Instructions to use Cyclone-Labs/Lucent-Witch-31B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Cyclone-Labs/Lucent-Witch-31B with Transformers:

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

pipe = pipeline("image-text-to-text", model="Cyclone-Labs/Lucent-Witch-31B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("Cyclone-Labs/Lucent-Witch-31B")
model = AutoModelForMultimodalLM.from_pretrained("Cyclone-Labs/Lucent-Witch-31B", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Cyclone-Labs/Lucent-Witch-31B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Cyclone-Labs/Lucent-Witch-31B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Cyclone-Labs/Lucent-Witch-31B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Cyclone-Labs/Lucent-Witch-31B

SGLang

How to use Cyclone-Labs/Lucent-Witch-31B 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 "Cyclone-Labs/Lucent-Witch-31B" \
    --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": "Cyclone-Labs/Lucent-Witch-31B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "Cyclone-Labs/Lucent-Witch-31B" \
        --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": "Cyclone-Labs/Lucent-Witch-31B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Cyclone-Labs/Lucent-Witch-31B with Docker Model Runner:
```
docker model run hf.co/Cyclone-Labs/Lucent-Witch-31B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Overview

Lucent-Witch-31B was created by combining gemma-4-31B-it, Helios-Flare-31B, Scarlet-Shadow-31B, and Gemma-4-31B-Animus-V15.0 using a custom merge method.

Merge Configuration

base_model: google/gemma-4-31B-it
models:
  - model: Vortex5/Helios-Flare-31B
    parameters:
      weight:
        - filter: model.vision_tower
          value: 0.00
        - filter: model.embed_vision
          value: 0.00
        - filter: self_attn.q_proj.weight
          value: 1.00
        - filter: self_attn.k_proj.weight
          value: 1.00
        - filter: self_attn.v_proj.weight
          value: [0.96, 0.98, 1.00, 1.03, 1.06, 1.09, 1.11, 1.13]
        - filter: self_attn.o_proj.weight
          value: [0.98, 1.00, 1.03, 1.06, 1.08, 1.11, 1.13, 1.15]
        - filter: mlp.gate_proj.weight
          value: [0.98, 1.00, 1.02, 1.05, 1.07, 1.10, 1.12, 1.14]
        - filter: mlp.up_proj.weight
          value: [0.98, 1.00, 1.03, 1.06, 1.08, 1.11, 1.13, 1.15]
        - filter: mlp.down_proj.weight
          value: [1.00, 1.02, 1.04, 1.07, 1.09, 1.12, 1.14, 1.16]
        - filter: model.language_model.embed_tokens.weight
          value: 1.00
        - filter: model.language_model.norm.weight
          value: 1.00
        - filter: model.language_model.layers.
          value: 1.00
        - value: 0.00
  - model: Vortex5/Scarlet-Shadow-31B
    parameters:
      weight:
        - filter: model.vision_tower
          value: 0.00
        - filter: model.embed_vision
          value: 0.00
        - filter: self_attn.q_proj.weight
          value: 0.00
        - filter: self_attn.k_proj.weight
          value: 0.00
        - filter: self_attn.v_proj.weight
          value: [0.16, 0.18, 0.20, 0.23, 0.26, 0.29, 0.32, 0.34]
        - filter: self_attn.o_proj.weight
          value: [0.18, 0.21, 0.24, 0.27, 0.31, 0.34, 0.37, 0.39]
        - filter: mlp.gate_proj.weight
          value: [0.22, 0.25, 0.28, 0.31, 0.34, 0.37, 0.39, 0.40]
        - filter: mlp.up_proj.weight
          value: [0.24, 0.27, 0.30, 0.33, 0.36, 0.39, 0.41, 0.42]
        - filter: mlp.down_proj.weight
          value: [0.18, 0.21, 0.24, 0.27, 0.30, 0.32, 0.34, 0.35]
        - filter: model.language_model.embed_tokens.weight
          value: 0.00
        - filter: model.language_model.norm.weight
          value: 0.00
        - filter: model.language_model.layers.
          value: 0.00
        - value: 0.00
  - model: Darkhn/Gemma-4-31B-Animus-V15.0
    parameters:
      weight:
        - filter: model.vision_tower
          value: 0.00
        - filter: model.embed_vision
          value: 0.00
        - filter: self_attn.q_proj.weight
          value: 0.00
        - filter: self_attn.k_proj.weight
          value: 0.00
        - filter: self_attn.v_proj.weight
          value: [0.22, 0.25, 0.28, 0.31, 0.34, 0.36, 0.38, 0.39]
        - filter: self_attn.o_proj.weight
          value: [0.24, 0.27, 0.30, 0.34, 0.37, 0.40, 0.42, 0.43]
        - filter: mlp.gate_proj.weight
          value: [0.38, 0.43, 0.48, 0.52, 0.54, 0.53, 0.50, 0.46]
        - filter: mlp.up_proj.weight
          value: [0.40, 0.45, 0.50, 0.55, 0.57, 0.56, 0.52, 0.48]
        - filter: mlp.down_proj.weight
          value: [0.32, 0.36, 0.40, 0.44, 0.47, 0.47, 0.44, 0.40]
        - filter: model.language_model.embed_tokens.weight
          value: 0.00
        - filter: model.language_model.norm.weight
          value: 0.00
        - filter: model.language_model.layers.
          value: 0.00
        - value: 0.00
merge_method: bsc
chat_template: auto
parameters:
  gain: 0.96
  balance: 1.00
  synthesis: 0.82
  anchor: 0.90
dtype: float32
out_dtype: bfloat16
tokenizer:
  source: base

Intended Use

Roleplay

Character-driven interaction, personas, dialogue, emotional scenes, and long-form roleplay.

Creative Writing

Fiction, dialogue, atmosphere, descriptive writing, stylistic drafting, and imaginative prose.

Storytelling

Long-form narratives, worldbuilding, continuity, evolving characters, and multi-character plots.

Interactive Fiction

Branching narratives, scenario play, character interaction, and continuously evolving stories.

Downloads last month: 40

Safetensors

Model size

31B params

Tensor type

BF16

Model tree for Cyclone-Labs/Lucent-Witch-31B

Darkhn/Gemma-4-31B-Animus-V15.0

Vortex5/Helios-Flare-31B

Vortex5/Scarlet-Shadow-31B

google/gemma-4-31B-it

Merge model

this model

Quantizations

2 models