Instructions to use royallab/MN-LooseCannon-12B-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use royallab/MN-LooseCannon-12B-v2 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="royallab/MN-LooseCannon-12B-v2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("royallab/MN-LooseCannon-12B-v2")
model = AutoModelForCausalLM.from_pretrained("royallab/MN-LooseCannon-12B-v2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use royallab/MN-LooseCannon-12B-v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "royallab/MN-LooseCannon-12B-v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "royallab/MN-LooseCannon-12B-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/royallab/MN-LooseCannon-12B-v2

SGLang

How to use royallab/MN-LooseCannon-12B-v2 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "royallab/MN-LooseCannon-12B-v2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "royallab/MN-LooseCannon-12B-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "royallab/MN-LooseCannon-12B-v2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "royallab/MN-LooseCannon-12B-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use royallab/MN-LooseCannon-12B-v2 with Docker Model Runner:
```
docker model run hf.co/royallab/MN-LooseCannon-12B-v2
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

MN-LooseCannon-12B-v2

This is a Mistral Nemo-based model consisting of a merge between:

This merge was performed with permission from the v1 creator (NGalrion). THe merge was performed in 2 steps.

The first is an intermediate starcannon v3 merge, but using magnum-v2.5-12b-kto:

models:
    - model: anthracite-org/magnum-v2.5-12b-kto
      parameters:
        density: 0.3
        weight: 0.5
    - model: nothingiisreal/MN-12B-Celeste-V1.9
      parameters:
        density: 0.7
        weight: 0.5

merge_method: ties
base_model: nothingiisreal/MN-12B-Celeste-V1.9
parameters:
    normalize: true
    int8_mask: true
dtype: bfloat16

The second is a merge between the intermediate starcannon v3 and Lyra, which reproduces LooseCannon.

models:
  - model: ./MN-LooseCannon-12B-v2-step1
    parameters:
      density: 0.3
      weight: 0.75
  - model: Sao10K/MN-12B-Lyra-v1
    parameters:
      density: 0.7
      weight: 0.25

merge_method: ties
base_model: ./MN-LooseCannon-12B-v2-step1
parameters:
  normalize: true
dtype: bfloat16

Usage

This model will follow the ChatML instruct format:

<|im_start|>system
You are a helpful assistant.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

It can also follow the base Mistral Nemo instruct format, but ChatML is recommended.

Bias, Risks, and Limitations

The model will show biases similar to those observed in niche roleplaying forums on the Internet, besides those exhibited by the base model. It is not intended for supplying factual information or advice in any form.