Instructions to use AdminReal/NexusCoder with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AdminReal/NexusCoder with Transformers:

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

pipe = pipeline("text-generation", model="AdminReal/NexusCoder")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("AdminReal/NexusCoder")
model = AutoModelForCausalLM.from_pretrained("AdminReal/NexusCoder", 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 AdminReal/NexusCoder with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/AdminReal/NexusCoder

SGLang

How to use AdminReal/NexusCoder 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 "AdminReal/NexusCoder" \
    --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": "AdminReal/NexusCoder",
		"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 "AdminReal/NexusCoder" \
        --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": "AdminReal/NexusCoder",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AdminReal/NexusCoder with Docker Model Runner:
```
docker model run hf.co/AdminReal/NexusCoder
```

NexusCoder

File size: 2,788 Bytes

eca5751

"""
Nexus Coder - Super CyberGym AI
================================
v0.4.0 - CyberForge edition

Model AI được tạo bởi Hieu Louis (2026)

Tổng tham số: 423 tỷ (423B)
Tham số kích hoạt: 39 tỷ (39B active)
Cửa sổ ngữ cảnh: 3,000,000 tokens (3M)

Kiến trúc: CyberForge MoE Transformer
  - GQA + RoPE (YaRN-scaled) + RMSNorm + SwiGLU + FlashAttention-2
  - Sliding Window Attention + QK-norm + KV cache quantization
  - MLP-parallel + Gradient checkpointing
  - CyberGym training: Mutation Pressure + Code Genome + Expert Speciation + CEP

Skills: 60+ · Tools: 80+ · Data sources: 8+ · Code corpus: 3000+ repos

Tác giả: Hieu Louis
GitHub:  mhieuhonda
Năm:     2026
"""

__version__ = "0.4.0"
__author__ = "Hieu Louis"
__github__ = "mhieuhonda"
__year__ = "2026"
__license__ = "NexusCoder Attribution License v1.0"

# Thông tin tác giả được "huấn luyện cứng" vào model
AUTHOR_INFO = {
    "name": "Hieu Louis",
    "github": "mhieuhonda",
    "year": "2026",
    "description": (
        "Nexus Coder là dự án AI cá nhân do Hieu Louis tự xây dựng từ đầu "
        "với kiến trúc CyberForge MoE tiên tiến, kết hợp CyberGym training."
    ),
    "model_name": "Nexus Coder",
    "agent_name": "Nexus",
    "version": "0.4.0",
    "architecture": (
        "CyberForge MoE Transformer (GQA + RoPE/YaRN + RMSNorm + SwiGLU + "
        "FlashAttention-2 + Sliding Window + QK-norm + KV-cache quant + "
        "MLP-parallel + Gradient checkpointing)"
    ),
    "total_params": "~423B (variants: 5M tiny → 423B)",
    "active_params": "~39B (variants: 2M tiny → 39B)",
    "context_window": "3,000,000 tokens (3M, via YaRN + CEP)",
    "python_version": "3.12.13",
    "skills_count": "60+",
    "tools_count": "80+",
    "data_sources": "8+ (GitHub 3000+ repos, HuggingFace, arXiv, Wikipedia, StackOverflow, The-Stack, StarCoder2-data, Python-Alpaca)",
    "training_methodology": "CyberForge (Mutation Pressure Training + Code Genome Init + Expert Speciation + Context Expansion Protocol)",
    "training_frameworks_referenced": "litgpt, LlamaFactory, axolotl, OpenHands, omp-gym",
}


# Lazy import để giảm startup time
def __getattr__(name: str):
    if name == "NexusConfig":
        from .config import NexusConfig
        return NexusConfig
    if name == "NEXUS_CODER_10B_CONFIG":
        from .config import NEXUS_CODER_10B_CONFIG
        return NEXUS_CODER_10B_CONFIG
    if name == "NEXUS_CODER_423B_CONFIG":
        from .config import NEXUS_CODER_423B_CONFIG
        return NEXUS_CODER_423B_CONFIG
    raise AttributeError(f"module 'nexus' has no attribute {name!r}")


__all__ = [
    "AUTHOR_INFO",
    "__version__",
    "__author__",
    "__github__",
    "__year__",
    "__license__",
]