MITRE V15 Tactic Bert Case Based

🚀 MITRE-v15-tactic-bert-case-based

This is a fine-tuned model derived from mitre-bert-base-cased on the MITRE ATT&CK version 15 procedure dataset. On the evaluation dataset, it achieves the following performance metrics:

• Loss: 0.057
• Accuracy: 0.87

✨ Features

• Text Classification: This fine-tuned model can be used for text classification, aiming to identify the tactic that a sentence belongs to within the MITRE ATT&CK framework. A single sentence or attack may fall into multiple tactics.
• Cybersecurity Focus: Primarily fine-tuned for text classification in the field of cybersecurity.

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

Basic Usage

You can use the model with Tensorflow as follows:

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
model_id = "sarahwei/MITRE-tactic-bert-case-based"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForSequenceClassification.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    # device_map="auto",
)
question = 'An attacker performs a SQL injection.'
input_ids = tokenizer(question,return_tensors="pt")
outputs = model(**input_ids)
logits = outputs.logits
sigmoid = torch.nn.Sigmoid()
probs = sigmoid(logits.squeeze().cpu())
predictions = np.zeros(probs.shape)
predictions[np.where(probs >= 0.5)] = 1
predicted_labels = [model.config.id2label[idx] for idx, label in enumerate(predictions) if label == 1.0]

Advanced Usage

There is no advanced usage example in the original document.

📚 Documentation

Intended uses & limitations

You can utilize the fine-tuned model for text classification, specifically to identify the tactic that a sentence belongs to in the MITRE ATT&CK framework. Note that a sentence or an attack might be associated with multiple tactics.

It's important to note that this model is mainly fine-tuned for text classification in the context of cybersecurity. It may not yield satisfactory results if the input sentence is not related to attacks.

Training procedure

Training parameter

Training results

📄 License

This model is licensed under the Apache 2.0 license.