This document is relevant for: Inf1, Inf2, Trn1, Trn2, Trn3

TensorBoard

TensorBoard integration with AWS Neuron provides powerful visualization and debugging capabilities for machine learning workloads. The Neuron TensorBoard plugins enable developers to monitor training progress, analyze model performance, and debug compilation issues through familiar TensorBoard interfaces.

TensorBoard for Trn1

TensorBoard Plugin for NeuronX (Trn1)

Comprehensive guide for using the TensorBoard Neuron plugin on Trn1 instances, including installation, configuration, and advanced visualization features.

Profiling PyTorch NeuronX (torch-neuronx) with TensorBoard

Step-by-step tutorial for monitoring PyTorch training progress on Trn1 instances using TensorBoard scalars, metrics visualization, and performance tracking.

This document is relevant for: Inf1, Inf2, Trn1, Trn2, Trn3