CTR PI0.5 Pick Dual Bottles — CTR (10K)
This public inference checkpoint is the verified final checkpoint from CTR training Experiment E527, which is linked from the core experiment data table.
- Architecture: PI0.5 LoRA
- Task: RoboTwin Pick Dual Bottles
- Training variant: CTR
- Final optimizer update: 10000
- Dataset: Shiki42/ctr-pick-dual-bottles-uniform50-20260917
- Dataset revision: 017086144b5d9bfa0ec72bdd8666a8f25688200a
- Source Experiment: E527
Uses only the user-reviewed, revision-pinned CTR IdleMask dataset approved by E362. Older CTR IdleMask checkpoints remain deprecated.
The repository contains inference parameters, processors/assets, resolved configuration, and a generated SHA256SUMS manifest. Optimizer state, data-loader state, logs, credentials, and training caches are intentionally excluded.
Results remain governed by the source Experiment and its separate evaluation Experiments. Publication does not upgrade a reported or audit-pending result to archived/confirmed.