# robodhruv/visualnav-transformer

Official code and checkpoint release for mobile robot foundation models: GNM, ViNT, and NoMaD.

Repository: https://github.com/robodhruv/visualnav-transformer
Canonical: https://ross.abutalabs.com/products/visualnav-transformer
Homepage: http://general-navigation-models.github.io
Language: Python
License: MIT
License Family: permissive
Last push: 2024-09-15T06:36:11+00:00

## Health v2 (maintenance only)
Score: 19/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 83
- inputs: {"age_days": 1170, "days_push": 717, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1294, forks 200 (observed 2026-08-28T04:04:16.383370+00:00)

## What it is
Official code and pre-trained checkpoints for the GNM, ViNT, and NoMaD family of general-purpose goal-conditioned visual navigation policies for mobile robots. It includes training and fine-tuning scripts, dataset processing tools, and ROS-based deployment scripts for robots like TurtleBot2 and LoCoBot.

## Use cases
- deploy a pretrained visual navigation policy on a mobile robot
- fine-tune a navigation foundation model on custom robot trajectory data
- convert rosbag demo trajectories into topological maps for navigation
- run zero-shot goal-directed navigation across different robot embodiments
- use NoMaD for autonomous exploration of an environment
- train goal-conditioned navigation models from images

## When to choose
- you need a ready-to-deploy visual navigation policy for a ROS-supported mobile robot
- you want to fine-tune a cross-embodiment navigation foundation model on your own data
- you are doing research on generalist robot navigation policies

## When to avoid
- you need navigation for autonomous cars or outdoor vehicles with GPS/HD maps
- your robot does not support ROS or camera-based control
- you need a production-grade navigation stack with SLAM and obstacle avoidance guarantees

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, robotics, computer-vision, llm-training
- domain: robotics, machine-learning, autonomous-vehicles, computer-vision
- platform: python
- tags: visual-navigation, foundation-model, goal-conditioned-policy, diffusion-policy, ros, turtlebot, cross-embodiment, pretrained-checkpoints, linux, gpu

## Member repositories
- robodhruv/visualnav-transformer (main) score 19

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:16.383370+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:54:06.438016+00:00, confidence not recorded.
  - readme: https://github.com/robodhruv/visualnav-transformer (fetched 2026-08-28T04:04:16.383370+00:00, sha f8693ff82ea8)
  - homepage: http://general-navigation-models.github.io (fetched 2026-08-29T12:10:55.033919+00:00, sha 8f95f16e9208)
- Data as of 2026-08-30T08:39:29.467469+00:00.