Change TensorDataset construct.

Author: Oceania2018

src/TensorFlowNET.Core/Data/DatasetManager.cs

@@ -17,10 +17,7 @@ public IDatasetV2 from_generator<T>(IEnumerable<T> generator, TF_DataType[] outp
         public IDatasetV2 from_tensor(NDArray tensors)
             => new TensorDataset(tensors);
 
-        public IDatasetV2 from_tensor(Tensor features, Tensor labels)
-            => new TensorDataset(features, labels);
-
-        public IDatasetV2 from_tensor(Tensor tensors)
+        public IDatasetV2 from_tensor(Tensors tensors)
             => new TensorDataset(tensors);
 
         public IDatasetV2 from_tensor_slices(Tensor features, Tensor labels)

src/TensorFlowNET.Core/Data/TensorDataset.cs

@@ -9,16 +9,9 @@ namespace Tensorflow
     /// </summary>
     public class TensorDataset : DatasetSource
     {
-        public TensorDataset(Tensor feature, Tensor label)
+        public TensorDataset(Tensors elements)
         {
-            _tensors = new[] { feature, label };
-            structure = _tensors.Select(x => x.ToTensorSpec()).ToArray();
-
-            variant_tensor = ops.tensor_dataset(_tensors, output_shapes);
-        }
-        public TensorDataset(Tensor element)
-        {
-            _tensors = new[] { element };
+            _tensors = elements;
             var batched_spec = _tensors.Select(x => x.ToTensorSpec()).ToArray();
             structure = batched_spec.Select(x => x._unbatch()).ToArray();
 

src/TensorFlowNET.Keras/Engine/DataAdapters/TensorLikeDataAdapter.cs

@@ -26,10 +26,15 @@ public TensorLikeDataAdapter(DataAdapterArgs args)
             var _partial_batch_size = num_samples % batch_size;
 
             var indices_dataset = tf.data.Dataset.range(1);
-            indices_dataset = indices_dataset.repeat();
+            indices_dataset = indices_dataset.repeat(args.Epochs);
             indices_dataset = indices_dataset.map(permutation).prefetch(1);
             indices_dataset = indices_dataset.flat_map(slice_batch_indices);
-            dataset = slice_inputs(indices_dataset, args.X, args.Y);
+            var elements = new Tensors();
+            if (args.X != null)
+                elements.Add(args.X);
+            if (args.Y != null)
+                elements.Add(args.Y);
+            dataset = slice_inputs(indices_dataset, elements);
         }
 
         Tensor permutation(Tensor tensor)
@@ -54,9 +59,9 @@ IDatasetV2 slice_batch_indices(Tensor indices)
             return flat_dataset;
         }
 
-        IDatasetV2 slice_inputs(IDatasetV2 indices_dataset, Tensor x, Tensor y)
+        IDatasetV2 slice_inputs(IDatasetV2 indices_dataset, Tensors elements)
         {
-            var dataset2 = tf.data.Dataset.from_tensor(x, y).repeat();
+            var dataset2 = tf.data.Dataset.from_tensor(elements).repeat();
             var dataset = tf.data.Dataset.zip(indices_dataset, dataset2);
 
             dataset = dataset.map((batch, data) =>

src/TensorFlowNET.Keras/Engine/Model.Predict.cs

@@ -1,4 +1,5 @@
-using System;
+using NumSharp;
+using System;
 using Tensorflow.Keras.ArgsDefinition;
 using Tensorflow.Keras.Engine.DataAdapters;
 
@@ -21,7 +22,7 @@ public partial class Model
         /// <param name="use_multiprocessing"></param>
         /// <returns></returns>
         public Tensor predict(Tensor x,
-            int batch_size = 32,
+            int batch_size = -1,
             int verbose = 0,
             int steps = -1,
             int max_queue_size = 10,