This article explains TensorFlow Core—the low-level, composable APIs for developers who need fine-grained control: researchers, framework authors, and power users extending Keras. It outlines use cases (custom layers/losses/optimizers, bespoke training loops, distribution with DTensor, advanced m... Read more
Hackernoon - Sep 09
This in-depth guide explains how to effectively use tf.function in TensorFlow 2 to convert Python code into performant, portable dataflow graphs. You’ll learn the differences between eager execution and graph mode, the rules of tracing, how to control retracing, and best practices like using inpu... Read more
Hackernoon - Aug 13
This guide walks you through using the TensorFlow Profiler with TensorBoard to identify and fix GPU performance bottlenecks in both single- and multi-GPU setups. It covers profiling workflows, diagnosing input pipeline issues, increasing GPU utilization, reducing kernel launch delays, optimizing ... Read more
Hackernoon - Aug 12
TensorFlow’s Grappler is a built-in graph optimization system that speeds up model execution by simplifying computations, optimizing memory use, and improving hardware efficiency. It offers a range of optimizers—like constant folding, pruning, remapping, and auto mixed precision—that you can togg... Read more
Hackernoon - Aug 12
This guide walks you through using mixed precision in Keras with TensorFlow to accelerate model training while reducing memory usage. By combining float16 or bfloat16 with float32 for key computations, you can achieve up to 3x faster performance on modern GPUs, TPUs, and Intel CPUs without sacrif... Read more
Hackernoon - Aug 12
This tutorial walks you through fine-tuning a ResNet-18 model from TensorFlow’s Model Garden for classifying images in the CIFAR-10 dataset. You’ll learn how to set up the environment, configure the model, prepare the dataset, visualize training data, and run training/evaluation steps. By the end... Read more
Hackernoon - Aug 12
The TensorFlow Model Garden is Google’s open-source hub for high-performance machine learning models in vision and NLP, offering both official and research implementations. It provides a powerful training experiment framework, specialized ML operations, and Orbit for streamlined custom training l... Read more
Hackernoon - Aug 12
This guide walks you through training machine learning models on Google Cloud TPUs and TPU Pods using TensorFlow. You’ll learn how to initialize TPUs, apply manual device placement, use tf.distribute.TPUStrategy, and train models via both Keras high-level APIs and custom training loops. The tutor... Read more
Hackernoon - Aug 12
