Instructions to use koifishlabs/Kurogo-Gemma-4-E2B-text-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 koifishlabs/Kurogo-Gemma-4-E2B-text-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 koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf koifishlabs/Kurogo-Gemma-4-E2B-text-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 koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf koifishlabs/Kurogo-Gemma-4-E2B-text-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 koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf koifishlabs/Kurogo-Gemma-4-E2B-text-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 koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M

Use Docker

docker model run hf.co/koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "koifishlabs/Kurogo-Gemma-4-E2B-text-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": "koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M

Ollama
How to use koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF with Ollama:
```
ollama run hf.co/koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M
```
Unsloth Desktop

How to use koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M

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": "koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF with Docker Model Runner:
```
docker model run hf.co/koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M
```

Lemonade

How to use koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Kurogo-Gemma-4-E2B-text-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use koifishlabs/Kurogo-Gemma-4-E2B-text-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 koifishlabs/Kurogo-Gemma-4-E2B-text-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 koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf koifishlabs/Kurogo-Gemma-4-E2B-text-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 "koifishlabs/Kurogo-Gemma-4-E2B-text-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"

Kurogo · Gemma 4 E2B (text, Q4_K_M, GGUF)

This is a text-only, Q4_K_M quantized GGUF build of Google's gemma-4-E2B-it, packaged for Kurogo — a fully offline, privacy-first AI app for iOS and Android. Everything runs on the user's device; no data ever leaves the phone.

What's in this repo

File	Format	Quantization	Size
`gemma-4-E2B-it-text-Q4_K_M.gguf`	GGUF (text-only)	Q4_K_M	~2.9 GiB

The original gemma-4-E2B-it is a multimodal model with vision and audio capabilities. This build ships only the text weights — no mmproj (multimodal projector), no vision encoder, no audio encoder. Kurogo uses the model exclusively for text generation, so the multimodal components are intentionally excluded to keep the on-device footprint small.

Quantization

Q4_K_M — 4-bit K-quants, "medium" preset. Good balance of size and quality for on-device inference. Specifically uses Q6_K for attention.wv and feed_forward.w2 tensors and Q4_K for the rest, per llama.cpp's standard K-quant layout.

How Kurogo uses it

Loaded via llama.rn (a React Native binding for llama.cpp):

import { initLlama } from "llama.rn";

const context = await initLlama({
  model: "/path/to/gemma-4-E2B-it-text-Q4_K_M.gguf",
  n_ctx: 32768,           // tiered by device RAM (2k–32k)
  n_gpu_layers: 99,
  cache_type_k: "q8_0",
  cache_type_v: "q8_0",
});

Context length is automatically tiered by device RAM (2k on ≤4 GB devices up to 32k on ≥8 GB).

Use cases

Conversational chat with the user's local AI agent
Wiki page generation and editing (markdown out)
Wiki search and summarization across the user's local notes
URL-based content import (tweets, articles → wiki pages)

All inference happens on-device — no API keys, no network calls, no telemetry.

License

This model is released under the Gemma Terms of Use. By downloading or using this model, you agree to those terms.

Attribution

Quantization: Standard llama.cpp Q4_K_M conversion
Packaged by: Koi Fish Labs for Kurogo

Model tree for koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF

Base model

google/gemma-4-E2B

Finetuned

google/gemma-4-E2B-it

Quantized

(334)

this model

koifishlabs
/

Kurogo-Gemma-4-E2B-text-GGUF

Kurogo · Gemma 4 E2B (text, Q4_K_M, GGUF)

What's in this repo

Quantization

How Kurogo uses it

Use cases

License

Attribution

Links

Model tree for koifishlabs/Kurogo-Gemma-4-E2B-text-GGUF