# Get the first available GPU
DEVICE_ID_LIST = GPUtil.getFirstAvailable()
DEVICE_ID = DEVICE_ID_LIST[0] # grab first element from list
# Set CUDA_VISIBLE_DEVICES to mask out all other GPUs than the first available device id
os.environ["CUDA_VISIBLE_DEVICES"] = str(DEVICE_ID)
except FileNotFoundError:
print("GPU not found")
if __name__ == '__main__':
select_gpu()
# constant.LIMIT_MEMORY = True
(x_train, y_train), (x_test, y_test) = cifar10.load_data()
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
clf = ImageClassifier(path='/home/haifeng/cifar10', verbose=True)
clf.fit(x_train, y_train, time_limit=12 * 60 * 60)
clf.final_fit(x_train, y_train, x_test, y_test)
y = clf.evaluate(x_test, y_test)
print(y)
# MLP for Pima Indians Dataset with 10-fold cross validation
# scores = clf.cross_validate(X, Y, 10)
# print(scores)
# print(np.mean(scores))
# print(np.std(scores))
# split into input (X) and output (Y) variables
# define 10-fold cross validation test harness
os.environ["CUDA_VISIBLE_DEVICES"] = str(DEVICE_ID)
except FileNotFoundError:
print("GPU not found")
if __name__ == '__main__':
select_gpu()
# constant.LIMIT_MEMORY = True
(x_train, y_train), (x_test, y_test) = cifar10.load_data()
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
kernel_lambda = 0.01
for i in range(4):
clf = ImageClassifier(searcher_type='bayesian', path='/home/haifeng/lambda', verbose=True,
searcher_args={'trainer_args': {'max_iter_num': 10},
'default_model_len': 10,
'kernel_lambda': kernel_lambda
})
clf.fit(x_train, y_train, time_limit=3*60*60)
# clf.final_fit(x_train, y_train, x_test, y_test)
y = clf.evaluate(x_test, y_test)
print(kernel_lambda, y)
kernel_lambda *= 10
# MLP for Pima Indians Dataset with 10-fold cross validation
# scores = clf.cross_validate(X, Y, 10)
# print(scores)
# print(np.mean(scores))
# print(np.std(scores))
# split into input (X) and output (Y) variables
# Set CUDA_VISIBLE_DEVICES to mask out all other GPUs than the first available device id
os.environ["CUDA_VISIBLE_DEVICES"] = str(DEVICE_ID)
except FileNotFoundError:
print("GPU not found")
if __name__ == '__main__':
select_gpu()
# constant.LIMIT_MEMORY = True
(x_train, y_train), (x_test, y_test) = cifar10.load_data()
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
trainer_args = {'max_iter_num': 0}
clf = ImageClassifier(searcher_type='bayesian',
path='/Users/haifeng/cifar102',
verbose=True,
searcher_args={'trainer_args': {
'max_iter_num': 0
}})
clf.fit(x_train, y_train, time_limit=12 * 60 * 60)
# clf.final_fit(x_train, y_train)
y = clf.evaluate(x_test, y_test)
print(y)
# MLP for Pima Indians Dataset with 10-fold cross validation
# scores = clf.cross_validate(X, Y, 10)
# print(scores)
# print(np.mean(scores))
# print(np.std(scores))
# split into input (X) and output (Y) variables
# define 10-fold cross validation test harness
from keras.datasets import mnist
from autokeras.classifier import ImageClassifier
from autokeras.constant import Constant
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, ))
Constant.SEARCH_MAX_ITER = 0
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("GPU not found")
if __name__ == '__main__':
select_gpu()
# constant.LIMIT_MEMORY = True
(x_train, y_train), (x_test, y_test) = cifar10.load_data()
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
beta = 0.01
for i in range(4):
clf = ImageClassifier(searcher_type='bayesian',
path='/Users/haifeng/beta',
verbose=True,
searcher_args={
'trainer_args': {
'max_iter_num': 10
},
'default_model_len': 10,
'beta': beta
})
clf.fit(x_train, y_train, time_limit=30)
# clf.final_fit(x_train, y_train, x_test, y_test)
y = clf.evaluate(x_test, y_test)
print(beta, y)
beta *= 10
# MLP for Pima Indians Dataset with 10-fold cross validation
# scores = clf.cross_validate(X, Y, 10)
# print(scores)
# print(np.mean(scores))
# print(np.std(scores))
except FileNotFoundError:
print("GPU not found")
if __name__ == '__main__':
# select_gpu()
os.environ["CUDA_VISIBLE_DEVICES"] = str(0)
constant.LIMIT_MEMORY = True
