用发电机复位 kernel

Question

Why my code with generator with any batch_sizes will reset and my Ram is going to fill

import some important libraries

import tensorflow as tf
import pandas as pd
import matplotlib.pyplot as plt

load and some spliting data

cifar10_data = tf.keras.datasets.cifar10

(train_images, train_labels), (test_images, test_labels) = cifar10_data.load_data()

CLASS_NAMES= ['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']

validation_images, validation_labels = train_images[:5000], train_labels[:5000]
train_images, train_labels = train_images[5000:], train_labels[5000:]

using tf.data form and build some pairs of data

train_ds = tf.data.Dataset.from_tensor_slices((train_images, train_labels))
test_ds = tf.data.Dataset.from_tensor_slices((test_images, test_labels))
validation_ds = tf.data.Dataset.from_tensor_slices((validation_images, validation_labels))

define a preprocessing

def process_images(image, label, size=227):
    # Normalize images to have a mean of 0 and standard deviation of 1
    image = tf.image.per_image_standardization(image)
    # Resize images from 32x32 to 277x277
    image = tf.image.resize(image, (227,227))
    return image, label

using tf.data for understanding size of data

train_ds_size = tf.data.experimental.cardinality(train_ds).numpy()
test_ds_size = tf.data.experimental.cardinality(test_ds).numpy()
validation_ds_size = tf.data.experimental.cardinality(validation_ds).numpy()

print("Training data size:", train_ds_size)
print("Test data size:", test_ds_size)
print("Validation data size:", validation_ds_size)

using tf.data methods for generating data in batch size = 64

train_ds = (train_ds
                  .map(process_images)
                  .shuffle(buffer_size=train_ds_size)
                  .batch(batch_size=64, drop_remainder=True))
test_ds = (test_ds
                  .map(process_images)
                  .shuffle(buffer_size=train_ds_size)
                  .batch(batch_size=64, drop_remainder=True))
validation_ds = (validation_ds
                  .map(process_images)
                  .shuffle(buffer_size=train_ds_size)
                  .batch(batch_size=64, drop_remainder=True))

define the model

model = tf.keras.models.Sequential([
    tf.keras.layers.Conv2D(filters=96, kernel_size=(11,11), strides=(4,4), activation='relu', input_shape=(227,227,3)),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D(pool_size=(3,3), strides=(2,2)),
    tf.keras.layers.Conv2D(filters=256, kernel_size=(5,5), strides=(1,1), activation='relu', padding="same"),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D(pool_size=(3,3), strides=(2,2)),
    tf.keras.layers.Conv2D(filters=384, kernel_size=(3,3), strides=(1,1), activation='relu', padding="same"),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.Conv2D(filters=384, kernel_size=(3,3), strides=(1,1), activation='relu', padding="same"),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.Conv2D(filters=256, kernel_size=(3,3), strides=(1,1), activation='relu', padding="same"),
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.MaxPool2D(pool_size=(3,3), strides=(2,2)),
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(4096, activation='relu'),
    tf.keras.layers.Dropout(0.5),
    tf.keras.layers.Dense(4096, activation='relu'),
    tf.keras.layers.Dropout(0.5),
    tf.keras.layers.Dense(10, activation='softmax')
])

compile the model

model.compile(loss='sparse_categorical_crossentropy', optimizer=tf.optimizers.SGD(lr=0.001), metrics=['accuracy'])
# model.summary()

fit the model on dataset

history = model.fit(train_ds,
          epochs=1,
          validation_data=validation_ds, verbose=1,
          validation_freq=1)

How can I use generator like this code without problem actually I need to use a generator in my code to solve memory problem but I don't know how to use this type of generator

Answer 1

you must reduce shuffle buffer size.

Answer 2

Its just cause of stack of dense layers with so many units (neurons) that will lead into overflow and OOM and as estimated for this model, the dense layers will contain 37752832 and 16781312 trainable parameters which is really enormous model.

So try out again with less units for dense layers, notice the most important thing in convolution models is that the dense layers are just for classifying the extracted feature maps, so its not needed to define dense layers with so many units, so emphasis on defining best model based on convolution base.