---
base_model: SynamicTechnologies/CYBERT
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: Cyber-ThreaD/CyBERT-APTNER
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# Cyber-ThreaD/CyBERT-APTNER

This model is a fine-tuned version of [SynamicTechnologies/CYBERT](https://huggingface.co/SynamicTechnologies/CYBERT) on the [APTNER](https://github.com/wangxuren/APTNER) dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4543
- Precision: 0.4372
- Recall: 0.4293
- F1: 0.4332
- Accuracy: 0.9043

It achieves the following results on the prediction set:
- Loss: 0.3602
- Precision: 0.5489
- Recall: 0.5125
- F1: 0.5301
- Accuracy: 0.9220


## 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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0

### Training results

| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.8182        | 0.59  | 500  | 0.5569          | 0.5219    | 0.2431 | 0.3317 | 0.9014   |
| 0.5357        | 1.19  | 1000 | 0.4980          | 0.4625    | 0.2895 | 0.3561 | 0.9024   |
| 0.4417        | 1.78  | 1500 | 0.4773          | 0.4029    | 0.3572 | 0.3787 | 0.9016   |
| 0.394         | 2.37  | 2000 | 0.4840          | 0.3697    | 0.3943 | 0.3816 | 0.8943   |
| 0.3534        | 2.97  | 2500 | 0.4742          | 0.3586    | 0.4437 | 0.3966 | 0.8914   |
| 0.3048        | 3.56  | 3000 | 0.4543          | 0.4372    | 0.4293 | 0.4332 | 0.9043   |
| 0.2992        | 4.15  | 3500 | 0.4846          | 0.3587    | 0.4392 | 0.3949 | 0.8907   |
| 0.2675        | 4.74  | 4000 | 0.4760          | 0.4100    | 0.4530 | 0.4304 | 0.9000   |
| 0.2454        | 5.34  | 4500 | 0.4702          | 0.4123    | 0.4407 | 0.4260 | 0.9014   |
| 0.2391        | 5.93  | 5000 | 0.4743          | 0.3957    | 0.4638 | 0.4270 | 0.8979   |
| 0.2088        | 6.52  | 5500 | 0.4778          | 0.4224    | 0.4485 | 0.4351 | 0.9038   |
| 0.2076        | 7.12  | 6000 | 0.5050          | 0.3736    | 0.4644 | 0.4140 | 0.8930   |
| 0.1946        | 7.71  | 6500 | 0.4964          | 0.4009    | 0.4599 | 0.4284 | 0.8977   |
| 0.1808        | 8.3   | 7000 | 0.4878          | 0.4226    | 0.4554 | 0.4384 | 0.9028   |
| 0.1683        | 8.9   | 7500 | 0.4947          | 0.3954    | 0.4626 | 0.4264 | 0.8976   |
| 0.1681        | 9.49  | 8000 | 0.4916          | 0.4081    | 0.4662 | 0.4352 | 0.9001   |


### Framework versions

- Transformers 4.36.0.dev0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1


### Citing & Authors

If you use the model kindly cite the following work

```
@inproceedings{deka2024attacker,
  title={AttackER: Towards Enhancing Cyber-Attack Attribution with a Named Entity Recognition Dataset},
  author={Deka, Pritam and Rajapaksha, Sampath and Rani, Ruby and Almutairi, Amirah and Karafili, Erisa},
  booktitle={International Conference on Web Information Systems Engineering},
  pages={255--270},
  year={2024},
  organization={Springer}
}

```