OmniClaw-V2-Q8_0

Local staging package for OmniClaw-V2-Q8_0.gguf from:

LuffyTheFox/OmniClaw-Qwen3.5-9B-Claude-4.6-Opus-Uncensored-v2-GGUF

Recommended runtime settings

• Temperature: 0.7
• Top K Sampling: 20
• Top P Sampling: 0.8
• Min P Sampling: 0
• Presence Penalty: 1.5
• Seed: 3407 or 42
• System prompt: see system_prompt.txt

Files

• OmniClaw-V2-Q8_0.gguf: expected model file, not included until downloaded.
• config.json: copied from the upstream model config.
• system_prompt.txt: system message requested for this model.
• Modelfile: Ollama import file for the GGUF.

Ollama import

After OmniClaw-V2-Q8_0.gguf is present in this directory:

ollama create omniclaw-v2-q8_0 -f Modelfile

Hugging Face upload

Set a target repo and token, then run:

TARGET_REPO=username/OmniClaw-V2-Q8_0 ./upload_to_hf.sh