You’ll train a simple MLP on MNIST using TensorFlow Core plus DTensor in a data-parallel setup: create a one-dimensional mesh (“batch”), keep model weights replicated (DVariables), shard the global batch across devices via pack/repack, and run a standard loop with tf.GradientTape, custom Adam, and accuracy/loss metrics. The code shows how mesh/layout choices propagate through ops, how to write DTensor-aware layers, and how to evaluate/plot results. Saving is limited today—DTensor models must be fully replicated to export, and saved models lose DTensor annotations.You’ll train a simple MLP on MNIST using TensorFlow Core plus DTensor in a data-parallel setup: create a one-dimensional mesh (“batch”), keep model weights replicated (DVariables), shard the global batch across devices via pack/repack, and run a standard loop with tf.GradientTape, custom Adam, and accuracy/loss metrics. The code shows how mesh/layout choices propagate through ops, how to write DTensor-aware layers, and how to evaluate/plot results. Saving is limited today—DTensor models must be fully replicated to export, and saved models lose DTensor annotations.

Data Parallel MNIST with DTensor and TensorFlow Core

2025/09/09 16:00

Content Overview

  • Introduction
  • Overview of data parallel training with DTensor
  • Setup
  • The MNIST Dataset
  • Preprocessing the data
  • Build the MLP
  • The dense layer
  • The MLP sequential model
  • Training metrics
  • Optimizer
  • Data packing
  • Training
  • Performance evaluation
  • Saving your model
  • Conclusion

\ \ \

Introduction

This notebook uses the TensorFlow Core low-level APIs and DTensor to demonstrate a data-parallel distributed training example.

Visit the Core APIs overview to learn more about TensorFlow Core and its intended use cases. Refer to the DTensor Overview guide and Distributed Training with DTensors tutorial to learn more about DTensor.

This example uses the same model and optimizer as those shown in the Multilayer Perceptrons tutorial. See this tutorial first to get comfortable with writing an end-to-end machine learning workflow with the Core APIs.

\

:::tip Note: DTensor is still an experimental TensorFlow API which means that its features are available for testing, and it is intended for use in test environments only.

:::

\

Overview of data parallel training with DTensor

Before building an MLP that supports distribution, take a moment to explore the fundamentals of DTensor for data parallel training.

DTensor allows you to run distributed training across devices to improve efficiency, reliability and scalability. DTensor distributes the program and tensors according to the sharding directives through a procedure called Single program, multiple data (SPMD) expansion. A variable of a DTensor aware layer is created as dtensor.DVariable, and the constructors of DTensor aware layer objects take additional Layout inputs in addition to the usual layer parameters.

The main ideas for data parallel training are as follows:

  • Model variables are replicated on N devices each.
  • A global batch is split into N per-replica batches.
  • Each per-replica batch is trained on the replica device.
  • The gradient is reduced before weight up data is collectively performed on all replicas.
  • Data parallel training provides nearly linear speed with respect to the number of devices

Setup

DTensor is part of TensorFlow 2.9.0 release.

\

#!pip install --quiet --upgrade --pre tensorflow 

\

import matplotlib from matplotlib import pyplot as plt # Preset Matplotlib figure sizes. matplotlib.rcParams['figure.figsize'] = [9, 6] 

\

import tensorflow as tf import tensorflow_datasets as tfds from tensorflow.experimental import dtensor print(tf.__version__) # Set random seed for reproducible results  tf.random.set_seed(22) 

\

2024-08-15 02:49:40.914029: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:485] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered 2024-08-15 02:49:40.935518: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:8454] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered 2024-08-15 02:49:40.941702: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1452] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered 2.17.0 

Configure 8 virtual CPUs for this experiment. DTensor can also be used with GPU or TPU devices. Given that this notebook uses virtual devices, the speedup gained from distributed training is not noticeable.

