import tensorflow as tf
from keras.applications import Xception
from keras.utils import multi_gpu_model
import numpy as np

num_samples = 1000
height = 224
width = 224
num_classes = 1000

# 实例化基础模型（或者「模版」模型）。
# 我们推荐在 CPU 设备范围内做此操作，
# 这样模型的权重就会存储在 CPU 内存中。
# 否则它们会存储在 GPU 上，而完全被共享。
with tf.device('/cpu:0'):
 model = Xception(weights=None,
   input_shape=(height, width, 3),
   classes=num_classes)

# 复制模型到 8 个 GPU 上。
# 这假设你的机器有 8 个可用 GPU。
parallel_model = multi_gpu_model(model, gpus=8)
parallel_model.compile(loss='categorical_crossentropy',
   optimizer='rmsprop')

# 生成虚拟数据
x = np.random.random((num_samples, height, width, 3))
y = np.random.random((num_samples, num_classes))

# 这个 `fit` 调用将分布在 8 个 GPU 上。
# 由于 batch size 是 256, 每个 GPU 将处理 32 个样本。
parallel_model.fit(x, y, epochs=20, batch_size=256)

# 通过模版模型存储模型（共享相同权重）：
model.save('my_model.h5')

class ParallelModelCheckpoint(ModelCheckpoint):
 def __init__(self,model,filepath, monitor='val_loss', verbose=0,
   save_best_only=False, save_weights_only=False,
   mode='auto', period=1):
 self.single_model = model
 super(ParallelModelCheckpoint,self).__init__(filepath, monitor, verbose,save_best_only, save_weights_only,mode, period)

 def set_model(self, model):
 super(ParallelModelCheckpoint,self).set_model(self.single_model)

checkpoint = ParallelModelCheckpoint(original_model)

# Model where a shared LSTM is used to encode two different sequences in parallel
input_a = keras.Input(shape=(140, 256))
input_b = keras.Input(shape=(140, 256))

shared_lstm = keras.layers.LSTM(64)

# Process the first sequence on one GPU
with tf.device_scope('/gpu:0'):
 encoded_a = shared_lstm(tweet_a)
# Process the next sequence on another GPU
with tf.device_scope('/gpu:1'):
 encoded_b = shared_lstm(tweet_b)

# Concatenate results on CPU
with tf.device_scope('/cpu:0'):
 merged_vector = keras.layers.concatenate([encoded_a, encoded_b],
      axis=-1)

server = tf.train.Server.create_local_server()
sess = tf.Session(server.target)

from keras import backend as K
K.set_session(sess)