Instructions to use ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B with Transformers:

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

pipe = pipeline("text-generation", model="ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B")
model = AutoModelForCausalLM.from_pretrained("ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B", device_map="auto")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B

SGLang

How to use ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B 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 "ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B" \
    --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": "ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B",
		"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 "ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B" \
        --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": "ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B with Docker Model Runner:
```
docker model run hf.co/ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

ZeroXClem/Astral-Fusion-Neural-Happy-L3.1-8B

Astral-Fusion-Neural-Happy-L3.1-8B is a hybrid model blending magic, creativity, and dynamic storytelling. It excels in instruction-following, immersive roleplaying, and magical narrative generation. This model fuses the finest qualities of Astral-Fusion (Llama 3.0), NIHAPPY (Llama 3.1), and NeuralMahou (Llama 3.0). ✨🚀

🌟 Family Tree

This model is a merger of the following:

ProdeusUnity/Astral-Fusion-8b-v0.0: A celestial, multi-purpose model that excels in instruction-following and role-playing. It merges storytelling and instructive content with a cosmic, mysterious flair.
Arkana08/NIHAPPY-L3.1-8B-v0.09: A dynamic role-playing model, ideal for interactive storytelling, balancing creativity and intricate reasoning to produce compelling narratives.
lemon07r/llama-3-NeuralMahou-8b: Infuses magical storytelling and creative imagination, perfect for generating fantasy-rich, immersive experiences with precision.

🧬 Model Lineage

A: ProdeusUnity/Astral-Fusion-8b-v0.0

A fusion of several instructive and storytelling models like Celeste-Stable and Sao10K L3-8B-Stheno, this model provides a well-rounded approach to instruction-following while also being capable of generating cosmic-inspired narratives.
Strong in roleplaying and narrative creation, with a focus on maintaining coherence in long-form outputs.

B: Arkana08/NIHAPPY-L3.1-8B-v0.09

Merged with a focus on dynamic storytelling and rule-following. NIHAPPY excels in interactive roleplay situations where logical reasoning meets creative imagination.
A perfect balance between storytelling and structured reasoning, making it ideal for immersive game worlds, character-driven dialogues, and dynamic experiences.

C: lemon07r/llama-3-NeuralMahou-8b

This model brings in a magical essence with its creative and fantasy-driven approach to language generation. Infused with NeuralMahou's fantasy-themed elements, it’s perfect for magical narratives and imaginative storytelling.
Trained to handle both technical and creative tasks, making it a versatile addition to this fusion.

🛠️ Merge Details

This model was merged using the Model Stock merge method, carefully integrating the unique strengths of all three models to create a cohesive, powerful system. Here’s the configuration used:

base_model: ProdeusUnity/Astral-Fusion-8b-v0.0
dtype: bfloat16
merge_method: model_stock
slices:
  - sources:
      - layer_range: [0, 32]
        model: Arkana08/NIHAPPY-L3.1-8B-v0.09
      - layer_range: [0, 32]
        model: lemon07r/llama-3-NeuralMahou-8b
      - layer_range: [0, 32]
        model: ProdeusUnity/Astral-Fusion-8b-v0.0

🎯 Key Features & Capabilities

1. Instruction Following & Reasoning:

Rooted in Astral-Fusion, this model excels in precise instruction-following and handling logical tasks. It maintains a celestial vibe while delivering clear, structured outputs.

2. Roleplay & Storytelling:

With the NIHAPPY model integrated, this model is perfect for crafting dynamic narratives, handling complex character dialogues, and maintaining immersive storylines in roleplaying contexts.

3. Magic & Imagination:

Thanks to NeuralMahou, this model shines in creative, magical storytelling. Whether you're crafting fantasy worlds or seeking magical responses, this model delivers imagination-rich outputs that blend logic and creativity seamlessly.

🚀 Use Case & Applications

Immersive Storytelling: Create fantasy-rich narratives, generate dynamic roleplaying characters, and write cosmic-inspired tales that blend magic, science, and imagination.
Interactive Roleplay: With NIHAPPY’s dynamic capabilities, the model is perfect for AI-based NPC dialogues, interactive worlds, and even game development.
Instructional & Long-Form Texts: Perfect for writing long-form instructional content, this model can switch effortlessly between creative and instructive modes.