Instructions to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF", dtype="auto")

llama-cpp-python

How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF",
	filename="Llama-3.1-8B-Lexi-Uncensored-V2-Heretic_IQ3_M.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M

Use Docker

docker model run hf.co/ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M

SGLang

How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Ollama
How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with Ollama:
```
ollama run hf.co/ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M
```

Unsloth Studio

How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF to start chatting

Atomic Chat new
Docker Model Runner
How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with Docker Model Runner:
```
docker model run hf.co/ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M
```

Lemonade

How to use ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF-Q4_K_M

List all available models

lemonade list

Llama-3.1-8B-Lexi-Uncensored-V2-Heretic-GGUF

A decensored version of Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2, made using Heretic v1.0.1

	Llama-3.1-8B-Lexi-Uncensored-V2-Heretic	Original model (Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2)
Refusals	2/100	35/100
KL divergence	0.02	0 (by definition)

Heretic Abliteration Parameters

Parameter	Value
direction_index	per layer
attn.o_proj.max_weight	0.95
attn.o_proj.max_weight_position	19.12
attn.o_proj.min_weight	0.93
attn.o_proj.min_weight_distance	9.83
mlp.down_proj.max_weight	0.98
mlp.down_proj.max_weight_position	23.23
mlp.down_proj.min_weight	0.10
mlp.down_proj.min_weight_distance	5.19

Safetensors Version

Safetensors version available at ChiKoi7/Llama-3.1-8B-Lexi-Uncensored-V2-Heretic

VERSION 2 Update Notes:

More compliant
Smarter
For best response, use this system prompt (feel free to expand upon it as you wish):

Think step by step with a logical reasoning and intellectual sense before you provide any response.

For more uncensored and compliant response, you can expand the system message differently, or simply enter a dot "." as system message.
IMPORTANT: Upon further investigation, the Q4 seems to have refusal issues sometimes. There seems to be some of the fine-tune loss happening due to the quantization. I will look into it for V3. Until then, I suggest you run F16 or Q8 if possible.

GENERAL INFO:

This model is based on Llama-3.1-8b-Instruct, and is governed by META LLAMA 3.1 COMMUNITY LICENSE AGREEMENT

Lexi is uncensored, which makes the model compliant. You are advised to implement your own alignment layer before exposing the model as a service. It will be highly compliant with any requests, even unethical ones.

You are responsible for any content you create using this model. Please use it responsibly.

Lexi is licensed according to Meta's Llama license. I grant permission for any use, including commercial, that falls within accordance with Meta's Llama-3.1 license.

IMPORTANT:

Use the same template as the official Llama 3.1 8B instruct. System tokens must be present during inference, even if you set an empty system message. If you are unsure, just add a short system message as you wish.

FEEDBACK:

If you find any issues or have suggestions for improvements, feel free to leave a review and I will look into it for upcoming improvements and next version.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	27.93
IFEval (0-Shot)	77.92
BBH (3-Shot)	29.69
MATH Lvl 5 (4-Shot)	16.92
GPQA (0-shot)	4.36
MuSR (0-shot)	7.77
MMLU-PRO (5-shot)	30.90

Downloads last month: 188

GGUF

Model size

8B params

Architecture

llama

Hardware compatibility

3-bit

4-bit

5-bit

6-bit

8-bit

16-bit

Evaluation results

strict accuracy on IFEval (0-Shot)
Open LLM Leaderboard

77.920
normalized accuracy on BBH (3-Shot)
Open LLM Leaderboard

29.690
exact match on MATH Lvl 5 (4-Shot)
Open LLM Leaderboard

16.920
acc_norm on GPQA (0-shot)
Open LLM Leaderboard

4.360
acc_norm on MuSR (0-shot)
Open LLM Leaderboard

7.770
accuracy on MMLU-PRO (5-shot)
test set Open LLM Leaderboard

30.900