tensorflow/models · error · ValueError

Batch size must be divisible by number of replicas : {}

Error message

Batch size must be divisible by number of replicas : {}

What it means

Error "Batch size must be divisible by number of replicas : {}" thrown in tensorflow/models.

Source

Thrown at official/legacy/image_classification/resnet/resnet_runnable.py:39

from official.legacy.image_classification.resnet import resnet_model
from official.modeling import grad_utils
from official.modeling import performance
from official.utils.flags import core as flags_core


class ResnetRunnable(orbit.StandardTrainer, orbit.StandardEvaluator):
  """Implements the training and evaluation APIs for Resnet model."""

  def __init__(self, flags_obj, time_callback, epoch_steps):
    self.strategy = tf.distribute.get_strategy()
    self.flags_obj = flags_obj
    self.dtype = flags_core.get_tf_dtype(flags_obj)
    self.time_callback = time_callback

    # Input pipeline related
    batch_size = flags_obj.batch_size
    if batch_size % self.strategy.num_replicas_in_sync != 0:
      raise ValueError(
          'Batch size must be divisible by number of replicas : {}'.format(
              self.strategy.num_replicas_in_sync))

    # As auto rebatching is not supported in
    # `distribute_datasets_from_function()` API, which is
    # required when cloning dataset to multiple workers in eager mode,
    # we use per-replica batch size.
    self.batch_size = int(batch_size / self.strategy.num_replicas_in_sync)

    if self.flags_obj.use_synthetic_data:
      self.input_fn = common.get_synth_input_fn(
          height=imagenet_preprocessing.DEFAULT_IMAGE_SIZE,
          width=imagenet_preprocessing.DEFAULT_IMAGE_SIZE,
          num_channels=imagenet_preprocessing.NUM_CHANNELS,
          num_classes=imagenet_preprocessing.NUM_CLASSES,
          dtype=self.dtype,
          drop_remainder=True)
    else:

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/legacy/image_classification/resnet/resnet_runnable.py:39 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of tensorflow/models@e006f5f0d5 (2026-08-24). Data as JSON: /api/errors/117f70792bb46c33. Report an issue: GitHub.