Datasets:
Download tests/test_tabletop.py from Nabidnur/RoboRACE-Bench: direct link, hf CLI and curl.
- Browser
- Download file 554 Bytes
-
https://huggingface.co/datasets/Nabidnur/RoboRACE-Bench/resolve/main/tests/test_tabletop.py
- Command line
-
hf download hf://datasets/Nabidnur/RoboRACE-Bench/tests/test_tabletop.py
-
curl -L -o test_tabletop.py https://huggingface.co/datasets/Nabidnur/RoboRACE-Bench/resolve/main/tests/test_tabletop.py
554 Bytes
| import pytest | |
| pytest.importorskip("mujoco") | |
| from robotics.sim.tabletop import make_suite # noqa: E402 | |
| def test_suite_balanced_and_renders(): | |
| suite = make_suite(n_per_task=4, seed=3, size=64) | |
| assert len(suite) == 8 and sum(s.success for s in suite) == 4 | |
| assert all(s.image.shape == (64, 64, 3) and s.image.std() > 1 for s in suite) | |
| def test_many_seeds_place_without_error(): | |
| for seed in range(30): # regression: crowded tables used to raise "could not place object" | |
| make_suite(n_per_task=6, seed=seed, size=8, render=False) | |