Optimize TensorFlow Models For Deployment with TensorRT

10 calificaciones
ofrecido por
Coursera Project Network
En este proyecto guiado, tú:

Optimize Tensorflow models using TensorRT (TF-TRT)

Use TF-TRT to optimize several deep learning models at FP32, FP16, and INT8 precision

Observe how tuning TF-TRT parameters affects performance and inference throughput

Clock1.5 hours
CloudNo se necesita descarga
VideoVideo de pantalla dividida
Comment DotsInglés (English)
LaptopSolo escritorio

This is a hands-on, guided project on optimizing your TensorFlow models for inference with NVIDIA's TensorRT. By the end of this 1.5 hour long project, you will be able to optimize Tensorflow models using the TensorFlow integration of NVIDIA's TensorRT (TF-TRT), use TF-TRT to optimize several deep learning models at FP32, FP16, and INT8 precision, and observe how tuning TF-TRT parameters affects performance and inference throughput. Prerequisites: In order to successfully complete this project, you should be competent in Python programming, understand deep learning and what inference is, and have experience building deep learning models in TensorFlow and its Keras API. Note: This course works best for learners who are based in the North America region. We’re currently working on providing the same experience in other regions.

Habilidades que desarrollarás

Deep LearningNVIDIA TensorRT (TF-TRT)Python ProgrammingTensorflowkeras

Aprende paso a paso

En un video que se reproduce en una pantalla dividida con tu área de trabajo, tu instructor te guiará en cada paso:

  1. Introduction and Project Overview

  2. Setup your TensorFlow and TensorRT Runtime

  3. Load the Data and Pre-trained InceptionV3 Model

  4. Create batched Input

  5. Load the TensorFlow SavedModel

  6. Get Baseline for Prediction Throughput and Accuracy

  7. Convert a TensorFlow saved model into a TF-TRT Float32 Graph

  8. Benchmark TF-TRT Float32

  9. Convert to TF-TRT Float16 and Benchmark

  10. Converting to TF-TRT INT8

Cómo funcionan los proyectos guiados

Tu espacio de trabajo es un escritorio virtual directamente en tu navegador, no requiere descarga.

En un video de pantalla dividida, tu instructor te guía paso a paso

Preguntas Frecuentes

Preguntas Frecuentes

¿Tienes más preguntas? Visita el Centro de Ayuda al Alumno.