下面要實(shí)現(xiàn)的功能是：g1和g2并聯(lián)揩悄，placeholder輸入x是3.0, g1實(shí)現(xiàn)系y=3*x，g2實(shí)現(xiàn)y+3, 最后輸出12

文件model_b.py如下：

import tensorflow as tf

from tensorflow.python.framework import graph_util

from tensorflow.python.tools import saved_model_utils

MODEL_SAVE_PATH = "./models/" # 保存模型的路徑

with tf.Graph().as_default() as g2:

????????????????input1 = tf.placeholder(tf.float32,name='g2_input')

????????????????data = tf.Variable(3.)

????????????????mul = tf.add(input1,data)

????????????????tf.identity(mul,name='g2_output')

????????????????init = tf.global_variables_initializer()

????????????????saver = tf.train.Saver()

????????????????with tf.Session(graph=g2) as sess:

????????????????????????????sess.run(init)

????????????????????????????g1def = graph_util.convert_variables_to_constants(sess,sess.graph_def,["g2_output"],

????????????????????????????????????????????????????????????????variable_names_whitelist=None,variable_names_blacklist=None)

? ? ? ? ? ? ? ? ? ? ? ? ? ? #tf.train.write_graph(g1def, MODEL_SAVE_PATH, 'model_g2.pb', as_text=False)

? ? ? ? ? ? ? ? ? ? ? ? ? ?? saver.save(sess, "./models/g2_model.ckpt")

文件model_combined.py如下：

import tensorflow as tf

from tensorflow.python.framework import graph_util

from tensorflow.python.tools import saved_model_utils

MODEL_SAVE_PATH = "./models/" # 保存模型的路徑

#g1和g2并聯(lián)态蒂，輸入x是3.0, g1實(shí)現(xiàn)系y=3*x椒涯，g2實(shí)現(xiàn)y+3, 最后輸出12

with tf.Graph().as_default() as g1:

????????????input1 = tf.placeholder(tf.float32,name='g1_input')

????????????data = tf.Variable(3.)

????????????mul = tf.multiply(input1,data)

????????????tf.identity(mul,name='g1_output')

????????????init = tf.global_variables_initializer()

????????????with tf.Session(graph=g1) as sess:

????????????????????????sess.run(init)

????????????????????????g1def = graph_util.convert_variables_to_constants(sess, sess.graph_def,["g1_output"],

????????????????????????????????????????????????variable_names_whitelist=None,

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? variable_names_blacklist=None)

with tf.Graph().as_default() as g2:

? ? ? ? ? ? ? ? ? with tf.Session(graph=g2) as sess:

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?? saver=tf.train.import_meta_graph('./models/g2_model.ckpt.meta')

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? saver.restore(sess, './models/g2_model.ckpt')

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? g2def = graph_util.convert_variables_to_constants(sess,sess.graph_def,["g2_output"])

##------------------------------------------------------------

with tf.Graph().as_default() as g_combined:

????????????with tf.Session(graph=g_combined) as sess:

????????????????????????x = tf.placeholder(tf.float32, name="my_input")

????????????????????????y = tf.import_graph_def(g1def, input_map={"g1_input:0": x}, return_elements=["g1_output:0"])

????????????????????????z, = tf.import_graph_def(g2def, input_map={"g2_input:0": y}, return_elements=["g2_output:0"])

????????????????????????tf.identity(z, "my_output")

????????????????????????print(sess.run(z,feed_dict={'my_input:0':3.}))

????????????????????????#保存1

????????????????????????#g_combineddef = graph_util.convert_variables_to_constants(sess, sess.graph_def, ["my_output"])

????????????????????????#tf.train.write_graph(g_combineddef, MODEL_SAVE_PATH, 'my_model.pb', as_text=False)

????????????????????????#保存2

????????????????????????#? tf.saved_model.simple_save(sess,

????????????????????????# ? "./modelbase",

????????????????????????# ? inputs={"my_input": x},

????????????????????????# ? outputs={"my_output": z})