sentiment-unipelt-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: 0.3423
• Accuracy: 0.8922
• Precision: 0.8749
• Recall: 0.8612
• F1: 0.8676

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
0.5459	1.0	122	0.5022	0.7243	0.6616	0.6474	0.6529
0.4447	2.0	244	0.4753	0.7694	0.7557	0.8069	0.7549
0.3589	3.0	366	0.3433	0.8571	0.8440	0.7989	0.8164
0.3034	4.0	488	0.3063	0.8697	0.8520	0.8253	0.8368
0.2691	5.0	610	0.3038	0.8622	0.8342	0.8325	0.8333
0.2555	6.0	732	0.3117	0.8521	0.8186	0.8454	0.8293
0.2429	7.0	854	0.3002	0.8546	0.8217	0.8397	0.8296
0.223	8.0	976	0.2853	0.8822	0.8596	0.8542	0.8568
0.2089	9.0	1098	0.2856	0.8672	0.8436	0.8310	0.8369
0.1959	10.0	1220	0.2882	0.8847	0.8679	0.8484	0.8573
0.1804	11.0	1342	0.2876	0.8797	0.8560	0.8524	0.8541
0.1659	12.0	1464	0.3127	0.8697	0.8503	0.8278	0.8377
0.1591	13.0	1586	0.3219	0.8747	0.8568	0.8338	0.8440
0.1585	14.0	1708	0.3209	0.8947	0.8690	0.8830	0.8755
0.1512	15.0	1830	0.3213	0.8847	0.8663	0.8509	0.8580
0.14	16.0	1952	0.3231	0.8897	0.8649	0.8720	0.8683
0.1255	17.0	2074	0.3452	0.8897	0.8743	0.8545	0.8635
0.1296	18.0	2196	0.3402	0.8922	0.8749	0.8612	0.8676
0.1254	19.0	2318	0.3436	0.8922	0.8749	0.8612	0.8676
0.128	20.0	2440	0.3423	0.8922	0.8749	0.8612	0.8676

Framework versions

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