In [ ]:

%matplotlib inline

%matplotlib inline

Tensorflow image dataset classification model example II¶

An example of a tensorflow network that classifies IndoorScenes images into 67 classes using tensorflow Sequential model.

In [ ]:

import os
import logging
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
os.environ['ABSL_MIN_LOG_LEVEL'] = '3'
# logging.getLogger('tensorflow').setLevel(logging.ERROR)

import tensorflow

import openml
import openml_tensorflow
import pandas as pd
from sklearn import preprocessing
from tensorflow.keras.preprocessing.image import ImageDataGenerator
import tensorflow as tf
from tensorflow.keras import datasets, layers, models
import logging
from keras import regularizers

import warnings
warnings.simplefilter(action='ignore', category=FutureWarning)

import pandas as pd
pd.options.mode.chained_assignment = None  # default='warn'

import os import logging os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3' os.environ['ABSL_MIN_LOG_LEVEL'] = '3' # logging.getLogger('tensorflow').setLevel(logging.ERROR) import tensorflow import openml import openml_tensorflow import pandas as pd from sklearn import preprocessing from tensorflow.keras.preprocessing.image import ImageDataGenerator import tensorflow as tf from tensorflow.keras import datasets, layers, models import logging from keras import regularizers import warnings warnings.simplefilter(action='ignore', category=FutureWarning) import pandas as pd pd.options.mode.chained_assignment = None # default='warn'

Enable logging in order to observe the progress while running the example.

In [ ]:

openml.config.logger.setLevel(logging.DEBUG)

openml.config.logger.setLevel(logging.DEBUG)

In [ ]:

openml.config.apikey = 'KEY'

openml.config.apikey = 'KEY'

In [ ]:

openml_tensorflow.config.epoch = 1 #  small epoch for test runs

IMG_SIZE = (128, 128)
IMG_SHAPE = IMG_SIZE + (3,)
base_learning_rate = 0.0001

# Toy example
datagen = ImageDataGenerator(
    rotation_range=25,
    width_shift_range=0.01,
    height_shift_range=0.01,
    brightness_range=(0.9, 1.1),
    zoom_range=0.1,
    horizontal_flip=True,
    vertical_flip=True,
)

openml_tensorflow.config.datagen = datagen
openml_tensorflow.config.dir = openml.config.get_cache_directory()+'/datasets/45936/Images/'
openml_tensorflow.config.x_col = "Filename"
openml_tensorflow.config.y_col = 'Class_encoded'
openml_tensorflow.config.datagen = datagen
openml_tensorflow.config.batch_size = 32
openml_tensorflow.config.class_mode = "categorical"
openml_tensorflow.config.perform_validation = True

kwargs = {
    'callbacks': tf.keras.callbacks.EarlyStopping(monitor='auc', patience=5),
    'verbose': 2
}
openml_tensorflow.config.kwargs = kwargs

openml_tensorflow.config.epoch = 1 # small epoch for test runs IMG_SIZE = (128, 128) IMG_SHAPE = IMG_SIZE + (3,) base_learning_rate = 0.0001 # Toy example datagen = ImageDataGenerator( rotation_range=25, width_shift_range=0.01, height_shift_range=0.01, brightness_range=(0.9, 1.1), zoom_range=0.1, horizontal_flip=True, vertical_flip=True, ) openml_tensorflow.config.datagen = datagen openml_tensorflow.config.dir = openml.config.get_cache_directory()+'/datasets/45936/Images/' openml_tensorflow.config.x_col = "Filename" openml_tensorflow.config.y_col = 'Class_encoded' openml_tensorflow.config.datagen = datagen openml_tensorflow.config.batch_size = 32 openml_tensorflow.config.class_mode = "categorical" openml_tensorflow.config.perform_validation = True kwargs = { 'callbacks': tf.keras.callbacks.EarlyStopping(monitor='auc', patience=5), 'verbose': 2 } openml_tensorflow.config.kwargs = kwargs

Large CNN

In [ ]:

IMG_SIZE = 128
NUM_CLASSES = 67

# Example tensorflow image classification model. You can do better :)
model = models.Sequential()

# 4 VGG-like CNN blocks
model.add(layers.Conv2D(32, (3, 3), activation='relu', padding='same',
                        input_shape=(IMG_SIZE, IMG_SIZE, 3)))
model.add(layers.BatchNormalization())
model.add(layers.Conv2D(32, (3, 3), activation='relu', padding='same'))
model.add(layers.BatchNormalization())
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Dropout(0.2))


model.add(layers.Conv2D(64, (3, 3), activation='relu', padding='same'))
model.add(layers.BatchNormalization())
model.add(layers.Conv2D(64, (3, 3), activation='relu', padding='same'))
model.add(layers.BatchNormalization())
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Dropout(0.3))


model.add(layers.Conv2D(128, (3, 3), activation='relu', padding='same'))
model.add(layers.BatchNormalization())
model.add(layers.Conv2D(128, (3, 3), activation='relu', padding='same'))
model.add(layers.BatchNormalization())
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Dropout(0.4))

model.add(layers.Conv2D(256, (3, 3), activation='relu', padding='same'))
model.add(layers.BatchNormalization())
model.add(layers.Conv2D(256, (3, 3), activation='relu', padding='same'))
model.add(layers.BatchNormalization())
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Dropout(0.5))

