Consolidate MapDataset function.

src/TensorFlowNET.Console/Tensorflow.Console.csproj

@@ -2,7 +2,7 @@
 
   <PropertyGroup>
     <OutputType>Exe</OutputType>
-    <TargetFramework>netcoreapp3.1</TargetFramework>
+    <TargetFramework>net5.0</TargetFramework>
     <RootNamespace>Tensorflow</RootNamespace>
     <AssemblyName>Tensorflow</AssemblyName>
     <Platforms>AnyCPU;x64</Platforms>

src/TensorFlowNET.Core/APIs/c_api.cs

@@ -43,7 +43,7 @@ namespace Tensorflow
     /// </summary>
     public partial class c_api
     {
-        public const string TensorFlowLibName = @"D:\Projects\tensorflow-haiping\bazel-bin\tensorflow\tensorflow";
+        public const string TensorFlowLibName = "tensorflow";
 
         public static string StringPiece(IntPtr handle)
         {

src/TensorFlowNET.Core/APIs/tf.strings.cs

@@ -24,6 +24,30 @@ namespace Tensorflow
         {
             string_ops ops = new string_ops();
 
+            /// <summary>
+            /// Converts all uppercase characters into their respective lowercase replacements.
+            /// </summary>
+            /// <param name="input"></param>
+            /// <param name="encoding"></param>
+            /// <param name="name"></param>
+            /// <returns></returns>
+            public Tensor lower(Tensor input, string encoding = "", string name = null)
+                => ops.lower(input: input, encoding: encoding, name: name);
+
+            /// <summary>
+            /// 
+            /// </summary>
+            /// <param name="input"></param>
+            /// <param name="pattern"></param>
+            /// <param name="rewrite"></param>
+            /// <param name="replace_global"></param>
+            /// <param name="name"></param>
+            /// <returns></returns>
+            public Tensor regex_replace(Tensor input, string pattern, string rewrite,
+                bool replace_global = true, string name = null)
+                => ops.regex_replace(input, pattern, rewrite,
+                    replace_global: replace_global, name: name);
+
             /// <summary>
             /// Return substrings from `Tensor` of strings.
             /// </summary>

src/TensorFlowNET.Core/Data/DatasetV2.cs

@@ -14,6 +14,7 @@ namespace Tensorflow
     public class DatasetV2 : IDatasetV2
     {
         protected dataset_ops ops = new dataset_ops();
+        public string[] class_names { get; set; }
         public Tensor variant_tensor { get; set; }
 
         public TensorSpec[] structure { get; set; }
@@ -54,7 +55,7 @@ namespace Tensorflow
         public IDatasetV2 optimize(string[] optimizations, string[] optimization_configs)
             => new OptimizeDataset(this, optimizations, optimization_configs: optimization_configs);
 
-        public IDatasetV2 map(Func<Tensor, Tensor> map_func,
+        public IDatasetV2 map(Func<Tensors, Tensors> map_func,
             bool use_inter_op_parallelism = true,
             bool preserve_cardinality = true,
             bool use_legacy_function = false)
@@ -64,7 +65,7 @@ namespace Tensorflow
                 preserve_cardinality: preserve_cardinality,
                 use_legacy_function: use_legacy_function);
 
-        public IDatasetV2 map(Func<Tensors, Tensors> map_func, int num_parallel_calls = -1)
+        public IDatasetV2 map(Func<Tensors, Tensors> map_func, int num_parallel_calls)
             => new ParallelMapDataset(this, map_func, num_parallel_calls: num_parallel_calls);
 
         public IDatasetV2 flat_map(Func<Tensor, IDatasetV2> map_func)

src/TensorFlowNET.Core/Data/IDatasetV2.cs

@@ -6,6 +6,8 @@ namespace Tensorflow
 {
     public interface IDatasetV2 : IEnumerable<(Tensor, Tensor)>
     {
+        string[] class_names { get; set; }
+
         Tensor variant_tensor { get; set; }
 
         TensorShape[] output_shapes { get; }
@@ -62,13 +64,13 @@ namespace Tensorflow
 
         IDatasetV2 optimize(string[] optimizations, string[] optimization_configs);
 
