roberta-base-indonesian-522M-with-haryo-dataset-30

This model is a fine-tuned version of cahya/roberta-base-indonesian-522M on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1761

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss
0.3933	1.0	31	0.4178
0.229	2.0	62	0.1942
0.17	3.0	93	0.2521
0.1559	4.0	124	0.1898
0.1404	5.0	155	0.2164
0.1138	6.0	186	0.1761
0.0865	7.0	217	0.1983
0.0808	8.0	248	0.2002
0.0567	9.0	279	0.3364
0.0724	10.0	310	0.3041
0.0613	11.0	341	0.2723
0.0516	12.0	372	0.2526
0.0436	13.0	403	0.2437
0.0359	14.0	434	0.3047
0.0309	15.0	465	0.2784
0.0275	16.0	496	0.3484
0.0231	17.0	527	0.3357
0.0225	18.0	558	0.3214
0.0182	19.0	589	0.3677
0.0278	20.0	620	0.3681
0.0167	21.0	651	0.3603
0.0165	22.0	682	0.3550
0.0135	23.0	713	0.3776
0.0118	24.0	744	0.4496
0.0145	25.0	775	0.3838
0.012	26.0	806	0.3790
0.0115	27.0	837	0.3788
0.0109	28.0	868	0.3906
0.0114	29.0	899	0.4072
0.0112	30.0	930	0.3991

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1