distilbert-base-multilingual-cased-multilabel-indonesian-hate-speech-new

This model is a fine-tuned version of distilbert/distilbert-base-multilingual-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4308
• F1: 0.7237
• Roc Auc: 0.8286
• Accuracy: 0.6372

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.3684	1.0	659	0.2877	0.6267	0.7611	0.5059
0.2774	2.0	1318	0.2626	0.6599	0.7804	0.5522
0.2315	3.0	1977	0.2808	0.6678	0.7768	0.5942
0.1583	4.0	2636	0.2750	0.7008	0.8130	0.6000
0.1257	5.0	3295	0.3320	0.6867	0.7887	0.6160
0.11	6.0	3954	0.3322	0.7114	0.8170	0.6226
0.0686	7.0	4613	0.3504	0.7123	0.8199	0.6241
0.0549	8.0	5272	0.3812	0.7151	0.8230	0.6276
0.0482	9.0	5931	0.4221	0.7075	0.8093	0.6380
0.0313	10.0	6590	0.4275	0.7130	0.8175	0.6317
0.0279	11.0	7249	0.4308	0.7237	0.8286	0.6372
0.0224	12.0	7908	0.4549	0.7222	0.8256	0.6439
0.0146	13.0	8567	0.4581	0.7235	0.8275	0.6418
0.0143	14.0	9226	0.4685	0.7182	0.8246	0.6363
0.0118	15.0	9885	0.4715	0.7182	0.8237	0.6384

Framework versions

• Transformers 4.51.3
• Pytorch 2.7.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.1