[Error] [carb.gym.plugin] Function fillActorRootStateTensor is not implemented


I’m new to Flex.

Flex does not require convex meshes for collisions, right?
(Flex is able to use triangle meshes directly)

I changed the command line option from’–physix’ to’–flex’ and ran my script(’ Reinforcement Learning Examples’), but I got the following error.
Could you give me some advice?

(rlgpu) @Fossa:~/wsp/isaacgym/python/rlgpu$ python train.py --task=FrankaCabinet --flex
Importing module ‘gym_37’ (/home/wsp/isaacgym/python/isaacgym/_bindings/linux-x86_64/gym_37.so)
Setting GYM_USD_PLUG_INFO_PATH to /home/wsp/isaacgym/python/isaacgym/_bindings/linux-x86_64/usd/plugInfo.json
PyTorch version 1.8.1
Device count 1
Using /home/.cache/torch_extensions as PyTorch extensions root…
Emitting ninja build file /home/.cache/torch_extensions/gymtorch/build.ninja…
Building extension module gymtorch…
Allowing ninja to set a default number of workers… (overridable by setting the environment variable MAX_JOBS=N)
ninja: no work to do.
Loading extension module gymtorch…
Importing module ‘rlgpu_37’ (/home/wsp/isaacgym/python/isaacgym/_bindings/linux-x86_64/rlgpu_37.so)
WARNING: Using Flex with GPU instead of PHYSX!
Setting seed: 7581
Physics Engine: FleX
Physics Device: cuda:0
GPU Pipeline: enabled
MESA-INTEL: warning: Performance support disabled, consider sysctl dev.i915.perf_stream_paranoid=0

num franka bodies: 10
num franka dofs: 9
num cabinet bodies: 5
num cabinet dofs: 4
[Error] [carb.gym.plugin] Function fillActorRootStateTensor is not implemented
[Error] [carb.gym.plugin] Function fillDofStateTensor is not implemented
[Error] [carb.gym.plugin] Function setActorRootStateTensorIndexed is not implemented
[Error] [carb.gym.plugin] Function setDofPositionTargetTensorIndexed is not implemented
[Error] [carb.gym.plugin] Function setDofStateTensorIndexed is not implemented

Driver Version: 460.91.03 CUDA Version: 11.2

Hi @DDPG7,

As the RL examples use the Tensor APIs, they aren’t compatible with Flex, I’m afraid. While we do have a limited set of Tensor APIs available for Flex, they aren’t sufficient to run the RL examples.

I’ll also note that you seem to have a potential Vulkan issue with an integrated GPU being used for rendering. Make sure that you’re able to successfully run vulkaninfo and that you are rendering using the NVIDIA GPU. See Cannot Run the Examples for some info on that.

Take care,

All right.