Centre d'aide Centre d'aide

La traduction de cette page n'est pas à jour. Cliquez ici pour voir la dernière version en anglais.

Boucles d’apprentissage personnalisées de Deep Learning

Personnaliser les boucles d’apprentissage du Deep Learning et les fonctions de perte

Si la fonction trainingOptions ne propose pas les options d’apprentissage dont vous avez besoin pour votre tâche ou que les couches de sortie personnalisées ne supportent pas les fonctions de perte dont vous avez besoin, vous pouvez définir une boucle d’apprentissage personnalisée. Pour les réseaux qui ne peuvent pas être créés avec des graphes de couches, vous pouvez définir des réseaux personnalisés comme une fonction. Pour en savoir plus, veuillez consulter Define Custom Training Loops, Loss Functions, and Networks.

Fonctions

développer tout

Réseaux personnalisés

`dlnetwork`	Deep learning network for custom training loops (depuis R2019b)
`resetState`	Reset state parameters of neural network
`plot`	Plot neural network architecture
`addInputLayer`	Add input layer to network (depuis R2022b)
`addLayers`	Add layers to layer graph or network
`removeLayers`	Remove layers from layer graph or network
`connectLayers`	Connect layers in layer graph or network
`disconnectLayers`	Disconnect layers in layer graph or network
`replaceLayer`	Replace layer in layer graph or network
`summary`	Print network summary (depuis R2022b)
`initialize`	Initialize learnable and state parameters of a `dlnetwork` (depuis R2021a)
`networkDataLayout`	Deep learning network data layout for learnable parameter initialization (depuis R2022b)

Boucles d’apprentissage personnalisées

`forward`	Compute deep learning network output for training (depuis R2019b)
`predict`	Compute deep learning network output for inference (depuis R2019b)
`adamupdate`	Update parameters using adaptive moment estimation (Adam) (depuis R2019b)
`rmspropupdate`	Update parameters using root mean squared propagation (RMSProp) (depuis R2019b)
`sgdmupdate`	Update parameters using stochastic gradient descent with momentum (SGDM) (depuis R2019b)
`dlupdate`	Update parameters using custom function (depuis R2019b)
`minibatchqueue`	Create mini-batches for deep learning (depuis R2020b)
`onehotencode`	Encode data labels into one-hot vectors (depuis R2020b)
`onehotdecode`	Decode probability vectors into class labels (depuis R2020b)
`padsequences`	Pad or truncate sequence data to same length (depuis R2021a)
`trainingProgressMonitor`	Monitor and plot training progress for deep learning custom training loops (depuis R2022b)

Différenciation automatique

`dlarray`	Deep learning array for customization (depuis R2019b)
`dlgradient`	Compute gradients for custom training loops using automatic differentiation (depuis R2019b)
`dlfeval`	Evaluate deep learning model for custom training loops (depuis R2019b)
`dims`	Étiquettes des dimensions de `dlarray` (depuis R2019b)
`finddim`	Find dimensions with specified label (depuis R2019b)
`stripdims`	Remove `dlarray` data format (depuis R2019b)
`extractdata`	Extract data from `dlarray` (depuis R2019b)
`isdlarray`	Check if object is `dlarray` (depuis R2020b)
`functionToLayerGraph`	(To be removed) Convert deep learning model function to a layer graph (depuis R2019b)

Opérations de Deep Learning

`dlconv`	Deep learning convolution (depuis R2019b)
`dltranspconv`	Deep learning transposed convolution (depuis R2019b)
`lstm`	Long short-term memory (depuis R2019b)
`gru`	Gated recurrent unit (depuis R2020a)
`attention`	Dot-product attention (depuis R2022b)
`embed`	Embed discrete data (depuis R2020b)
`fullyconnect`	Sum all weighted input data and apply a bias (depuis R2019b)
`dlode45`	Deep learning solution of nonstiff ordinary differential equation (ODE) (depuis R2021b)
`relu`	Apply rectified linear unit activation (depuis R2019b)
`leakyrelu`	Apply leaky rectified linear unit activation (depuis R2019b)
`gelu`	Apply Gaussian error linear unit (GELU) activation (depuis R2022b)
`batchnorm`	Normalize data across all observations for each channel independently (depuis R2019b)
`crosschannelnorm`	Cross channel square-normalize using local responses (depuis R2020a)
`groupnorm`	Normalize data across grouped subsets of channels for each observation independently (depuis R2020b)
`instancenorm`	Normalize across each channel for each observation independently (depuis R2021a)
`layernorm`	Normalize data across all channels for each observation independently (depuis R2021a)
`avgpool`	Pool data to average values over spatial dimensions (depuis R2019b)
`maxpool`	Pool data to maximum value (depuis R2019b)
`maxunpool`	Unpool the output of a maximum pooling operation (depuis R2019b)
`softmax`	Apply softmax activation to channel dimension (depuis R2019b)
`sigmoid`	Appliquer l’activation sigmoïde (depuis R2019b)