-        IDatasetV2 map(Func<Tensor, Tensor> map_func,
+        IDatasetV2 map(Func<Tensors, Tensors> map_func,
             bool use_inter_op_parallelism = true,
             bool preserve_cardinality = true,
             bool use_legacy_function = false);
 
         IDatasetV2 map(Func<Tensors, Tensors> map_func,
-            int num_parallel_calls = -1);
+            int num_parallel_calls);
 
         IDatasetV2 flat_map(Func<Tensor, IDatasetV2> map_func);
 

src/TensorFlowNET.Core/Data/MapDataset.cs

@@ -10,7 +10,7 @@ namespace Tensorflow
     public class MapDataset : UnaryDataset
     {
         public MapDataset(IDatasetV2 input_dataset,
-            Func<Tensor, Tensor> map_func,
+            Func<Tensors, Tensors> map_func,
             bool use_inter_op_parallelism = true,
             bool preserve_cardinality = false,
             bool use_legacy_function = false) : base(input_dataset)

src/TensorFlowNET.Core/Framework/Models/TensorSpec.cs

@@ -15,7 +15,7 @@ namespace Tensorflow.Framework.Models
             if (_shape.ndim == 0)
                 throw new ValueError("Unbatching a tensor is only supported for rank >= 1");
 
-            return new TensorSpec(_shape.dims[1..], _dtype);
+            return new TensorSpec(_shape.dims.Skip(1).ToArray(), _dtype);
         }
 
         public TensorSpec _batch(int dim = -1)

src/TensorFlowNET.Core/Gradients/image_grad.cs

@@ -30,7 +30,7 @@ namespace Tensorflow.Gradients
             var shape = new TensorShape(image.shape.Skip(1).Take(2).ToArray());
             Tensor image_shape = null;
             if (shape.is_fully_defined())
-                image_shape = constant_op.constant(image.shape[1..3]);
+                image_shape = constant_op.constant(image.shape.Skip(1).Take(2).ToArray());
             else
                 image_shape = array_ops.shape(image)["1:3"];
 

src/TensorFlowNET.Core/Keras/ArgsDefinition/Preprocessing/PreprocessingLayerArgs.cs

@@ -0,0 +1,10 @@
+using System;
+using System.Collections.Generic;
+using System.Text;
+
+namespace Tensorflow.Keras.ArgsDefinition
+{
+    public class PreprocessingLayerArgs : LayerArgs
+    {
+    }
+}

src/TensorFlowNET.Core/Keras/ArgsDefinition/Preprocessing/TextVectorizationArgs.cs

@@ -0,0 +1,15 @@
+using System;
+using System.Collections.Generic;
+using System.Text;
+
+namespace Tensorflow.Keras.ArgsDefinition
+{
+    public class TextVectorizationArgs : PreprocessingLayerArgs
+    {
+        public Func<Tensor, Tensor> Standardize { get; set; }
+        public string Split { get; set; } = "standardize";
+        public int MaxTokens { get; set; } = -1;
+        public string OutputMode { get; set; } = "int";
+        public int OutputSequenceLength { get; set; } = -1;
+    }
+}

src/TensorFlowNET.Core/Operations/string_ops.cs

@@ -20,6 +20,54 @@ namespace Tensorflow
 {
     public class string_ops
     {
+        public Tensor lower(Tensor input, string encoding = "", string name = null)
+        {
+            if (tf.Context.executing_eagerly())
+            {
+                var results = tf.Runner.TFE_FastPathExecute(tf.Context, tf.Context.DeviceName,
+                    "StringLower", name,
+                    null,
+                    input, encoding);
+
+                return results[0];
+            }
+
+            var _op = tf.OpDefLib._apply_op_helper("StringLower", name: name, args: new
+            {
+                input,
+                encoding
+            });
+
+            return _op.output;
+        }
+
+        public Tensor regex_replace(Tensor input, string pattern, string rewrite,
+                bool replace_global = true, string name = null)
+        {
+            if (tf.Context.executing_eagerly())
+            {
+                var results = tf.Runner.TFE_FastPathExecute(tf.Context, tf.Context.DeviceName,
+                    "StaticRegexReplace", name,
+                    null,
+                    input, 
+                    "pattern", pattern,
+                    "rewrite", rewrite, 
+                    "replace_global", replace_global);
+
+                return results[0];
+            }
+
+            var _op = tf.OpDefLib._apply_op_helper("StaticRegexReplace", name: name, args: new
+            {
+                input,
+                pattern,
+                rewrite,
+                replace_global
+            });
+
+            return _op.output;
+        }
+
         /// <summary>
         /// Return substrings from `Tensor` of strings.
         /// </summary>

