Tensorflow Gpu Support，检查 tf 是否正在使用我的 GPU？

Question

我尝试获得 tensorflows gpu 支持并安装了所有必需的软件，例如 cuda... 所以现在我想检查我的代码是否真的使用了我的 gpu，但我不太确定这意味着什么（由于“找到具有属性的设备 0”而不确定）：

print(tf.config.experimental.list_physical_devices())

输出是：

2.3.0
2020-12-11 13:28:30.220505: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library nvcuda.dll
2020-12-11 13:28:30.249113: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties: 
pciBusID: 0000:01:00.0 name: GeForce RTX 2070 with Max-Q Design computeCapability: 7.5
coreClock: 1.185GHz coreCount: 36 deviceMemorySize: 8.00GiB deviceMemoryBandwidth: 357.69GiB/s
2020-12-11 13:28:30.249473: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudart64_101.dll
2020-12-11 13:28:30.253079: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cublas64_10.dll
2020-12-11 13:28:30.256500: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cufft64_10.dll
2020-12-11 13:28:30.257517: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library curand64_10.dll
2020-12-11 13:28:30.261527: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cusolver64_10.dll
2020-12-11 13:28:30.263554: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cusparse64_10.dll
2020-12-11 13:28:30.271417: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudnn64_7.dll
2020-12-11 13:28:30.271623: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
[PhysicalDevice(name='/physical_device:CPU:0', device_type='CPU'), PhysicalDevice(name='/physical_device:XLA_CPU:0', device_type='XLA_CPU'), PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU'), PhysicalDevice(name='/physical_device:XLA_GPU:0', device_type='XLA_GPU')]```

Answer 1

你安装什么pip install tensorflow或pip install tensorflow-gpu？

如果是后者，从tf.config.experimental.list_physical_devices()的输出看，你的GPU正在使用，因为tensorflow可以找到你的GeForce RTX 2070 GPU并成功打开tensorflow使用GPU所需的所有库，所以不用担心它

给你一个我安装前者的电脑的例子，输出是这样的：

2020-12-11 20:40:16.336824: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll 
not found
2020-12-11 20:40:16.337402: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2020-12-11 20:40:19.716047: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library nvcuda.dll
2020-12-11 20:40:20.813638: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties: 
pciBusID: 0000:01:00.0 name: GeForce RTX 2060 computeCapability: 7.5
coreClock: 1.2GHz coreCount: 30 deviceMemorySize: 6.00GiB deviceMemoryBandwidth: 245.91GiB/s
2020-12-11 20:40:20.814470: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll 
not found
2020-12-11 20:40:20.814951: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cublas64_10.dll'; dlerror: cublas64_10.dll not found
2020-12-11 20:40:20.815328: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cufft64_10.dll'; dlerror: cufft64_10.dll not 
found
2020-12-11 20:40:20.816220: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'curand64_10.dll'; dlerror: curand64_10.dll not found
2020-12-11 20:40:20.816770: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cusolver64_10.dll'; dlerror: cusolver64_10.dll not found
2020-12-11 20:40:20.818276: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cusparse64_10.dll'; dlerror: cusparse64_10.dll not found
2020-12-11 20:40:20.818777: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'cudnn64_7.dll'; dlerror: cudnn64_7.dll not found
2020-12-11 20:40:20.818963: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1753] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
[PhysicalDevice(name='/physical_device:CPU:0', device_type='CPU'), PhysicalDevice(name='/physical_device:XLA_CPU:0', device_type='XLA_CPU'), PhysicalDevice(name='/physical_device:XLA_GPU:0', device_type='XLA_GPU')]

使用日志检查：

import tensorflow as tf
tf.debugging.set_log_device_placement(True)

# Create some tensors
a = tf.constant([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]])
b = tf.constant([[1.0, 2.0], [3.0, 4.0], [5.0, 6.0]])
c = tf.matmul(a, b)

print(c)

如果它打印出这样的东西，那么它就会在 GPU 中运行（如果以 CPU:0 结尾，那么 CPU 正在使用）：

Executing op MatMul in device /job:localhost/replica:0/task:0/device:GPU:0
tf.Tensor(
[[22. 28.]
 [49. 64.]], shape=(2, 2), dtype=float32)