en-phishing-email-classifier

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6750

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss
0.3976	1.0	137	0.3507
0.2556	2.0	274	0.3787
0.1948	3.0	411	0.4024
0.1544	4.0	548	0.3626
0.1267	5.0	685	0.3511
0.0988	6.0	822	0.3824
0.08	7.0	959	0.4497
0.0657	8.0	1096	0.4339
0.0573	9.0	1233	0.5017
0.0472	10.0	1370	0.5168
0.0467	11.0	1507	0.5324
0.0416	12.0	1644	0.6004
0.0386	13.0	1781	0.5630
0.0375	14.0	1918	0.5628
0.0307	15.0	2055	0.5649
0.0284	16.0	2192	0.6346
0.0274	17.0	2329	0.6234
0.0261	18.0	2466	0.6284
0.0238	19.0	2603	0.6398
0.0239	20.0	2740	0.6649
0.0227	21.0	2877	0.6750

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1