src/TensorFlowNET.Keras/Engine/CombinerPreprocessingLayer.cs

@@ -0,0 +1,18 @@
+using System;
+using System.Collections.Generic;
+using System.Text;
+using Tensorflow.Keras.ArgsDefinition;
+
+namespace Tensorflow.Keras.Engine
+{
+    public class CombinerPreprocessingLayer : Layer
+    {
+        PreprocessingLayerArgs args;
+
+        public CombinerPreprocessingLayer(PreprocessingLayerArgs args)
+            : base(args)
+        {
+            
+        }
+    }
+}

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

@@ -39,7 +39,7 @@ namespace Tensorflow.Keras.Engine.DataAdapters
             dataset = slice_inputs(indices_dataset, inputs);
         }
 
-        Tensor permutation(Tensor tensor)
+        Tensors permutation(Tensors tensor)
         {
             var indices = math_ops.range(num_samples, dtype: dtypes.int64);
             if (args.Shuffle)
@@ -82,7 +82,7 @@ namespace Tensorflow.Keras.Engine.DataAdapters
                     .Select(x => gen_array_ops.gather_v2(x, indices, 0))
                     .ToArray();
                 return new Tensors(results);
-            });
+            }, -1);
 
             return dataset.with_options(new DatasetOptions { });
         }

src/TensorFlowNET.Keras/Engine/MetricsContainer.cs

@@ -62,8 +62,8 @@ namespace Tensorflow.Keras.Engine
             {
                 var y_t_rank = y_t.rank;
                 var y_p_rank = y_p.rank;
-                var y_t_last_dim = y_t.shape[^1];
-                var y_p_last_dim = y_p.shape[^1];
+                var y_t_last_dim = y_t.shape[y_t.shape.Length - 1];
+                var y_p_last_dim = y_p.shape[y_p.shape.Length - 1];
 
                 bool is_binary = y_p_last_dim == 1;
                 bool is_sparse_categorical = (y_t_rank < y_p_rank || y_t_last_dim == 1) && y_p_last_dim > 1;

src/TensorFlowNET.Keras/Engine/Sequential.cs

@@ -14,6 +14,7 @@
    limitations under the License.
 ******************************************************************************/
 
+using System.Linq;
 using System.Collections.Generic;
 using Tensorflow.Keras.ArgsDefinition;
 using Tensorflow.Keras.Layers;
@@ -103,7 +104,7 @@ namespace Tensorflow.Keras.Engine
                 if (set_inputs)
                 {
                     // If an input layer (placeholder) is available.
-                    outputs = layer.InboundNodes[^1].Outputs;
+                    outputs = layer.InboundNodes.Last().Outputs;
                     inputs = layer_utils.get_source_inputs(outputs[0]);
                     built = true;
                     _has_explicit_input_shape = true;

src/TensorFlowNET.Keras/KerasInterface.cs

@@ -11,6 +11,7 @@ using Tensorflow.Keras.Metrics;
 using Tensorflow.Keras.Models;
 using Tensorflow.Keras.Optimizers;
 using Tensorflow.Keras.Saving;
+using Tensorflow.Keras.Utils;
 
 namespace Tensorflow.Keras
 {
@@ -27,6 +28,7 @@ namespace Tensorflow.Keras
         public OptimizerApi optimizers { get; } = new OptimizerApi();
         public MetricsApi metrics { get; } = new MetricsApi();
         public ModelsApi models { get; } = new ModelsApi();
+        public KerasUtils utils { get; } = new KerasUtils();
 
         public Sequential Sequential(List<ILayer> layers = null,
                 string name = null)
@@ -73,7 +75,7 @@ namespace Tensorflow.Keras
                 Tensor tensor = null)
         {
             if (batch_input_shape != null)
-                shape = batch_input_shape.dims[1..];
+                shape = batch_input_shape.dims.Skip(1).ToArray();
 
             var args = new InputLayerArgs
             {

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

@@ -42,7 +42,7 @@ namespace Tensorflow.Keras.Layers
             if (BatchInputShape != null)
             {
                 args.BatchSize = BatchInputShape.dims[0];
-                args.InputShape = BatchInputShape.dims[1..];
+                args.InputShape = BatchInputShape.dims.Skip(1).ToArray();
             }
 
