October 2018
Beginner
362 pages
9h 32m
English
Now that we have our GPUs set up for deep learning, let's learn how to utilize them. In TensorFlow, GPUs are represented as strings:
Distributed training is the practice of training a network across several GPUs, and it's becoming an increasingly common way to train models in the AI field. TensorFlow, Keras, and PyTorch all support distributed training.
Logging is an operation in TensorFlow to assign a particular set of commands to a particular GPU or CPU on your system. With logging, we can also parallelize our operations, meaning that we can distribute training across several ...
Read now
Unlock full access