modernbert-grade-classifier

This model is a fine-tuned version of answerdotai/ModernBERT-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0005
• F1: 0.6175

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1
1.9568	1.0	197	1.4912	0.4107
1.3756	2.0	394	1.2853	0.4907
1.1388	3.0	591	1.1807	0.5524
0.9889	4.0	788	1.1218	0.5586
0.8512	5.0	985	1.0889	0.5767
0.7781	6.0	1182	1.0464	0.5901
0.6999	7.0	1379	1.0223	0.5967
0.639	8.0	1576	1.0122	0.6144
0.6315	9.0	1773	1.0074	0.6095
0.585	10.0	1970	1.0041	0.6127
0.5777	11.0	2167	1.0024	0.6121
0.5512	12.0	2364	0.9948	0.6160
0.5521	13.0	2561	1.0011	0.6176
0.5451	14.0	2758	0.9990	0.6147
0.5502	15.0	2955	1.0005	0.6175

Framework versions

• Transformers 4.48.0.dev0
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.21.0