hotepfederales/hotep-llm-merged

Hotep Intelligence V12 — Legacy Merged Model

⚠️ Legacy: This is the V12 (Kemet line) merged model based on Qwen2.5-7B-Instruct. Production has moved to Kush V3 (Llama 3.1 8B, 100/100 eval score).

Hotep Intelligence LLM — V12 Kemet, fine-tuned on Qwen2.5-7B-Instruct. Archived for reference.

Model Details

Parameter	Value
Base model	Qwen/Qwen2.5-7B-Instruct
Training	568 SFT + 728 DPO
Status	Legacy (superseded by Kush V2 → Kush V3)

Current Production Model

Use hotepfederales/hotep-kush-v3 instead:

• Llama 3.1 8B base (no Qwen dependency)
• 1,594 SFT + 791 IPO pairs
• 100/100 quality score, 0% rubric leakage
• Available as Q8_0 GGUF for Ollama

Usage

# Use current production model instead
ollama run hotepfederales/hotep-llm-kush-v3

Live at: https://askhotep.ai | https://t.me/hotep_llm_bot