通过并行计算和云进行深度学习

在本地使用多个 GPU 或在云中扩展深度学习，并以交互方式或批处理作业的形式训练多个网络

使用 Parallel Computing Toolbox™ 在多个 GPU、集群和云中训练深度网络。在本地使用多个 GPU 或在云中扩展深度学习，并以交互方式或批处理作业的形式训练多个网络。要了解有关选项的信息，请参阅Scale Up Deep Learning in Parallel, on GPUs, and in the Cloud。

主题

Scale Up Deep Learning in Parallel, on GPUs, and in the Cloud
Explore options for deep learning with MATLAB^® in parallel and using multiple GPUs, locally or in the cloud.
Deep Learning with MATLAB on Multiple GPUs
Speed up deep neural network training using multiple GPUs locally or in the cloud.
Deep Learning with Big Data
Train deep neural networks using large amounts of data.
Deep Learning in the Cloud
Access MATLAB in the cloud for deep learning.
Use Experiment Manager to Train Networks in Parallel
Run multiple simultaneous trials or one trial at a time on multiple workers.
Run Custom Training Loops on a GPU and in Parallel
Speed up custom training loops by running on a GPU, in parallel using multiple GPUs, or on a cluster.
Cloud AI Workflow Using the Deep Learning Container

An example workflow for training, importing data, and optimizing a deep neural network in the cloud using the Deep Learning Container.
- 步骤 1: Train Network on Amazon Web Services Using MATLAB Deep Learning Container
- 步骤 2: Use Amazon S3 Buckets with MATLAB Deep Learning Container
- 步骤 3: Use Experiment Manager in the Cloud with MATLAB Deep Learning Container

精选示例

使用自动并行支持功能在云中训练网络

此示例说明如何利用 MATLAB 对并行训练的自动支持来训练卷积神经网络。深度学习训练通常需要几小时或几天。借助并行计算，您可以在本地或云集群中使用多个图形处理单元 (GPU) 来加快训练速度。如果您可以使用具有多个 GPU 的计算机，则可以在数据的本地副本上完成此示例。如果要使用更多资源，则可以将深度学习训练扩展到云。要了解有关并行训练选项的详细信息，请参阅Scale Up Deep Learning in Parallel, on GPUs, and in the Cloud。此示例逐步指导您如何利用 MATLAB 的自动并行支持功能在云集群中训练深度学习网络。

打开实时脚本

使用自动多 GPU 支持训练网络

此示例说明如何使用自动并行支持在本地计算机上使用多个 GPU 进行深度学习训练。

打开实时脚本

Create a Deep Learning Experiment for Classification

Train a deep learning network for classification using Experiment Manager.

打开示例

Create a Deep Learning Experiment for Regression

Train a deep learning network for regression using Experiment Manager.

打开示例

并行训练深度学习网络

此示例说明如何在本地计算机上运行多个深度学习试验。使用此示例作为模板，您可以修改网络层和训练选项，以满足您的具体应用需要。无论您有一个还是多个 GPU，都可以使用这种方法。如果您只有一个 GPU，网络会在后台逐个进行训练。本示例中的方法使您能够在进行深度学习试验时继续使用 MATLAB®。

打开实时脚本

使用 parfor 训练多个深度学习网络

此示例说明如何使用 parfor 循环对训练选项执行参数扫描。

打开实时脚本

使用 parfeval 训练多个深度学习网络

此示例说明如何使用 parfeval 对深度学习网络的网络架构深度执行参数扫描，并在训练期间检索数据。

打开实时脚本

将深度学习数据上传到云

此示例说明如何将数据上传到 Amazon S3 存储桶。

打开实时脚本

Train Network in Parallel with Custom Training Loop

Set up a custom training loop to train a network in parallel. In this example, parallel workers train on portions of the overall mini-batch. If you have a GPU, then training happens on the GPU. During training, a DataQueue object sends training progress information back to the MATLAB client.

打开实时脚本

Train Network Using Federated Learning

Train a network using federated learning. Federated learning is a technique that enables you to train a network in a distributed, decentralized way [1].

打开实时脚本

将深度学习批处理作业发送到集群

此示例说明如何将深度学习训练批处理作业发送到集群，以便您可以在训练过程中继续工作或者关闭 MATLAB。

打开实时脚本

Train Network on Amazon Web Services Using MATLAB Deep Learning Container

Train a deep learning network in the cloud using MATLAB® on an Amazon EC2® instance.

打开实时脚本

Use Amazon S3 Buckets with MATLAB Deep Learning Container

Train your deep learning model with training data stored in an Amazon S3™ Bucket and save the trained model to the cloud.

打开实时脚本

Use Experiment Manager in the Cloud with MATLAB Deep Learning Container

Fine-tune your deep learning network by using Experiment Manager in the cloud. Make use of multiple NVIDIA® high-performance GPUs on an AWS® EC2 instance to run multiple experiments in parallel. Tune the hyperparameters of your network and try different network architectures. You can sweep through a range of hyperparameters automatically and save the results for each variation. Compare the results of your experiments to find the best network.

打开实时脚本