             // moved to base class

src/TensorFlowNET.Keras/Layers/LayersApi.cs

@@ -9,6 +9,8 @@ namespace Tensorflow.Keras.Layers
 {
     public partial class LayersApi
     {
+        public Preprocessing preprocessing { get; } = new Preprocessing();
+
         /// <summary>
         /// Functional interface for the batch normalization layer.
         /// http://arxiv.org/abs/1502.03167

src/TensorFlowNET.Keras/Layers/Preprocessing/TextVectorization.cs

@@ -0,0 +1,20 @@
+using System;
+using System.Collections.Generic;
+using System.Text;
+using Tensorflow.Keras.ArgsDefinition;
+using Tensorflow.Keras.Engine;
+
+namespace Tensorflow.Keras.Layers
+{
+    public class TextVectorization : CombinerPreprocessingLayer
+    {
+        TextVectorizationArgs args;
+
+        public TextVectorization(TextVectorizationArgs args)
+            : base(args)
+        {
+            args.DType = TF_DataType.TF_STRING;
+            // string standardize = "lower_and_strip_punctuation",
+        }
+    }
+}

src/TensorFlowNET.Keras/Layers/Reshaping/Flatten.cs

@@ -1,4 +1,5 @@
 using System;
+using System.Linq;
 using Tensorflow.Framework;
 using Tensorflow.Keras.ArgsDefinition;
 using Tensorflow.Keras.Engine;
@@ -45,7 +46,7 @@ namespace Tensorflow.Keras.Layers
                     return array_ops.reshape(inputs, new[] { batch_dim, -1 });
                 }
 
-                var non_batch_dims = ((int[])input_shape)[1..];
+                var non_batch_dims = ((int[])input_shape).Skip(1).ToArray();
                 var num = 1;
                 if (non_batch_dims.Length > 0)
                 {

src/TensorFlowNET.Keras/Layers/Reshaping/Reshape.cs

@@ -37,7 +37,7 @@ namespace Tensorflow.Keras.Layers
 
         public override TensorShape ComputeOutputShape(TensorShape input_shape)
         {
-            if (input_shape.dims[1..].Contains(-1))
+            if (input_shape.dims.Skip(1).Contains(-1))
             {
                 throw new NotImplementedException("");
             }

src/TensorFlowNET.Keras/Preprocessings/DatasetUtils.get_training_or_validation_split.cs

@@ -1,4 +1,5 @@
 using System;
+using System.Linq;
 
 namespace Tensorflow.Keras.Preprocessings
 {
@@ -17,18 +18,21 @@ namespace Tensorflow.Keras.Preprocessings
             float validation_split,
             string subset)
         {
+            if (string.IsNullOrEmpty(subset))
+                return (samples, labels);
+
             var num_val_samples = Convert.ToInt32(samples.Length * validation_split);
             if (subset == "training")
             {
                 Console.WriteLine($"Using {samples.Length - num_val_samples} files for training.");
-                samples = samples[..^num_val_samples];
-                labels = labels[..^num_val_samples];
+                samples = samples.Take(samples.Length - num_val_samples).ToArray();
+                labels = labels.Take(labels.Length - num_val_samples).ToArray();
             }
             else if (subset == "validation")
             {
                 Console.WriteLine($"Using {num_val_samples} files for validation.");
-                samples = samples[(samples.Length - num_val_samples)..];
-                labels = labels[(labels.Length - num_val_samples)..];
+                samples = samples.Skip(samples.Length - num_val_samples).ToArray();
+                labels = labels.Skip(labels.Length - num_val_samples).ToArray();
             }
             else
                 throw new NotImplementedException("");

src/TensorFlowNET.Keras/Preprocessings/DatasetUtils.index_directory.cs

@@ -21,40 +21,41 @@ namespace Tensorflow.Keras.Preprocessings
         /// file_paths, labels, class_names
         /// </returns>
         public (string[], int[], string[]) index_directory(string directory,
+            string labels,
             string[] formats = null,
             string[] class_names = null,
             bool shuffle = true,
             int? seed = null,
             bool follow_links = false)
         {
-            var labels = new List<int>();
+            var label_list = new List<int>();
             var file_paths = new List<string>();
 
