Instructions to use electron-rare/spikingkiki-27b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use electron-rare/spikingkiki-27b with Transformers:

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

pipe = pipeline("text-generation", model="electron-rare/spikingkiki-27b")
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, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("electron-rare/spikingkiki-27b")
model = AutoModelForImageTextToText.from_pretrained("electron-rare/spikingkiki-27b")
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

vLLM

How to use electron-rare/spikingkiki-27b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "electron-rare/spikingkiki-27b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "electron-rare/spikingkiki-27b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/electron-rare/spikingkiki-27b

SGLang

How to use electron-rare/spikingkiki-27b 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 "electron-rare/spikingkiki-27b" \
    --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": "electron-rare/spikingkiki-27b",
		"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 "electron-rare/spikingkiki-27b" \
        --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": "electron-rare/spikingkiki-27b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use electron-rare/spikingkiki-27b with Docker Model Runner:
```
docker model run hf.co/electron-rare/spikingkiki-27b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

SpikingKiKi 27B

Spiking Neural Network (SNN) variant of Qwen3.5-27B class, using Leaky Integrate-and-Fire (LIF) neurons in selected layers. Research artifact from Hypneum Lab (hypneum-lab GitHub org).

Base architecture: Qwen3.5 dense (~27B params)
Modification: LIF spiking neurons replace standard activations in attention/FFN
Metadata: see lif_metadata.json for per-layer LIF parameters (threshold, decay, refractory period)
Status: research-stage; not benchmarked end-to-end

Files

model-*.safetensors (12 shards) — full weights
config.json — Qwen3.5 architecture config
lif_metadata.json — LIF layer parameters
chat_template.jinja — standard Qwen3.5 chat template

Research context

Part of the SpikingKiki line exploring biologically-inspired sparse activation for energy-efficient inference. Related to the GENIAL framework (see hypneum-lab on GitHub).

Org context

This repo is part of electron-rare, the legacy research org. Production ailiance artifacts (gateway, mascarade LoRAs) live at Ailiance-fr (post-2026-05-11 carve-out).

Downloads last month: 180

Safetensors

Model size

27B params

Tensor type

BF16

Model tree for electron-rare/spikingkiki-27b

Quantizations

2 models