.. _neuronperf_rn: NeuronPerf 1.x Release Notes ============================ :ref:`NeuronPerf ` is a lightweight Python library with a simple API that enables fast measurements of performance when running models using Neuron. .. contents:: Table of Contents :local: :depth: 1 NeuronPerf release [1.8.55.0] ----------------------------- Date: 04/01/2024 * Minor updates NeuronPerf release [1.8.15.0] ----------------------------- Date: 12/21/2023 * Minor updates NeuronPerf release [1.8.7.0] ---------------------------- Date: 8/28/2023 * Minor updates NeuronPerf release [1.7.0.0] ---------------------------- Date: 3/28/2023 * Adds trn1/inf2 support for PyTorch and TensorFlow 2.x * Uses new IMDSv2 for obtaining instance types NeuronPerf release [1.6.0.0] ---------------------------- Date: 11/23/2022 * New Evaluation + metrics API (see NeuronPerf Evaluation Guide) * Support map and iterable-type torch datasets * Support custom torch DataLoader args via dataloader_kwargs * New get_report_by_tag utility to identify specific configurations * Python 3.7+ now default from 3.6 * Pricing and sizing info updated for inf1 + trn1 Bug fixes * GPU inputs are now moved correctly NeuronPerf release [1.3.0.0] ---------------------------- Date: 04/29/2022 * Minor updates NeuronPerf release [1.2.0.0] ---------------------------- Date: 03/25/2022 * Initial release of NeuronPerf * Supports PyTorch, TensorFlow, and Apache MXNet. * Supports customizable JSON and CSV reports * See :ref:`neuronperf` for more information.