ner-cyber-bert

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

• Loss: 0.1597

• F1: 0.6882

• Classification Report: precision recall f1-score support

  Indicator 0.77 0.81 0.79 270 Malware 0.70 0.79 0.74 238

Organization 0.71 0.50 0.59 133 System 0.56 0.53 0.54 236 Vulnerability 0.89 0.80 0.84 10

micro avg       0.69      0.69      0.69       887
macro avg       0.73      0.69      0.70       887

weighted avg 0.69 0.69 0.68 887

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: 2
• eval_batch_size: 2
• 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Classification Report
0.0714	1.0	1332	0.0983	0.7004	precision recall f1-score support

Indicator       0.80      0.79      0.80       207
  Malware       0.83      0.62      0.71       252

Organization 0.55 0.58 0.57 91 System 0.66 0.64 0.65 179 Vulnerability 0.83 0.56 0.67 9

micro avg       0.74      0.67      0.70       738
macro avg       0.74      0.64      0.68       738

weighted avg 0.75 0.67 0.70 738 | | 0.0829 | 2.0 | 2664 | 0.1173 | 0.7266 | precision recall f1-score support

Indicator       0.78      0.75      0.77       207
  Malware       0.89      0.68      0.77       252

Organization 0.67 0.43 0.52 91 System 0.68 0.73 0.70 179 Vulnerability 0.88 0.78 0.82 9

micro avg       0.77      0.68      0.73       738
macro avg       0.78      0.67      0.72       738

weighted avg 0.78 0.68 0.72 738 | | 0.0295 | 3.0 | 3996 | 0.1451 | 0.7130 | precision recall f1-score support

Indicator       0.72      0.77      0.75       207
  Malware       0.89      0.62      0.73       252

Organization 0.69 0.45 0.55 91 System 0.70 0.75 0.73 179 Vulnerability 0.64 0.78 0.70 9

micro avg       0.76      0.67      0.71       738
macro avg       0.73      0.67      0.69       738

weighted avg 0.77 0.67 0.71 738 | | 0.0228 | 4.0 | 5328 | 0.1244 | 0.7087 | precision recall f1-score support

Indicator       0.71      0.87      0.78       207
  Malware       0.91      0.62      0.74       252

Organization 0.47 0.62 0.53 91 System 0.68 0.69 0.69 179 Vulnerability 0.86 0.67 0.75 9

micro avg       0.71      0.71      0.71       738
macro avg       0.72      0.69      0.70       738

weighted avg 0.74 0.71 0.71 738 | | 0.0309 | 5.0 | 6660 | 0.1340 | 0.7458 | precision recall f1-score support

Indicator       0.76      0.88      0.82       207
  Malware       0.83      0.73      0.78       252

Organization 0.64 0.54 0.58 91 System 0.68 0.71 0.69 179 Vulnerability 0.70 0.78 0.74 9

micro avg       0.75      0.75      0.75       738
macro avg       0.72      0.73      0.72       738

weighted avg 0.75 0.75 0.74 738 | | 0.0068 | 6.0 | 7992 | 0.1710 | 0.7143 | precision recall f1-score support

Indicator       0.73      0.86      0.79       207
  Malware       0.92      0.60      0.73       252

Organization 0.57 0.51 0.53 91 System 0.66 0.71 0.68 179 Vulnerability 0.88 0.78 0.82 9

micro avg       0.74      0.69      0.71       738
macro avg       0.75      0.69      0.71       738

weighted avg 0.76 0.69 0.71 738 | | 0.0033 | 7.0 | 9324 | 0.1669 | 0.7265 | precision recall f1-score support

Indicator       0.70      0.84      0.76       207
  Malware       0.84      0.73      0.78       252

Organization 0.55 0.56 0.55 91 System 0.68 0.73 0.71 179 Vulnerability 0.58 0.78 0.67 9

micro avg       0.71      0.74      0.73       738
macro avg       0.67      0.73      0.69       738

weighted avg 0.72 0.74 0.73 738 | | 0.0003 | 8.0 | 10656 | 0.1820 | 0.7214 | precision recall f1-score support

Indicator       0.65      0.86      0.74       207
  Malware       0.88      0.67      0.76       252

Organization 0.64 0.52 0.57 91 System 0.67 0.78 0.72 179 Vulnerability 0.64 0.78 0.70 9

micro avg       0.71      0.73      0.72       738
macro avg       0.69      0.72      0.70       738

weighted avg 0.73 0.73 0.72 738 | | 0.0001 | 9.0 | 11988 | 0.1766 | 0.7270 | precision recall f1-score support

Indicator       0.73      0.83      0.77       207
  Malware       0.85      0.71      0.78       252

Organization 0.54 0.55 0.55 91 System 0.69 0.72 0.70 179 Vulnerability 0.54 0.78 0.64 9

micro avg       0.73      0.73      0.73       738
macro avg       0.67      0.72      0.69       738

weighted avg 0.73 0.73 0.73 738 | | 0.0018 | 10.0 | 13320 | 0.1781 | 0.7251 | precision recall f1-score support

Indicator       0.71      0.86      0.78       207
  Malware       0.85      0.69      0.76       252

Organization 0.54 0.55 0.54 91 System 0.67 0.74 0.70 179 Vulnerability 0.58 0.78 0.67 9

micro avg       0.72      0.73      0.73       738
macro avg       0.67      0.72      0.69       738

weighted avg 0.73 0.73 0.73 738 |

Framework versions

• Transformers 4.52.2
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1