Downloads · 30 days
483
4% of all-time downloads
Andycurrent/DarkIdol-Llama-3.1-8B-Instruct-1.3-Uncensored_GGUF
DarkIdol-Llama-3.1-8B-Instruct-1.3-Uncensored_GGUF is a machine learning model from Andycurrent. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as llama3.
This repository contains the DarkIdol-Llama-3.1-8B-Instruct-1.3-Uncensored model — an instruction-oriented 8B parameter variant designed for users who want a highly responsive, lightly-restricted assistant for offline…
Downloads · 30 days
483
4% of all-time downloads
All-time downloads
13.4K
Public
Repo size
49.1 GB
Likes
8
Public
Click a slice to open those files.
.gguf49.1 GB · 100%
From the Hugging Face model README
This repository contains the DarkIdol-Llama-3.1-8B-Instruct-1.3-Uncensored model — an instruction-oriented 8B parameter variant designed for users who want a highly responsive, lightly-restricted assistant for offline or local workflows.
DarkIdol is part of a series of locally-focused instruction-tuned language models that emphasize: User-directed alignment Minimal artificial guardrails High responsiveness in conversational and reasoning tasks Strong support for multi-step thinking and long interactions
This version (1.3) aims to provide an assistant that feels adaptable and direct, suited for power users, tinkerers, developers, and those running LLMs privately.
The model uses a ChatML-like interaction structure:
<|im_start|>system
{system message}
<|im_end|>
<|im_start|>user
{your prompt here}
<|im_end|>
<|im_start|>assistant
This template provides clean separation between roles and improves controllability during conversation.
Special thanks to: The creators and maintainers of the base Llama-3.1 architecture The open-source ecosystems supporting training, quantization, and deployment Community contributors who provide feedback and testing across diverse hardware setups