Instructions to use awai-network/mishima with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use awai-network/mishima 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 awai-network/mishima:TQ1_0
# Run inference directly in the terminal:
llama cli -hf awai-network/mishima:TQ1_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf awai-network/mishima:TQ1_0
# Run inference directly in the terminal:
llama cli -hf awai-network/mishima:TQ1_0

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 awai-network/mishima:TQ1_0
# Run inference directly in the terminal:
./llama-cli -hf awai-network/mishima:TQ1_0

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 awai-network/mishima:TQ1_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf awai-network/mishima:TQ1_0

Use Docker

docker model run hf.co/awai-network/mishima:TQ1_0

LM Studio
Jan

vLLM

How to use awai-network/mishima with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/awai-network/mishima:TQ1_0

Ollama
How to use awai-network/mishima with Ollama:
```
ollama run hf.co/awai-network/mishima:TQ1_0
```
Unsloth Desktop

How to use awai-network/mishima with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf awai-network/mishima:TQ1_0

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": "awai-network/mishima:TQ1_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use awai-network/mishima with Docker Model Runner:
```
docker model run hf.co/awai-network/mishima:TQ1_0
```

Lemonade

How to use awai-network/mishima with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull awai-network/mishima:TQ1_0

Run and chat with the model

lemonade run user.mishima-TQ1_0

List all available models

lemonade list

Hermes Agent

How to use awai-network/mishima with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf awai-network/mishima:TQ1_0

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 awai-network/mishima:TQ1_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use awai-network/mishima with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf awai-network/mishima:TQ1_0

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 "awai-network/mishima:TQ1_0" \
  --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"

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Mishima

Mishima is the public Murakumo name for the exact Ternary-Bonsai-2-27B-PTQ1_0.gguf checkpoint served by api.murakumo.cloud.

This repository mirrors the checkpoint unchanged from prism-ml/Ternary-Bonsai-2-27B-gguf. It is not a fine-tune by awai.network.

Field	Value
Murakumo model ID	`mishima`
Format	GGUF, PTQ1_0 ternary
Parameters	27.36B
File	`Ternary-Bonsai-2-27B-PTQ1_0.gguf`
Size	5,946,648,928 bytes
SHA-256	`53107f530aa52eb00912263ab1ee29bd199261c87cd7b4ad4ca1318c1fe33ee3`
Training context declared by the checkpoint	262,144 tokens
Current standard Murakumo serving window	32,768 tokens (some edge heads use 8,192)

Hosted API

curl https://api.murakumo.cloud/v1/chat/completions \
  -H 'content-type: application/json' \
  -d '{
    "model": "mishima",
    "messages": [{"role": "user", "content": "Hello"}],
    "max_tokens": 128
  }'

The endpoint is OpenAI-compatible. Capacity and per-head context are runtime properties and may differ from the checkpoint's native context declaration.

Local use

The deployed Murakumo runtime uses Prism ML's PTQ1_0-capable llama.cpp fork. Use a runtime that explicitly supports the PTQ1_0 ternary metadata; a loader that does not support it should refuse the file instead of interpreting it as an ordinary GGUF quantization.

Attribution

Created using Bonsai by Prism ML. The checkpoint and its upstream notices are Apache-2.0 licensed. See NOTICE.txt and the upstream model card for full architecture, evaluation, and citation details.

Downloads last month: -

GGUF

Model size

27B params

Architecture

qwen35

Hardware compatibility

1-bit

Model tree for awai-network/mishima

Base model

Qwen/Qwen3.8-27B

Quantized

(1202)

this model