             var class_dirs = Directory.GetDirectories(directory);
-            class_names = class_dirs.Select(x => x.Split(Path.DirectorySeparatorChar)[^1]).ToArray();
+            class_names = class_dirs.Select(x => x.Split(Path.DirectorySeparatorChar).Last()).ToArray();
 
             for (var label = 0; label < class_dirs.Length; label++)
             {
                 var files = Directory.GetFiles(class_dirs[label]);
                 file_paths.AddRange(files);
-                labels.AddRange(Enumerable.Range(0, files.Length).Select(x => label));
+                label_list.AddRange(Enumerable.Range(0, files.Length).Select(x => label));
             }
 
-            var return_labels = labels.Select(x => x).ToArray();
+            var return_labels = label_list.Select(x => x).ToArray();
             var return_file_paths = file_paths.Select(x => x).ToArray();
 
             if (shuffle)
             {
                 if (!seed.HasValue)
                     seed = np.random.randint((long)1e6);
-                var random_index = np.arange(labels.Count);
+                var random_index = np.arange(label_list.Count);
                 var rng = np.random.RandomState(seed.Value);
                 rng.shuffle(random_index);
                 var index = random_index.ToArray<int>();
 
-                for (int i = 0; i < labels.Count; i++)
+                for (int i = 0; i < label_list.Count; i++)
                 {
-                    return_labels[i] = labels[index[i]];
+                    return_labels[i] = label_list[index[i]];
                     return_file_paths[i] = file_paths[index[i]];
                 }
             }

src/TensorFlowNET.Keras/Preprocessings/Preprocessing.cs

@@ -1,4 +1,7 @@
-using Tensorflow.Keras.Preprocessings;
+using System;
+using Tensorflow.Keras.ArgsDefinition;
+using Tensorflow.Keras.Layers;
+using Tensorflow.Keras.Preprocessings;
 
 namespace Tensorflow.Keras
 {
@@ -6,5 +9,18 @@ namespace Tensorflow.Keras
     {
         public Sequence sequence => new Sequence();
         public DatasetUtils dataset_utils => new DatasetUtils();
+
+        public TextVectorization TextVectorization(Func<Tensor, Tensor> standardize = null,
+            string split = "standardize",
+            int max_tokens = -1,
+            string output_mode = "int",
+            int output_sequence_length = -1) => new TextVectorization(new TextVectorizationArgs
+            {
+                Standardize = standardize,
+                Split = split,
+                MaxTokens = max_tokens,
+                OutputMode = output_mode,
+                OutputSequenceLength = output_sequence_length
+            });
     }
 }

src/TensorFlowNET.Keras/Preprocessings/Preprocessing.image_dataset_from_directory.cs

@@ -43,6 +43,7 @@ namespace Tensorflow.Keras
                 num_channels = 3;
             
             var (image_paths, label_list, class_name_list) = keras.preprocessing.dataset_utils.index_directory(directory,
+                labels,
                 formats: WHITELIST_FORMATS,
                 class_names: class_names,
                 shuffle: shuffle,
@@ -64,13 +65,30 @@ namespace Tensorflow.Keras
             string[] class_names = null,
             int batch_size = 32,
             bool shuffle = true,
+            int max_length = -1,
             int? seed = null,
             float validation_split = 0.2f,
-            string subset = null)
+            string subset = null,
+            bool follow_links = false)
         {
-           
+            var (file_paths, label_list, class_name_list) = dataset_utils.index_directory(
+               directory,
+               labels,
+               formats: new[] { ".txt" },
+               class_names: class_names,
+               shuffle: shuffle,
+               seed: seed,
+               follow_links: follow_links);
 
-            return null;
+            (file_paths, label_list) = dataset_utils.get_training_or_validation_split(
+                file_paths, label_list, validation_split, subset);
+
+            var dataset = paths_and_labels_to_dataset(file_paths, label_list, label_mode, class_name_list.Length);
+            if (shuffle)
+                dataset = dataset.shuffle(batch_size * 8, seed: seed);
+            dataset = dataset.batch(batch_size);
+            dataset.class_names = class_name_list;
+            return dataset;
         }
     }
 }

src/TensorFlowNET.Keras/Preprocessings/Preprocessing.paths_and_labels_to_dataset.cs

