'''
Created on 2018年8月25日
@author: moonlit
'''
from keras.datasets import mnist
from autokeras.image_supervised import ImageClassifier
if __name__ == '__main__':
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train = x_train.reshape(x_train.shape + (1, ))
x_test = x_test.reshape(x_test.shape + (1, ))
clf = ImageClassifier(verbose=True)
clf.fit(x_train, y_train, time_limit=12 * 60 * 60)
clf.final_fit(x_train, y_train, x_test, y_test, retrain=True)
y = clf.evaluate(x_test, y_test)
print(y)
print(files)
x_train, y_train = load_image_dataset(csv_file_path=train_labels,
images_path=train_path)
print(x_train.shape)
print(y_train.shape)
x_val, y_val = load_image_dataset(csv_file_path=validation_labels,
images_path=validation_path)
print(x_val.shape)
print(y_val.shape)
# Searching for the Best Model
clf = ImageClassifier(verbose=True,
searcher_args={'trainer_args': {
'max_iter_num': 25
}})
clf.fit(x_train, y_train,
time_limit=4 * 60 * 60) # default time_limit is 24 hours
# Fitting the Best Model Found During Search
clf.final_fit(x_train,
y_train,
x_val,
y_val,
retrain=True,
trainer_args={'max_iter_num': 10})
# Evaluating the Model
print(clf.evaluate(x_val, y_val))
if test_csv_file_path is not None and test_images_path is not None:
x_train, y_train = load_image_dataset(
csv_file_path=train_csv_file_path,
images_path=train_images_path)
else:
sys.exit("No train data input!!")
if test_csv_file_path is not None and test_images_path is not None:
x_test, y_test = load_image_dataset(
csv_file_path=test_csv_file_path, images_path=test_images_path)
do_evaluate = True
else:
do_evaluate = False
if agent_percept == 'image' and agent_func_interpert_input == 'autokeras':
clf = ImageClassifier(verbose=True)
clf.fit(x_train, y_train, time_limit=autokeras_timeout)
clf.final_fit(x_train, y_train, x_test, y_test, retrain=True)
else:
sys.exit("No valid percept type or INTERPERT-INPUT function!!")
if do_evaluate is True:
y = clf.evaluate(x_test, y_test)
print('test data evaluate = ' + str(y))
clf.load_searcher().load_best_model().produce_keras_model().save(
'my_model.h5')
keras_model = load_model('my_model.h5')
if keras_output_activation == 'softmax':
#
# Beispiel für die Benutzung von AutoKeras
#
from keras.datasets import mnist
from autokeras.image_supervised import ImageClassifier
# Wichtig: ansonsten RuntimeError
if __name__ == '__main__':
# Laden der MNIST Daten
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train = x_train.reshape(x_train.shape + (1, ))
x_test = x_test.reshape(x_test.shape + (1, ))
# Instanzierung des ImageClassifiers von AutoKeras
clf = ImageClassifier(verbose=True,
searcher_args={'trainer_args': {
'max_iter_num': 3
}})
# Ähnlich wie bei Keras, wird hier eine fit() Funktion benutzt.
# Der Parameter time_limit (in Sekunden) gibt an, wie lange AutoKeras nach den optimalen Modell suchen soll
clf.fit(x_train, y_train, time_limit=60 * 10)
# Wenn ein Model gefunden wurde, wird es erneut trainiert
clf.final_fit(x_train, y_train, x_test, y_test, retrain=True)
# Ähnlich Keras kann hier die predict()-Funktion aufgerufen werden
results = clf.predict(x_test)
print(results)
# 코드 추가 필요 : Windows를 사용하는 경우
# 참조 페이지 : https://github.com/jhfjhfj1/autokeras/issues/76
# if you are on windows, torch with CUDA and multiprocessing do not seem to work well together.
# Also please try to wrap your code in trainalutokeras_raw.py in:
# if __name__ == "__main__"
if __name__ == '__main__':
# 1. Load data
(x_train, y_train), (x_test, y_test) = cifar10.load_data()
# 2. Define classifier : 경로를 설정하지 않으면 default 경로로 저장됨
# pip 버전 default 경로 : (AutoKeras를 실행한 위치 root)\tmp\autokeras
# git 버전 default경로 : (AutoKeras를 설치한 위치 username)\AppData\Local\Temp\autokeras
clf = ImageClassifier(verbose=True,
searcher_args={'trainer_args': {
'max_iter_num': 5
}})
#clf = ImageClassifier(verbose=True, path='d:/tmp/autokeras/', searcher_args={'trainer_args':{'max_iter_num':5}})
# 3. Fitting
# time_limit : 초단위, 시간이 지나면 작동을 자동으로 멈춥니다.
clf.fit(x_train, y_train, time_limit=24 * 60 * 60)
# 3-1. Load saved model (3번 항목 실행후 3 주석처리 필요)
# if you reloaded your saved clf, y_encoder & data_transformer should be defined like following.
from autokeras.preprocessor import OneHotEncoder, DataTransformer
from autokeras.constant import Constant
clf.y_encoder = OneHotEncoder()
clf.y_encoder.fit(y_train)
clf.data_transformer = DataTransformer(x_train,
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Thu Sep 27 16:46:35 2018
@author: Mohamed Laradji
"""
from autokeras.image_supervised import ImageClassifier, load_image_dataset
x_train, y_train = load_image_dataset(csv_file_path="data/train/labels.csv",
images_path="data/train")
x_val, y_val = load_image_dataset(csv_file_path="data/val/labels.csv",
images_path="data/val")
if __name__ == '__main__':
clf = ImageClassifier(verbose=True)
clf.fit(x_train, y_train,
time_limit=12 * 60 * 60) # Currently results in errors.
clf.final_fit(x_train, y_train, x_val, y_val, retrain=True)
y = clf.evaluate(x_val, y_val)