Isaac Lab wrapper's multi-agent state retrieval with gymnasium 1.0

CHANGELOG.md

@@ -29,6 +29,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 - Model state dictionary initialization for composite Gymnasium spaces in JAX
 - Add missing `reduction` parameter to Gaussian model instantiator
 - Optax's learning rate schedulers integration in JAX implementation
+- Isaac Lab wrapper's multi-agent state retrieval with gymnasium 1.0
 
 ### Removed
 - Remove OpenAI Gym (`gym`) from dependencies and source code. **skrl** continues to support gym environments,

skrl/envs/wrappers/jax/isaaclab_envs.py

@@ -207,7 +207,10 @@ def state(self) -> Union[np.ndarray, jax.Array, None]:
         :return: State
         :rtype: np.ndarray, jax.Array or None
         """
-        state = self._env.state()
+        try:
+            state = self._env.state()
+        except AttributeError:  # 'OrderEnforcing' object has no attribute 'state'
+            state = self._unwrapped.state()
         if state is not None:
             state = flatten_tensorized_space(tensorize_space(next(iter(self.state_spaces.values())), state))
             return _torch2jax(state, self._jax)

skrl/envs/wrappers/torch/isaaclab_envs.py

@@ -148,7 +148,10 @@ def state(self) -> torch.Tensor:
         :return: State
         :rtype: torch.Tensor
         """
-        state = self._env.state()
+        try:
+            state = self._env.state()
+        except AttributeError:  # 'OrderEnforcing' object has no attribute 'state'
+            state = self._unwrapped.state()
         if state is not None:
             return flatten_tensorized_space(tensorize_space(next(iter(self.state_spaces.values())), state))
         return state