Fashion-MNIST VAE¶
-
class
deepobs.tensorflow.testproblems.fmnist_vae.fmnist_vae(batch_size, l2_reg=None)[source]¶ DeepOBS test problem class for a variational autoencoder (VAE) on Fashion-MNIST.
The network has been adapted from the here and consists of an encoder:
- With three convolutional layers with each
64filters. - Using a leaky ReLU activation function with \(\alpha = 0.3\)
- Dropout layers after each convolutional layer with a rate of
0.2.
and an decoder:
- With two dense layers with
24and49units and leaky ReLU activation. - With three deconvolutional layers with each
64filters. - Dropout layers after the first two deconvolutional layer with a rate of
0.2. - A final dense layer with
28 x 28units and sigmoid activation.
No regularization is used.
Parameters: - batch_size (int) -- Batch size to use.
- l2_reg (float) -- No L2-Regularization (weight decay) is used in this
test problem. Defaults to
Noneand any input here is ignored.
-
dataset¶ The DeepOBS data set class for Fashion-MNIST.
-
train_init_op¶ A tensorflow operation initializing the test problem for the training phase.
-
train_eval_init_op¶ A tensorflow operation initializing the test problem for evaluating on training data.
-
test_init_op¶ A tensorflow operation initializing the test problem for evaluating on test data.
-
losses¶ A tf.Tensor of shape (batch_size, ) containing the per-example loss values.
-
regularizer¶ A scalar tf.Tensor containing a regularization term. Will always be
0.0since no regularizer is used.
- With three convolutional layers with each