malware_data.py

import openai
import json
import os

openai.api_key = os.getenv('LLAMA_API_KEY')

def generate_synthetic_data_via_api(num_samples=100):
    synthetic_data = []

    for _ in range(num_samples):
        prompt = (
            "Generate a synthetic malware event with the following details in this format:\n"
            "1. Event ID: <a unique identifier>\n"
            "2. Timestamp: <hour between 0 and 23>\n"
            "3. Severity: <Low, Medium, High>\n"
            "4. Description: <Brief description of the event>\n"
            "5. Is the event unusual?: <yes or no>"
        )

        try:
            response = openai.Completion.create(
                engine="text-davinci-003",
                prompt=prompt,
                max_tokens=100,
                temperature=0.7
            )

            response_text = response.choices[0].text.strip()

            lines = response_text.split("\n")
            if len(lines) < 5:
                print(f"Incomplete response, skipping: {response_text}")
                continue

            event_id = lines[0].split(":")[1].strip()
            timestamp = int(lines[1].split(":")[1].strip())
            severity = lines[2].split(":")[1].strip()
            description = lines[3].split(":")[1].strip()
            is_unusual = lines[4].split(":")[1].strip()

            label = 1 if is_unusual.lower() == "yes" else 0

            synthetic_data.append({
                'event_id': event_id,
                'timestamp': timestamp,
                'severity': severity,
                'description': description,
                'label': label
            })

        except Exception as e:
            print(f"Error processing response: {str(e)}")

    with open('malware_data.json', 'w') as f:
        json.dump(synthetic_data, f, indent=4)

    print(f"{num_samples} samples generated and saved to malware_data.json")

if __name__ == "__main__":
    generate_synthetic_data_via_api(num_samples=100)