Several fixes to move the actor_server and learner_server code from the maniskill environment to the real robot environment.

Co-authored-by: Adil Zouitine <adilzouitinegm@gmail.com>

lerobot/configs/env/so100_real.yaml

@@ -0,0 +1,27 @@
+# @package _global_
+
+fps: 30
+
+env:
+  name: real_world
+  task: null
+  state_dim: 6
+  action_dim: 6
+  fps: ${fps}
+  device: mps
+  
+  wrapper:
+    crop_params_dict:
+      observation.images.laptop: [58, 89, 357, 455]
+      observation.images.phone: [3, 4, 471, 633]
+    resize_size: [128, 128]
+    control_time_s: 20
+    reset_follower_pos: true
+    use_relative_joint_positions: true
+    reset_time_s: 10
+    display_cameras: false
+    delta_action: 0.1
+
+  reward_classifier:
+    pretrained_path: outputs/classifier/checkpoints/best/pretrained_model
+    config_path: lerobot/configs/policy/hilserl_classifier.yaml