Skip to content

Shadow Dip v1.0.0 reference dataset

Choose a tag to compare

View all tags
@cristpierce cristpierce released this 14 Aug 23:07
shadow-dip-v1.0.0
684c6e8
This commit was signed with the committer’s verified signature.
cristpierce
SSH Key Fingerprint: kmKHxkZstpKLS25Vd6a8HMSlB4z1XoFAsIrJAUh3jYs
Verified
Learn about vigilant mode.

Team-authored synthetic G1 reference dataset for the Ultimate Bots Trial 03 SuperSONIC challenge.

Evidence:

  • 30 motions: 22 train/rehearsal, 4 selection-validation, 4 untouched final-test
  • 30/30 reference validation passes, 0 failures, 0 warnings, 0 self-contacts
  • Public Ubuntu/Python 3.11 regeneration and cross-platform semantic comparison passed
  • Source commit: 684c6e8
  • Manifest SHA-256: 1b2045380e09e6276c5ac4ff4c2bb1c7bd5903a974940f9928d7351b5f90a5d1
  • Validation SHA-256: 5aedeedee8d775c34c0b5c67f235591829da281cfa9385b8b8a15b8c10a6b999
  • Archive SHA-256: 94099f031b8a0b5ea36c809e705f77088342a6b54d73f9735508b146841c1370
  • Reference video SHA-256: d9f6f4284e5cecbc80349d050786b2c876a26f1a93dd4ba6e3da8f9149efe0c3

The MP4 is a labelled kinematic target reference, not policy output. This release contains no trained checkpoint, ONNX policy, stock/fine-tuned simulator result, or BONES-SEED motion/derivative. Those are not claimed.

Motion Data by Bones Studio.

Assets 4
Loading