# Pooling and one dense layer + output layer
model.add(layers.GlobalAveragePooling2D())
model.add(layers.Dense(192, activation='relu', kernel_regularizer=regularizers.L2(1e-4)))
model.add(layers.BatchNormalization())
model.add(layers.Dropout(0.25))
model.add(layers.Dense(NUM_CLASSES, activation='softmax'))

model.compile(
    optimizer='adam',
    loss='categorical_crossentropy',
    metrics=['AUC'])

IMG_SIZE = 128 NUM_CLASSES = 67 # Example tensorflow image classification model. You can do better :) model = models.Sequential() # 4 VGG-like CNN blocks model.add(layers.Conv2D(32, (3, 3), activation='relu', padding='same', input_shape=(IMG_SIZE, IMG_SIZE, 3))) model.add(layers.BatchNormalization()) model.add(layers.Conv2D(32, (3, 3), activation='relu', padding='same')) model.add(layers.BatchNormalization()) model.add(layers.MaxPooling2D((2, 2))) model.add(layers.Dropout(0.2)) model.add(layers.Conv2D(64, (3, 3), activation='relu', padding='same')) model.add(layers.BatchNormalization()) model.add(layers.Conv2D(64, (3, 3), activation='relu', padding='same')) model.add(layers.BatchNormalization()) model.add(layers.MaxPooling2D((2, 2))) model.add(layers.Dropout(0.3)) model.add(layers.Conv2D(128, (3, 3), activation='relu', padding='same')) model.add(layers.BatchNormalization()) model.add(layers.Conv2D(128, (3, 3), activation='relu', padding='same')) model.add(layers.BatchNormalization()) model.add(layers.MaxPooling2D((2, 2))) model.add(layers.Dropout(0.4)) model.add(layers.Conv2D(256, (3, 3), activation='relu', padding='same')) model.add(layers.BatchNormalization()) model.add(layers.Conv2D(256, (3, 3), activation='relu', padding='same')) model.add(layers.BatchNormalization()) model.add(layers.MaxPooling2D((2, 2))) model.add(layers.Dropout(0.5)) # Pooling and one dense layer + output layer model.add(layers.GlobalAveragePooling2D()) model.add(layers.Dense(192, activation='relu', kernel_regularizer=regularizers.L2(1e-4))) model.add(layers.BatchNormalization()) model.add(layers.Dropout(0.25)) model.add(layers.Dense(NUM_CLASSES, activation='softmax')) model.compile( optimizer='adam', loss='categorical_crossentropy', metrics=['AUC'])

In [ ]:

# Download the OpenML task for the Indoorscenes dataset.

# task = openml.tasks.get_task(362065)
task = openml.tasks.get_task(362070)

# Run the Keras model on the task (requires an API key).
run = openml.runs.run_model_on_task(model, task, avoid_duplicate_runs=False)

# If you want to publish the run with the onnx file, 
# then you must call openml_tensorflow.add_onnx_to_run() immediately before run.publish(). 
# When you publish, onnx file of last trained model is uploaded. 
# Careful to not call this function when another run_model_on_task is called in between, 
# as during publish later, only the last trained model (from last run_model_on_task call) is uploaded.   
run = openml_tensorflow.add_onnx_to_run(run)

run.publish()

print('URL for run: %s/run/%d?api_key=%s' % (openml.config.server, run.run_id, openml.config.apikey))

# Download the OpenML task for the Indoorscenes dataset. # task = openml.tasks.get_task(362065) task = openml.tasks.get_task(362070) # Run the Keras model on the task (requires an API key). run = openml.runs.run_model_on_task(model, task, avoid_duplicate_runs=False) # If you want to publish the run with the onnx file, # then you must call openml_tensorflow.add_onnx_to_run() immediately before run.publish(). # When you publish, onnx file of last trained model is uploaded. # Careful to not call this function when another run_model_on_task is called in between, # as during publish later, only the last trained model (from last run_model_on_task call) is uploaded. run = openml_tensorflow.add_onnx_to_run(run) run.publish() print('URL for run: %s/run/%d?api_key=%s' % (openml.config.server, run.run_id, openml.config.apikey))

In [ ]:

# Visualize model in netron

from urllib.request import urlretrieve

published_run = openml.runs.get_run(run.run_id)
url = 'https://api.openml.org/data/download/{}/model.onnx'.format(published_run.output_files['onnx_model'])

file_path, _ = urlretrieve(url, 'model.onnx')

import netron
# Visualize the ONNX model using Netron
netron.start(file_path)

# Visualize model in netron from urllib.request import urlretrieve published_run = openml.runs.get_run(run.run_id) url = 'https://api.openml.org/data/download/{}/model.onnx'.format(published_run.output_files['onnx_model']) file_path, _ = urlretrieve(url, 'model.onnx') import netron # Visualize the ONNX model using Netron netron.start(file_path)