Change coupling defaults to usable defaults

bayesflow/networks/coupling_flow/coupling_flow.py

@@ -39,9 +39,9 @@ class CouplingFlow(InferenceNetwork):
     def __init__(
         self,
         depth: int = 6,
-        subnet: str = "default",
+        subnet: str | keras.Layer = "mlp",
         transform: str = "affine",
-        permutation: str | None = None,
+        permutation: str | None = "random",
         use_actnorm: bool = True,
         base_distribution: str = "normal",
         **kwargs

bayesflow/networks/coupling_flow/couplings/dual_coupling.py

@@ -10,7 +10,7 @@
 
 @register_keras_serializable(package="bayesflow.networks.coupling_flow")
 class DualCoupling(InvertibleLayer):
-    def __init__(self, subnet: str = "resnet", transform: str = "affine", **kwargs):
+    def __init__(self, subnet: str = "mlp", transform: str = "affine", **kwargs):
         super().__init__(**keras_kwargs(kwargs))
         self.coupling1 = SingleCoupling(subnet, transform, **kwargs)
         self.coupling2 = SingleCoupling(subnet, transform, **kwargs)

bayesflow/networks/coupling_flow/couplings/single_coupling.py

@@ -17,7 +17,7 @@ class SingleCoupling(InvertibleLayer):
     """
     def __init__(
         self,
-        subnet: str = "resnet",
+        subnet: str = "mlp",
         transform: str = "affine",
         **kwargs
     ):

bayesflow/utils/dispatch/find_network.py

@@ -15,7 +15,7 @@ def _(name: str, **kwargs):
         case "mlp" | "default":
             from bayesflow.networks import MLP
             network = MLP(**kwargs)
-        #TODO - remove, since MLP encompasses the functionality
+        # TODO - remove, since MLP encompasses the functionality
         case "resnet":
             from bayesflow.networks import ResNet
             network = ResNet(**kwargs)