Opérations de perte de Deep Learning

`sigmoid`	Appliquer l’activation sigmoïde (depuis R2019b)
`crossentropy`	Cross-entropy loss for classification tasks (depuis R2019b)
`l1loss`	L₁ loss for regression tasks (depuis R2021b)
`l2loss`	L₂ loss for regression tasks (depuis R2021b)
`huber`	Huber loss for regression tasks (depuis R2021a)
`mse`	Half mean squared error (depuis R2019b)
`ctc`	Connectionist temporal classification (CTC) loss for unaligned sequence classification (depuis R2021a)

Accélération des fonctions de Deep Learning

`dlaccelerate`	Accelerate deep learning function for custom training loops (depuis R2021a)
`AcceleratedFunction`	Accelerated deep learning function (depuis R2021a)
`clearCache`	Clear accelerated deep learning function trace cache (depuis R2021a)

Rubriques

Boucles d’apprentissage personnalisées

Train Deep Learning Model in MATLAB
Learn how to training deep learning models in MATLAB^®.
Define Custom Training Loops, Loss Functions, and Networks
Learn how to define and customize deep learning training loops, loss functions, and networks using automatic differentiation.
Train Network Using Custom Training Loop
This example shows how to train a network that classifies handwritten digits with a custom learning rate schedule.
Specify Training Options in Custom Training Loop
Learn how to specify common training options in a custom training loop.
Define Model Loss Function for Custom Training Loop
Learn how to define a model loss function for a custom training loop.
Update Batch Normalization Statistics in Custom Training Loop
This example shows how to update the network state in a custom training loop.
Make Predictions Using dlnetwork Object
This example shows how to make predictions using a dlnetwork object by splitting data into mini-batches.
Monitor Custom Training Loop Progress
Track and plot custom training loop progress.
Train Network with Multiple Outputs
This example shows how to train a deep learning network with multiple outputs that predict both labels and angles of rotations of handwritten digits.
Classify Videos Using Deep Learning with Custom Training Loop
This example shows how to create a network for video classification by combining a pretrained image classification model and a sequence classification network.
Train Image Classification Network Robust to Adversarial Examples
This example shows how to train a neural network that is robust to adversarial examples using fast gradient sign method (FGSM) adversarial training.
Train Neural ODE Network
This example shows how to train an augmented neural ordinary differential equation (ODE) network.
Train Robust Deep Learning Network with Jacobian Regularization
Train a neural network that is robust to adversarial examples using a Jacobian regularization scheme.
Solve Ordinary Differential Equation Using Neural Network
This example shows how to solve an ordinary differential equation (ODE) using a neural network.
Assemble Multiple-Output Network for Prediction
This example shows how to assemble a multiple output network for prediction.
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.

Fonctions de modèle

Train Network Using Model Function
This example shows how to create and train a deep learning network by using functions rather than a layer graph or a dlnetwork.
Update Batch Normalization Statistics Using Model Function
This example shows how to update the network state in a network defined as a function.
Make Predictions Using Model Function
This example shows how to make predictions using a model function by splitting data into mini-batches.
Initialize Learnable Parameters for Model Function
Learn how to initialize learnable parameters for custom training loops using a model function.
Train Latent ODE Network with Irregularly Sampled Time-Series Data
This example shows how to train a latent ordinary differential equation (ODE) autoencoder with time-series data that is sampled at irregular time intervals.
Multivariate Time Series Anomaly Detection Using Graph Neural Network
This example shows how to detect anomalies in multivariate time series data using a graph neural network (GNN).

Différenciation automatique

List of Functions with dlarray Support
View the list of functions that support dlarray objects.
Automatic Differentiation Background
Learn how automatic differentiation works.
Use Automatic Differentiation In Deep Learning Toolbox
How to use automatic differentiation in deep learning.

Accélération des fonctions de Deep Learning

Deep Learning Function Acceleration for Custom Training Loops
Accelerate model functions and model loss functions for custom training loops by caching and reusing traces.
Accelerate Custom Training Loop Functions
This example shows how to accelerate deep learning custom training loop and prediction functions.
Check Accelerated Deep Learning Function Outputs
This example shows how to check that the outputs of accelerated functions match the outputs of the underlying function.
Evaluate Performance of Accelerated Deep Learning Function
This example shows how to evaluate the performance gains of using an accelerated function.

