이제 하이퍼파라미터 튜닝은 케라스 튜너에게 맡기세요

이제 발표도 인공지능에게 맡기세요

하 하 하

정승환 대표님 및 라이온로켓분들 감사합니다.

숨막히는 대본

허 허 허

...

​

​

​

​

model = Sequential()

model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(28, 28, 1)))

model.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))

model.add(Flatten())

model.add(Dense(20, activation='relu'))

model.add(Dropout(0.2))

model.add(Dense(10, activation='softmax'))

model.compile(loss='categorical_crossentropy', optimizer=Adam(0.001))

model.summary()

​

​

​

​

model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(28, 28, 1)))

​

​

​

​

model.add(Dense(20, activation='relu'))

model.add(Dropout(0.2))

​

​

​

​

​

model.add(Dense(10, activation='softmax'))

​

​

​

​

​

model.compile(loss='categorical_crossentropy', optimizer=Adam(0.001))

​

​

​

​

​

• 첫번째 Conv2D의 필터 수 : L1_NUM_FILTERS
• 두번째 Conv2D의 필터 수 : L2_NUM_FILTERS
• Dense - Dropout 은닉층의 반복횟수 : NUM_LAYERS
• Dense 레이어의 출력뉴런 수 : NUM_DIMS
• Dropout 레이어의 드랍율 : DROPOUT_RATE
• 최적화기의 학습률 : LR

​

​

​

​

LR = Choice('learning_rate', [0.001, 0.0005, 0.0001], group='optimizer')

DROPOUT_RATE = Linear('dropout_rate', 0.0, 0.5, 5, group='dense')

NUM_DIMS = Range('num_dims', 8, 32, 8, group='dense')

NUM_LAYERS = Range('num_layers', 1, 3, group='dense')

L2_NUM_FILTERS = Range('l2_num_filters', 8, 64, 8, group='cnn')

L1_NUM_FILTERS = Range('l1_num_filters', 8, 64, 8, group='cnn')

​

​

​

​

​

​

def model_fn():

LR = Choice('learning_rate', [0.001, 0.0005, 0.0001], group='optimizer')

DROPOUT_RATE = Linear('dropout_rate', 0.0, 0.5, 5, group='dense')

NUM_DIMS = Range('num_dims', 8, 32, 8, group='dense')

NUM_LAYERS = Range('num_layers', 1, 3, group='dense')

L2_NUM_FILTERS = Range('l2_num_filters', 8, 64, 8, group='cnn')

L1_NUM_FILTERS = Range('l1_num_filters', 8, 64, 8, group='cnn')

​

​

​

​

model = Sequential()

model.add(Conv2D(L1_NUM_FILTERS, kernel_size=(3, 3), activation='relu'))

model.add(Conv2D(L2_NUM_FILTERS, kernel_size=(3, 3), activation='relu'))

model.add(Flatten())

for _ in range(NUM_LAYERS):

model.add(Dense(NUM_DIMS, activation='relu'))

model.add(Dropout(DROPOUT_RATE))

model.add(Dense(10, activation='softmax'))

model.compile(loss='categorical_crossentropy', optimizer=Adam(LR))

� return model

​

​

​

​

​

tuner = Tuner(model_fn, 'val_accuracy' epoch_budget=500, max_epochs=5)

tuner.search(train_data,validation_data=validation_data)

​

짝 짝 짝