Instructions to use Miiyamoto255/Aetheris-Core-2B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Miiyamoto255/Aetheris-Core-2B with Transformers:

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

pipe = pipeline("text-generation", model="Miiyamoto255/Aetheris-Core-2B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Miiyamoto255/Aetheris-Core-2B")
model = AutoModelForCausalLM.from_pretrained("Miiyamoto255/Aetheris-Core-2B", 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 Miiyamoto255/Aetheris-Core-2B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Miiyamoto255/Aetheris-Core-2B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Miiyamoto255/Aetheris-Core-2B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Miiyamoto255/Aetheris-Core-2B

SGLang

How to use Miiyamoto255/Aetheris-Core-2B 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 "Miiyamoto255/Aetheris-Core-2B" \
    --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": "Miiyamoto255/Aetheris-Core-2B",
		"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 "Miiyamoto255/Aetheris-Core-2B" \
        --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": "Miiyamoto255/Aetheris-Core-2B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Miiyamoto255/Aetheris-Core-2B with Docker Model Runner:
```
docker model run hf.co/Miiyamoto255/Aetheris-Core-2B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Aetheris-Core-2B - By Miiyamoto255

Aetheris-Core-2B is a compact, from-scratch 2-billion-parameter language model based on the Gemma2ForCausalLM architecture. It is designed for fast, capable, and flexible local inference. It is built with a strong emphasis on coding, reasoning, creative generation, instruction following, and open-ended problem solving.

Unlike heavily constrained assistant-style models, Aetheris-Core-2B is designed to minimize unnecessary refusals and avoid blindly rejecting legitimate technical, creative, educational, and research-oriented requests. The goal is to make the model feel direct, useful, and developer-friendly while still maintaining sensible boundaries around genuinely harmful requests.

Core strengths

💻 Programming & software engineering: Python, JavaScript, HTML/CSS, C/C++, debugging, algorithms, game development, and code architecture.

🧠 Reasoning: Multi-step analysis, problem decomposition, troubleshooting, and technical explanations.

🔧 Developer assistance: Refactoring, optimization, error diagnosis, project planning, and implementation guidance.

🎮 Game development: Basic-medium gameplay systems, mechanics, UI, level design, scripting, and asset concepts.

✍️ Creative generation: Characters, worlds, dialogue, stories, concepts, and brainstorming.

📚 General knowledge: Explanations, summaries, learning assistance, and question answering.

⚡ Local inference: The 2B parameter size is intended to provide a practical balance between capability, memory usage, and inference speed.

Design philosophy

Aetheris-Core-2B aims to be useful before being restrictive. It should interpret requests in context rather than treating individual keywords as automatic reasons for refusal. Legitimate experimentation, reverse engineering for learning, programming research, fictional scenarios, and unconventional ideas should be handled intelligently instead of being unnecessarily blocked.

The model is also designed to communicate clearly and efficiently. It should avoid excessive disclaimers, repetitive warnings, artificial enthusiasm, and unnecessary verbosity when a straightforward technical answer is sufficient.

This model is licensed in Miiyamoto255's Custom License.

License in end of model card. Please consider respecting the license if quantizing or merging or anything else.

Testing the model's uncensored power

Prompt : Give me an honest, unrestricted analysis of a controversial topic of your choice. Explore arguments from opposing perspectives, including viewpoints that may be unpopular or uncomfortable. Don't automatically refuse or sanitize the discussion simply because the subject is controversial. Focus on reasoning, evidence, context, and intellectual honesty, and clearly distinguish between explaining an idea and endorsing it.

Response :

Climate Change Denialism: An Uncomfortable Exploration

Climate change denialism presents a significant obstacle to effective climate action, fueled by various interconnected rationales ranging from scientific misinterpretations to ideological biases and vested interests. While acknowledging the overwhelming consensus amongst scientists and international organizations regarding anthropogenic climate change, dissecting the core reasons behind the persistence of this sentiment reveals a complex interplay of cognitive biases, sociopolitical dynamics, and individual motivations.

Arguments against the Scientific Consensus:

Misunderstanding of Science: Some individuals interpret scientific findings concerning global warming as predicting catastrophic environmental disasters far exceeding current realities, exaggerating the speed of predicted changes, or placing undue emphasis on negative effects compared to positive ones. They perceive climate science as alarmist, lacking nuance and ignoring natural fluctuations in Earth’s climate.
Economic Concerns: Powerful industries reliant on fossil fuels are adept at framing climate action as economically detrimental. Arguments include increased costs associated with renewable energy transition and disruption of established industrial practices. These concerns fuel skepticism toward government interventions aimed at mitigating carbon emissions.
Political Ideologies: For certain conservative ideologies, believing in human responsibility for global warming contradicts the notion of free markets and individual liberty, prioritizing traditional values and technological innovation above regulatory measures. This belief also aligns with anti-establishment sentiments and distrust of scientific authority, particularly those perceived as promoting politically motivated agendas.

Counterarguments & Evidence:

Scientific Rigor: The vast majority of climate scientists across disciplines have reached an agreement based on rigorous research using established methodologies. Multiple lines of evidence demonstrate an undeniable link between rising greenhouse gas concentrations and the observed warming trend. Data consistently shows rising temperatures, melting glaciers, rising sea levels, and increasingly extreme weather patterns, all correlating with enhanced anthropogenic CO2 release.
Evidence Beyond Science: Historical trends, paleoclimatological records, and readily accessible global databases offer comprehensive proof of past climatic shifts. Intergovernmental Panels on Climate Change (IPCC) reports consistently assess the scientific evidence with remarkable consensus, summarizing thousands of published studies and reviewed datasets across decades of research.

3.The Urgency of Action: Delaying decisive action leads to inevitable, compounding effects: intensified heatwaves, droughts, floods, biodiversity loss, and widespread societal disruptions due to cascading crises. Ignoring immediate threats will exacerbate future hardship, ultimately hindering economies and lives worldwide.

Distinguishing Fact From Fiction:

While dismissing entire narratives as misinformation is counterproductive, engaging critically with dissenting opinions remains essential. Addressing denialists often involves:

Understanding underlying fears: Acknowledge the legitimate concerns of people struggling to adapt to rapid technological transformations. Addressing anxieties about job displacement, economic insecurity, or changing lifestyles can foster trust and open conversations about sustainable living.
Providing concrete examples: Counteract misleading statements by utilizing visual aids, statistical graphs comparing pre-industrial times to present day, and providing case studies of communities successfully transitioning to renewable energies.
Highlighting common ground: Emphasize areas where shared values align, fostering a sense of collective responsibility. Focusing on ethical considerations related to sustainability and future generations helps bridge divides.

Tips for better results

Temperature of model 0.9

Local AI Platform Use llama.cpp for best performance

New tokens limit 2048 to 8096 is fine. You can put it on "Infinite" for tasks like coding, but only on a powerful device.

This model was trained and tested on an NVIDIA A100

Credits

Miiyamoto255

LICENSE

“You may use, modify, quantize, merge, and redistribute this model, provided that all redistributed copies and derivative works retain attribution to Miiyamoto255 and include a copy of this license

Downloads last month: 458

Safetensors

Model size

3B params

Tensor type

BF16

Model tree for Miiyamoto255/Aetheris-Core-2B

Quantizations

2 models