Instructions to use stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small",
	filename="gemma-3-12b-it-q4_0_s.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S
# Run inference directly in the terminal:
llama-cli -hf stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S
# Run inference directly in the terminal:
llama-cli -hf stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S

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 stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S
# Run inference directly in the terminal:
./llama-cli -hf stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S

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 stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S
# Run inference directly in the terminal:
./build/bin/llama-cli -hf stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S

Use Docker

docker model run hf.co/stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S

LM Studio
Jan
Ollama
How to use stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small with Ollama:
```
ollama run hf.co/stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S
```

Unsloth Studio new

How to use stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small 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 stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small 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 stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small to start chatting

Docker Model Runner
How to use stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small with Docker Model Runner:
```
docker model run hf.co/stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S
```

Lemonade

How to use stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small:Q4_0_S

Run and chat with the model

lemonade run user.google-gemma-3-12b-it-qat-q4_0-gguf-small-Q4_0_S

List all available models

lemonade list

This is a "self" merge of https://huggingface.co/google/gemma-3-12b-it-qat-q4_0-gguf and https://huggingface.co/bartowski/google_gemma-3-12b-it-GGUF.

The official QAT weights released by google use fp16 (instead of Q6_K) for the embeddings table, which makes this model take a significant extra amount of memory (and storage) compared to what Q4_0 quants are supposed to take. Instead of quantizing the table myself, I extracted it from Bartowski's quantized models because I thought using imatrix quants would give better quality (it doesn't, imatrix isn't used for token embeddings).

Here are some perplexity measurements:

Model	File size ↓	PPL (wiki.text.raw) ↓	Hellaswag, 4k tasks ↑
iQ3_xs (bartowski)	5.21 GB	10.0755 +/- 0.08024	---
This model	6.89 GB	9.2637 +/- 0.07216	72.925% [71.5366M, 74.2794%]
Q4_0 (bartowski)	6.91 GB	9.5589 +/- 0.07527	73.125% [71.7295%, 74.4761%]
QAT Q4_0 (google)	8.07 GB	9.2565 +/- 0.07212	72.850% [71.4505%, 74.2056%]
Q5_K_S (bartowski)	8.23 GB	9.8540 +/- 0.08016	---

Note that this model ends up smaller than the Q4_0 from Bartowski. This is because llama.cpp sets some tensors to Q4_1 when quantizing models to Q4_0 with imatrix, but this is a static quant. I don't understand why Q5_K_S is performing worse on that test than the default Q4_0, I wasn't expecting this outcome. This merge seems to be a good balance between model size and perplexity. I believe this is representative to the overall quality of the model.

Downloads last month: 785

GGUF

Model size

12B params

Architecture

gemma3

Hardware compatibility

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 2 Ask for provider support

Model tree for stduhpf/google-gemma-3-12b-it-qat-q4_0-gguf-small

Base model

google/gemma-3-12b-pt

Finetuned

google/gemma-3-12b-it

Quantized

bartowski/google_gemma-3-12b-it-GGUF

Quantized

(1)

this model