@@ -1,4 +1,5 @@
 using System;
+using System.IO;
 using static Tensorflow.Binding;
 
 namespace Tensorflow.Keras
@@ -34,5 +35,31 @@ namespace Tensorflow.Keras
             // img.set_shape((image_size[0], image_size[1], num_channels));
             return img;
         }
+
+        public IDatasetV2 paths_and_labels_to_dataset(string[] image_paths,
+            int[] labels,
+            string label_mode,
+            int num_classes,
+            int max_length = -1)
+        {
+            var path_ds = tf.data.Dataset.from_tensor_slices(image_paths);
+            var string_ds = path_ds.map(x => path_to_string_content(x, max_length));
+
+            if (label_mode == "int")
+            {
+                var label_ds = dataset_utils.labels_to_dataset(labels, label_mode, num_classes);
+                string_ds = tf.data.Dataset.zip(string_ds, label_ds);
+            }
+
+            return string_ds;
+        }
+
+        Tensor path_to_string_content(Tensor path, int max_length)
+        {
+            var txt = tf.io.read_file(path);
+            if (max_length > -1)
+                txt = tf.strings.substr(txt, 0, max_length);
+            return txt;
+        }
     }
 }

src/TensorFlowNET.Keras/Tensorflow.Keras.csproj

@@ -6,7 +6,7 @@
     <LangVersion>8.0</LangVersion>
     <RootNamespace>Tensorflow.Keras</RootNamespace>
     <Platforms>AnyCPU;x64</Platforms>
-    <Version>0.4.0</Version>
+    <Version>0.4.1</Version>
     <Authors>Haiping Chen</Authors>
     <Product>Keras for .NET</Product>
     <Copyright>Apache 2.0, Haiping Chen 2020</Copyright>
@@ -34,8 +34,8 @@ Keras is an API designed for human beings, not machines. Keras follows best prac
     <RepositoryType>Git</RepositoryType>
     <SignAssembly>true</SignAssembly>
     <AssemblyOriginatorKeyFile>Open.snk</AssemblyOriginatorKeyFile>
-    <AssemblyVersion>0.4.0.0</AssemblyVersion>
-    <FileVersion>0.4.0.0</FileVersion>
+    <AssemblyVersion>0.4.1.0</AssemblyVersion>
+    <FileVersion>0.4.1.0</FileVersion>
     <PackageLicenseFile>LICENSE</PackageLicenseFile>
   </PropertyGroup>
 

src/TensorFlowNet.Benchmarks/Tensorflow.Benchmark.csproj

@@ -2,7 +2,7 @@
 
   <PropertyGroup>
     <OutputType>Exe</OutputType>
-    <TargetFramework>netcoreapp3.1</TargetFramework>
+    <TargetFramework>net5.0</TargetFramework>
     <Platforms>AnyCPU;x64</Platforms>
   </PropertyGroup>
 

tensorflowlib/README.md

@@ -56,7 +56,7 @@ Set ENV `BAZEL_VC=C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\
 
 1. Build static library
 
-`bazel build --output_base=C:/tmp/tfcompilation build --config=opt //tensorflow:tensorflow`
+`bazel build --output_base=C:/tmp/tfcompilation --config=opt //tensorflow:tensorflow`
 
 2. Build pip package
 
@@ -70,6 +70,16 @@ Set ENV `BAZEL_VC=C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\
 
 `pip install C:/tmp/tensorflow_pkg/tensorflow-1.15.0-cp36-cp36m-win_amd64.whl`
 
+### Build from source for MacOS
+
+```shell
+$ cd /usr/local/lib/bazel/bin
+$ curl -LO https://release.bazel.build/3.7.2/release/bazel-3.7.2-darwin-x86_64
+$ chmod +x bazel-3.7.2-darwin-x86_64
+$ cd ~/Projects/tensorflow
+$ bazel build --config=opt //tensorflow:tensorflow
+```
+
 ### Build specific version for tf.net
 
 https://github.com/SciSharp/tensorflow