当我尝试将 model 保存到 tflite 时，Model object 没有属性“_is_graph_network”

Question

Tensorflow version - 1.14.0 Python version - 3.7.5

This is the model I created

from tensorflow import keras
import tensorflow.python.keras.backend as K
from tensorflow.python.keras import callbacks
from tensorflow.python.keras import Sequential
from tensorflow.python.keras.models import Model
from tensorflow.python.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout, BatchNormalization, GlobalAveragePooling2D
from tensorflow.python.keras.preprocessing.image import ImageDataGenerator
from tensorflow.python.keras.callbacks import ModelCheckpoint, EarlyStopping


def create_model_v1():
     model = keras.Sequential()

     model.add(Conv2D(filters = 64, kernel_size = 3, padding='same', activation = 'relu', input_shape=(img_rows, img_cols, color_type)))
     model.add(MaxPooling2D(pool_size = 2))
     model.add(Conv2D(filters = 128, padding='same', kernel_size = 3, activation = 'relu'))
     model.add(MaxPooling2D(pool_size = 2))def create_model_v1():

model_v1 = create_model_v1()

history_v1 = model_v1.fit(x_train, y_train, 
      validation_data=(x_test, y_test),callbacks=callbacks,
      epochs=nb_epoch, batch_size=batch_size, verbose=1)

This is the code for exporting into tflite:

keras_file = 'saved_models/history1.h5'
keras.models.save_model(history_v1, keras_file)

converter = tf.lite.TocoConverter.from_keras_model_file(keras_file)
tflite_model = converter.convert()
open('linear.tflite', 'wb').write(tflite_model)

This is the error:

'History' object has no attribute '_is_graph_network'

Answer 1

You need to pass model_v1 to converter, not history_v1 .

https://www.tensorflow.org/lite/convert/python_api#converting_a_keras_model_