fix core

tensorlayer/layers/core/core_paddle.py

@@ -2,14 +2,14 @@
 # -*- coding: utf-8 -*-
 
 import copy, six
-import tensorlayer as tl
 from .common import str2act
 from paddle.fluid import framework
 from paddle.fluid.dygraph import Layer
 from paddle.fluid.framework import in_dygraph_mode
+from paddle.fluid.dygraph.base import program_desc_tracing_guard, param_guard
+from paddle.fluid.dygraph import parallel_helper
 
-
-_global_layer_name_dict = {}  # TODO: better implementation?
+_global_layer_name_dict = {}
 
 
 class Module(Layer):
@@ -54,7 +54,10 @@ class Module(Layer):
             self.act = act
 
         # Layer building state
-        # self._built = False
+        self._built = False
+
+        # paddl_built
+        self._paddle_built = False
 
         # Layer nodes state
         self._nodes = []
@@ -160,6 +163,31 @@ class Module(Layer):
     def forward(self, *inputs, **kwargs):
         raise Exception("The forward method must be implemented by inherited class")
 
+    def __call__(self, *inputs, **kwargs):
+        with param_guard(self._parameters), param_guard(self._buffers):
+            for forward_pre_hook in self._forward_pre_hooks.values():
+                hook_result = forward_pre_hook(self, inputs)
+                if hook_result is not None:
+                    if not isinstance(hook_result, tuple):
+                        hook_result = (hook_result, )
+                    inputs = hook_result
+
+            if not self._paddle_built:
+                with program_desc_tracing_guard(False):
+                    self._build_once(*inputs, **kwargs)
+                    if parallel_helper._is_data_parallel_mode():
+                        parallel_helper._broadcast_parameters(
+                            self._parameters.values())
+                self._paddle_built = True
+
+            outputs = self.forward(*inputs, **kwargs)
+
+            for forward_post_hook in self._forward_post_hooks.values():
+                hook_result = forward_post_hook(self, inputs, outputs)
+                if hook_result is not None:
+                    outputs = hook_result
+
+            return outputs
 
     def _get_weights(self, var_name, shape, init=None, trainable=True):
         if var_name in ["filters", "weights"]:

tensorlayer/package_info.py

@@ -2,8 +2,8 @@
 # -*- coding: utf-8 -*-
 """Deep learning and Reinforcement learning library for Researchers and Engineers."""
 
-MAJOR = 2
-MINOR = 2
+MAJOR = 3
+MINOR = 0
 PATCH = 0
 PRE_RELEASE = ''
 # Use the following formatting: (major, minor, patch, prerelease)