legal-bert-lter-small-finetuned

This model is a fine-tuned version of AnhNam/Luong_modernBert_ft on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4593
• F1: 0.7823

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: 1e-06
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• 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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1
0.3443	1.0	7	0.5216	0.5071
0.434	2.0	14	0.5117	0.5071
1.1626	3.0	21	0.5010	0.5071
1.6782	4.0	28	0.4867	0.7016
0.4249	5.0	35	0.4777	0.7016
0.9495	6.0	42	0.4716	0.7016
0.5363	7.0	49	0.4685	0.7823
1.0944	8.0	56	0.4669	0.7823
1.1263	9.0	63	0.4637	0.7823
0.8751	10.0	70	0.4626	0.7823
0.4535	11.0	77	0.4611	0.7823
0.5308	12.0	84	0.4596	0.7823
0.8087	13.0	91	0.4595	0.7823
0.4981	14.0	98	0.4592	0.7823
0.698	15.0	105	0.4593	0.7823

Framework versions

• Transformers 4.48.2
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0