用如下代码可检测tensorflow的能使用设备情况:
import os
from tensorflow.python.client import device_lib
os.environ["TF_CPP_MIN_LOG_LEVEL"] = "99"
if __name__ == "__main__":
print(device_lib.list_local_devices())
输出:
[name: "/cpu:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 5641726301910989523
]
可知目前系统可用设备 为CPU
在Python环境中输入:
import numpy
import tensorflow as tf
a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
c = tf.matmul(a, b)
sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
print(sess.run(c))
之后就会出现详细的信息:
Device mapping:
/job:localhost/replica:0/task:0/device:GPU:0 -> device: 0, name: Tesla K40c, pci bus
id: 0000:05:00.0
b: /job:localhost/replica:0/task:0/device:GPU:0
a: /job:localhost/replica:0/task:0/device:GPU:0
MatMul: /job:localhost/replica:0/task:0/device:GPU:0
[[ 22. 28.]
[ 49. 64.]]
就能看到是GPU在工作还是CPU再工作了。
检查tensorflow的版本情况:
pip3 list
各应用版本为:
1 2 |
|
升级了tensorflow版本,忘记了升级tensorflow-gpu版本,现在两个版本有代差,而tensorflow默认选择版本高的CPU版本来计算了。
升级tensorflow-gpu:
pip3 install --index-url http://pypi.douban.com/simple --trusted-host pypi.douban.com --upgrade tensorflow-gpu
再次检测可用设备情况,结果如下:
2018-09-04 10:51:22.996654: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1405] Found device 0 with properties: name: GeForce GTX 1060 5GB major: 6 minor: 1 memoryClockRate(GHz): 1.7085 pciBusID: 0000:01:00.0 totalMemory: 4.94GiB freeMemory: 4.23GiB 2018-09-04 10:51:22.996666: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1484] Adding visible gpu devices: 0 2018-09-04 10:51:23.189923: I tensorflow/core/common_runtime/gpu/gpu_device.cc:965] Device interconnect StreamExecutor with strength 1 edge matrix: 2018-09-04 10:51:23.189953: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0 2018-09-04 10:51:23.189959: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] 0: N 2018-09-04 10:51:23.190105: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1097] Created TensorFlow device (/device:GPU:0 with 3969 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1060 5GB, pci bus id: 0000:01:00.0, compute capability: 6.1) [name: "/device:CPU:0" device_type: "CPU" memory_limit: 268435456 locality { } incarnation: 15761951721866580392 , name: "/device:GPU:0" device_type: "GPU" memory_limit: 4161994752 locality { bus_id: 1 links { } } incarnation: 16066826866269340415 physical_device_desc: "device: 0, name: GeForce GTX 1060 5GB, pci bus id: 0000:01:00.0, compute capability: 6.1" ]
设置使用GPU:
使用 tf.device('/gpu:1') 指定Session在第二块GPU上运行:
import tensorflow as tf
with tf.device('/gpu:1'):
v1 = tf.constant([1.0, 2.0, 3.0], shape=[3], name='v1')
v2 = tf.constant([1.0, 2.0, 3.0], shape=[3], name='v2')
sumV12 = v1 + v2
with tf.Session(config=tf.ConfigProto(log_device_placement=True)) as sess:
设置使用cpu:
tensorflow中不同的GPU使用/gpu:0和/gpu:1区分,而CPU不区分设备号,统一使用 /cpu:0
import tensorflow as tf
with tf.device('/cpu:0'):
v1 = tf.constant([1.0, 2.0, 3.0], shape=[3], name='v1')
v2 = tf.constant([1.0, 2.0, 3.0], shape=[3], name='v2')
sumV12 = v1 + v2
with tf.Session(config=tf.ConfigProto(log_device_placement=True)) as sess:
print sess.run(sumV12)