ResourceVariable.sparse_read()

src/TensorFlowNET.Core/Binding.Util.cs

@@ -129,6 +129,7 @@ namespace Tensorflow
             }
         }
 
+        [DebuggerStepThrough]
         [DebuggerNonUserCode()] // with "Just My Code" enabled this lets the debugger break at the origin of the exception
         public static TOut tf_with<TIn, TOut>(TIn py, Func<TIn, TOut> action) where TIn : IObjectLife
         {

src/TensorFlowNET.Core/Keras/Engine/Sequential.cs

@@ -56,9 +56,9 @@ namespace Tensorflow.Keras.Engine
                     {
                         // Instantiate an input layer.
                         var x = keras.layers.Input(
-                          batch_shape: batch_shape,
-                          dtype: dtype,
-                          name: layer.name + "_input");
+                              batch_shape: batch_shape,
+                              dtype: dtype,
+                              name: layer.name + "_input");
 
                         // This will build the current layer
                         // and create the node connecting the current layer
@@ -71,7 +71,7 @@ namespace Tensorflow.Keras.Engine
                 if (set_inputs)
                 {
                     // If an input layer (placeholder) is available.
-                    // outputs = layer._inbound_nodes;
+                    // outputs = layer.inbound_nodes;
                 }
 
             }

src/TensorFlowNET.Core/Keras/Layers/Layer.cs

@@ -106,6 +106,7 @@ namespace Tensorflow.Keras.Layers
             VariableScope scope = null)
         {
             var input_list = inputs;
+            var input = inputs[0];
             Tensor outputs = null;
 
             // We will attempt to build a TF graph if & only if all inputs are symbolic.
@@ -139,6 +140,7 @@ namespace Tensorflow.Keras.Layers
                     _maybe_build(inputs[0]);
 
                     outputs = call(inputs[0], training: training);
+                    (input, outputs) = _set_connectivity_metadata_(input, outputs);
                     _handle_activity_regularization(inputs[0], outputs);
                     _set_mask_metadata(inputs[0], outputs, null);
                 });
@@ -147,6 +149,12 @@ namespace Tensorflow.Keras.Layers
             return outputs;
         }
 
+        private (Tensor, Tensor) _set_connectivity_metadata_(Tensor inputs, Tensor outputs)
+        {
+            //_add_inbound_node(input_tensors: inputs, output_tensors: outputs);
+            return (inputs, outputs);
+        }
+
         private void _handle_activity_regularization(Tensor inputs, Tensor outputs)
         {
             //if(_activity_regularizer != null)

src/TensorFlowNET.Core/Operations/array_ops.py.cs

@@ -605,8 +605,9 @@ namespace Tensorflow
             if (axis != 0)
                 return gen_array_ops.gather_v2(@params, indices, axis, name: name);
 
-            if (@params is ResourceVariable variable)
-                return variable.sparse_read();
+            if (@params is ResourceVariable variable &&
+                indices is Tensor indices_tensor)
+                return variable.sparse_read(indices_tensor, name);
 
             return gen_array_ops.gather_v2(@params, indices, axis, name: name);
         }

src/TensorFlowNET.Core/Operations/gen_resource_variable_ops.cs

@@ -73,5 +73,20 @@ namespace Tensorflow
 
             return _op.output;
         }
+
+        public static Tensor resource_gather(Tensor resource, Tensor indices, TF_DataType dtype, 
+            int batch_dims = 0, bool validate_indices = true, string name = null)
+        {
+            var _op = _op_def_lib._apply_op_helper("ResourceGather", name, new
+            {
+                resource,
+                indices,
+                dtype,
+                batch_dims,
+                validate_indices
+            });
+
+            return _op.output;
+        }
     }
 }

src/TensorFlowNET.Core/Variables/ResourceVariable.cs

@@ -216,6 +216,18 @@ namespace Tensorflow
             _dtype = dtypes.as_tf_dtype((DataType)_handle.op.get_attr("dtype"));
         }
 
+        public Tensor sparse_read(Tensor indices, string name = "Gather")
+        {
+            return tf_with(ops.name_scope(name), scope =>
+            {
+                name = scope;
+                var value = gen_resource_variable_ops.resource_gather(
+                    _handle, indices, dtype: _dtype, name: name);
+
+                return array_ops.identity(value);
+            });
+        }
+
         public override string ToString()
         {
             return $"tf.ResourceVariable '{name}' shape={shape} dtype={dtype}";