Question

我正在尝试通过编码一些简单的问题来学习张量流：我试图使用直接采样蒙特卡罗方法找到pi的值。

使用for loop执行此操作时，运行时间比我想象的要长得多。我已经看过其他关于类似事情的帖子，我试图按照解决方案，但我认为我仍然必须做错事。

以下是我的代码：

import tensorflow as tf
import numpy as np
import time

n_trials = 50000

tf.reset_default_graph()


x = tf.random_uniform(shape=(), name='x')
y = tf.random_uniform(shape=(), name='y')
r = tf.sqrt(x**2 + y**2)

hit = tf.Variable(0, name='hit')

# perform the monte carlo step
is_inside = tf.cast(tf.less(r, 1), tf.int32)
hit_op = hit.assign_add(is_inside) 

with tf.Session() as sess:
    init_op = tf.global_variables_initializer()
    sess.run(init_op)

    # Make sure no new nodes are added to the graph
    sess.graph.finalize()

    start = time.time()   

    # Run monte carlo trials  -- This is very slow
    for _ in range(n_trials):
        sess.run(hit_op)

    hits = hit.eval()
    print("Pi is {}".format(4*hits/n_trials))
    print("Tensorflow operation took {:.2f} s".format((time.time()-start)))

>>> Pi is 3.15208
>>> Tensorflow operation took 8.98 s

相比之下，在numpy中执行for loop类型解决方案要快一个数量级

start = time.time()   
hits = [ 1 if np.sqrt(np.sum(np.square(np.random.uniform(size=2)))) < 1 else 0 for _ in range(n_trials) ]
a = 0
for hit in hits:
    a+=hit
print("numpy operation took {:.2f} s".format((time.time()-start)))
print("Pi is {}".format(4*a/n_trials))

>>> Pi is 3.14032
>>> numpy operation took 0.75 s

下面附有各种试验的总体执行时间差异图。

请注意：我的问题不是“如何最快地执行此任务”，我认识到有更有效的方法来计算Pi。我只使用它作为基准测试工具来检查张量流对我熟悉的东西（numpy）的性能。

Answer 1

简单，session.run有很多开销，并且它不是设计用于那种方式。通常，例如一个神经网络，你会调用一个session.run进行大量矩阵的十几次乘法运算，然后这个0.2 ms就完全无关紧要了。至于你的情况，你可能想要这样的东西。它比我机器上的numpy版本快5倍。

顺便说一下，你在numpy中完成同样的事情。如果使用loop来减少np.sum而不是np.sum则会慢得多。

    import tensorflow as tf
    import numpy as np
    import time

    n_trials = 50000

    tf.reset_default_graph()

    x = tf.random_uniform(shape=(n_trials,), name='x')
    y = tf.random_uniform(shape=(), name='y')
    r = tf.sqrt(x**2 + y**2)

    hit = tf.Variable(0, name='hit')

    # perform the monte carlo step
    is_inside = tf.cast(tf.less(r, 1), tf.int32)
    hit2= tf.reduce_sum(is_inside)
        #hit_op = hit.assign_add(is_inside) 

    with tf.Session() as sess:
    #    init_op = tf.global_variables_initializer()
        sess.run(tf.initialize_all_variables())

        # Make sure no new nodes are added to the graph
        sess.graph.finalize()

        start = time.time()   

        # Run monte carlo trials  -- This is very slow
        #for _ in range(n_trials):
        sess.run(hit2)

        hits = hit2.eval()
        print("Pi is {}".format(4*hits/n_trials))
        print("Tensorflow operation took {:.2f} s".format((time.time()-start)))

Answer 2

速度缓慢与sess.run中Python和Tensorflow之间的一些通信开销有关，它在循环内执行多次。我建议使用tf.while_loop来执行Tensorflow中的计算。这比numpy更好。

import tensorflow as tf
import numpy as np
import time

n_trials = 50000

tf.reset_default_graph()

hit = tf.Variable(0, name='hit')

def body(ctr):
    x = tf.random_uniform(shape=[2], name='x')
    r = tf.sqrt(tf.reduce_sum(tf.square(x))
    is_inside = tf.cond(tf.less(r,1), lambda: tf.constant(1), lambda: tf.constant(0))
    hit_op = hit.assign_add(is_inside)
    with tf.control_dependencies([hit_op]):
        return ctr + 1

def condition(ctr):
    return ctr < n_trials

with tf.Session() as sess:
    tf.global_variables_initializer().run()
    result = tf.while_loop(condition, body, [tf.constant(0)])

    start = time.time()
    sess.run(result)

    hits = hit.eval()
    print("Pi is {}".format(4.*hits/n_trials))
    print("Tensorflow operation took {:.2f} s".format((time.time()-start)))

评估Tensorflow操作在循环中非常慢

2 个答案: