nico248000000000/Qwen3.8-27B-cyber

VISIONConcurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 14, 2026License:otherArchitecture:Transformer0.0K Featherless Exclusive Cold

nico248000000000/Qwen3.8-27B-cyber is an instruction-tuned 27 billion parameter language model based on Qwen/Qwen3.8-27B, specifically fine-tuned for cybersecurity applications. It excels at providing concrete answers for offensive, defensive, GRC, architecture, SOC/DFIR, and RSSI procedures. This model is designed to assist with authorized defensive work, tabletop exercises, and control design within a 2048-token context window.

Loading preview...

Overview

nico248000000000/Qwen3.8-27B-cyber is a specialized instruction-tuned language model, building upon the Qwen/Qwen3.8-27B base model. It has been fine-tuned using LoRA/QLoRA (rank 8, alpha 16) on a curated dataset of cybersecurity procedures, including pentesting, red teaming, SOC/DFIR operations, cloud and identity security, and GRC (ISO, NIST, NIS2, DORA) frameworks. The model maintains the multimodal capabilities of its base, supporting vision and video inputs, though these layers were frozen during text SFT.

Key Capabilities

  • Cybersecurity Specialization: Provides concrete, actionable answers across various cybersecurity domains, including offensive security (without weaponized payloads), defensive strategies, governance, risk, and compliance (GRC), security architecture, and Security Operations Center (SOC)/Digital Forensics and Incident Response (DFIR).
  • Instruction-Tuned: Optimized to follow instructions for tasks like control design, detection engineering, and audit readiness.
  • Performance: Achieved a significant reduction in training loss (-93.5%) and evaluation loss (-24.2%) compared to the base model, indicating effective specialization.
  • Multimodal Support: Retains the ability to process vision and video modalities, which can be leveraged by loading the merged Transformers folder or GGUF + mmproj.

Good For

  • Authorized defensive cybersecurity work.
  • Tabletop exercises and incident response simulations.
  • Designing security controls and detection mechanisms.
  • Explaining attack techniques and security concepts in a practical manner.
  • Assisting with audit readiness and compliance questions related to standards like ISO, NIST, NIS2, and DORA.