Instructions to use mrysh/gemma-4-e4b-it-iq1_s with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use mrysh/gemma-4-e4b-it-iq1_s 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 mrysh/gemma-4-e4b-it-iq1_s:F16
# Run inference directly in the terminal:
llama cli -hf mrysh/gemma-4-e4b-it-iq1_s:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf mrysh/gemma-4-e4b-it-iq1_s:F16
# Run inference directly in the terminal:
llama cli -hf mrysh/gemma-4-e4b-it-iq1_s:F16

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 mrysh/gemma-4-e4b-it-iq1_s:F16
# Run inference directly in the terminal:
./llama-cli -hf mrysh/gemma-4-e4b-it-iq1_s:F16

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 mrysh/gemma-4-e4b-it-iq1_s:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf mrysh/gemma-4-e4b-it-iq1_s:F16

Use Docker

docker model run hf.co/mrysh/gemma-4-e4b-it-iq1_s:F16

Llama new
LM Studio
Jan
Ollama
How to use mrysh/gemma-4-e4b-it-iq1_s with Ollama:
```
ollama run hf.co/mrysh/gemma-4-e4b-it-iq1_s:F16
```
Unsloth Desktop

How to use mrysh/gemma-4-e4b-it-iq1_s with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mrysh/gemma-4-e4b-it-iq1_s:F16

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "mrysh/gemma-4-e4b-it-iq1_s:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use mrysh/gemma-4-e4b-it-iq1_s with Docker Model Runner:
```
docker model run hf.co/mrysh/gemma-4-e4b-it-iq1_s:F16
```

Lemonade

How to use mrysh/gemma-4-e4b-it-iq1_s with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull mrysh/gemma-4-e4b-it-iq1_s:F16

Run and chat with the model

lemonade run user.gemma-4-e4b-it-iq1_s-F16

List all available models

lemonade list

Hermes Agent

How to use mrysh/gemma-4-e4b-it-iq1_s with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mrysh/gemma-4-e4b-it-iq1_s:F16

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 mrysh/gemma-4-e4b-it-iq1_s:F16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use mrysh/gemma-4-e4b-it-iq1_s with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mrysh/gemma-4-e4b-it-iq1_s:F16

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 "mrysh/gemma-4-e4b-it-iq1_s:F16" \
  --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"

gemma-4-e4b-it - iq1_s

Quantized version of google/gemma-4-E4B-it made for a quantization benchmark (4 models x 8 techniques, evaluated with vLLM on BFCL, Energy-QA, hiring-bias and MedQA).

method: iq1_s, about 1.5 bits, group size None
library: 0693dbfb915f
calibration samples: 512
size on disk: 3.095 GB

GGUF IQ1_S (about 1.56 bits per weight, importance matrix from a mixed calibration set). Text only (the vision/audio part is not quantized and not used). Runs in vLLM with the vllm-gguf-plugin or in llama.cpp. Expect a large quality loss: it is the extreme end of the benchmark.

Evaluated as a text model (vLLM, thinking off). Same license as the base model; see the base model card.

Downloads last month: 313

GGUF

Model size

8B params

Architecture

gemma4

Hardware compatibility

1-bit

Inference Providers NEW

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

Model tree for mrysh/gemma-4-e4b-it-iq1_s

Base model

google/gemma-4-E4B

Finetuned

google/gemma-4-E4B-it

Quantized

(421)

this model