sentiment-seq_bn-rf16-0

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: 0.3129
• Accuracy: 0.8747
• Precision: 0.8479
• Recall: 0.8513
• F1: 0.8496

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
0.5551	1.0	122	0.4975	0.7168	0.6485	0.6271	0.6337
0.4753	2.0	244	0.4682	0.7419	0.7118	0.7474	0.7171
0.412	3.0	366	0.3882	0.8271	0.7994	0.7676	0.7804
0.3498	4.0	488	0.3691	0.8421	0.8098	0.8083	0.8091
0.3361	5.0	610	0.3795	0.8145	0.7789	0.8113	0.7897
0.3081	6.0	732	0.4142	0.7970	0.7665	0.8089	0.7761
0.2918	7.0	854	0.3555	0.8471	0.8130	0.8393	0.8235
0.2739	8.0	976	0.3317	0.8647	0.8439	0.8217	0.8315
0.2586	9.0	1098	0.3596	0.8571	0.8243	0.8464	0.8336
0.2509	10.0	1220	0.3299	0.8622	0.8326	0.8375	0.8349
0.2468	11.0	1342	0.3224	0.8672	0.8393	0.8410	0.8402
0.2372	12.0	1464	0.3294	0.8571	0.8251	0.8389	0.8314
0.2305	13.0	1586	0.3134	0.8697	0.8449	0.8378	0.8412
0.2249	14.0	1708	0.3225	0.8697	0.8399	0.8528	0.8458
0.2193	15.0	1830	0.3188	0.8747	0.8471	0.8538	0.8503
0.2061	16.0	1952	0.3392	0.8521	0.8186	0.8404	0.8278
0.21	17.0	2074	0.3122	0.8797	0.8560	0.8524	0.8541
0.2112	18.0	2196	0.3332	0.8546	0.8216	0.8422	0.8303
0.2002	19.0	2318	0.3121	0.8772	0.8524	0.8506	0.8515
0.2041	20.0	2440	0.3129	0.8747	0.8479	0.8513	0.8496

Framework versions

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