Solve the pitfalls of mixing tensorflow and keras.

from  import ImageDataGenerator
from  import Sequential
from  import Conv2D, MaxPooling2D, BatchNormalization
from  import Activation, Dropout, Flatten, Dense
 
#Каталог с данными для обучения
train_dir = 'train'
# Каталог с данными для проверки
val_dir = 'val'
# Каталог с данными для тестирования
test_dir = 'val'
 
# Размеры изображения
img_width, img_height = 800, 800
# Размерность тензора на основе изображения для входных данных в нейронную сеть
# backend Tensorflow, channels_last
input_shape = (img_width, img_height, 3)
# Количество эпох
epochs = 1
# Размер мини-выборки
batch_size = 4
# Количество изображений для обучения
nb_train_samples = 300
# Количество изображений для проверки
nb_validation_samples = 25
# Количество изображений для тестирования
nb_test_samples = 25
 
model = Sequential()
 
(Conv2D(32, (7, 7), padding="same", input_shape=input_shape))
(BatchNormalization())
(Activation('tanh'))
(MaxPooling2D(pool_size=(10, 10)))
 
(Conv2D(64, (5, 5), padding="same"))
(BatchNormalization())
(Activation('tanh'))
(MaxPooling2D(pool_size=(10, 10)))
 
(Flatten())
(Dense(512))
(Activation('relu'))
(Dropout(0.5))
(Dense(10, activation='softmax'))
 
(loss='categorical_crossentropy',
              optimizer="Nadam",
              metrics=['accuracy'])
print(())
datagen = ImageDataGenerator(rescale=1. / 255)
 
train_generator = datagen.flow_from_directory(
    train_dir,
    target_size=(img_width, img_height),
    batch_size=batch_size,
    class_mode='categorical')
 
val_generator = datagen.flow_from_directory(
    val_dir,
    target_size=(img_width, img_height),
    batch_size=batch_size,
    class_mode='categorical')
 
test_generator = datagen.flow_from_directory(
    test_dir,
    target_size=(img_width, img_height),
    batch_size=batch_size,
    class_mode='categorical')
 
model.fit_generator(
    train_generator,
    steps_per_epoch=nb_train_samples // batch_size,
    epochs=epochs,
    validation_data=val_generator,
    validation_steps=nb_validation_samples // batch_size)
 
print('Сохраняем сеть')
 
("grib.h5")
print("Сохранение завершено!")

from  import ImageDataGenerator
from  import Sequential
from  import Conv2D, MaxPooling2D, BatchNormalization
from  import Activation, Dropout, Flatten, Dense
from  import load_model
 
print("Загрузка сети")
model = load_model("grib.h5")
print("Загрузка завершена!")

 import ImageDataGenerator
 import Sequential
 import Conv2D, MaxPooling2D, BatchNormalization
 import Activation, Dropout, Flatten, Dense

(tf.global_variables_initializer())

import cv2
import numpy as np
from  import load_model
from  import get_labels
from  import preprocess_input
import time
import os
import tensorflow as tf
from  import gfile
 
["CUDA_VISIBLE_DEVICES"] = "-1"
 
emotion_labels = get_labels('fer2013')
emotion_target_size = (64,64)
#emotion_model_path = './models/emotion_model.hdf5'
#emotion_classifier = load_model(emotion_model_path)
#emotion_target_size = emotion_classifier.input_shape[1:3]
 
path = '/mnt/nas/cv_data/emotion/test'
filelist = (path)
total_num = len(filelist)
timeall = 0
n = 0
 
sess = ()
#(tf.global_variables_initializer())
with ("./trans_model/emotion_mode.pb", 'rb') as f:
    graph_def = ()
    graph_def.ParseFromString(())
    .as_default()
    tf.import_graph_def(graph_def, name='')
 
    pred = .get_tensor_by_name("predictions/Softmax:0")
 
    ######################img##########################
    for item in filelist:
        if (item == '.DS_Store') | (item == ''):
            continue
        src = ((path), item)
        bgr_image = (src)
        gray_image = (bgr_image, cv2.COLOR_BGR2GRAY)
        gray_face = gray_image
        try:
            gray_face = (gray_face, (emotion_target_size))
        except:
            continue
 
        gray_face = preprocess_input(gray_face, True)
        gray_face = np.expand_dims(gray_face, 0)
        gray_face = np.expand_dims(gray_face, -1)
 
        input = .get_tensor_by_name('input_1:0')
        res = (pred, {input: gray_face})
        print("src:", src)
 
        emotion_probability = (res[0])
        emotion_label_arg = (res[0])
        emotion_text = emotion_labels[emotion_label_arg]
        print("predict:", res[0], ",prob:", emotion_probability, ",label:", emotion_label_arg, ",text:",emotion_text)