MNIST Data Set

class deepobs.tensorflow.datasets.mnist.mnist(batch_size, train_eval_size=10000)[source]

DeepOBS data set class for the MNIST data set.

Parameters:
  • batch_size (int) -- The mini-batch size to use. Note that, if batch_size is not a divider of the dataset size (60 000 for train, 10 000 for test) the remainder is dropped in each epoch (after shuffling).
  • train_eval_size (int) -- Size of the train eval data set. Defaults to 10 000 the size of the test set.
batch

A tuple (x, y) of tensors, yielding batches of MNIST images (x with shape (batch_size, 28, 28, 1)) and corresponding one-hot label vectors (y with shape (batch_size, 10)). Executing these tensors raises a tf.errors.OutOfRangeError after one epoch.

train_init_op

A tensorflow operation initializing the dataset for the training phase.

train_eval_init_op

A tensorflow operation initializing the testproblem for evaluating on training data.

valid_init_op

A tensorflow operation initializing the testproblem for evaluating on validation data.

test_init_op

A tensorflow operation initializing the testproblem for evaluating on test data.

phase

A string-value tf.Variable that is set to train, train_eval, valid, or test, depending on the current phase. This can be used by testproblems to adapt their behavior to this phase.