Instructions to use Archi-medes/LabGuide_Preview with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Archi-medes/LabGuide_Preview with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Archi-medes/LabGuide_Preview",
	filename="LabGuide_Preview_LFM2.gguf",
)

llm.create_chat_completion(
	messages = "{\n    \"question\": \"What is my name?\",\n    \"context\": \"My name is Clara and I live in Berkeley.\"\n}"
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Archi-medes/LabGuide_Preview with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Archi-medes/LabGuide_Preview
# Run inference directly in the terminal:
llama-cli -hf Archi-medes/LabGuide_Preview

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Archi-medes/LabGuide_Preview
# Run inference directly in the terminal:
llama-cli -hf Archi-medes/LabGuide_Preview

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf Archi-medes/LabGuide_Preview
# Run inference directly in the terminal:
./llama-cli -hf Archi-medes/LabGuide_Preview

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf Archi-medes/LabGuide_Preview
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Archi-medes/LabGuide_Preview

Use Docker

docker model run hf.co/Archi-medes/LabGuide_Preview

LM Studio
Jan
Ollama
How to use Archi-medes/LabGuide_Preview with Ollama:
```
ollama run hf.co/Archi-medes/LabGuide_Preview
```

Unsloth Studio new

How to use Archi-medes/LabGuide_Preview with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Archi-medes/LabGuide_Preview to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Archi-medes/LabGuide_Preview to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Archi-medes/LabGuide_Preview to start chatting

Pi new

How to use Archi-medes/LabGuide_Preview with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Archi-medes/LabGuide_Preview

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "LabGuide_Preview"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Archi-medes/LabGuide_Preview with Docker Model Runner:
```
docker model run hf.co/Archi-medes/LabGuide_Preview
```

Lemonade

How to use Archi-medes/LabGuide_Preview with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Archi-medes/LabGuide_Preview

Run and chat with the model

lemonade run user.LabGuide_Preview-{{QUANT_TAG}}

List all available models

lemonade list

LabGuide Preview Model

Model Summary

The LabGuide Preview Model is a demonstration release built entirely with Madlab, using its synthetic dataset generator and training workflow.
It is based on LiquidAI/LFM2-700M, adapted to showcase Madlab’s end-to-end capabilities for dataset creation, model training, and assistant deployment.

This model illustrates how applications can leverage Madlab to train their own assistants in a reproducible and accessible way.
It is not intended for production use, but rather as a preview for contributors, collaborators, and community feedback.

Training Data

Source: Synthetic dataset generated entirely with Madlab’s dataset generator.
Purpose: Designed to demonstrate Madlab’s ability to produce structured, reproducible training data.
Scope: Preview-scale dataset, not representative of real-world or production-ready corpora.

Training Process

Framework: Madlab training pipeline.
Base Model: LiquidAI/LFM2-700M.
Workflow: Synthetic dataset generation → Madlab training loop → Magic Judge Evaluation → Preview model release.
Objective: Demonstrate Madlab’s integrated workflow for building application-specific assistants.

Intended Uses

Contributor onboarding and workflow validation.
Demonstration of Madlab’s synthetic dataset generator and training pipeline.
Benchmarking and experimentation in controlled preview settings.

Limitations

Demo-only: Not suitable for production or deployment in real-world applications.
Synthetic data: Training data is fully synthetic and may not reflect natural language distributions.
Preview scale: Model performance is illustrative, not optimized for accuracy or robustness.

Ethical Considerations

This model is provided for demonstration and educational purposes.
It should not be used in applications where accuracy, safety, or reliability are critical.
Contributors are encouraged to treat outputs as illustrative examples only.

Acknowledgements

Base model: LiquidAI/LFM2-700M.
Built and trained with Madlab.

Downloads last month: 14

Safetensors

Model size

0.7B params

Tensor type

F32

Model tree for Archi-medes/LabGuide_Preview

Base model

LiquidAI/LFM2-700M

Finetuned

(19)

this model

Archi-medes
/

LabGuide_Preview