Skip to content

Conversation

@rrzhang139
Copy link

Running GPU benchmark output:

python -m mani_skill.examples.benchmarking.gpu_sim -e "PegHitBall-v1" -n=4096 -o=state
# -------------------------------------------------------------------------- #
Task ID: PegHitBall-v1, 4096 parallel environments, sim_backend=physx_cuda
obs_mode=state, control_mode=pd_joint_delta_pos
render_mode=rgb_array, sensor_details=RGBD(128x128)
sim_freq=120, control_freq=60
observation space: Box(-inf, inf, (4096, 42), float32)
(single) action space: Box(-1.0, 1.0, (8,), float32)
# -------------------------------------------------------------------------- #
start recording env.step metrics
env.step: 70894.176 steps/s, 17.308 parallel steps/s, 1000 steps in 57.776s
    CPU mem: 11058.938 MB, GPU mem: N/A
start recording env.step+env.reset metrics
env.step+env.reset: 57858.325 steps/s, 14.126 parallel steps/s, 1000 steps in 70.794s
    CPU mem: 11081.500 MB, GPU mem: N/A

Wandb link: https://wandb.ai/rzhang139/ManiSkill/runs/w4l5z58x/overview

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant