| { |
| "schema_version": 1, |
| "kind": "dvidia.training-release-ledger", |
| "repository": "https://github.com/Dvidia-Inference/dvidia-training", |
| "source_commit": "1718c3b7791114649f037150b08153927d5eff5b", |
| "code_license": "MIT", |
| "footage_license": "MIT", |
| "footage_origin": "Authored synthetic schematic recordings and native simulator telemetry; no human or private footage.", |
| "pilot_scope": "Small CPU software pilot; one exact native held-out scene, no physical robot qualification.", |
| "runs": [ |
| { |
| "id": "lab-native-integration-first", |
| "status": "failed_export", |
| "reason": "New trainer initially wrote movement_model.json while the orchestrator expected movement_head.json; no capsule or completed download was produced.", |
| "test_use": "Synthetic fixture verification; no architecture or weight tuning followed this integration failure." |
| }, |
| { |
| "id": "standalone-visual", |
| "status": "complete", |
| "path": "visual-training/run.json", |
| "evidence": "Clean base package install without MuJoCo; actual encoded clips and verified data-only result." |
| }, |
| { |
| "id": "standalone-native", |
| "status": "complete", |
| "path": "native-training/run.json", |
| "evidence": "Optional engine environment; actual aligned native supervision, candidate capsule, separate one-scene qualification and identical-target open-jaw control." |
| } |
| ], |
| "qualification_receipt": "qualification/qualification.json", |
| "hardware_minimum_verified": false, |
| "historical_studies": "Earlier releases and frozen studies were preserved; these fixtures are synthetic product integration checks.", |
| "software_checks": { |
| "tested_runtime_commit": "d1f0cbcfb296432b4ad0338dfd75e3f285b62d84", |
| "workflow_url": "https://github.com/Dvidia-Inference/dvidia-training/actions/runs/37740260434", |
| "conclusion": "success", |
| "jobs_passed": 8, |
| "visual_platforms": [ |
| "Linux", |
| "macOS", |
| "Windows" |
| ], |
| "python_versions": [ |
| "3.12", |
| "3.13" |
| ], |
| "native_arm_scope": "Linux synthetic one-scene pilot", |
| "container_scope": "Linux Docker actual example/training/inspection under --network none", |
| "release_changes_after_test": "Documentation only; runtime and tests unchanged." |
| }, |
| "release_url": "https://github.com/Dvidia-Inference/dvidia-training/releases/tag/v0.1.0" |
| } |
|
|