\

def configure_virtual_cpus(ncpu):   phy_devices = tf.config.list_physical_devices('CPU')   tf.config.set_logical_device_configuration(phy_devices[0], [         tf.config.LogicalDeviceConfiguration(),     ] * ncpu)  configure_virtual_cpus(8)  DEVICES = [f'CPU:{i}' for i in range(8)] devices = tf.config.list_logical_devices('CPU') device_names = [d.name for d in devices] device_names 

\

WARNING: All log messages before absl::InitializeLog() is called are written to STDERR I0000 00:00:1723690183.661893  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.665603  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.669301  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.672556  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.683679  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.687589  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.691101  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.694059  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.696961  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.700515  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.704018  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690183.706976  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.934382  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.936519  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.938569  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.940700  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.942765  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.944750  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.946705  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.948674  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.950629  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.952626  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.954710  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.956738  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.995780  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.997864  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690184.999851  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.001859  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See mo ['/device:CPU:0',  '/device:CPU:1',  '/device:CPU:2',  '/device:CPU:3',  '/device:CPU:4',  '/device:CPU:5',  '/device:CPU:6',  '/device:CPU:7'] re at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.003740  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.005715  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.007659  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.009659  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.011546  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.014055  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.016445  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723690185.018866  157397 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 

The MNIST Dataset

The dataset is available from TensorFlow Datasets. Split the data into training and testing sets. Only use 5000 examples for training and testing to save time.

\

train_data, test_data = tfds.load("mnist", split=['train[:5000]', 'test[:5000]'], batch_size=128, as_supervised=True) 

Preprocessing the data

Preprocess the data by reshaping it to be 2-dimensional and by rescaling it to fit into the unit interval, [0,1].

\

def preprocess(x, y):   # Reshaping the data   x = tf.reshape(x, shape=[-1, 784])   # Rescaling the data   x = x/255   return x, y  train_data, test_data = train_data.map(preprocess), test_data.map(preprocess) 

Build the MLP

Build an MLP model with DTensor aware layers.

The dense layer

Start by creating a dense layer module that supports DTensor. The dtensor.call_with_layout function can be used to call a function that takes in a DTensor input and produces a DTensor output. This is useful for initializing a DTensor variable, dtensor.DVariable, with a TensorFlow supported function.

\

class DenseLayer(tf.Module):    def __init__(self, in_dim, out_dim, weight_layout, activation=tf.identity):     super().__init__()     # Initialize dimensions and the activation function     self.in_dim, self.out_dim = in_dim, out_dim     self.activation = activation      # Initialize the DTensor weights using the Xavier scheme     uniform_initializer = tf.function(tf.random.stateless_uniform)     xavier_lim = tf.sqrt(6.)/tf.sqrt(tf.cast(self.in_dim + self.out_dim, tf.float32))     self.w = dtensor.DVariable(       dtensor.call_with_layout(           uniform_initializer, weight_layout,           shape=(self.in_dim, self.out_dim), seed=(22, 23),           minval=-xavier_lim, maxval=xavier_lim))      # Initialize the bias with the zeros     bias_layout = weight_layout.delete([0])     self.b = dtensor.DVariable(       dtensor.call_with_layout(tf.zeros, bias_layout, shape=[out_dim]))    def __call__(self, x):     # Compute the forward pass     z = tf.add(tf.matmul(x, self.w), self.b)     return self.activation(z) 

The MLP sequential model

Now create an MLP module that executes the dense layers sequentially.

\

class MLP(tf.Module):    def __init__(self, layers):     self.layers = layers    def __call__(self, x, preds=False):      # Execute the model's layers sequentially     for layer in self.layers:       x = layer(x)     return x 

Performing "data-parallel" training with DTensor is equivalent to tf.distribute.MirroredStrategy. To do this each device will run the same model on a shard of the data batch. So you'll need the following:

  • dtensor.Mesh with a single "batch" dimension
  • dtensor.Layout for all the weights that replicates them across the mesh (using dtensor.UNSHARDED for each axis)
  • dtensor.Layout for the data that splits the batch dimension across the mesh

Create a DTensor mesh that consists of a single batch dimension, where each device becomes a replica that receives a shard from the global batch. Use this mesh to instantiate an MLP mode with the following architecture:

Forward Pass: ReLU(784 x 700) x ReLU(700 x 500) x Softmax(500 x 10)

