Repository navigation
Build and publish the first offline cable training product - #9
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The simulation proposal now has a runnable CPU product: CableEndReach-v0 supplies seeded native MuJoCo cable dynamics, bounded endpoint forces, continuous-dwell verification, complete state continuation and explicit invalid-run reporting. An offline CLI tunes a small feedback controller, exports hashed simulation-only skill packs and saves an interactive replay. This release publishes the exact source and measured evidence as the seventh research entry.
Frozen evaluation: selected and baseline controllers each pass 10/10 seeds; zero force and early release each pass 0/10. The selected controller takes slightly more simulated completion time than the baseline. The report discloses repeated evaluation during development, timing boundaries, ideal attachment and privileged observations. No physical calibration, gripping, GPU throughput or arbitrary-robot qualification is claimed.
Validation: 23 Python tests pass; a native macOS network-denial run passes 3/3 with a libc EPERM self-test; unpacked pack code reproduces three complete trajectories and final diagnostics exactly. The site build, validator and all 21 Node tests pass, including source/manifest hashes and static links. Replay controls were verified in the browser.
The public archive excludes virtual environments and development runs. Dependencies retain their own terms; original product code is MIT. Publishing uses the existing dedicated research Worker and does not change the main DVIDIA site.