[FIXED] ValueError: Dimensions must be equal, but are 96 and 256 in tpu on tensorflow

Issue

I am trying to create a mnist gan which will use tpu. I copied the gan code from here.

Then i made some of my own modifications to run the code on tpu.for making changes i followed this tutorial which shows how to us tpu on tensorflow on tensorflow website.

but thats not working and raising an error here is my code.

# -*- coding: utf-8 -*-
"""Untitled13.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1gbHDaCeFUCGDkkNgAPjGFQIDvZ5NxVfr
"""

# Commented out IPython magic to ensure Python compatibility.
# %tensorflow_version 2.x

import tensorflow as tf
import numpy as np

resolver = tf.distribute.cluster_resolver.TPUClusterResolver(tpu='')
tf.config.experimental_connect_to_cluster(resolver)
# This is the TPU initialization code that has to be at the beginning.
tf.tpu.experimental.initialize_tpu_system(resolver)
print("All devices: ", tf.config.list_logical_devices('TPU'))

strategy = tf.distribute.TPUStrategy(resolver)

import glob
import matplotlib.pyplot as plt
import numpy as np
import os
import PIL
from tensorflow.keras import layers
import time

from IPython import display

(train_images, train_labels), (_, _) = tf.keras.datasets.mnist.load_data()

train_images = train_images.reshape(train_images.shape[0], 28, 28, 1).astype('float32')
train_images = (train_images - 127.5) / 127.5  # Normalize the images to [-1, 1]

BUFFER_SIZE = 60000
BATCH_SIZE = 256

# Batch and shuffle the data
train_dataset = tf.data.Dataset.from_tensor_slices(train_images).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)

def make_generator_model():
    model = tf.keras.Sequential()
    model.add(layers.Dense(7*7*256, use_bias=False, input_shape=(100,)))
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())

    model.add(layers.Reshape((7, 7, 256)))
    assert model.output_shape == (None, 7, 7, 256)  # Note: None is the batch size

    model.add(layers.Conv2DTranspose(128, (5, 5), strides=(1, 1), padding='same', use_bias=False))
    assert model.output_shape == (None, 7, 7, 128)
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())

    model.add(layers.Conv2DTranspose(64, (5, 5), strides=(2, 2), padding='same', use_bias=False))
    assert model.output_shape == (None, 14, 14, 64)
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())

    model.add(layers.Conv2DTranspose(1, (5, 5), strides=(2, 2), padding='same', use_bias=False, activation='tanh'))
    assert model.output_shape == (None, 28, 28, 1)

    return model

def make_discriminator_model():
    model = tf.keras.Sequential()
    model.add(layers.Conv2D(64, (5, 5), strides=(2, 2), padding='same',
                                     input_shape=[28, 28, 1]))
    model.add(layers.LeakyReLU())
    model.add(layers.Dropout(0.3))

    model.add(layers.Conv2D(128, (5, 5), strides=(2, 2), padding='same'))
    model.add(layers.LeakyReLU())
    model.add(layers.Dropout(0.3))

    model.add(layers.Flatten())
    model.add(layers.Dense(1))

    return model

# This method returns a helper function to compute cross entropy loss
cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True, reduction=tf.keras.losses.Reduction.NONE)

EPOCHS = 50
noise_dim = 100
num_examples_to_generate = 16

# You will reuse this seed overtime (so it's easier)
# to visualize progress in the animated GIF)
seed = tf.random.normal([num_examples_to_generate, noise_dim])

def generate_and_save_images(model, epoch, test_input):
  # Notice `training` is set to False.
  # This is so all layers run in inference mode (batchnorm).
  predictions = model(test_input, training=False)

  fig = plt.figure(figsize=(4, 4))

  for i in range(predictions.shape[0]):
      plt.subplot(4, 4, i+1)
      plt.imshow(predictions[i, :, :, 0] * 127.5 + 127.5, cmap='gray')
      plt.axis('off')

  plt.savefig('image_at_epoch_{:04d}.png'.format(epoch))
  plt.show()

def train(dataset, epochs):
  for epoch in range(epochs):
    start = time.time()

    for image_batch in (dataset):
      strategy.run(train_step, args=(image_batch,))