\

mesh = dtensor.create_mesh([("batch", 8)], devices=DEVICES) weight_layout = dtensor.Layout([dtensor.UNSHARDED, dtensor.UNSHARDED], mesh)  input_size = 784 hidden_layer_1_size = 700 hidden_layer_2_size = 500 hidden_layer_2_size = 10  mlp_model = MLP([     DenseLayer(in_dim=input_size, out_dim=hidden_layer_1_size,                 weight_layout=weight_layout,                activation=tf.nn.relu),     DenseLayer(in_dim=hidden_layer_1_size , out_dim=hidden_layer_2_size,                weight_layout=weight_layout,                activation=tf.nn.relu),     DenseLayer(in_dim=hidden_layer_2_size, out_dim=hidden_layer_2_size,                 weight_layout=weight_layout)]) 

Training metrics

Use the cross-entropy loss function and accuracy metric for training.

\

def cross_entropy_loss(y_pred, y):   # Compute cross entropy loss with a sparse operation   sparse_ce = tf.nn.sparse_softmax_cross_entropy_with_logits(labels=y, logits=y_pred)   return tf.reduce_mean(sparse_ce)  def accuracy(y_pred, y):   # Compute accuracy after extracting class predictions   class_preds = tf.argmax(y_pred, axis=1)   is_equal = tf.equal(y, class_preds)   return tf.reduce_mean(tf.cast(is_equal, tf.float32)) 

Optimizer

Using an optimizer can result in significantly faster convergence compared to standard gradient descent. The Adam optimizer is implemented below and has been configured to be compatible with DTensor. In order to use Keras optimizers with DTensor, refer to the experimentaltf.keras.dtensor.experimental.optimizers module.

\

class Adam(tf.Module):      def __init__(self, model_vars, learning_rate=1e-3, beta_1=0.9, beta_2=0.999, ep=1e-7):       # Initialize optimizer parameters and variable slots       self.model_vars = model_vars       self.beta_1 = beta_1       self.beta_2 = beta_2       self.learning_rate = learning_rate       self.ep = ep       self.t = 1.       self.v_dvar, self.s_dvar = [], []       # Initialize optimizer variable slots       for var in model_vars:         v = dtensor.DVariable(dtensor.call_with_layout(tf.zeros, var.layout, shape=var.shape))         s = dtensor.DVariable(dtensor.call_with_layout(tf.zeros, var.layout, shape=var.shape))         self.v_dvar.append(v)         self.s_dvar.append(s)      def apply_gradients(self, grads):       # Update the model variables given their gradients       for i, (d_var, var) in enumerate(zip(grads, self.model_vars)):         self.v_dvar[i].assign(self.beta_1*self.v_dvar[i] + (1-self.beta_1)*d_var)         self.s_dvar[i].assign(self.beta_2*self.s_dvar[i] + (1-self.beta_2)*tf.square(d_var))         v_dvar_bc = self.v_dvar[i]/(1-(self.beta_1**self.t))         s_dvar_bc = self.s_dvar[i]/(1-(self.beta_2**self.t))         var.assign_sub(self.learning_rate*(v_dvar_bc/(tf.sqrt(s_dvar_bc) + self.ep)))       self.t += 1.       return 

Data packing

Start by writing a helper function for transferring data to the device. This function should use dtensor.pack to send (and only send) the shard of the global batch that is intended for a replica to the device backing the replica. For simplicity, assume a single-client application.

Next, write a function that uses this helper function to pack the training data batches into DTensors sharded along the batch (first) axis. This ensures that DTensor evenly distributes the training data to the 'batch' mesh dimension. Note that in DTensor, the batch size always refers to the global batch size; therefore, the batch size should be chosen such that it can be divided evenly by the size of the batch mesh dimension. Additional DTensor APIs to simplify tf.data integration are planned, so please stay tuned.

