Instructions to use NeuraCraft/Lance-AI with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use NeuraCraft/Lance-AI with Transformers:

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

pipe = pipeline("text-generation", model="NeuraCraft/Lance-AI", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("NeuraCraft/Lance-AI", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use NeuraCraft/Lance-AI with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/NeuraCraft/Lance-AI

SGLang

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

Docker Model Runner
How to use NeuraCraft/Lance-AI with Docker Model Runner:
```
docker model run hf.co/NeuraCraft/Lance-AI
```

Lance AI – We are the Future

🚀 Lance AI is a custom-built text generation model, designed to serve as the foundation for a more advanced AI. Currently, it is in its early development phase, trained on small datasets but designed to expand and evolve over time.

🌟 Key Features

✅ Custom-built architecture (Not based on GPT-2/GPT-3) ✅ Supports Hugging Face's transformers ✅ Small-scale model with room for growth ✅ Lightweight, efficient, and optimized for local and cloud inference ✅ Planned real-time internet access & vision capabilities

📥 Installation & Setup

You can load Lance AI using transformers:

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "NeuraCraft/Lance-AI" tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True)

input_text = "The future of AI is" inputs = tokenizer(input_text, return_tensors="pt") outputs = model.generate(**inputs, max_length=250)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

🛠 How to Use Lance AI

1️⃣ Direct Text Generation

Lance AI can generate text from simple prompts:

prompt = "In the year 2050, humanity discovered" inputs = tokenizer(prompt, return_tensors="pt") output = model.generate(**inputs, max_length=50)

print(tokenizer.decode(output[0], skip_special_tokens=True))

2️⃣ Fine-tuning for Custom Applications

You can fine-tune Lance AI for your own dataset using Hugging Face’s Trainer API.

from transformers import Trainer, TrainingArguments

training_args = TrainingArguments( output_dir="./lance_ai_finetuned", per_device_train_batch_size=8, per_device_eval_batch_size=8, num_train_epochs=3, save_steps=500 )

trainer = Trainer( model=model, args=training_args, train_dataset=your_dataset, eval_dataset=your_eval_dataset )

trainer.train()

📊 Performance & Evaluation

Lance AI is currently in its early stages, and performance is being actively tested. Initial evaluations focus on: 🔹 Perplexity (PPL) – Measures text coherence 🔹 Text Generation Quality – Manual evaluation for fluency and relevance 🔹 Token Accuracy – Predicts the next token based on input text

✅ Planned Enhancements

🔹 Larger training datasets for improved fluency 🔹 Real-time browsing for knowledge updates 🔹 Vision integration for multimodal AI

🚀 Future Roadmap

Lance AI is just getting started! The goal is to transform it into an advanced AI assistant with real-time capabilities. 📅 Planned Features:

🔜 Larger model with better efficiency

🔜 Internet browsing for real-time knowledge updates

🔜 Image and video generation capabilities

🔜 AI-powered PC automation

🏗 Development & Contributions

Lance AI is being developed by NeuraCraft. Contributions, suggestions, and testing feedback are welcome!

📬 Contact & Updates:

Developer: NeuraCraft

Project Status: 🚧 In Development

Follow for updates: Coming soon

Downloads last month: 1,838

Safetensors

Model size

0.4B params

Tensor type

BF16

Collection including NeuraCraft/Lance-AI

Lance AI

Collection

Discover Lance AI – Our Most Advanced Models • 2 items • Updated 2 days ago