!pip install -q kaggle

!mkdir -p ~/.kaggle
!cp kaggle.json ~/.kaggle/

!chmod 600 /root/.kaggle/kaggle.json

!kaggle datasets list 

!kaggle datasets download -d yaswanthgali/sport-celebrity-image-classification

!unzip "/content/sport-celebrity-image-classification.zip" -d "/content/drive/MyDrive/Kaggle Datasets"

dir = "/content/drive/MyDrive/Kaggle Datasets/Sports-celebrity images"

from tensorflow.keras.preprocessing.image import ImageDataGenerator
# All images will be rescaled by 1./255.
datagen = ImageDataGenerator( rescale = 1.0/255. , validation_split=0.1, 
                             rotation_range=20,
                             width_shift_range=0.2,
                             height_shift_range=0.2,
                             shear_range=60,
                             horizontal_flip=True)
train_generator = datagen.flow_from_directory(
    directory=dir, target_size=(100, 100), color_mode='rgb', class_mode='categorical', batch_size=32, 
    shuffle=True, seed=2022, subset='training')

validation_generator = datagen.flow_from_directory(
    directory=dir, target_size=(100, 100), color_mode='rgb', class_mode='categorical', 
    batch_size=32, shuffle=True, seed=2022, subset='validation')

import tensorflow as tf

tf.random.set_seed(2022)
model = tf.keras.models.Sequential([
    tf.keras.layers.Conv2D(16, (3,3), activation='relu', input_shape=(100, 100, 3)),
    tf.keras.layers.MaxPooling2D(2,1),
    tf.keras.layers.Flatten(), 
    tf.keras.layers.Dense(1024, activation='relu'), 
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.Dense(512, activation='relu'), 
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.Dense(4, activation='softmax')  
])
from tensorflow.keras.callbacks import EarlyStopping
monitor = EarlyStopping(monitor='val_loss', min_delta=1e-3, patience=5,verbose=3,restore_best_weights=True)
model.compile(optimizer='adam',loss='categorical_crossentropy', metrics = ['accuracy'])

history = model.fit(train_generator, validation_data=validation_generator, epochs=50,
                    verbose=2, callbacks=[monitor])