legal-bert-Federal-Regulations

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

• Loss: 0.6277
• Accuracy: 0.7154
• Precision: 0.7483
• Recall: 0.7154
• F1: 0.7248
• Roc Auc: 0.7853
• Confusion Matrix: [[2451, 934], [445, 1016]]

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 16
• 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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Roc Auc	Confusion Matrix
0.606	1.0	600	0.6376	0.6605	0.7333	0.6605	0.6745	0.7564	[[2110, 1275], [370, 1091]]
0.5312	2.0	1200	0.5504	0.7418	0.7473	0.7418	0.7442	0.7829	[[2708, 677], [574, 887]]
0.4563	3.0	1800	0.6277	0.7154	0.7483	0.7154	0.7248	0.7853	[[2451, 934], [445, 1016]]

Framework versions

• Transformers 4.46.1
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.1