Skip to main content

A lightweight companion package for the Sim2Sim-OnePass public research release.

Project description

Sim2Sim-OnePass Public Release

This curated layer is the fast path through the repository. It is designed so a new visitor can understand the claim, watch the evidence, inspect the canonical PASS artifacts, and run the shortest validation path without digging through the full development history.

Public Surfaces

This repository is the primary public artifact. It is accompanied by:

  • website: https://anthony-olevester.github.io/Robotics_sim-to-sim_OnePass/
  • PyPI companion package: https://pypi.org/project/sim2sim-onepass/

Those surfaces help people browse the same public claim in different ways. This root README stays repo-first.

Command Capability Levels

Command Category Requirements
info Standalone package install only
quickstart Standalone package install only
repo-map Standalone package install only
results-summary Standalone package install only
visual-index Standalone package install only
docs Standalone package install only
check-env Standalone package install only
quick-sanity Dataset-dependent paired datasets, or --demo for the tiny built-in fixture
rollout-check Dataset-dependent paired datasets, model, norm file, and optional extra sim2sim-onepass[rollout]
alignment-gate Full repo / simulator dependent full curated repo layout plus simulator dependencies
alignment-report Full repo / simulator dependent full curated repo layout plus simulator dependencies and workflow files

Pick Your Route

If you want to... Open this
See the project in one screen RESULTS_SUMMARY.md
Watch the visual evidence first VISUAL_INDEX.md
Run the shortest validation path QUICKSTART.md
Inspect the packaged PASS bundle outputs/canonical_pass/
Understand how the repo is organized REPO_MAP.md
See the public website https://anthony-olevester.github.io/Robotics_sim-to-sim_OnePass/
Install the companion package https://pypi.org/project/sim2sim-onepass/

See It Before You Read It

Triptych preview Rollout figure
Triptych preview Rollout figure

Open the full visual package here:

The Claim In Plain Terms

After enforcing deterministic cross-simulator alignment between PyBullet and MuJoCo, this repo learns a residual next-state correction that:

  • reduces Bullet-to-MuJoCo one-step physical error,
  • remains stable under long-horizon rollout checks,
  • passes holdout and alignment gates,
  • and can be inspected visually through replay videos, triptychs, and overlay plots.

Canonical Story

  1. Deterministic paired plans are executed in PyBullet and MuJoCo.
  2. A strict alignment gate blocks training if reset state, timing, or first-step consistency drift.
  3. A residual model predicts the MuJoCo-minus-Bullet next-state gap.
  4. Hard-mode stress evaluation verifies one-step accuracy, holdouts, and rollout stability.
  5. Behavioral acceptance exports replay videos, triptychs, and overlay plots for inspection.

Canonical Evidence

  • Quantitative anchor: packaged in outputs/canonical_pass/source_stress_report.md and outputs/canonical_pass/source_stress_metrics.json
  • Visual anchor: packaged in outputs/canonical_pass/source_behavioral_report.md and the copied videos and preview images
  • Public-facing copied bundle: outputs/canonical_pass/
  • Provenance map: configs/canonical_sources.json

Exact Quickstart

This public repo is optimized for inspection first. Open QUICKSTART.md for the shortest path through the packaged evidence and for command references preserved from the source workspace.

Companion Package

The repository also publishes a lightweight companion package named sim2sim-onepass.

  • install from PyPI: python -m pip install sim2sim-onepass
  • optional rollout extra: python -m pip install "sim2sim-onepass[rollout]"
  • package notes: PYPI_PACKAGE_NOTES.md

The package is not the full simulator stack. It provides CLI access to docs, results navigation, environment checks, and selected curated utilities around this repo.

What Lives Where

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

sim2sim_onepass-0.1.1.tar.gz (23.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

sim2sim_onepass-0.1.1-py3-none-any.whl (18.9 kB view details)

Uploaded Python 3

File details

Details for the file sim2sim_onepass-0.1.1.tar.gz.

File metadata

  • Download URL: sim2sim_onepass-0.1.1.tar.gz
  • Upload date:
  • Size: 23.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.10

File hashes

Hashes for sim2sim_onepass-0.1.1.tar.gz
Algorithm Hash digest
SHA256 726be9231287d896673a2c92f7d9d2eda72d9253a78ece2a1a9b9c8bb27c460b
MD5 adba97e34a7a9dc2c63039a85315d33c
BLAKE2b-256 be7eb72d4230df96dbf7d2da5d644262b712c43760b2b3e54871f4d0ca778c2a

See more details on using hashes here.

File details

Details for the file sim2sim_onepass-0.1.1-py3-none-any.whl.

File metadata

File hashes

Hashes for sim2sim_onepass-0.1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 e7c3fe14a94f8646a6e62de8b627eb657784a665c69c712588cb7dac1463fc02
MD5 cca55a98b314e3e349ad313f05fa8ace
BLAKE2b-256 01ac774990a838977943e6bc6ccb378f7243c159a623e772d9355fb6b7f0fd99

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page