Modelpublic

bob-bob-bob-3/classification-hate-speech-DE-2

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face

0likes4downloads

Model Card

classification-hate-speech-DE-2

This model is a fine-tuned version of indolem/indobertweet-base-uncased on the None dataset. It achieves the following results on the evaluation set:

—Loss: 3.0522
—F1 macro: 0.3925
—Weighted: 0.5885
—Balanced accuracy: 0.5201

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
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 14

Training results

Training Loss	Epoch	Step	Validation Loss	F1 macro	Weighted	Balanced accuracy
1.3119	1.0	152	1.0717	0.4111	0.6682	0.4701
0.717	2.0	304	1.2404	0.3989	0.6063	0.5128
0.4124	3.0	456	1.5531	0.4029	0.6116	0.5286
0.1543	4.0	608	2.4557	0.3507	0.5427	0.5053
0.0445	5.0	760	2.6602	0.3950	0.5707	0.5254
0.1482	6.0	912	2.8551	0.3990	0.5643	0.5533
0.0016	7.0	1064	2.6333	0.4016	0.6096	0.5128
0.0012	8.0	1216	2.7488	0.3970	0.6019	0.5206
0.0008	9.0	1368	2.8494	0.3989	0.6025	0.5203
0.0005	10.0	1520	3.0943	0.3886	0.5799	0.5228
0.0005	11.0	1672	3.0410	0.3896	0.5872	0.5132
0.0006	12.0	1824	3.0730	0.4022	0.5912	0.5379
0.0004	13.0	1976	3.0387	0.3928	0.5910	0.5206
0.0005	14.0	2128	3.0522	0.3925	0.5885	0.5201

Framework versions

—Transformers 4.41.2
—Pytorch 2.3.0+cu121
—Datasets 2.20.0
—Tokenizers 0.19.1