Instructions to use barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M

Use Docker

docker model run hf.co/barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use barozp/gemma-4-19b-a4b-it-REAP-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "barozp/gemma-4-19b-a4b-it-REAP-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "barozp/gemma-4-19b-a4b-it-REAP-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M

Ollama
How to use barozp/gemma-4-19b-a4b-it-REAP-GGUF with Ollama:
```
ollama run hf.co/barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M
```

Unsloth Studio

How to use barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for barozp/gemma-4-19b-a4b-it-REAP-GGUF to start chatting

How to use barozp/gemma-4-19b-a4b-it-REAP-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf barozp/gemma-4-19b-a4b-it-REAP-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": "barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use barozp/gemma-4-19b-a4b-it-REAP-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M

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 "barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M" \
  --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 barozp/gemma-4-19b-a4b-it-REAP-GGUF with Docker Model Runner:
```
docker model run hf.co/barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M
```

Lemonade

How to use barozp/gemma-4-19b-a4b-it-REAP-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.gemma-4-19b-a4b-it-REAP-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-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 barozp/gemma-4-19b-a4b-it-REAP-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

gemma-4-19b-a4b-it-REAP — GGUF Quantizations

GGUF quantizations of 0xSero/gemma-4-19b-a4b-it-REAP, a 30% expert-pruned variant of google/gemma-4-26b-a4b-it using the REAP (Router-weighted Expert Activation Pruning) method.

Available Files

File	Quant	Size	BPW	Description
`gemma-4-19b-a4b-it-REAP-BF16.gguf`	BF16	~36 GB	16.0	Full precision, for re-quantization
`gemma-4-19b-a4b-it-REAP-Q8_0.gguf`	Q8_0	~19 GB	8.0	Near-lossless, large file
`gemma-4-19b-a4b-it-REAP-Q6_K.gguf`	Q6_K	~15 GB	6.56	Near-lossless, recommended for high quality
`gemma-4-19b-a4b-it-REAP-Q5_K_M.gguf`	Q5_K_M	~13 GB	5.68	High quality, larger size
`gemma-4-19b-a4b-it-REAP-Q5_K_S.gguf`	Q5_K_S	~12 GB	5.52	High quality, slightly smaller
`gemma-4-19b-a4b-it-REAP-Q4_K_M.gguf`	Q4_K_M	~12 GB	4.89	Recommended — best quality/size balance
`gemma-4-19b-a4b-it-REAP-Q4_K_S.gguf`	Q4_K_S	~11 GB	4.63	4-bit small
`gemma-4-19b-a4b-it-REAP-Q3_K_L.gguf`	Q3_K_L	~9.5 GB	4.27	3-bit large
`gemma-4-19b-a4b-it-REAP-Q3_K_M.gguf`	Q3_K_M	~9 GB	3.91	3-bit medium
`gemma-4-19b-a4b-it-REAP-Q3_K_S.gguf`	Q3_K_S	~8.5 GB	3.66	3-bit small
`gemma-4-19b-a4b-it-REAP-Q2_K.gguf`	Q2_K	~7.5 GB	2.96	Smallest size, lowest quality

Model Details

Property	Value
Architecture	Gemma 4 (hybrid sliding/full attention MoE)
Parameters	19.02B total / ~4B active per token
Experts	90 total / 8 active per token
Context Length	262,144 tokens
Original dtype	BF16
Quantization tool	llama.cpp
License	Gemma

Quantization Process

# 1. Convert BF16 SafeTensors → GGUF
python convert_hf_to_gguf.py 0xSero/gemma-4-19b-a4b-it-REAP \
  --outfile gemma-4-19b-a4b-it-REAP-BF16.gguf \
  --outtype bf16

# 2. Quantize (example: Q4_K_M)
llama-quantize gemma-4-19b-a4b-it-REAP-BF16.gguf \
  gemma-4-19b-a4b-it-REAP-Q4_K_M.gguf Q4_K_M

Usage

llama.cpp

llama-cli \
  -m gemma-4-19b-a4b-it-REAP-Q4_K_M.gguf \
  -ngl 99 -c 4096 \
  -p "Your prompt here"

llama-server (OpenAI-compatible API)

llama-server \
  -m gemma-4-19b-a4b-it-REAP-Q4_K_M.gguf \
  -ngl 99 -c 4096 \
  --port 8080

LM Studio / Jan / Ollama

Download the .gguf file and load it directly in your preferred local inference UI.

Hardware Requirements

Config	VRAM / RAM
Full GPU (Q4_K_M, recommended)	14+ GB VRAM
Hybrid CPU+GPU (Q4_K_M)	8 GB VRAM + 8 GB RAM
CPU only (Q4_K_M)	16+ GB RAM

About the Original Model

0xSero/gemma-4-19b-a4b-it-REAP applies REAP expert pruning (arXiv:2510.13999) to remove 30% of MoE experts (38 of 128 per layer) from Gemma 4 26B-A4B-it, while preserving routing behavior. Active parameters per token remain unchanged at ~4B. The result is a ~31% smaller model with near-identical generation quality across coding, math, and reasoning benchmarks.