Keras Gpu Tensorflow

TensorFlow recommends running estimates on a variety of kinds of devices, including CPU plus GPU. They are designated including string identifiers for instance:

• "/device: CPU:0": This CPU of our machine.
• "/GPU:0": Short-hand characters for the initial GPU of your machine that is evident over TensorFlow.
• "/job:localhost/replica:0/task:0/device:GPU:1": completely qualified name of the second GPU of our machine which is evident to TensorFlow.

If a TensorFlow operation has both CPU and GPU implementations, by default, the GPU devices will be given priority when the operation is allocated to a device. For instance, tf.matmul holds both CPU and GPU kernels. Upon a system including devices CPU:0 also GPU:0, the GPU:0 device will be chosen to run tf.matmul except you explicitly ask running that on a different device.

Improve your knowledge in Keras from scratch using Keras Online Course.