Python 张量流：tf.随机均匀vs.tf.常数

所以，我一直在用tensorflow耗尽内存。为了了解发生了什么，我尝试做一个简单的矩阵加法：

C=a+B

，其中

a

和

B

是大小随机的矩阵（1500015000）

我在这里使用了这个非常简单的示例代码，只是对它做了一些修改：

这是我的修改：

import sys 
import numpy as np
import tensorflow as tf
from datetime import datetime

device_name = sys.argv[1]  # Choose device from cmd line. Options: gpu or cpu
shape = int(sys.argv[2])
if device_name == "gpu":
    device_name = "/gpu:0"
else:
    device_name = "/cpu:0"


with tf.device(device_name):
    a = np.random.rand(shape,shape)
    b = np.random.rand(shape,shape)
    sum_operation = tf.add(a,b)


startTime = datetime.now()
with tf.Session(config=tf.ConfigProto(log_device_placement=True)) as session:
    result = session.run(sum_operation)
    print(result)

# It can be hard to see the results on the terminal with lots of output -- add some newlines to improve readability.
print("\n" * 5)
print("Shape:", shape, "Device:", device_name)
print("Time taken:", datetime.now() - startTime)

print("\n" * 5)

我运行了：

python测试文件.pygpu 15000

你认为我为什么会收到这个错误

Traceback (most recent call last):
  File "test_file.py", line 28, in <module>
    result = session.run(sum_operation)
  File "/root/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 789, in run
    run_metadata_ptr)
  File "/root/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 997, in _run
    feed_dict_string, options, run_metadata)
  File "/root/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1132, in _do_run
    target_list, options, run_metadata)
  File "/root/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1139, in _do_call
    return fn(*args)
  File "/root/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1117, in _run_fn
    self._extend_graph()
  File "/root/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1161, in _extend_graph
    add_shapes=self._add_shapes)
  File "/root/anaconda2/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 2324, in _as_graph_def
    raise ValueError("GraphDef cannot be larger than 2GB.")
ValueError: GraphDef cannot be larger than 2GB.

不走运。同样的错误。我可能做错了什么

但具体来说——我试图将无速度与张量流速度与我的一些更复杂的计算（数据集大得多）进行比较。所以，我最终尝试将相同的随机矩阵输入到theano和tensorflow，我猜这种方法不起作用

---

为什么

tf.random\u uniform

有效，而

tf.constant

方法无效？

tf.constant将值嵌入到图形（2GB）限制中，而tf.random\u uniform使用您拥有的任何内存（无限制）动态分配它。tf.constant将值嵌入到图形（2GB）限制中，而tf.random_uniform使用您拥有的任何内存动态分配它（没有限制）

with tf.device(device_name):
    a = np.random.rand(shape,shape)
    b = np.random.rand(shape,shape)
    A = tf.constant(a)
    B = tf.constant(b)
    sum_operation = tf.add(A,B)