Download task/single-software-execution/cli/blender/task-43/task-43.json from HongchengGao/EngiWorld-Tasks: direct link, hf CLI and curl.
- Browser
- Download file 2.33 kB
-
https://huggingface.co/datasets/HongchengGao/EngiWorld-Tasks/resolve/main/task/single-software-execution/cli/blender/task-43/task-43.json
- Command line
-
hf download hf://datasets/HongchengGao/EngiWorld-Tasks/task/single-software-execution/cli/blender/task-43/task-43.json
-
curl -L -o task-43.json https://huggingface.co/datasets/HongchengGao/EngiWorld-Tasks/resolve/main/task/single-software-execution/cli/blender/task-43/task-43.json
2.33 kB
| { | |
| "id": "single-software-execution--cli--blender--task-43--ubuntu", | |
| "snapshot": "Blender-4.2.3", | |
| "instruction": "Find the camera orientation that maximizes the silhouette area of `Target`, render `/home/user/Desktop/output/optimal.png`, and save `/home/user/Desktop/answer.blend`. ## Requirements 1. `/home/user/Desktop/answer.blend` exists and opens in Blender. 2. Launch Blender by running `blender` from a terminal. On this machine, use `/opt/blender/blender` for Blender command-line or background runs.\n\nScene contains a `Target` object (mesh) centered near the origin. 3. Scene contains a `Camera` object (type CAMERA) as the active `scene.camera`. 4. `/home/user/Desktop/output/optimal.png` exists. 5. The rendered PNG is exactly 256x256. 6. The submitted render has silhouette area (alpha > 0.5 pixel count) >= 1000 pixels (object is visible, not edge-on). 7. The submitted camera orientation yields a silhouette area within 2% of the true analytical optimum (camera on the +Y axis at distance 5, looking at the origin, orthographic).", | |
| "source": "https://docs.blender.org/manual/en/4.1/; https://docs.blender.org/api/4.1/; https://developer.blender.org/docs/release_notes/4.1/; https://www.loc.gov/preservation/digital/formats/fdd/fdd000507.shtml", | |
| "config": [ | |
| { | |
| "type": "upload_file", | |
| "parameters": { | |
| "files": [ | |
| { | |
| "local_path": "task-43/init_file/scene.blend", | |
| "path": "/home/user/Desktop/scene.blend" | |
| } | |
| ] | |
| } | |
| } | |
| ], | |
| "trajectory": "trajectories/", | |
| "related_apps": [ | |
| "blender" | |
| ], | |
| "artifact_paths": [ | |
| "/home/user/Desktop/answer.blend", | |
| "/home/user/Desktop/output/optimal.png" | |
| ], | |
| "evaluator": { | |
| "postconfig": [ | |
| { | |
| "type": "upload_file", | |
| "parameters": { | |
| "files": [ | |
| { | |
| "local_path": "task-43/eval.py", | |
| "path": "/home/user/Desktop/eval.py" | |
| } | |
| ] | |
| } | |
| } | |
| ], | |
| "func": "exact_match", | |
| "result": { | |
| "type": "vm_command_line", | |
| "command": "python3 /home/user/Desktop/eval.py", | |
| "shell": "true" | |
| }, | |
| "expected": { | |
| "type": "rule", | |
| "rules": { | |
| "expected": "True\n" | |
| } | |
| } | |
| }, | |
| "proxy": false, | |
| "fixed_ip": false, | |
| "possibility_of_env_change": "low" | |
| } | |