itdonmatteratall

import numpy as np
import matplotlib.pyplot as plt
from keras.models import Sequential
from keras.layers.convolutional import Conv2D
from keras.layers.core import Dense, Dropout, Activation, Flatten
from keras.layers.convolutional import Convolution2D, MaxPooling2D, ZeroPadding2D
from keras.layers.normalization import BatchNormalization
#from sklearn.preprocessing import LabelEncoder
#from sklearn.preprocessing import OneHotEncoder
import copy
import cv2
import pickle
#from keras.optimizers import Adam
#from keras.preprocessing.image import ImageDataGenerator
from keras.callbacks import EarlyStopping

dim = 200

print("[INFO] Importing Data...")
pickle_in = open("train_xarr.pickle","rb")
train_xarr = pickle.load(pickle_in)

pickle_in = open("val_xarr.pickle","rb")
val_xarr = pickle.load(pickle_in)

#pickle_in = open("test_xarr.pickle","rb")
#test_xarr = pickle.load(pickle_in)

pickle_in = open("train_yarr.pickle","rb")
train_yarr = pickle.load(pickle_in)

pickle_in = open("val_yarr.pickle","rb")
val_yarr = pickle.load(pickle_in)

#pickle_in = open("test_yarr.pickle","rb")
#test_yarr = pickle.load(pickle_in)

train_xarr = train_xarr.reshape([-1, dim, dim,1])
val_xarr = val_xarr.reshape([-1, dim, dim,1])
#test_xarr = test_xarr.reshape([-1, dim, dim,1])

model = Sequential()
model.add(Conv2D(64, 11, strides=4))
model.add(ZeroPadding2D(2))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=3, strides=2))

model.add(Conv2D(192, 5))
model.add(ZeroPadding2D(2))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=3, strides=2))

model.add(Conv2D(384, 3))
model.add(ZeroPadding2D(1))
model.add(Activation('relu'))

model.add(Conv2D(256, 3))
model.add(ZeroPadding2D(1))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=3, strides=2))

model.add(Flatten())
model.add(Dropout(0.5))
model.add(Dense(4096, input_shape=(6 * 6 * 256, )))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(4096))
model.add(Activation('relu'))
model.add(Dense(70))
model.add(Activation('softmax'))

EPOCHS = 20
#INIT_LR = 1e-3
#BS = 50
#opt = Adam(lr=INIT_LR, decay=INIT_LR / EPOCHS)
model.compile(loss="categorical_crossentropy", optimizer='rmsprop',
	metrics=["accuracy"])
es = EarlyStopping(monitor='val_loss', min_delta=0, patience=0, verbose=0, mode='auto', baseline=None, restore_best_weights=False)
#aug = ImageDataGenerator(rotation_range=30, width_shift_range=0.1,
#	height_shift_range=0.1, shear_range=0.2, zoom_range=0.2,
#	horizontal_flip=True, fill_mode="nearest")

#steps_per_epoch=len(train_xarr) // BS
#validation_data=(val_xarr, val_yarr)
print("[INFO] training network...")
history = model.fit(train_xarr, train_yarr, epochs=EPOCHS, validation_data=(val_xarr, val_yarr), batch_size = 10, verbose=1)
print("できた")

#print (model.summary())

"""
pickle_in = open("test_xarr.pickle","rb")
test_xarr = pickle.load(pickle_in)
pickle_in = open("test_yarr.pickle","rb")
test_yarr = pickle.load(pickle_in)
test_xarr = test_xarr.reshape([-1, dim, dim,1])

error_and_stuff_i_think = history.history

#test_loss, test_acc = model.evaluate(test_xarr, test_yarr)

train_predicts = model.predict_proba(train_xarr)

#yuh = test_yarr-predicts
"""
model.save('aircraft_model.h5')