# datawhalechina/every-embodied

仅需Python基础，从0构建自己的具身智能机器人；从0逐步构建VLA/OpenVLA/SmolVLA/Pi0， 深入理解具身智能

Repository: https://github.com/datawhalechina/every-embodied
Canonical: https://ross.abutalabs.com/products/every-embodied
Language: Python
License: NOASSERTION
License Family: other
Topics: embodied-agent, embodied-ai, embodied-intelligence, openvla, smolvla, vision-language-action-model
Last push: 2026-08-24T12:50:39+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 35, longevity 45
- inputs: {"age_days": 633, "days_push": 9, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3359, forks 340 (observed 2026-08-28T04:07:58.032960+00:00)

## What it is
An open-source Chinese-language tutorial course by Datawhale that teaches embodied AI from scratch, guiding learners with only basic Python knowledge to build their own embodied intelligent robots. It covers building vision-language-action models such as VLA, OpenVLA, SmolVLA, and Pi0 step by step to develop deep understanding of embodied intelligence.

## Use cases
- learn embodied AI from scratch with only Python basics
- build my own embodied intelligence robot
- understand how VLA and OpenVLA models work
- step-by-step tutorial for building SmolVLA and Pi0
- get started with vision-language-action models
- find a structured course on embodied intelligence
- reproduce state-of-the-art embodied AI models

## When to choose
- you are a beginner with basic Python wanting a guided path into embodied AI and robotics
- you want to understand and implement VLA-family models like OpenVLA, SmolVLA, and Pi0 from the ground up
- you prefer structured, team-based open-source learning materials in Chinese

## When to avoid
- you need production-ready robot control software rather than educational material
- you need an English-language course or polished commercial documentation
- you are looking for a deployable robotics framework or SDK instead of tutorials

## Facets
- artifact type: learning-resource
- maturity: active
- function: machine-learning, deep-learning, agent-framework, simulation, developer-tools
- domain: artificial-intelligence, robotics, deep-learning, tutorials, education
- platform: python, cross-platform
- tags: embodied-ai, vision-language-action, vla, openvla, smolvla, pi0, robotics, open-source-course, datawhale, chinese-language

## Member repositories
- datawhalechina/every-embodied (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:58.032960+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 2026-08-29T18:40:53.123268+00:00, confidence not recorded.
  - readme: https://github.com/datawhalechina/every-embodied (fetched 2026-08-28T04:07:58.032960+00:00, sha 23e832ead4c2)
- Data as of 2026-08-30T08:39:29.467469+00:00.