static void Main()
{
GradientLog.OutputWriter = Console.Out;
GradientSetup.UseEnvironmentFromVariable();
Tensor a = tf.constant(5.0, name: "a");
Tensor b = tf.constant(10.0, name: "b");
Tensor sum = tf.add(a, b, name: "sum");
Tensor div = tf.div(a, b, name: "div");
dynamic config = config_pb2.ConfigProto();
// unless this is set, tensorflow-gpu consumes all of GPU memory
// don't set it if you don't want you training to crash due to random OOM in the middle
config.gpu_options.allow_growth = true;
Session sess = Session.NewDyn(config: config);
sess.UseSelf(session => {
var writer = new FileWriter(".", session.graph);
Console.WriteLine($"a = {session.run(a)}");
Console.WriteLine($"b = {session.run(b)}");
Console.WriteLine($"a + b = {session.run(sum)}");
Console.WriteLine($"a / b = {session.run(div)}");
writer.close();
session.close();
});
}
static void Main()
{
Console.Title = nameof(ResNetSampleProgram);
GradientLog.OutputWriter = Console.Out;
GradientSetup.UseEnvironmentFromVariable();
Run();
}
static int Main(string[] args)
{
GradientLog.OutputWriter = Console.Out;
GradientSetup.UseEnvironmentFromVariable();
// required before using PythonEngine
GradientSetup.EnsureInitialized();
np = PythonEngine.ImportModule("numpy");
return(ConsoleCommandDispatcher.DispatchCommand(
ConsoleCommandDispatcher.FindCommandsInSameAssemblyAs(typeof(LinearSvmProgram)),
args, Console.Out));
}
static int Main(string[] args)
{
GradientLog.OutputWriter = Console.Out;
GradientSetup.UseEnvironmentFromVariable();
// ported from https://github.com/sherjilozair/char-rnn-tensorflow
return(Parser.Default.ParseArguments <CharRNNTrainingParameters, CharRNNSamplingParameters>(args)
.MapResult(
(CharRNNTrainingParameters train) => Train(train),
(CharRNNSamplingParameters sample) => Sample(sample),
_ => 1));
}
static int Main(string[] args)
{
Console.Title = "GPT-2";
GradientSetup.OptInToUsageDataCollection();
GradientSetup.UseEnvironmentFromVariable();
// force Gradient initialization
tensorflow.tf.no_op();
return(ConsoleCommandDispatcher.DispatchCommand(
ConsoleCommandDispatcher.FindCommandsInSameAssemblyAs(typeof(Gpt2Program)),
args, Console.Out));
}
public static int Main(string[] args)
{
GradientSetup.OptInToUsageDataCollection();
GradientSetup.UseEnvironmentFromVariable();
dynamic config = config_pb2.ConfigProto();
config.gpu_options.allow_growth = true;
tf.keras.backend.set_session(Session.NewDyn(config: config));
return(ConsoleCommandDispatcher.DispatchCommand(
ConsoleCommandDispatcher.FindCommandsInSameAssemblyAs(typeof(CSharpOrNotProgram)),
args, Console.Out));
}
static void Main()
{
GradientLog.OutputWriter = Console.Out;
GradientSetup.UseEnvironmentFromVariable();
// requires Internet connection
(dynamic train, dynamic test) = tf.keras.datasets.fashion_mnist.load_data();
// will be able to do (trainImages, trainLabels) = train;
ndarray trainImages = train.Item1;
ndarray trainLabels = train.Item2;
ndarray testImages = test.Item1;
ndarray testLabels = test.Item2;
bool loaded = 60000 == trainImages.Length;
Debug.Assert(loaded);
var model = new Sequential(new Layer[] {
// will be able to do: new Flatten(kwargs: new { input_shape = (28, 28) }),
new Flatten(kwargs: new PythonDict <string, object> {
static void Main()
{
GradientLog.OutputWriter = Console.Out;
GradientSetup.UseEnvironmentFromVariable();
var input = tf.placeholder(tf.float32, new TensorShape(null, 1), name: "x");
var output = tf.placeholder(tf.float32, new TensorShape(null, 1), name: "y");
var hiddenLayer = tf.layers.dense(input, hiddenSize,
activation: tf.sigmoid_fn,
kernel_initializer: new ones_initializer(),
bias_initializer: new random_uniform_initializer(minval: -x1, maxval: -x0),
name: "hidden");
var model = tf.layers.dense(hiddenLayer, units: 1, name: "output");
var cost = tf.losses.mean_squared_error(output, model);
var training = new GradientDescentOptimizer(learning_rate: learningRate).minimize(cost);
dynamic init = tf.global_variables_initializer();
new Session().UseSelf(session => {
session.run(new[] { init });
foreach (int iteration in Enumerable.Range(0, iterations))
{
var(trainInputs, trainOutputs) = GenerateTestValues();
var iterationDataset = new PythonDict <dynamic, object> {
[input] = trainInputs,
[output] = trainOutputs,
};
session.run(new[] { training }, feed_dict: iterationDataset);
if (iteration % 100 == 99)
{
Console.WriteLine($"cost = {session.run(new[] { cost }, feed_dict: iterationDataset)}");
}
}
var(testInputs, testOutputs) = GenerateTestValues();
var testValues = session.run(new[] { model }, feed_dict: new PythonDict <dynamic, object> {
[input] = testInputs,
});
new variable_scope("hidden", reuse: true).UseSelf(_ => {
Variable w = tf.get_variable("kernel");
Variable b = tf.get_variable("bias");
Console.WriteLine("hidden:");
Console.WriteLine($"kernel= {w.eval()}");
Console.WriteLine($"bias = {b.eval()}");
});
new variable_scope("output", reuse: true).UseSelf(_ => {
Variable w = tf.get_variable("kernel");
Variable b = tf.get_variable("bias");
Console.WriteLine("hidden:");
Console.WriteLine($"kernel= {w.eval()}");
Console.WriteLine($"bias = {b.eval()}");
});
});
}
