bert_model_out

This model is a fine-tuned version of beomi/kcbert-base on the unsmile_data dataset. It achieves the following results on the evaluation set:

• Loss: 0.1441
• Irap: 0.8751

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: 64
• eval_batch_size: 64
• 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: 7

Training results

Training Loss	Epoch	Step	Validation Loss	Irap
No log	1.0	235	0.1474	0.8568
No log	2.0	470	0.1315	0.8660
0.1689	3.0	705	0.1232	0.8803
0.1689	4.0	940	0.1313	0.8811
0.0728	5.0	1175	0.1362	0.8760
0.0728	6.0	1410	0.1423	0.8761
0.0403	7.0	1645	0.1441	0.8751

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.0
• Tokenizers 0.21.0