Instructions to use dahara1/gemma-4-12B-it-qat-assistant_gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use dahara1/gemma-4-12B-it-qat-assistant_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 dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16
# Run inference directly in the terminal:
llama cli -hf dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16
# Run inference directly in the terminal:
llama cli -hf dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

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 dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16
# Run inference directly in the terminal:
./llama-cli -hf dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

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 dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

Use Docker

docker model run hf.co/dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

LM Studio
Jan
Ollama
How to use dahara1/gemma-4-12B-it-qat-assistant_gguf with Ollama:
```
ollama run hf.co/dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16
```

Unsloth Studio

How to use dahara1/gemma-4-12B-it-qat-assistant_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 dahara1/gemma-4-12B-it-qat-assistant_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 dahara1/gemma-4-12B-it-qat-assistant_gguf to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for dahara1/gemma-4-12B-it-qat-assistant_gguf to start chatting

How to use dahara1/gemma-4-12B-it-qat-assistant_gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

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": "dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use dahara1/gemma-4-12B-it-qat-assistant_gguf with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Docker Model Runner
How to use dahara1/gemma-4-12B-it-qat-assistant_gguf with Docker Model Runner:
```
docker model run hf.co/dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16
```

Lemonade

How to use dahara1/gemma-4-12B-it-qat-assistant_gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

Run and chat with the model

lemonade run user.gemma-4-12B-it-qat-assistant_gguf-BF16

List all available models

lemonade list

Hermes Agent

How to use dahara1/gemma-4-12B-it-qat-assistant_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 dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

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 dahara1/gemma-4-12B-it-qat-assistant_gguf:BF16

Run Hermes

hermes

Atomic Chat

dahara1 commited on Jun 10

Commit

b978df4

verified ·

1 Parent(s): ee18475

Update README.md

Browse files

Files changed (1) hide show

README.md +5 -1

README.md CHANGED Viewed

@@ -10,15 +10,19 @@ tags:
 - mtp
 ---
-Gemma4は小さいモデルに先行させる事で推論を高速化するMTP(Multi-Token Prediction)という仕組みがあります。
 このモデルは[dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix](https://huggingface.co/dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix)をMTPで動かすための小モデルです。
 現時点ではQ4_0またはQ8_0が良さそうに見えます
 Gemma4 has a mechanism called MTP (Multi-Token Prediction) that speeds up inference by running a smaller model before the main model.
 This model is a small model for running [dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix](https://huggingface.co/dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix) with MTP.
 At this point, Q4_0 or Q8_0 looks promising.
 # Benchmark Reslut (RTX 4060ti(16GB))
 [am17an/mtp-bench.py](https://gist.github.com/am17an/228edfb84ed082aa88e3865d6fa27090)
 [mtp_bench_jp.py](https://huggingface.co/dahara1/gemma-4-12B-it-qat-assistant/raw/main/mtp_bench_jp.py) japanese translated version by dahara1.

 - mtp
 ---
+Gemma4は小さいモデルに先行させる事で推論を高速化するMTP(Multi-Token Prediction)という仕組みがあります。
+500MB程度のメモリ使用量が増えますが1.5超の速度向上が期待できます
 このモデルは[dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix](https://huggingface.co/dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix)をMTPで動かすための小モデルです。
 現時点ではQ4_0またはQ8_0が良さそうに見えます
 Gemma4 has a mechanism called MTP (Multi-Token Prediction) that speeds up inference by running a smaller model before the main model.
+While memory usage will increase by approximately 500MB, a speed improvement of over 1.5x can be expected.
 This model is a small model for running [dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix](https://huggingface.co/dahara1/gemma-4-12B-it-qat-UD-japanese-imatrix) with MTP.
 At this point, Q4_0 or Q8_0 looks promising.
 # Benchmark Reslut (RTX 4060ti(16GB))
+私が使用したベンチマークスクリプトは以下です
+The benchmark script I used is as follows:
 [am17an/mtp-bench.py](https://gist.github.com/am17an/228edfb84ed082aa88e3865d6fa27090)
 [mtp_bench_jp.py](https://huggingface.co/dahara1/gemma-4-12B-it-qat-assistant/raw/main/mtp_bench_jp.py) japanese translated version by dahara1.