indobert-click-judges

This model is a fine-tuned version of indobenchmark/indobert-base-p1 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6266

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: 0.003
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
0.5924	1.0	57	0.8505
0.4860	2.0	114	0.8379
0.4696	3.0	171	1.1534
0.4481	4.0	228	0.8031
0.4778	5.0	285	0.7002
0.4221	6.0	342	0.7191
0.4045	7.0	399	0.7007
0.4360	8.0	456	0.8092
0.4138	9.0	513	0.6848
0.3627	10.0	570	0.7568
0.3379	11.0	627	0.6879
0.3474	12.0	684	0.7341
0.3369	13.0	741	0.6409
0.2920	14.0	798	0.6682
0.1993	15.0	855	0.6469
0.1975	16.0	912	0.6547
0.1987	17.0	969	0.6035
0.1596	18.0	1026	0.6234
0.1643	19.0	1083	0.6223
0.1399	20.0	1140	0.6266

Framework versions

• Transformers 5.14.0
• Pytorch 2.12.1+cu126
• Datasets 5.0.0
• Tokenizers 0.22.2