integrated metaworld tasks into the framework

alr_envs/meta/README.MD

@@ -0,0 +1,26 @@
+# MetaWorld Wrappers
+
+These are the Environment Wrappers for selected [Metaworld](https://meta-world.github.io/) environments in order to use our Motion Primitive gym interface with them. 
+All Metaworld environments have a 39 dimensional observation space with the same structure. The tasks differ only in the objective and the initial observations that are randomized. 
+Unused observations are zeroed out. E.g. for `Button-Press-v2` the observation mask looks the following:
+```python
+    return np.hstack([
+        # Current observation
+        [False] * 3,  # end-effector position
+        [False] * 1,  # normalized gripper open distance
+        [True] * 3,  # main object position
+        [False] * 4,  # main object quaternion
+        [False] * 3,  # secondary object position
+        [False] * 4,  # secondary object quaternion
+        # Previous observation
+        [False] * 3,  # previous end-effector position
+        [False] * 1,  # previous normalized gripper open distance
+        [False] * 3,  # previous main object position
+        [False] * 4,  # previous main object quaternion
+        [False] * 3,  # previous second object position
+        [False] * 4,  # previous second object quaternion
+        # Goal
+        [True] * 3,  # goal position
+    ])
+```
+For other tasks only the boolean values have to be adjusted accordingly.

alr_envs/meta/__init__.py

@@ -0,0 +1 @@
+from alr_envs.meta import button_press

alr_envs/meta/button_press.py

@@ -0,0 +1,48 @@
+from typing import Tuple, Union
+
+import numpy as np
+
+from mp_env_api import MPEnvWrapper
+
+
+class MPWrapper(MPEnvWrapper):
+
+    @property
+    def active_obs(self):
+        # This structure is the same for all metaworld environments.
+        # Only the observations which change could differ
+        return np.hstack([
+            # Current observation
+            [False] * 3,  # end-effector position
+            [False] * 1,  # normalized gripper open distance
+            [True] * 3,  # main object position
+            [False] * 4,  # main object quaternion
+            [False] * 3,  # secondary object position
+            [False] * 4,  # secondary object quaternion
+            # Previous observation
+            # TODO: Include previous values? According to their source they might be wrong for the first iteration.
+            [False] * 3,  # previous end-effector position
+            [False] * 1,  # previous normalized gripper open distance
+            [False] * 3,  # previous main object position
+            [False] * 4,  # previous main object quaternion
+            [False] * 3,  # previous second object position
+            [False] * 4,  # previous second object quaternion
+            # Goal
+            [True] * 3,  # goal position
+        ])
+
+    @property
+    def current_pos(self) -> Union[float, int, np.ndarray]:
+        return self.env.physics.named.data.qpos[:]
+
+    @property
+    def current_vel(self) -> Union[float, int, np.ndarray, Tuple]:
+        return self.env.physics.named.data.qvel[:]
+
+    @property
+    def goal_pos(self) -> Union[float, int, np.ndarray, Tuple]:
+        raise ValueError("Goal position is not available and has to be learnt based on the environment.")
+
+    @property
+    def dt(self) -> Union[float, int]:
+        return self.env.dt