\

def repack_local_tensor(x, layout):   # Repacks a local Tensor-like to a DTensor with layout   # This function assumes a single-client application   x = tf.convert_to_tensor(x)   sharded_dims = []    # For every sharded dimension, use tf.split to split the along the dimension.   # The result is a nested list of split-tensors in queue[0].   queue = [x]   for axis, dim in enumerate(layout.sharding_specs):     if dim == dtensor.UNSHARDED:       continue     num_splits = layout.shape[axis]     queue = tf.nest.map_structure(lambda x: tf.split(x, num_splits, axis=axis), queue)     sharded_dims.append(dim)    # Now you can build the list of component tensors by looking up the location in   # the nested list of split-tensors created in queue[0].   components = []   for locations in layout.mesh.local_device_locations():     t = queue[0]     for dim in sharded_dims:       split_index = locations[dim]  # Only valid on single-client mesh.       t = t[split_index]     components.append(t)    return dtensor.pack(components, layout)  def repack_batch(x, y, mesh):   # Pack training data batches into DTensors along the batch axis   x = repack_local_tensor(x, layout=dtensor.Layout(['batch', dtensor.UNSHARDED], mesh))   y = repack_local_tensor(y, layout=dtensor.Layout(['batch'], mesh))   return x, y 

Training

Write a traceable function that executes a single training step given a batch of data. This function does not require any special DTensor annotations. Also write a function that executes a test step and returns the appropriate performance metrics.

\

@tf.function def train_step(model, x_batch, y_batch, loss, metric, optimizer):   # Execute a single training step   with tf.GradientTape() as tape:     y_pred = model(x_batch)     batch_loss = loss(y_pred, y_batch)   # Compute gradients and update the model's parameters   grads = tape.gradient(batch_loss, model.trainable_variables)   optimizer.apply_gradients(grads)   # Return batch loss and accuracy   batch_acc = metric(y_pred, y_batch)   return batch_loss, batch_acc  @tf.function def test_step(model, x_batch, y_batch, loss, metric):   # Execute a single testing step   y_pred = model(x_batch)   batch_loss = loss(y_pred, y_batch)   batch_acc = metric(y_pred, y_batch)   return batch_loss, batch_acc 

Now, train the MLP model for 3 epochs with a batch size of 128.

\

# Initialize the training loop parameters and structures epochs = 3 batch_size = 128 train_losses, test_losses = [], [] train_accs, test_accs = [], [] optimizer = Adam(mlp_model.trainable_variables)  # Format training loop for epoch in range(epochs):   batch_losses_train, batch_accs_train = [], []   batch_losses_test, batch_accs_test = [], []    # Iterate through training data   for x_batch, y_batch in train_data:     x_batch, y_batch = repack_batch(x_batch, y_batch, mesh)     batch_loss, batch_acc = train_step(mlp_model, x_batch, y_batch, cross_entropy_loss, accuracy, optimizer)    # Keep track of batch-level training performance     batch_losses_train.append(batch_loss)     batch_accs_train.append(batch_acc)    # Iterate through testing data   for x_batch, y_batch in test_data:     x_batch, y_batch = repack_batch(x_batch, y_batch, mesh)     batch_loss, batch_acc = test_step(mlp_model, x_batch, y_batch, cross_entropy_loss, accuracy)     # Keep track of batch-level testing     batch_losses_test.append(batch_loss)     batch_accs_test.append(batch_acc)  # Keep track of epoch-level model performance   train_loss, train_acc = tf.reduce_mean(batch_losses_train), tf.reduce_mean(batch_accs_train)   test_loss, test_acc = tf.reduce_mean(batch_losses_test), tf.reduce_mean(batch_accs_test)   train_losses.append(train_loss)   train_accs.append(train_acc)   test_losses.append(test_loss)   test_accs.append(test_acc)   print(f"Epoch: {epoch}")   print(f"Training loss: {train_loss.numpy():.3f}, Training accuracy: {train_acc.numpy():.3f}")   print(f"Testing loss: {test_loss.numpy():.3f}, Testing accuracy: {test_acc.numpy():.3f}") 

\

Epoch: 0 Training loss: 1.850, Training accuracy: 0.343 Testing loss: 1.375, Testing accuracy: 0.504 Epoch: 1 Training loss: 1.028, Training accuracy: 0.674 Testing loss: 0.744, Testing accuracy: 0.782 Epoch: 2 Training loss: 0.578, Training accuracy: 0.839 Testing loss: 0.486, Testing accuracy: 0.869 

Performance evaluation

Start by writing a plotting function to visualize the model's loss and accuracy during training.