(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, ))
print(x_test.shape)
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
clf = ImageClassifier(searcher_type='random',
path='/tmp/mnist_random/',
verbose=False)
clf.fit(x_train, y_train, time_limit=12 * 60 * 60)
# clf.final_fit(x_train, y_train)
y = clf.evaluate(x_test, y_test)
print(y)
# MLP for Pima Indians Dataset with 10-fold cross validation
scores = clf.cross_validate(X, Y, 10)
print(scores)
print(np.mean(scores))
print(np.std(scores))
# split into input (X) and output (Y) variables
# define 10-fold cross validation test harness
import sys
from keras.datasets import mnist
import numpy as np
from autokeras import constant
from autokeras.classifier import ImageClassifier
if __name__ == '__main__':
constant.MAX_MODEL_NUM = 1
constant.MAX_ITER_NUM = 1
constant.EPOCHS_EACH = 1
(x_train, y_train), (x_test, y_test) = mnist.load_data()
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
clf = ImageClassifier(searcher_type=sys.argv[1],
path=sys.argv[2],
verbose=False)
clf.fit(x_train, y_train)
y = clf.evaluate(x_test, y_test)
# MLP for Pima Indians Dataset with 10-fold cross validation
scores = clf.cross_validate(X, Y, 2)
print(np.mean(scores))
print(np.std(scores))
# split into input (X) and output (Y) variables
# define 10-fold cross validation test harness
select_gpu()
# os.environ["CUDA_VISIBLE_DEVICES"] = str(3)
# constant.LIMIT_MEMORY = True
(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, ))
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
trainer_args = {
'max_iter_num': 12,
'batch_size': 128,
'optimizer': Adam,
'augment': False
}
clf = ImageClassifier(searcher_type='bayesian',
path='/home/haifeng/mnist',
verbose=True,
trainer_args=trainer_args)
clf.fit(x_train, y_train, time_limit=12 * 60 * 60)
clf.final_fit(x_train, y_train, x_test, y_test, trainer_args, retrain=True)
y = clf.evaluate(x_test, y_test)
print(y)
# MLP for Pima Indians Dataset with 10-fold cross validation
# model = clf.load_searcher().load_best_model()
clf.verbose = True
# scores = clf.cross_validate(X, Y, 10)
# print(scores)
# print(np.mean(scores))
# print(np.std(scores))
# split into input (X) and output (Y) variables
if __name__ == '__main__':
select_gpu()
# os.environ["CUDA_VISIBLE_DEVICES"] = str(3)
# constant.LIMIT_MEMORY = True
(x_train, y_train), (x_test, y_test) = fashion_mnist.load_data()
x_train = x_train.reshape(x_train.shape + (1, ))
x_test = x_test.reshape(x_test.shape + (1, ))
X = np.concatenate((x_train, x_test))
Y = np.concatenate((y_train, y_test))
clf = ImageClassifier(searcher_type='bayesian',
path='/home/haifeng/fashion',
verbose=True,
searcher_args={
'trainer_args': {
'max_iter_num': 10,
'augment': False
},
'default_model_len': 3
})
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)
# MLP for Pima Indians Dataset with 10-fold cross validation
# scores = clf.cross_validate(X, Y, 10)
# print(scores)
# print(np.mean(scores))
# print(np.std(scores))
# Load MNIST data and AutoKeras library
from keras.datasets import mnist
from autokeras.classifier import ImageClassifier
# Set up test and training set
(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,
searcher_args={'trainer_args': {
'max_iter_num': 5
}})
clf.fit(x_train, y_train, time_limit=5 * 60 * 60)
clf.final_fit(x_train,
y_train,
x_test,
y_test,
retrain=False,
trainer_args={'max_iter_num': 10})
y = clf.evaluate(x_test, y_test)
print(y * 100)
best_model = clf.load_searcher().load_best_model()
print(best_model.n_layers)
from torchvision import models
from autokeras.classifier import load_image_dataset
from autokeras.classifier import ImageClassifier
if __name__ == '__main__':
x_train, y_train = load_image_dataset(csv_file_path="train2/label.csv",
images_path="train2")
print(x_train.shape)
print(y_train.shape)
x_test, y_test = load_image_dataset(csv_file_path="test2/label.csv",
images_path="test2")
print(x_test.shape)
print(y_test.shape)
clf = ImageClassifier()
clf.fit(x_train, y_train)
results = clf.predict(x_test)