sainikhiljuluri2015/Foundation-Sec-Cybersecurity-8B-Merged
sainikhiljuluri2015/Foundation-Sec-Cybersecurity-8B-Merged is an 8 billion parameter language model fine-tuned by sainikhiljuluri2015, based on fdtn-ai/Foundation-Sec-8B. This model is specifically optimized for a wide range of cybersecurity tasks, including threat analysis, incident response, and vulnerability assessment. It was trained on approximately 50,000 cybersecurity instruction-response pairs, making it highly specialized for security-related applications.
Loading preview...
Overview
This model, sainikhiljuluri2015/Foundation-Sec-Cybersecurity-8B-Merged, is an 8 billion parameter language model that has been fine-tuned specifically for cybersecurity applications. It is based on the fdtn-ai/Foundation-Sec-8B model, with LoRA weights merged into the base for streamlined deployment. The training involved approximately 50,000 instruction-response pairs sourced from specialized cybersecurity datasets, including the Trendyol Cybersecurity Dataset, Fenrir v2.0 Dataset, and Primus-Instruct.
Key Capabilities
This model excels in various cybersecurity functions, offering specialized support for:
- Threat analysis and classification
- Security alert triage
- Incident response guidance
- Malware analysis
- MITRE ATT&CK mapping
- Vulnerability assessment
- SQL injection detection
- Phishing analysis
- CVE knowledge
- Security best practices
Training Details
The model underwent 2 epochs of training with a learning rate of 2e-4 and a maximum sequence length of 1024. LoRA was applied with a rank of 16 and an alpha of 32, targeting 7 key modules (q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj).
Good For
This model is ideal for developers and security professionals who require a specialized language model for automating or assisting with cybersecurity-specific tasks. Its fine-tuning on relevant datasets makes it particularly effective for generating insights, classifying threats, and providing guidance within the cybersecurity domain.