Instructions to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with Transformers:

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF", dtype="auto")

llama-cpp-python

How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF",
	filename="Gemma-4-Harmonia-31B-uncensored-heretic-BF16.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-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 llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-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 llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M

Use Docker

docker model run hf.co/llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with Ollama:
```
ollama run hf.co/llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M
```

Unsloth Studio

How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-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 llmfan46/Gemma-4-Harmonia-31B-uncensored-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 llmfan46/Gemma-4-Harmonia-31B-uncensored-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 llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF to start chatting

How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with Docker Model Runner:
```
docker model run hf.co/llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M
```

Lemonade

How to use llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull llmfan46/Gemma-4-Harmonia-31B-uncensored-heretic-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Gemma-4-Harmonia-31B-uncensored-heretic-GGUF-Q4_K_M

List all available models

lemonade list

Size differences

by Austriani - opened May 27

Discussion

Austriani

May 27

Hey, llmfan46, I want to tell you that there is some kind of error - Q3_K_M weights 26.1GB, while Q4_K_M only 21.2GB. Is it some kind of error? Or names just mixed up? I would like you to fix this up if possible.

IrisColt

about 1 month ago

Wow, you uncensored Harmonia! The original version was plagued with constant refusals, so this is such a pleasant surprise. Thanks a ton!!!

IrisColt

about 1 month ago

Oof, those sizes are wild... like they have nothing to do with what you'd expect a 31B to weigh at each quant.

IrisColt

about 1 month ago

Are you using er... a slightly different layout from mainline?

IrisColt

about 1 month ago

That extra size is just enough to blow past my 3090's VRAM at max context, and performance goes straight in the toilet, heh

llmfan46

Owner about 1 month ago

•

edited about 1 month ago

The reason is mentioned in the model card, here:

"For the K-quants below, selected Gemma 4 attention and FFN tensors are kept at higher precision where useful.

These GGUFs preserve key Gemma 4 attention projection tensors at higher precision.

Q6_K, Q5_K_M, Q5_K_S, Q4_K_M, Q4_K_S Q3_K_LandQ3_K_Mkeep the main attention projection tensors asQ8_0`:
    attn_q
    attn_k
    attn_v
    attn_output

This helps preserve Gemma 4’s attention path at higher precision, especially for lower-bit quants, while avoiding large file-size increases from unnecessarily up-quantizing the largest MoE expert tensors."

I will redo them now sith standard sizes.

IrisColt

about 1 month ago

•

edited about 1 month ago

OMG, did you just redo the quants? Thanks so much!

That said, if you think it's best to keep those key Gemma 4 attention projection tensors at higher precision, I'm all for it. I can just shorten my context.

BingoBird

about 1 month ago

People quantize their expectations. While size is what really matters, people will avoid going to q3_k_xl for example, even when it gives better results than q4_K_S.

llmfan46

Owner about 1 month ago

OMG, did you just redo the quants? Thanks so much!

That said, if you think it's best to keep those key Gemma 4 attention projection tensors at higher precision, I'm all for it. I can just shorten my context.

Yeah? But I just redid them because you said that performance was greatly decreased.

BingoBird

about 1 month ago

•

edited about 1 month ago

Is this "I want it smaller but I want it larger"?

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment