RefVariable, variable_scope

src/TensorFlowNET.Core/Eager/Context.cs

@@ -2,7 +2,7 @@
 using System.Collections.Generic;
 using System.Text;
 
-namespace Tensorflow.Eager
+namespace Tensorflow
 {
     public class Context
     {

src/TensorFlowNET.Core/Graphs/Graph.cs

@@ -152,6 +152,11 @@ namespace Tensorflow
             return false;
         }
 
+        public string get_name_scope()
+        {
+            return _name_stack;
+        }
+
         public string name_scope(string name)
         {
             string new_stack = "";

src/TensorFlowNET.Core/Operations/OpDefLibrary.cs

@@ -10,7 +10,7 @@ namespace Tensorflow
 {
     public class OpDefLibrary
     {
-        public unsafe Operation _apply_op_helper(string op_type_name, string name = "", Dictionary<string, object> keywords = null)
+        public Operation _apply_op_helper(string op_type_name, string name = "", Dictionary<string, object> keywords = null)
         {
             var g = ops.get_default_graph();
             var op_def = g.GetOpDef(op_type_name);

src/TensorFlowNET.Core/Variables/RefVariable.cs

@@ -8,22 +8,62 @@ namespace Tensorflow
     {
         public bool _in_graph_mode = true;
         public Tensor _initial_value;
+        public string _graph_key;
+        public bool _trainable;
+        public Tensor _variable;
 
-        public RefVariable(object initial_value, 
+        public RefVariable(object initial_value,
+            bool trainable = true,
+            List<string> collections = null,
+            bool validate_shape = true,
+            string caching_device = "",
             string name = "",
-            TF_DataType trainable = TF_DataType.DtInvalid,
-            bool validate_shape = true) : 
-            base(initial_value, name, trainable, validate_shape)
+            TF_DataType dtype = TF_DataType.DtInvalid) : 
+            base(initial_value, trainable, collections, validate_shape, caching_device, name, dtype)
         {
-            _init_from_args(initial_value, name, trainable);
+            _init_from_args(initial_value, trainable, collections, validate_shape, caching_device, name, dtype);
         }
 
         private void _init_from_args(object initial_value,
+            bool trainable = true,
+            List<string> collections = null,
+            bool validate_shape = true,
+            string caching_device = "",
             string name = "",
-            TF_DataType trainable = TF_DataType.DtInvalid)
+            TF_DataType dtype = TF_DataType.DtInvalid)
         {
-            name = ops.name_scope("", "Variable", initial_value);
-            _initial_value = ops.convert_to_tensor(initial_value, name: "initial_value");
+            if (initial_value is null)
+                throw new ValueError("initial_value must be specified.");
+
+            var init_from_fn = false;
+
+            if(collections == null)
+            {
+                collections = new List<string> { ops.GraphKeys.GLOBAL_VARIABLES };
+            }
+
+            // Store the graph key so optimizers know how to only retrieve variables from
+            // this graph.
+            _graph_key = ops.get_default_graph()._graph_key;
+
+            _trainable = trainable;
+            if (!collections.Contains(ops.GraphKeys.TRAINABLE_VARIABLES))
+                collections.Add(ops.GraphKeys.TRAINABLE_VARIABLES);
+
+            ops.init_scope();
+            name = new ops.name_scope(name, "Variable", init_from_fn ? new List<object>() : new List<object> { initial_value });
+            if (init_from_fn)
+            {
+
+            }
+            else
+            {
+                _initial_value = ops.convert_to_tensor(initial_value, name: "initial_value");
+            }
+
+            var shape = _initial_value.shape;
+            dtype = _initial_value.dtype;
+            _variable = gen_state_ops.variable_v2(shape, dtype, name);
         }
     }
 }

src/TensorFlowNET.Core/Variables/VariableV1.cs

@@ -16,7 +16,13 @@ namespace Tensorflow
     /// </summary>
     public class VariableV1
     {
-        public VariableV1(object initial_value, string name = "", TF_DataType trainable = TF_DataType.DtInvalid, bool validate_shape = true)
+        public VariableV1(object initial_value,
+            bool trainable = true,
+            List<string> collections = null,
+            bool validate_shape = true,
+            string caching_device = "",
+            string name = "",
+            TF_DataType dtype = TF_DataType.DtInvalid)
         {
 
         }

src/TensorFlowNET.Core/Variables/gen_state_ops.py.cs

@@ -0,0 +1,35 @@
+using System;
+using System.Collections.Generic;
+using System.Text;
+
+namespace Tensorflow
+{
+    public class gen_state_ops
+    {
+        public static OpDefLibrary _op_def_lib = new OpDefLibrary();
+
+        /// <summary>
+        /// Holds state in the form of a tensor that persists across steps.
+        /// Outputs a ref to the tensor state so it may be read or modified.
+        /// </summary>
+        /// <param name="shape">The shape of the variable tensor.</param>
+        /// <param name="dtype">The type of elements in the variable tensor.</param>
+        /// <param name="name"></param>
+        /// <param name="container"></param>
+        /// <param name="shared_name"></param>
+        /// <returns></returns>
+        public static Tensor variable_v2(long[] shape, TF_DataType dtype, string name = "", string container = "", string shared_name = "")
+        {
+            var keywords = new Dictionary<string, object>();
+            keywords.Add("dtype", dtype);
+            keywords.Add("shape", shape);
+
+            var _op = _op_def_lib._apply_op_helper("VariableV2", name: name, keywords: keywords);
+
+            var _result = _op.outputs;
+            var _inputs_flat = _op.inputs;
+
+            return new Tensor(_op, 0, dtype);
+        }
+    }
+}

src/TensorFlowNET.Core/Variables/variable_scope.py.cs

@@ -26,7 +26,9 @@ namespace Tensorflow
             }
             else
             {
-                return new RefVariable(initial_value);
+                return new RefVariable(initial_value, 
+                    name: name,
+                    dtype: dtype);
             }
         }
 

src/TensorFlowNET.Core/Variables/variables.py.cs

@@ -12,7 +12,7 @@ namespace Tensorflow
         /// <returns></returns>
         public static object trainable_variables()
         {
-            return ops.get_collection(ops.GraphKey.TRAINABLE_VARIABLES);
+            return ops.get_collection(ops.GraphKeys.TRAINABLE_VARIABLES);
         }
     }
 }

src/TensorFlowNET.Core/ops.GraphKeys.py.cs → src/TensorFlowNET.Core/ops.GraphKeys.cs

@@ -15,12 +15,18 @@ namespace Tensorflow
         /// specified, but it is also possible to pass an explicit list of
         /// variables.
         /// </summary>
-        public static class GraphKey
+        public static class GraphKeys
         {
             /// <summary>
             /// the subset of `Variable` objects that will be trained by an optimizer.
             /// </summary>
             public static string TRAINABLE_VARIABLES = "trainable_variables";
+
+            /// <summary>
+            /// Key to collect Variable objects that are global (shared across machines).
+            /// Default collection for all variables, except local ones.
+            /// </summary>
+            public static string GLOBAL_VARIABLES = "variables";
         }
     }
 }

src/TensorFlowNET.Core/ops.name_scope.cs

@@ -0,0 +1,44 @@
+using System;
+using System.Collections.Generic;
+using System.Text;
+
+namespace Tensorflow
+{
+    public partial class ops
+    {
+        public class name_scope
+        {
+            public string _name;
+            public string _default_name;
+            public object _values;
+            public Context _ctx;
+            public string _name_scope;
+
+            public name_scope(string name, string default_name, List<object> values)
+            {
+                _name = name;
+                _default_name = default_name;
+                _values = values;
+                _ctx = new Context();
+
+                _name_scope = __enter__();
+            }
+
+            public string __enter__()
+            {
+                if (String.IsNullOrEmpty(_name))
+                {
+                    _name = _default_name;
+                }
+
+                var g = get_default_graph();
+                return g.name_scope(_name);
+            }
+
+            public static implicit operator string(name_scope ns)
+            {
+                return ns._name_scope;
+            }
+        }
+    }
+}

src/TensorFlowNET.Core/ops.py.cs

@@ -97,20 +97,6 @@ namespace Tensorflow
             return node_def;
         }
 
-        public static string name_scope(string name, string default_name = "", object values = null)
-        {
-            string _name = "";
-            if (String.IsNullOrEmpty(name))
-            {
-                _name = default_name;
-            }
-
-            var g = get_default_graph();
-            var _name_scope = g.name_scope(_name);
-
-            return _name_scope;
-        }
-
         public static string _name_from_scope_name(string name)
         {
             if (name.EndsWith("/"))
@@ -123,6 +109,27 @@ namespace Tensorflow
             }
         }
 
+        /// <summary>
+        /// A context manager that lifts ops out of control-flow scopes and function-building graphs.
+        /// </summary>
+        /// <returns></returns>
+        public static void init_scope()
+        {
+            // Retrieve the active name scope: entering an `init_scope` preserves
+            // the name scope of the current context.
+            var default_graph = get_default_graph();
+            var scope = default_graph.get_name_scope();
+            if (!String.IsNullOrEmpty(scope) && !scope.EndsWith("/"))
+                // Names that end with trailing slashes are treated by `name_scope` as
+                // absolute.
+                scope += "/";
+            // inner_device_stack = default_graph._device_function_stack
+            // var outer_context = default_graph.as_default;
+
+            var outer_graph = get_default_graph();
+            // outer_device_stack = None
+        }
+
         public static int uid()
         {
             return 1;

src/TensorFlowNET.Core/tf.cs

@@ -1,10 +1,6 @@
 using System;
 using System.Collections.Generic;
-using System.Runtime.InteropServices;
 using System.Text;
-using TF_DataType = Tensorflow.DataType;
-using attr_value_pb2 = Tensorflow;
-using Tensorflow.Eager;
 
 namespace Tensorflow
 {