Instructions to use grenishrai/yoru-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 grenishrai/yoru-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 grenishrai/yoru-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf grenishrai/yoru-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 grenishrai/yoru-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf grenishrai/yoru-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 grenishrai/yoru-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf grenishrai/yoru-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 grenishrai/yoru-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf grenishrai/yoru-GGUF:Q4_K_M

Use Docker

docker model run hf.co/grenishrai/yoru-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use grenishrai/yoru-GGUF with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/grenishrai/yoru-GGUF:Q4_K_M

Ollama
How to use grenishrai/yoru-GGUF with Ollama:
```
ollama run hf.co/grenishrai/yoru-GGUF:Q4_K_M
```

Unsloth Studio

How to use grenishrai/yoru-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 grenishrai/yoru-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 grenishrai/yoru-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for grenishrai/yoru-GGUF to start chatting

Docker Model Runner
How to use grenishrai/yoru-GGUF with Docker Model Runner:
```
docker model run hf.co/grenishrai/yoru-GGUF:Q4_K_M
```

Lemonade

How to use grenishrai/yoru-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull grenishrai/yoru-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.yoru-GGUF-Q4_K_M

List all available models

lemonade list

Atomic Chat

Yoru GGUF

Yoru is a model in the Mori family. This repo is the Q4_K_M GGUF of the merged F16 checkpoint grenishrai/yoru.

Use this file with llama.cpp, LM Studio, Ollama, or any GGUF runner. For Transformers / PEFT, use the F16 repo instead.

It is a style / persona mix for casual internet / Gen Z chat, not a facts model.

Model Details

Name: Yoru
Family: Mori
This repo: grenishrai/yoru-GGUF
F16 weights: grenishrai/yoru
Original base: HuggingFaceTB/SmolLM2-1.7B-Instruct
Type: Q4_K_M GGUF of a QLoRA SFT merge
Architecture: Llama (SmolLM2), ~1.71B params, context 8192
Chat template: ChatML (<|im_start|> / <|im_end|>)
Language: English (internet slang)
License: MIT (this file; the F16 and base models have their own cards)
Dataset: grenishrai/genz-sft-dataset

File	Type	Size	Notes
`yoru-Q4_K_M.gguf`	Q4_K_M	~1.06 GB	only file in this repo; default

The F16 source is ~3.42 GB. This quant is the shareable / local-inference copy.

How to Use

Pass the same system prompt the data used, or the voice will slip.

You are a pure Gen Z speaker. Always reply in natural Gen Z slang and internet speech. Use words and phrases like: no cap, fr fr, lowkey, highkey, bet, rizz, mid, slay, periodt, it's giving, sus, oof, vibes, down horrendous, bussin, cooked, goated, main character, etc. Keep replies casual, short to medium length, and online. Never break character. Never explain the slang. Never sound formal or like a normal AI.

llama.cpp

llama-cli -hf grenishrai/yoru-GGUF \
  --sys "You are a pure Gen Z speaker. Always reply in natural Gen Z slang and internet speech. Use words and phrases like: no cap, fr fr, lowkey, highkey, bet, rizz, mid, slay, periodt, it's giving, sus, oof, vibes, down horrendous, bussin, cooked, goated, main character, etc. Keep replies casual, short to medium length, and online. Never break character. Never explain the slang. Never sound formal or like a normal AI." \
  -p "I barely slept and now I have to be a person today"

Local file:

llama-cli -m yoru-Q4_K_M.gguf \
  --sys "You are a pure Gen Z speaker. ..." \
  -p "I barely slept and now I have to be a person today"

LM Studio / Ollama

Import yoru-Q4_K_M.gguf. Set the chat template to ChatML if it is not picked up from the GGUF. Put the system prompt above in the system field.

Python (llama-cpp-python)

from llama_cpp import Llama

llm = Llama.from_pretrained(
    repo_id="grenishrai/yoru-GGUF",
    filename="yoru-Q4_K_M.gguf",
    n_ctx=8192,
)

system = (
    "You are a pure Gen Z speaker. Always reply in natural Gen Z slang and internet speech. "
    "Use words and phrases like: no cap, fr fr, lowkey, highkey, bet, rizz, mid, slay, periodt, "
    "it's giving, sus, oof, vibes, down horrendous, bussin, cooked, goated, main character, etc. "
    "Keep replies casual, short to medium length, and online. Never break character. "
    "Never explain the slang. Never sound formal or like a normal AI."
)

out = llm.create_chat_completion(
    messages=[
        {"role": "system", "content": system},
        {"role": "user", "content": "I barely slept and now I have to be a person today"},
    ],
    max_tokens=80,
    temperature=0.8,
    top_p=0.9,
)
print(out["choices"][0]["message"]["content"])

Uses

Same as the F16 model: short informal chat, persona experiments. Not for formal, medical, legal, or factual QA. Not a sample of real Gen Z speech.

Q4_K_M will be a bit less sharp than F16. If a reply looks off, check the same prompt on grenishrai/yoru before blaming the fine-tune.

Training / quant

Yoru was QLoRA-trained on SmolLM2-1.7B-Instruct, best checkpoint epoch 2, then merged to fp16. This file is that merge quantized with llama.cpp Q4_K_M. Training details live on the F16 card.

Citation

@misc{yoru-gguf-2026,
  title  = {Yoru (Mori) Q4_K_M GGUF},
  author = {grenishrai},
  year   = {2026},
  url    = {https://huggingface.co/grenishrai/yoru-GGUF},
  note   = {Quantized from https://huggingface.co/grenishrai/yoru}
}

Model Card Contact

Open an issue on the model repository.

Downloads last month: -

GGUF

Model size

2B params

Architecture

llama

Hardware compatibility

4-bit

Model tree for grenishrai/yoru-GGUF

Base model

HuggingFaceTB/SmolLM2-1.7B

Quantized

HuggingFaceTB/SmolLM2-1.7B-Instruct

Finetuned

grenishrai/yoru