\

def plot_metrics(train_metric, test_metric, metric_type):   # Visualize metrics vs training Epochs   plt.figure()   plt.plot(range(len(train_metric)), train_metric, label = f"Training {metric_type}")   plt.plot(range(len(test_metric)), test_metric, label = f"Testing {metric_type}")   plt.xlabel("Epochs")   plt.ylabel(metric_type)   plt.legend()   plt.title(f"{metric_type} vs Training Epochs"); 

\

plot_metrics(train_losses, test_losses, "Cross entropy loss") 

\

\

plot_metrics(train_accs, test_accs, "Accuracy") 

\

Saving your model

The integration of tf.saved_model and DTensor is still under development. As of TensorFlow 2.9.0, tf.saved_model only accepts DTensor models with fully replicated variables. As a workaround, you can convert a DTensor model to a fully replicated one by reloading a checkpoint. However, after a model is saved, all DTensor annotations are lost and the saved signatures can only be used with regular Tensors. This tutorial will be updated to showcase the integration once it is solidified.

Conclusion

This notebook provided an overview of distributed training with DTensor and the TensorFlow Core APIs. Here are a few more tips that may help:

  • The TensorFlow Core APIs can be used to build highly-configurable machine learning workflows with support for distributed training.
  • The DTensor concepts guide and Distributed training with DTensors tutorial contain the most up-to-date information about DTensor and its integrations.

For more examples of using the TensorFlow Core APIs, check out the guide. If you want to learn more about loading and preparing data, see the tutorials on image data loading or CSV data loading.

\n

\ \

:::info Originally published on the TensorFlow website, this article appears here under a new headline and is licensed under CC BY 4.0. Code samples shared under the Apache 2.0 License.

:::

\

Disclaimer: The articles reposted on this site are sourced from public platforms and are provided for informational purposes only. They do not necessarily reflect the views of MEXC. All rights remain with the original authors. If you believe any content infringes on third-party rights, please contact service@support.mexc.com for removal. MEXC makes no guarantees regarding the accuracy, completeness, or timeliness of the content and is not responsible for any actions taken based on the information provided. The content does not constitute financial, legal, or other professional advice, nor should it be considered a recommendation or endorsement by MEXC.

You May Also Like

Polygon Tops RWA Rankings With $1.1B in Tokenized Assets

Polygon Tops RWA Rankings With $1.1B in Tokenized Assets

The post Polygon Tops RWA Rankings With $1.1B in Tokenized Assets appeared on BitcoinEthereumNews.com. Key Notes A new report from Dune and RWA.xyz highlights Polygon’s role in the growing RWA sector. Polygon PoS currently holds $1.13 billion in RWA Total Value Locked (TVL) across 269 assets. The network holds a 62% market share of tokenized global bonds, driven by European money market funds. The Polygon POL $0.25 24h volatility: 1.4% Market cap: $2.64 B Vol. 24h: $106.17 M network is securing a significant position in the rapidly growing tokenization space, now holding over $1.13 billion in total value locked (TVL) from Real World Assets (RWAs). This development comes as the network continues to evolve, recently deploying its major “Rio” upgrade on the Amoy testnet to enhance future scaling capabilities. This information comes from a new joint report on the state of the RWA market published on Sept. 17 by blockchain analytics firm Dune and data platform RWA.xyz. The focus on RWAs is intensifying across the industry, coinciding with events like the ongoing Real-World Asset Summit in New York. Sandeep Nailwal, CEO of the Polygon Foundation, highlighted the findings via a post on X, noting that the TVL is spread across 269 assets and 2,900 holders on the Polygon PoS chain. The Dune and https://t.co/W6WSFlHoQF report on RWA is out and it shows that RWA is happening on Polygon. Here are a few highlights: – Leading in Global Bonds: Polygon holds 62% share of tokenized global bonds (driven by Spiko’s euro MMF and Cashlink euro issues) – Spiko U.S.… — Sandeep | CEO, Polygon Foundation (※,※) (@sandeepnailwal) September 17, 2025 Key Trends From the 2025 RWA Report The joint publication, titled “RWA REPORT 2025,” offers a comprehensive look into the tokenized asset landscape, which it states has grown 224% since the start of 2024. The report identifies several key trends driving this expansion. According to…
Share
BitcoinEthereumNews2025/09/18 00:40
Team Launches AI Tools to Boost KYC and Mainnet Migration for Investors

