Fix custom model documentation

docs/references/network_architectures.rst

@@ -38,7 +38,8 @@ You can also build your own encoder factory.
 
    # your own neural network
    class CustomEncoder(nn.Module):
-       def __init__(self, obsevation_shape, feature_size):
+       def __init__(self, observation_shape, feature_size):
+           super().__init__()
            self.feature_size = feature_size
            self.fc1 = nn.Linear(observation_shape[0], 64)
            self.fc2 = nn.Linear(64, feature_size)
@@ -72,7 +73,8 @@ controls.
 .. code-block:: python
 
    class CustomEncoderWithAction(nn.Module):
-       def __init__(self, obsevation_shape, action_size, feature_size):
+       def __init__(self, observation_shape, action_size, feature_size):
+           super().__init__()
            self.feature_size = feature_size
            self.fc1 = nn.Linear(observation_shape[0] + action_size, 64)
            self.fc2 = nn.Linear(64, feature_size)
@@ -90,7 +92,7 @@ controls.
        def create(self, observation_shape):
            return CustomEncoder(observation_shape, self.feature_size)
 
-       def create_with_action(observation_shape, action_size, discrete_action):
+       def create_with_action(self, observation_shape, action_size, discrete_action):
            return CustomEncoderWithAction(observation_shape, action_size, self.feature_size)
 
        @staticmethod

docs/tutorials/customize_neural_network.rst

@@ -114,3 +114,20 @@ Now, you can customize actor-critic algorithms.
       actor_encoder_factory=encoder_factory,
       critic_encoder_factory=encoder_factory,
   ).create()
+
+
+Make your models loadable
+-------------------------
+
+If you want ``load_learnable`` method to load the algorithm configuration including
+your encoder configuration, you need to register your encoder factory.
+
+.. code-block:: python
+
+   from d3rlpy.models.encoders import register_encoder_factory
+
+   # register your own encoder factory
+   register_encoder_factory(CustomEncoderFactory)
+
+   # load algorithm from d3
+   dqn = d3rlpy.load_learnable("model.d3")