Datasets:
Download masters/myosuite/baoding/meta.json from MLL-Lab/embodied-exam-data: direct link, hf CLI and curl.
- Browser
- Download file 2.19 kB
-
https://huggingface.co/datasets/MLL-Lab/embodied-exam-data/resolve/main/masters/myosuite/baoding/meta.json
- Command line
-
hf download hf://datasets/MLL-Lab/embodied-exam-data/masters/myosuite/baoding/meta.json
-
curl -L -o meta.json https://huggingface.co/datasets/MLL-Lab/embodied-exam-data/resolve/main/masters/myosuite/baoding/meta.json
2.19 kB
| { | |
| "suite": "myosuite", | |
| "task": "baoding", | |
| "family": "baoding", | |
| "source": "agent-success:claude_code-claude_opus_5_5-default/myosuite-baoding-i00__J6BJVtG", | |
| "source_detail": "Claude Opus 5.5 (Claude Code), trial myosuite-baoding-i00__J6BJVtG, 2026-09-28 (privileged mode; validated as a positive example in docs/benchmarks/myosuite.md)", | |
| "renderer": "Blender 4.2.23 LTS Cycles path tracing, OptiX on an RTX 4090, adaptive sampling (threshold 0.02, up to 96 spp), OpenImageDenoise with albedo and normal passes, motion blur 180 deg, AgX Medium High Contrast; indoor studio cove (no sky) with key/fill/rim/top area lights; render 2560x1440 at 30 fps; every geom is the simulator's own mesh at its recorded MuJoCo world pose (poses recorded per control step from the task's own replay, keyed at their exact times, linear between steps)", | |
| "rendering": "path-tracing", | |
| "width": 2560, | |
| "height": 1440, | |
| "fps": 30, | |
| "frames": 228, | |
| "duration_s": 7.6, | |
| "speed": 1.0, | |
| "success_check": true, | |
| "success_check_how": "the trajectory was replayed twice from the task's frozen instance in the task's own image with the exam's task repository's myosuite tooling; both replays ended in success at the same step (success), their per-step state digests were identical at every step, and the final state hash equals the benchmark's recorded one", | |
| "replay": { | |
| "control_steps": 215, | |
| "control_dt_s": 0.025, | |
| "sim_time_s": 5.375, | |
| "final_state_sha": "1a84408eaae929ce", | |
| "expected_final_state_sha": "1a84408eaae929ce", | |
| "replays_identical_per_step": true, | |
| "trajectory_sha256": "0b7b6599cbadd574052a46684016d877ac2ff324f5970ace9b7076c1e039bc88", | |
| "task_dirs": [ | |
| "myosuite-baoding-i00-privileged", | |
| "myosuite-baoding-i00-standard" | |
| ] | |
| }, | |
| "holds": { | |
| "start_s": 0.8, | |
| "end_s": 1.4, | |
| "note": "first and last recorded states held; the camera keeps moving, no extra simulation" | |
| }, | |
| "camera": { | |
| "az_deg": 90.0, | |
| "el_deg": 12.0, | |
| "lens_mm": 85.0, | |
| "target": [ | |
| -0.2738, | |
| -0.4544, | |
| 1.4161 | |
| ], | |
| "distance_m": 1.3981 | |
| }, | |
| "poster_frame": 186, | |
| "image": "eai-myosuite:0.1.1@sha256:6fd9c164bfe3d031ce0ee0238d4a1b4a985c2c99fa1009069a47cf5b01b88c09", | |
| "created": "2026-10-17T07:59:02Z" | |
| } |