phishing-binary-classification-bert

This model is a fine-tuned version of google-bert/bert-base-uncased on aisuko/phishing-binary-classification dataset. It achieves the following results on the evaluation set:

• Loss: 0.4878
• Accuracy: 0.82
• Auc: 0.919

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

aisuko/phishing-binary-classification dataset

Training procedure

Please check Kaggle notbebook FT Google Bert for Binary Classification

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Auc
0.6681	1.0	1250	0.6198	0.69	0.885
0.6185	2.0	2500	0.5813	0.712	0.897
0.5907	3.0	3750	0.5478	0.82	0.9
0.5693	4.0	5000	0.5267	0.815	0.908
0.5608	5.0	6250	0.5193	0.787	0.91
0.5486	6.0	7500	0.5168	0.769	0.915
0.5409	7.0	8750	0.5034	0.79	0.916
0.5338	8.0	10000	0.5016	0.784	0.918
0.5331	9.0	11250	0.4947	0.796	0.919
0.5308	10.0	12500	0.4878	0.82	0.919

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0