Exemples présentés

Train Generative Adversarial Network (GAN)

Train Generative Adversarial Network (GAN)

Train a generative adversarial network to generate images.

Ouvrir le live script

Train Conditional Generative Adversarial Network (CGAN)

Train Conditional Generative Adversarial Network (CGAN)

Train a conditional generative adversarial network to generate images.

Ouvrir le live script

Train Wasserstein GAN with Gradient Penalty (WGAN-GP)

Train Wasserstein GAN with Gradient Penalty (WGAN-GP)

Train a Wasserstein generative adversarial network with a gradient penalty (WGAN-GP) to generate images.

Ouvrir le live script

Train Fast Style Transfer Network

Train Fast Style Transfer Network

Train a network to transfer the style of an image to a second image. It is based on the architecture defined in [1].

Ouvrir le live script

Train Network Using Custom Training Loop

Train Network Using Custom Training Loop

Train a network that classifies handwritten digits with a custom learning rate schedule.

Ouvrir le live script

Train Network Using Model Function

Train Network Using Model Function

Create and train a deep learning network by using functions rather than a layer graph or a dlnetwork. The advantage of using functions is the flexibility to describe a wide variety of networks. The disadvantage is that you must complete more steps and prepare your data carefully. This example uses images of handwritten digits, with the dual objectives of classifying the digits and determining the angle of each digit from the vertical.

Ouvrir le live script

Train a Twin Network for Dimensionality Reduction

Train a Twin Network for Dimensionality Reduction

Train a twin neural network with shared weights to compare handwritten digits using dimensionality reduction.

Ouvrir le live script

Train a Twin Neural Network to Compare Images

Train a Twin Neural Network to Compare Images

Train a twin neural network with shared weights to identify similar images of handwritten characters.

Ouvrir le live script

Image Captioning Using Attention

Image Captioning Using Attention

Train a deep learning model for image captioning using attention.

Ouvrir le live script

Multilabel Text Classification Using Deep Learning

Multilabel Text Classification Using Deep Learning

Classify text data that has multiple independent labels.

Ouvrir le live script

Sequence-to-Sequence Translation Using Attention

Sequence-to-Sequence Translation Using Attention

Convert decimal strings to Roman numerals using a recurrent sequence-to-sequence encoder-decoder model with attention.

Ouvrir le live script

Train Network in Parallel with Custom Training Loop

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.

Ouvrir le live script

Train Variational Autoencoder (VAE) to Generate Images

Train Variational Autoencoder (VAE) to Generate Images

Train a deep learning variational autoencoder (VAE) to generate images.

Ouvrir le live script

Train Network Using Cyclical Learning Rate for Snapshot Ensembling

Train Network Using Cyclical Learning Rate for Snapshot Ensembling

Train a network to classify images of objects using a cyclical learning rate schedule and snapshot ensembling for better test accuracy. In the example, you learn how to use a cosine function for the learning rate schedule, take snapshots of the network during training to create a model ensemble, and add L2-norm regularization (weight decay) to the training loss.

Ouvrir le live script

Solve Partial Differential Equations Using Deep Learning

Solve Partial Differential Equations Using Deep Learning

Solve Burger's equation using deep learning.

Ouvrir le live script

Solve Partial Differential Equation with L-BFGS Method and Deep Learning

Solve Partial Differential Equation with L-BFGS Method and Deep Learning

Train a physics informed neural network (PINN) to numerically compute the solution of the Burger's equation by using the limited-memory BFGS (L-BFGS) algorithm.

Ouvrir le live script

Train Neural ODE Network

Train Neural ODE Network

Train an augmented neural ordinary differential equation (ODE) network.

Ouvrir le live script

Dynamical System Modeling Using Neural ODE

Dynamical System Modeling Using Neural ODE

Train a neural network with neural ordinary differential equations (ODEs) to learn the dynamics of a physical system.

Ouvrir le live script

Node Classification Using Graph Convolutional Network

Node Classification Using Graph Convolutional Network

Classify nodes in a graph using a graph convolutional network (GCN).

Ouvrir le live script

Multilabel Graph Classification Using Graph Attention Networks

Multilabel Graph Classification Using Graph Attention Networks

Classify graphs that have multiple independent labels using graph attention networks (GATs).

Ouvrir le live script

Train Network Using Federated Learning

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].

Ouvrir le live script