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.
Share Insights

You May Also Like

China Launches Cross-Border QR Code Payment Trial

China Launches Cross-Border QR Code Payment Trial

The post China Launches Cross-Border QR Code Payment Trial appeared on BitcoinEthereumNews.com. Key Points: Main event involves China initiating a cross-border QR code payment trial. Alipay and Ant International are key participants. Impact on financial security and regulatory focus on illicit finance. China’s central bank, led by Deputy Governor Lu Lei, initiated a trial of a unified cross-border QR code payment gateway with Alipay and Ant International as participants. This pilot addresses cross-border fund risks, aiming to enhance financial security amid rising money laundering through digital channels, despite muted crypto market reactions. China’s Cross-Border Payment Gateway Trial with Alipay The trial operation of a unified cross-border QR code payment gateway marks a milestone in China’s financial landscape. Prominent entities such as Alipay and Ant International are at the forefront, participating as the initial institutions in this venture. Lu Lei, Deputy Governor of the People’s Bank of China, highlighted the systemic risks posed by increased cross-border fund flows. Changes are expected in the dynamics of digital transactions, potentially enhancing transaction efficiency while tightening regulations around illicit finance. The initiative underscores China’s commitment to bolstering financial security amidst growing global fund movements. “The scale of cross-border fund flows is expanding, and the frequency is accelerating, providing opportunities for risks such as cross-border money laundering and terrorist financing. Some overseas illegal platforms transfer funds through channels such as virtual currencies and underground banks, creating a ‘resonance’ of risks at home and abroad, posing a challenge to China’s foreign exchange management and financial security.” — Lu Lei, Deputy Governor, People’s Bank of China Bitcoin and Impact of China’s Financial Initiatives Did you know? China’s latest initiative echoes the Payment Connect project of June 2025, furthering real-time cross-boundary remittances and expanding its influence on global financial systems. As of September 17, 2025, Bitcoin (BTC) stands at $115,748.72 with a market cap of $2.31 trillion, showing a 0.97%…
Share
BitcoinEthereumNews2025/09/18 05:28
Essential UK Stablecoin Regulations Align with US Approach – What You Need to Know

Essential UK Stablecoin Regulations Align with US Approach – What You Need to Know

BitcoinWorld Essential UK Stablecoin Regulations Align with US Approach – What You Need to Know Are you wondering how the UK’s new stablecoin regulations will affect your cryptocurrency activities? The Bank of England just announced that their approach to stablecoin regulations will closely mirror the United States’ framework. This crucial development signals a coordinated global effort to bring stability and security to the digital asset space. What Do the New Stablecoin Regulations Mean for You? According to Deputy Governor Sarah Breeden, the Bank of England plans to implement specific holding limits as part of their stablecoin regulations. These limits include temporary caps of £20,000 for individual users and £10 million for corporations. This approach ensures that the UK’s stablecoin regulations maintain consistency with international standards while protecting consumers. The alignment of stablecoin regulations between the UK and US represents a significant step toward global cryptocurrency standardization. Moreover, this coordinated effort helps prevent regulatory arbitrage and creates a more predictable environment for businesses operating in both markets. Why Are Stablecoin Regulations So Important? Stablecoin regulations serve multiple critical purposes in the cryptocurrency ecosystem. First, they provide consumer protection against potential market manipulation and fraud. Second, they establish clear guidelines for businesses operating in this space. Finally, proper stablecoin regulations help maintain financial stability by ensuring these digital assets don’t pose systemic risks. Consumer Protection: Limits prevent excessive exposure to single assets Market Confidence: Clear rules encourage institutional participation Financial Stability: Prevents systemic risks from unregulated growth International Cooperation: Aligned approaches reduce regulatory conflicts How Will These Stablecoin Regulations Impact the Market? The implementation of these stablecoin regulations will likely have immediate effects on how users and businesses interact with digital assets. The £20,000 individual limit means retail investors must diversify their stablecoin holdings across multiple providers or assets. Similarly, the £10 million corporate cap requires larger entities to implement sophisticated treasury management strategies. These stablecoin regulations also create opportunities for innovation in custody solutions and risk management tools. Financial technology companies can develop products that help users comply with the new requirements while maximizing their operational efficiency within the regulatory framework. What Challenges Do Stablecoin Regulations Present? While the alignment of stablecoin regulations between the UK and US provides clarity, it also introduces certain challenges. Market participants must adapt to new compliance requirements and reporting standards. Additionally, the temporary nature of the holding limits means businesses need flexible systems that can accommodate future regulatory changes. However, the benefits of having clear stablecoin regulations outweigh these transitional challenges. The framework provides much-needed certainty for investors and businesses alike, potentially accelerating mainstream adoption of digital assets. Key Takeaways from the New Stablecoin Regulations The Bank of England’s announcement about stablecoin regulations marks a pivotal moment for the cryptocurrency industry. By aligning with US approaches, the UK demonstrates its commitment to fostering a secure and innovative digital asset ecosystem. These stablecoin regulations balance innovation with necessary safeguards, creating a foundation for sustainable growth. As these stablecoin regulations take effect, market participants should prepare for increased compliance requirements while recognizing the long-term benefits of regulatory clarity. The coordinated approach between major financial centers sets a positive precedent for global cryptocurrency regulation. Frequently Asked Questions When will the new stablecoin regulations take effect? The Bank of England hasn’t announced a specific implementation date, but the framework is expected to be introduced in the coming months following further consultation with industry stakeholders. How do the UK stablecoin regulations compare to other countries? The UK’s approach closely mirrors US regulations, creating alignment between two major financial markets. This coordination helps prevent regulatory fragmentation and supports global cryptocurrency adoption. Can individuals hold more than £20,000 in stablecoins? The £20,000 limit applies per individual user per service provider. Users can potentially hold additional stablecoins with different regulated providers, though they should monitor their overall exposure. Will these regulations affect existing stablecoin holdings? Existing holdings will likely need to comply with the new limits once the regulations take effect. Users should prepare to adjust their portfolios accordingly during any transition period. Do these regulations apply to all types of stablecoins? The framework primarily targets fiat-backed stablecoins, which maintain reserves in traditional currencies. Other types of stablecoins may face different regulatory treatment based on their underlying structures. How will enforcement of these regulations work? The Bank of England and Financial Conduct Authority will jointly oversee compliance, with authorized firms required to implement systems that ensure adherence to the holding limits and other requirements. Found this analysis of stablecoin regulations helpful? Share this article with your network on social media to help others understand these important regulatory developments. Your shares help spread valuable information throughout the cryptocurrency community. To learn more about the latest cryptocurrency trends, explore our article on key developments shaping digital assets institutional adoption. This post Essential UK Stablecoin Regulations Align with US Approach – What You Need to Know first appeared on BitcoinWorld.
Share
Coinstats2025/11/06 03:55