Modelpublic

johannawawi/v3_balanced_dataset_fine-tuning-java-indo-sentiment-analysist-3-class

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes17downloads

Model Card

results_final

This model is a fine-tuned version of w11wo/indonesian-roberta-base-sentiment-classifier on the None dataset. It achieves the following results on the evaluation set:

—Loss: 1.4330
—Accuracy: 0.9
—F1 Macro: 0.9002
—F1 Weighted: 0.9003
—Precision Macro: 0.9019
—Recall Macro: 0.8999
—Precision Weighted: 0.9020
—Recall Weighted: 0.9

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: 5.827811728751637e-05
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: cosine
—lrschedulerwarmup_ratio: 0.1
—num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	F1 Weighted	Precision Macro	Recall Macro	Precision Weighted	Recall Weighted
0.0238	1.8182	500	1.4330	0.9	0.9002	0.9003	0.9019	0.8999	0.9020	0.9
0.002	3.6364	1000	1.0771	0.8891	0.8887	0.8888	0.8897	0.8889	0.8897	0.8891
0.0123	5.4545	1500	1.1399	0.8909	0.8908	0.8908	0.8913	0.8908	0.8913	0.8909
0.104	7.2727	2000	1.0762	0.8855	0.8852	0.8853	0.8859	0.8853	0.8859	0.8855

Framework versions

—Transformers 4.51.3
—Pytorch 2.6.0+cu124
—Datasets 2.14.4
—Tokenizers 0.21.1