    # Produce images for the GIF as you go
    display.clear_output(wait=True)
    generate_and_save_images(generator,
                             epoch + 1,
                             seed)

    # Save the model every 15 epochs
    if (epoch + 1) % 15 == 0:
      checkpoint.save(file_prefix = checkpoint_prefix)

    print ('Time for epoch {} is {} sec'.format(epoch + 1, time.time()-start))

  # Generate after the final epoch
  display.clear_output(wait=True)
  generate_and_save_images(generator,
                           epochs,
                           seed)

def generator_loss(fake_output):
    return cross_entropy(tf.ones_like(fake_output), fake_output)

def discriminator_loss(real_output, fake_output):
    real_loss = cross_entropy(tf.ones_like(real_output), real_output)
    fake_loss = cross_entropy(tf.zeros_like(fake_output), fake_output)
    total_loss = real_loss + fake_loss
    return total_loss

# Notice the use of `tf.function`
# This annotation causes the function to be "compiled".
@tf.function
def train_step(images):
    noise = tf.random.normal([BATCH_SIZE, noise_dim])

    with tf.GradientTape() as gen_tape, tf.GradientTape() as disc_tape:
      generated_images = generator(noise, training=True)

      real_output = discriminator(images, training=True)
      fake_output = discriminator(generated_images, training=True)

      fake_output_0 = discriminator(generated_images, training=True)

      gen_loss = generator_loss(fake_output_0)
      disc_loss = discriminator_loss(real_output, fake_output)

    gradients_of_generator = gen_tape.gradient(gen_loss, generator.trainable_variables)
    gradients_of_discriminator = disc_tape.gradient(disc_loss, discriminator.trainable_variables)

    generator_optimizer.apply_gradients(zip(gradients_of_generator, generator.trainable_variables))
    discriminator_optimizer.apply_gradients(zip(gradients_of_discriminator, discriminator.trainable_variables))

with strategy.scope():
  generator = make_generator_model()
  generator_optimizer = tf.keras.optimizers.Adam(1e-4)

  discriminator = make_discriminator_model()
  discriminator_optimizer = tf.keras.optimizers.Adam(1e-4)

  checkpoint_dir = './training_checkpoints'
  checkpoint_prefix = os.path.join(checkpoint_dir, "ckpt")
  checkpoint = tf.train.Checkpoint(generator_optimizer=generator_optimizer,
                                  discriminator_optimizer=discriminator_optimizer,
                                  generator=generator,
                                  discriminator=discriminator)
  
  train(train_dataset, EPOCHS)

and the final output is (not showing whole output cause i am in colab and i do not want copy output pf each cell one by one)

ValueError: Dimensions must be equal, but are 96 and 256 for '{{node add}} = AddV2[T=DT_FLOAT](binary_crossentropy/weighted_loss/Mul, binary_crossentropy_1/weighted_loss/Mul)' with input shapes: [96], [256].

Solution

The training data has 60000 instances, if you split them into batches of size 256 you are left a smaller batch of size 60000 % 256 which is 96. Keras also assumes this as a batch if you dont drop it. So in train_step for this batch of size 96, the shape of real_output will be (96, 1) and the shape of fake_output will be (256, 1). As you set reduction to None in cross_entropy loss, the shape will be retained, so shape of real_loss will (96,) and shape of fake_loss will be (256,) then adding them will definitely result in an error.

You may solve this problem this way -

# Let reduction param be default one which is 'auto'/'sum_over_batch_size' reduction type
cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True)

or

# Drop the remainder batch
train_dataset = tf.data.Dataset.from_tensor_slices(train_images).shuffle(BUFFER_SIZE).batch(BATCH_SIZE, drop_remainder=True)

Answered By - Vishwas Chepuri