Team Launches AI Tools to Boost KYC and Mainnet Migration for Investors

The post Team Launches AI Tools to Boost KYC and Mainnet Migration for Investors appeared on BitcoinEthereumNews.com. The Pi Network team has announced the implementation of upgrades to simplify verification and increase the pace of its Mainnet migration. This comes before the token unlock happening this December. Pi Network Integrates AI Tools to Boost KYC Process In a recent blog post, the Pi team said it has improved its KYC process with the same AI technology as Fast Track KYC. This will cut the number of applications waiting for human review by 50%. As a result, more Pioneers will be able to reach Mainnet eligibility sooner. Fast Track KYC was first introduced in September to help new and non-users set up a Mainnet wallet. This was in an effort to reduce the long wait times caused by the previous rule. The old rule required completing 30 mining sessions before qualifying for verification. Fast Track cannot enable migration on its own. However, it is now fully part of the Standard KYC process which allows access to Mainnet. This comes at a time when the network is set for another unlock in December. About 190 million tokens will unlock worth approximately $43 million at current estimates.  These updates will help more Pioneers finish their migration faster especially when there are fewer validators available. This integration allows Pi’s validation resources to serve as a platform utility. In the future, applications that need identity verification or human-verified participation can use this system. Team Releases Validator Rewards Update The Pi Network team provided an update about validator rewards. They expect to distribute the first rewards by the end of Q1 2026. This delay happened because they needed to analyze a large amount of data collected since 2021. Currently, 17.5 million users have completed the KYC process, and 15.7 million users have moved to the Mainnet. However, there are around 3 million users…
Share
BitcoinEthereumNews2025/12/06 16:08
Taiko Makes Chainlink Data Streams Its Official Oracle

Taiko Makes Chainlink Data Streams Its Official Oracle

The post Taiko Makes Chainlink Data Streams Its Official Oracle appeared on BitcoinEthereumNews.com. Key Notes Taiko has officially integrated Chainlink Data Streams for its Layer 2 network. The integration provides developers with high-speed market data to build advanced DeFi applications. The move aims to improve security and attract institutional adoption by using Chainlink’s established infrastructure. Taiko, an Ethereum-based ETH $4 514 24h volatility: 0.4% Market cap: $545.57 B Vol. 24h: $28.23 B Layer 2 rollup, has announced the integration of Chainlink LINK $23.26 24h volatility: 1.7% Market cap: $15.75 B Vol. 24h: $787.15 M Data Streams. The development comes as the underlying Ethereum network continues to see significant on-chain activity, including large sales from ETH whales. The partnership establishes Chainlink as the official oracle infrastructure for the network. It is designed to provide developers on the Taiko platform with reliable and high-speed market data, essential for building a wide range of decentralized finance (DeFi) applications, from complex derivatives platforms to more niche projects involving unique token governance models. According to the project’s official announcement on Sept. 17, the integration enables the creation of more advanced on-chain products that require high-quality, tamper-proof data to function securely. Taiko operates as a “based rollup,” which means it leverages Ethereum validators for transaction sequencing for strong decentralization. Boosting DeFi and Institutional Interest Oracles are fundamental services in the blockchain industry. They act as secure bridges that feed external, off-chain information to on-chain smart contracts. DeFi protocols, in particular, rely on oracles for accurate, real-time price feeds. Taiko leadership stated that using Chainlink’s infrastructure aligns with its goals. The team hopes the partnership will help attract institutional crypto investment and support the development of real-world applications, a goal that aligns with Chainlink’s broader mission to bring global data on-chain. Integrating real-world economic information is part of a broader industry trend. Just last week, Chainlink partnered with the Sei…
Share
BitcoinEthereumNews2025/09/18 03:34