You can deploy this model and start earning money today!

Thox-ai/
ThoxHeretic-27B

Name: ThoxHeretic-27B
Availability: InStock

Views

No views yet

ThoxHeretic-27B

Your AI. Your Data. Your Rules.

Multimodal GGUF redistribution of a DavidAU Qwen3.6-27B abliterated 'Heretic' tune. Not THOX-trained.

What this is

Redistribution of DavidAU/Qwen3.6-27B-Fable-Fusion-711-…-Heretic-…-GGUF; SHA-256 byte-identical to upstream.
Multimodal — needs bundled mmproj-f16.gguf; ctx 256K / 32K default.
Base 'Qwen3.6-27B' is self-flagged unverified in the source card — treat as unconfirmed.
Supersedes the interim ThoxMythos-27B-Unleashed scaffold. Uncensored — keep gated.

Intended use

Redistributed third-party weights for research and internal evaluation. Uncensored/abliterated behavior — deploy behind THOX guardrails only.

Prompt / chat template

text

<|im_start|>system
{system}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Usage

llama.cpp

bash

huggingface-cli download Thox-ai/ThoxHeretic-27B --include '*.gguf' --local-dir ./ThoxHeretic-27B
llama-cli -m ./ThoxHeretic-27B/model-Q4_K_M.gguf -p "Hello"

Multimodal — pass the bundled mmproj-*.gguf with --mmproj for image input.

Thox-ai/ThoxHeretic-27B