ML Community Day is November 9! Join us for updates from TensorFlow, JAX, and more Learn more


Class to synchronize, aggregate gradients and pass them to the optimizer.

Inherits From: Optimizer

This class is deprecated. For synchronous training, please use Distribution Strategies.

In a typical asynchronous training environment, it's common to have some stale gradients. For example, with a N-replica asynchronous training, gradients will be applied to the variables N times independently. Depending on each replica's training speed, some gradients might be calculated from copies of the variable from several steps back (N-1 steps on average). This optimizer avoids stale gradients by collecting gradients from all replicas, averaging them, then applying them to the variables in one shot, after which replicas can fetch the new variables and continue.

The following accumulators/queue are created:

  • N gradient accumulators, one per variable to train. Gradients are pushed to them and the chief worker will wait until enough gradients are collected and then average them before applying to variables. The accumulator will drop all stale gradients (more details in the accumulator op).
  • 1 token queue where the optimizer pushes the new global_step value after all variables are updated.

The following local variable is created:

  • sync_rep_local_step, one per replica. Compared against the global_step in each accumulator to check for staleness of the gradients.

The optimizer adds nodes to the graph to collect gradients and pause the trainers until variables are updated. For the Parameter Server job:

  1. An accumulator is created for each variable, and each replica pushes the gradients into the accumulators instead of directly applying them to the variables.
  2. Each accumulator averages once enough gradients (replicas_to_aggregate) have been accumulated.
  3. Apply the averaged gradients to the variables.
  4. Only after all variables have been updated, increment the global step.
  5. Only after step 4, pushes global_step in the token_queue, once for each worker replica. The workers can now fetch the global step, use it to update its local_step variable and start the next batch. Please note that some workers can consume multiple minibatches, while some may not consume even one. This is because each worker fetches minibatches as long as a token exists. If one worker is stuck for some reason and does not consume a token, another worker can use it.

For the replicas:

  1. Start a step: fetch variables and compute gradients.
  2. Once the gradients have been computed, push them into gradient accumulators. Each accumulator will check the staleness and drop the stale.
  3. After pushing all the gradients, dequeue an updated value of global_step from the token queue and record that step to its local_step variable. Note that this is effectively a barrier.
  4. Start the next batch.


# Create any optimizer to update the variables, say a simple SGD:
opt = GradientDescentOptimizer(learning_rate=0.1)

# Wrap the optimizer with sync_replicas_optimizer with 50 replicas: at each
# step the optimizer collects 50 gradients before applying to variables.
# Note that if you want to have 2 backup replicas, you can change
# total_num_replicas=52 and make sure this number matches how many physical
# replicas you started in your job.
opt = tf.compat.v1.train.SyncReplicasOptimizer(opt, replicas_to_aggregate=50,

# Some models have startup_delays to help stabilize the model but when using
# sync_replicas training, set it to 0.

# Now you can call `minimize()` or `compute_gradients()` and
# `apply_gradients()` normally
training_op = opt.minimize(total_loss, global_step=self.global_step)

# You can create the hook which handles initialization and queues.
sync_replicas_hook = opt.make_session_run_hook(is_chief)

In the training program, every worker will run the train_op as if not synchronized.

with training.MonitoredTrainingSession(
    master=workers[worker_id].target, is_chief=is_chief,
    hooks=[sync_replicas_hook]) as mon_sess:
  while not mon_sess.should_stop():

To use SyncReplicasOptimizer with an Estimator, you need to send sync_replicas_hook while calling the fit.

my_estimator = DNNClassifier(..., optimizer=opt), hooks=[sync_replicas_hook])

opt The actual optimizer that will be used to compute and apply the gradients. Must be one of the Optimizer classes.
replicas_to_aggregate number of replicas to aggregate for each variable update.
total_num_replicas Total number of tasks/workers/replicas, could be different from replicas_to_aggregate. If total_num_replicas > replicas_to_aggregate: it is backup_replicas + replicas_to_aggregate. If total_num_replicas < replicas_to_aggregate: Replicas compute multiple batches per update to variables.
variable_averages Optional ExponentialMovingAverage object, used to maintain moving averages for the variables passed in variables_to_average.
variables_to_average a list of variables that need to be averaged. Only needed if variable_averages is passed in.
use_locking If True use locks for update operation.
name string. Optional name of the returned operation.