🩺 Qwen3.6-35B Nightshift Heretic Uncensored Medic

👁️ Medical Vision AI for Emergency Triage, Radiology & Diagnostic Imaging

A specialized Qwen3.6-35B A3B MoE build for Ollama, designed for local multimodal AI workflows involving medical imaging, radiology, emergency triage, laboratory reports and clinical research.

Built as part of the 16 HEX Matrix / Eastern IT School model series, with a focus on direct local inference and reduced refusal behavior.

📌 MODEL OVERVIEW

• Base Model: Qwen3.6-35B-A3B-MoE
• Architecture: Mixture-of-Experts (MoE)
• Total Parameters: ~35B
• Active Parameters: ~3B per token
• Experts: 256
• Format: GGUF / Ollama
• Primary Modality: Text + Vision
• Use Case: Local medical AI, diagnostic image analysis, radiology and clinical research

Source checkpoint: Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive

👁️ NATIVE VISION & MEDICAL IMAGING

The Medic build is focused on multimodal workflows involving:

• X-ray and radiographic images
• CT and MRI imagery
• Ultrasound / USG
• Trauma and clinical photographs
• Laboratory reports and medical documentation
• Multi-image and multi-document analysis

Typical workflows include identifying visible abnormalities, organizing observations, generating diagnostic hypotheses and structuring information for further human review.

Important: This model is an experimental AI system for research and decision-support workflows. It is not a certified medical device and does not replace a qualified healthcare professional or clinical diagnosis.

🩺 EMERGENCY TRIAGE & DIAGNOSTIC WORKFLOWS

The model can be configured for structured clinical reasoning workflows such as:

1. Observations & Artifacts Identify visible findings, anomalies and potential image artifacts.

2. Diagnostic Hypotheses Organize possible pathological findings based on the available information.

3. Triage Classification Structure cases into practical priority levels such as STAT / Priority / Observation.

This workflow is intended for research, prototyping and local clinical-AI experimentation, not autonomous medical decision-making.

🔬 MEDICAL RESEARCH

Suitable for local research workflows involving:

• Radiology image analysis
• Medical image understanding
• Laboratory data interpretation
• Clinical document analysis
• Patient-history summarization
• Multimodal medical AI research
• Local diagnostic-agent prototypes
• Medical vision model evaluation

Running inference locally can help organizations and researchers maintain direct control over the data-processing environment.

👨‍💻 WHO IS IT FOR?

Clinicians & Medical Researchers

For experimental second-opinion and clinical-AI research workflows, especially where multimodal image and document analysis is required.

AI Engineers

For building local medical vision agents, triage pipelines and multimodal AI applications with Ollama.

Local AI Developers

For testing Qwen-based medical AI workflows without relying on a hosted inference API.

⚙️ 16 HEX ENGINE CONFIGURATION

The 16 HEX Matrix is the configuration and branding layer used by this model series.

Recommended sampling configuration:

Temperature: 0.1
Top-K:       16
Top-P:       0.3
Context:     32768

These settings favor a more constrained and deterministic generation profile. Actual output quality and reliability depend on the model, prompt, image quality, context and inference backend.

💻 HARDWARE & LOCAL INFERENCE

Hardware requirements depend on quantization, context length, KV cache, GPU offloading and inference configuration.

For practical local deployment, systems with approximately 24 GB or more of VRAM/unified memory are a useful starting point, while lower-memory systems may require partial CPU offloading or reduced context.

Designed for local execution through Ollama and compatible GGUF-based inference environments.

🔐 LOCAL AI & DATA CONTROL

Local inference keeps model processing within your own environment rather than requiring a hosted AI API.

This can be useful for research environments handling sensitive material, but local execution alone does not constitute HIPAA or GDPR compliance. Compliance depends on the complete technical, organizational and legal environment in which the system is deployed.

🚀 OLLAMA

Run the model locally with:

ollama run jikepjikep_16HEX/qwen3.6-35b-nightshift-heretic-uncensored-medic-a3b-moe-q4

🧬 MODEL FOCUS

Qwen3.6-35B → A3B MoE → Native Vision → Medical Imaging → Radiology → Emergency Triage → Laboratory Diagnostics → Local Clinical AI

Part of the 16 HEX / Eastern IT School Nightshift Heretic model series.

#MedicalAI #MedicalVision #Radiology #MedicalImaging #EmergencyTriage #ClinicalAI #Qwen35B #Qwen3 #MoE #VisionAI #Ollama #GGUF #LocalAI #MultimodalAI #16HEX