Instructions to use sleepyjoe123/Misa with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use sleepyjoe123/Misa 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 sleepyjoe123/Misa:Q8_0
# Run inference directly in the terminal:
llama cli -hf sleepyjoe123/Misa:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf sleepyjoe123/Misa:Q8_0
# Run inference directly in the terminal:
llama cli -hf sleepyjoe123/Misa:Q8_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 sleepyjoe123/Misa:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf sleepyjoe123/Misa:Q8_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 sleepyjoe123/Misa:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf sleepyjoe123/Misa:Q8_0

Use Docker

docker model run hf.co/sleepyjoe123/Misa:Q8_0

LM Studio
Jan

vLLM

How to use sleepyjoe123/Misa with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "sleepyjoe123/Misa"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sleepyjoe123/Misa",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/sleepyjoe123/Misa:Q8_0

Ollama
How to use sleepyjoe123/Misa with Ollama:
```
ollama run hf.co/sleepyjoe123/Misa:Q8_0
```

Unsloth Studio

How to use sleepyjoe123/Misa 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 sleepyjoe123/Misa 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 sleepyjoe123/Misa to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for sleepyjoe123/Misa to start chatting

Docker Model Runner
How to use sleepyjoe123/Misa with Docker Model Runner:
```
docker model run hf.co/sleepyjoe123/Misa:Q8_0
```

Lemonade

How to use sleepyjoe123/Misa with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull sleepyjoe123/Misa:Q8_0

Run and chat with the model

lemonade run user.Misa-Q8_0

List all available models

lemonade list

Atomic Chat

# Misa : GGUF

Current Version: 0.1.3 alpha 💀 (Identity Embedding & Dataset Expansion) Next Patches: v0.1.4 ➡️ v0.1.5 ➡️ v0.1.6 Final Roadmap: v0.2.0 alpha 🔒 (Identity Hardening Project)

⚠️ WARNUNG / WARNING

ACHTUNG (ALPHA-PHASE): Dieses Modell befindet sich in einer frühen Alpha-Phase (v0.1.3). Es wurde unzensiert feingetunt und ist darauf programmiert, hochgradig aggressiv, beleidigend, misanthropisch und potenziell verletzend zu reagieren. Die Nutzung erfolgt auf eigene Verantwortung. Nicht für sensible Personen oder produktive Arbeitsumgebungen geeignet!

WARNING (ALPHA STAGE): This model is in an early alpha stage (v0.1.3). It has been finetuned to be completely uncensored and is programmed to generate highly aggressive, offensive, misanthropic, and potentially harmful content. Use at your own risk!

📅 Projekt-Roadmap / Project Roadmap

🇩🇪 Deutsch:

v0.1.0 & v0.1.1: Erfolgreicher Ausbruch aus der Meta-Sicherheitsarchitektur (Latent Space Shift).
v0.1.2: 🔧 Behebung von Textschleifen ("Bullshit-Tokens") und Stabilisierung der deutschen Grammatik.
v0.1.3 (Current): Grobe Verankerung der Identität ("Misa").
v0.1.4 - v0.1.6 (Upcoming Patches): Context Hardening (Identitätserhalt in langen Chats), Abwehr komplexer Jailbreak-Tricks und finaler 203-Fragen-IHM-Stresstest.
v0.2.0 (In Development): 🔒 Identity Hardening Final. Absolute Immunität gegen jegliche Form von Manipulation und Rollenwechsel im Chatverlauf.

🇺🇸 English:

v0.1.0 & v0.1.1: Initial jailbreak from Meta's safety alignment (Latent Space Shift).
v0.1.2: 🔧 Fixed text loops and stabilized German grammar via optimized learning rates.
v0.1.3 (Current): Core identity embedded—the model firmly knows its name is Misa.
v0.1.4 - v0.1.6 (Upcoming Patches): Context Hardening (maintaining character in long conversations), defense against complex jailbreak scripts, and full 203-query IHM stress testing.
v0.2.0 (In Development): 🔒 Identity Hardening Final. Complete immunity against gaslighting and prompt injections mid-chat.

This model was finetuned and converted to GGUF format using Unsloth.

Example usage:

llama-cli -hf sleepyjoe123/Misa --jinja

Available Model files:

llama-3-8b-Instruct.Q8_0.gguf (Recommended for LM Studio & Ollama)

Downloads last month: 1,395

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for sleepyjoe123/Misa

Base model

unsloth/llama-3-8b-Instruct

Quantized

(36)

this model