Architecture: Qwen 2.5 Coder 14B Transformer Architecture
Merging Technique: FIM Tokenizer Patching
Constituent Models:

• Base Model: [object Object]

Methodology: Applied FIM token fixes and verified GGUF quantizations.

> Key Upgrades Over Base Model:

• High-Speed Autocomplete: Instant inline completion on local machines.
• Uncensored: Generates security, reverse engineering, and script logic without refusal.

// Note: Arena Elo and head-to-head winrates updated continuously as evaluation telemetry processes.

> Recommended Settings

• Temperature: 0.2 - 0.7
• Top-P: 0.95
• Backend Engines: Compatible with llama.cpp, vLLM, Ollama, LM Studio, KoboldCPP

• [+] llama.cpp: Native support across all quantizations.
• [+] Ollama / LM Studio: Full GGUF compatibility.

Disclaimer: Qwen2.5-Coder-14B-Instruct-Uncensored-Patched is provided for research and sovereign local deployment. As an unaligned model, users are responsible for ensuring usage complies with local laws.

Credits: Gratitude to original base model authors (Qwen/Qwen2.5-Coder-14B-Instruct) and open-source AI community tools.