koElectra_coupang_intent_v2

This model is a fine-tuned version of monologg/koElectra-base-v3-discriminator on the custom-intent-dataset dataset. It achieves the following results on the evaluation set:

• Loss: 0.9741
• Accuracy: 0.9994
• F1: 0.9994

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: 256
• eval_batch_size: 256
• 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
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 15
• mixed_precision_training: Native AMP
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0	0	3.3323	0.0351	0.0063
3.2686	1.8519	50	3.2054	0.2292	0.1957
2.7920	3.7037	100	2.6143	0.8146	0.7966
2.2238	5.5556	150	2.0090	0.9357	0.9208
1.7884	7.4074	200	1.5677	0.9789	0.9771
1.4640	9.2593	250	1.2716	0.9971	0.9971
1.2745	11.1111	300	1.0944	0.9988	0.9988
1.1624	12.9630	350	1.0038	0.9994	0.9994
1.1179	14.8148	400	0.9741	0.9994	0.9994

Framework versions

• Transformers 5.1.0
• Pytorch 2.9.1+cu128
• Datasets 4.5.0
• Tokenizers 0.22.2