modernbert-ct4a-11-no-aug

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

• Loss: 0.4779
• Accuracy: 0.9075
• F1: 0.7601
• Auc: 0.8332
• Accuracy Per Label: [0.9051094890510949, 0.9197080291970803, 0.8978102189781022]
• F1 Per Label: [0.7547169811320755, 0.7441860465116279, 0.78125]
• Auc Per Label: [0.853083853083853, 0.8031878031878031, 0.8433752141633354]

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: 8
• eval_batch_size: 8
• 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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Auc	Accuracy Per Label	F1 Per Label	Auc Per Label
No log	1.0	154	0.2880	0.8686	0.5709	0.7054	[0.8759124087591241, 0.8686131386861314, 0.8613138686131386]	[0.5853658536585366, 0.5, 0.6274509803921569]	[0.7172557172557175, 0.6685724185724186, 0.73043974871502]
No log	2.0	308	0.2495	0.8954	0.6899	0.7750	[0.8686131386861314, 0.9197080291970803, 0.8978102189781022]	[0.5714285714285714, 0.7317073170731707, 0.7666666666666667]	[0.7127512127512129, 0.7884615384615384, 0.8236721873215306]
No log	3.0	462	0.3805	0.9100	0.7742	0.8496	[0.9051094890510949, 0.9197080291970803, 0.9051094890510949]	[0.7450980392156863, 0.7659574468085106, 0.8115942028985508]	[0.8383575883575884, 0.8326403326403327, 0.8777841233580811]
0.1833	4.0	616	0.4659	0.9051	0.7504	0.8234	[0.9051094890510949, 0.9197080291970803, 0.8905109489051095]	[0.7450980392156863, 0.7441860465116279, 0.7619047619047619]	[0.8383575883575884, 0.8031878031878031, 0.8286693318103941]
0.1833	5.0	770	0.4779	0.9075	0.7601	0.8332	[0.9051094890510949, 0.9197080291970803, 0.8978102189781022]	[0.7547169811320755, 0.7441860465116279, 0.78125]	[0.853083853083853, 0.8031878031878031, 0.8433752141633354]

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.1
• Tokenizers 0.21.1