1. 程式人生 > >TensorFlow檢視GPU資訊

TensorFlow檢視GPU資訊

檢視是否有GPU

import tensorflow as tf

gpu_device_name = tf.test.gpu_device_name()
print(gpu_device_name)

輸出

/device:GPU:0

GPU是否可用

# 返回True或者False
tf.test.is_gpu_available()
from tensorflow.python.client import device_lib

# 列出所有的本地機器裝置
local_device_protos = device_lib.list_local_devices()
# 列印
#     print(local_device_protos)

# 只打印GPU裝置
[print(x) for x in local_device_protos if x.device_type == 'GPU']

輸出如下

name: "/device:GPU:0"
device_type: "GPU"
memory_limit: 10641965056
locality {
  bus_id: 1
  links {
  }
}
incarnation: 7203835317376623910
physical_device_desc: "device: 0, name: Tesla K80, pci bus id: 0000:00:04.0, compute capability: 3.7"