0.7847	0.7955
0.3349	10.0	1220	0.3608	0.8346	0.8074	0.7805	0.7917
0.3189	11.0	1342	0.3574	0.8396	0.8128	0.7890	0.7992
0.3121	12.0	1464	0.3547	0.8471	0.8175	0.8093	0.8132
0.3181	13.0	1586	0.3478	0.8521	0.8332	0.7979	0.8122
0.3092	14.0	1708	0.3435	0.8596	0.8374	0.8157	0.8253
0.3018	15.0	1830	0.3466	0.8546	0.8296	0.8121	0.8200
0.2955	16.0	1952	0.3365	0.8596	0.8347	0.8207	0.8272
0.2917	17.0	2074	0.3353	0.8596	0.8374	0.8157	0.8253
0.2956	18.0	2196	0.3379	0.8596	0.8360	0.8182	0.8262
0.2899	19.0	2318	0.3353	0.8647	0.8455	0.8192	0.8306
0.2885	20.0	2440	0.3356	0.8622	0.8399	0.8200	0.8289

Framework versions

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