feat/release_pip

Context

see issue : aidge#67

aidge_interop_torch/utils.py

@@ -98,7 +98,6 @@ class AidgeModule(torch.nn.Module):
         self.input_nodes = [None]
         self.scheduler = None
         self.optimizer = None
-        self.grad_compiled = False
 
     def set_optimizer(self, opt):
         self.optimizer = opt
@@ -118,13 +117,13 @@ class AidgeModule(torch.nn.Module):
 
                 # TODO: add a system to avoid creating a new node everytime
 
-                if self.input_nodes[0] == None:
+                if self.input_nodes[0] is None:
                     self.input_nodes[0] = aidge_core.Producer(
                         aidge_tensor, "Input_0")
 
                     # TODO: get datatype & backend from graph view
                     self.input_nodes[0].get_operator().set_datatype(
-                        aidge_core.DataType.Float32)
+                        aidge_core.dtype.float32)
                     self.input_nodes[0].get_operator().set_backend("cpu")
 
                     self.input_nodes[0].add_child(self._graph_view)
@@ -151,8 +150,6 @@ class AidgeModule(torch.nn.Module):
 
             @staticmethod
             def backward(ctx, grad_output):
-                if not self.grad_compiled: aidge_core.compile_gradient(self._graph_view)
-
                 if self.multi_outputs_flag:
                     raise RuntimeError(
                         "Backward is not possible if the model has multi-outputs")
@@ -264,7 +261,7 @@ class ContextNoBatchNormFuse:
 
 def wrap(torch_model: torch.nn.Module,
          input_size: Union[list, tuple],
-         opset_version: int = 11,
+         opset_version: int = 15,
          in_names: list = None,
          out_names: list = None,
          verbose: bool = False) -> AidgeModule: