bert-base-uncased-Federal-Regulations

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: 0.6193
• Accuracy: 0.7332
• Precision: 0.7510
• Recall: 0.7332
• F1: 0.7394
• Roc Auc: 0.7821
• Confusion Matrix: [[2590, 795], [498, 963]]

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: 4e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use 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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Roc Auc	Confusion Matrix
0.5846	1.0	600	0.6114	0.6620	0.7393	0.6620	0.6759	0.7668	[[2092, 1293], [345, 1116]]
0.5123	2.0	1200	0.5976	0.7210	0.7535	0.7210	0.7301	0.7848	[[2465, 920], [432, 1029]]
0.4449	3.0	1800	0.6193	0.7332	0.7510	0.7332	0.7394	0.7821	[[2590, 795], [498, 963]]

Framework versions

• Transformers 4.46.3
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3