NumPy arrays and TensorFlow Tensors的区别和联系

1，tensor的特点

• Tensors can be backed by accelerator memory (like GPU, TPU).
• Tensors are immutable

2，双向转换

• TensorFlow operations automatically convert NumPy ndarrays to Tensors.
• NumPy operations automatically convert Tensors to NumPy ndarrays

3，转换的代价

Tensors can be explicitly converted to NumPy ndarrays by invoking the .numpy() method on them. These conversions are typically cheap as the array and Tensor share the underlying memory representation if possible. However, sharing the underlying representation isn't always possible since the Tensor may be hosted in GPU memory while NumPy arrays are always backed by host memory, and the conversion will thus involve a copy from GPU to host memory.

4，使用tensor时如何测定和选择gpu

x = tf.random_uniform([3, 3])

print("Is there a GPU available: "),

print(tf.test.is_gpu_available())

print("Is the Tensor on GPU #0: "),

print(x.device.endswith('GPU:0'))

print(tf.test.is_built_with_cuda())

5，显式指定运行的xpu

import time

def time_matmul(x):

start = time.time()

for loop in range(10):

tf.matmul(x, x)

result = time.time()-start

print("10 loops: {:0.2f}ms".format(1000*result))

# Force execution on CPU

print("On CPU:")

with tf.device("CPU:0"):

x = tf.random_uniform([900, 900])

assert x.device.endswith("CPU:0")

time_matmul(x)

# Force execution on GPU #0 if available

if tf.test.is_gpu_available():

with tf.device("GPU:0"): # Or GPU:1 for the 2nd GPU, GPU:2 for the 3rd etc.

x = tf.random_uniform([1000, 1000])

assert x.device.endswith("GPU:0")

time_matmul(x)