🏔️ Alpamayo 1

This Repository is a fork of the original NVIDIA Alpamayo 1 project with an example zitification of the original inference test. This README has been modified with instructions to setup and run the zitified version over a NetFoundry Network.

Bridging Reasoning and Action Prediction for Generalizable Autonomous Driving

Note: Following the release of NVIDIA Alpamayo at CES 2026, Alpamayo-R1 has been renamed to Alpamayo 1.

Setup and Configure the Example

Create or use an existing ziti network with at least one NF hosted edge router. This can be accomplished using the NetFoundry Console.

1. Install prereqs

On an ubuntu 24.04 system with an NVidia GPU that meets the Alpamayo hardware requirements a. install prereqs

sudo apt update
sudo apt upgrade
sudo apt install git
#if not installed
curl -LsSf https://astral.sh/uv/install.sh | sh
export PATH="$HOME/.local/bin:$PATH"
sudo apt-get install -y nvidia-cuda-toolkit  

b. ## Build the Example On the linux system that will run the Client mkdir ~/repos cd repos git clone https://github.com/NVlabs/alpamayo.git

2. Create a client side Netfoundry Idenetity

Create and enroll a ziti identity place the identity json file in the ~/repos/alpamayo/src on the ubuntu system in step 1.

a. alpamayo_client01.json

3. Add a customer hosted edge router e.g. "us-east-1-alpamayo-edge01" to your network

4. Launch the customer hosted edge-router in aws region e.g. us-east-1

a. follow: https://support.netfoundry.io/hc/en-us/articles/ 360016342971-Deployment-Guide-for-AWS-Edge-Routers

5. Create a NF service named "alpqamayo-service" and use wildcard address "*.hf.co" and *.huggingface.co , protocol TCP and 443 as the

port. Assign the router identity e.g. us-east-1-alpamayo-edge01 as the hosting entity and forward address, protocol and port to yes.

5. Create a service policy to bind the identity to the NF service e.g.

6. Create a router policy and with the NF hosted edge-router and the alpamayo_client01 as the identity e.g.

7. Set up the virtual environment

uv venv ar1_venv
source ar1_venv/bin/activate
uv sync --active
uv pip uninstall hf-xet #Remove Hugging Face storage acceleration layer (Bypasses OpenZiti interception)

4. Authenticate with HuggingFace

The model requires access to gated resources. Request access here:

• 🤗 Physical AI AV Dataset
• 🤗 Alpamayo Model Weights

Then authenticate:

hf auth login

Get your token at: https://huggingface.co/settings/tokens

Running Inference

Test script

NOTE: This script will download both some example data (relatively small) and the model weights (22 GB). The latter can be particularly slow depending on network bandwidth. The script will download the model over openziti and the 22GB will be sent over your Cloud/NetFoundry network so be aware of associated data costs. Also note the model will be cached in ~/.cache/huggingface/hub in sub-directories after the first download.

python3 src/alpamayo_r1/ztest_inference.py --ziti-identity alpamayo_client01.json

In case you would like to obtain more trajectories and reasoning traces, please feel free to change the num_traj_samples=1 argument to a higher number (Line 60).

Interactive notebook

We provide a notebook with similar inference code at notebook/inference.ipynb.

Project Structure

alpamayo/
├── notebook/
│   └── inference.ipynb                  # Example notebook
├── src/
│   └── alpamayo_r1/
│       ├── action_space/
│       │   └── ...                      # Action space definitions
│       ├── diffusion/
│       │   └── ...                      # Diffusion model components
│       ├── geometry/
│       │   └── ...                      # Geometry utilities and modules
│       ├── models/
│       │   ├── ...                      # Model components and utils functions
│       ├── __init__.py                  # Package marker
│       ├── config.py                    # Model and experiment configuration
│       ├── helper.py                    # Utility functions
│       ├── load_physical_aiavdataset.py # Dataset loader
│       ├── test_inference.py            # Inference test script
        ├── ztest_inference.py           # zitified Inference test script
├── pyproject.toml                       # Project dependencies
└── uv.lock                              # Locked dependency versions

Troubleshooting

Flash Attention issues

The model uses Flash Attention 2 by default. If you encounter compatibility issues:

# Use PyTorch's scaled dot-product attention instead
config.attn_implementation = "sdpa"

License

Apache License 2.0 - see LICENSE for details.

Disclaimer

Alpamayo 1 is a pre-trained reasoning model designed to accelerate research and development in the autonomous vehicle (AV) domain. It is intended to serve as a foundation for a range of AV-related use cases-from instantiating an end-to-end backbone for autonomous driving to enabling reasoning-based auto-labeling tools. In short, it should be viewed as a building block for developing customized AV applications.

Important notes:

• Alpamayo 1 is provided solely for research, experimentation, and evaluation purposes.
• Alpamayo 1 is not a fully fledged driving stack. Among other limitations, it lacks access to critical real-world sensor inputs, does not incorporate required diverse and redundant safety mechanisms, and has not undergone automotive-grade validation for deployment.

By using this model, you acknowledge that it is a research tool intended to support scientific inquiry, benchmarking, and exploration—not a substitute for a certified AV stack. The developers and contributors disclaim any responsibility or liability for the use of the model or its outputs.