Co je Azure Machine Learning?What is Azure Machine Learning?
V tomto článku se dozvíte o Azure Machine Learning cloudovém prostředí, které můžete využít ke školení, nasazení, automatizaci, správě a sledování modelů ML.In this article, you learn about Azure Machine Learning, a cloud-based environment you can use to train, deploy, automate, manage, and track ML models.
Azure Machine Learning můžete použít pro libovolný druh strojového učení, od klasického ml až po hloubkové učení, pod dohledem a bez dohledu učení.Azure Machine Learning can be used for any kind of machine learning, from classical ml to deep learning, supervised, and unsupervised learning. Bez ohledu na to, jestli chcete psát kód v Pythonu nebo R pomocí sady SDK nebo pracovat s možnostmi žádného kódu a nízkého kódu v studiu, můžete v pracovní prostor Azure Machine Learning sestavovat, vyvíjet a sledovat modely strojového učení a hloubkového učení.Whether you prefer to write Python or R code with the SDK or work with no-code/low-code options in the studio, you can build, train, and track machine learning and deep-learning models in an Azure Machine Learning Workspace.
Spusťte školení na místním počítači a pak nahorizontální navýšení kapacity do cloudu.Start training on your local machine and then scale out to the cloud.
Služba také spolupracuje s oblíbenými nástroji pro hloubkové učení a posílením open source nástrojů, jako jsou PyTorch, TensorFlow, scikit-Learning a Ray RLlib.The service also interoperates with popular deep learning and reinforcement open-source tools such as PyTorch, TensorFlow, scikit-learn, and Ray RLlib.
Tip
Bezplatná zkušební verzeFree trial! Pokud ještě nemáte předplatné Azure, vytvořte si bezplatný účet před tím, než začnete.If you don’t have an Azure subscription, create a free account before you begin. Vyzkoušení bezplatné nebo placené verze Azure Machine Learning dnesTry the free or paid version of Azure Machine Learning today. Získáte kredity, které můžete využít na služby Azure.You get credits to spend on Azure services. Po jejich vyčerpání si můžete účet ponechat a používat bezplatné služby Azure.After they're used up, you can keep the account and use free Azure services. Nikdy vám nebudeme účtovat žádné poplatky, pokud si sami nezměníte nastavení a nezačnete používat placené služby.Your credit card is never charged unless you explicitly change your settings and ask to be charged.
Co je strojové učení?What is machine learning?
Strojové učení je technika datové vědy, která umožňuje počítačům s využitím existujících dat předvídat budoucí chování, výsledky a trendy.Machine learning is a data science technique that allows computers to use existing data to forecast future behaviors, outcomes, and trends. Pomocí strojového učení se počítače učí bez explicitního programování.By using machine learning, computers learn without being explicitly programmed.
Díky předpovědím neboli predikcím ze strojového učení můžou být aplikace a zařízení chytřejší.Forecasts or predictions from machine learning can make apps and devices smarter. Například při nakupování online pomůže strojové učení doporučit další produkty, které byste mohli chtít v závislosti na zakoupení.For example, when you shop online, machine learning helps recommend other products you might want based on what you've bought. Nebo při protažení platební karty terminálem strojové učení porovná transakci s databází transakcí a pomáhá detekovat podvody.Or when your credit card is swiped, machine learning compares the transaction to a database of transactions and helps detect fraud. A takový robotický vysavač na základě strojového učení zjišťuje, jestli už v místnosti dostatečně uklidil.And when your robot vacuum cleaner vacuums a room, machine learning helps it decide whether the job is done.
Nástroje pro strojové učení, které se vejdou na jednotlivé úlohyMachine learning tools to fit each task
Azure Machine Learning poskytuje vývojářům a odborníkům přes data, kteří potřebují pro své pracovní postupy pro strojové učení, včetně těchto nástrojů:Azure Machine Learning provides all the tools developers and data scientists need for their machine learning workflows, including:
Návrhář Azure Machine Learning: přetažení modulů pro sestavování experimentů a následné nasazení kanálů.The Azure Machine Learning designer: drag-n-drop modules to build your experiments and then deploy pipelines.
Jupyter poznámkové bloky: pomocí našich ukázkových poznámkových bloků nebo vytvořte vlastní poznámkové bloky a využijte naši sadu SDK pro ukázky Pythonu pro váš strojové učení.Jupyter notebooks: use our example notebooks or create your own notebooks to leverage our SDK for Python samples for your machine learning.
Skripty nebo notebooky r, ve kterých používáte sadu SDK pro R k psaní vlastního kódu nebo použití modulů R v návrháři.R scripts or notebooks in which you use the SDK for R to write your own code, or use the R modules in the designer.
Mnohé modely řešení (Preview) jsou sestavené na Azure Machine Learning a umožňují výuku, provozování a správu stovek nebo dokonce tisíců modelů strojového učení.The Many Models Solution Accelerator (preview) builds on Azure Machine Learning and enables you to train, operate, and manage hundreds or even thousands of machine learning models.
Open Source architektury, jako jsou PyTorch, TensorFlow a scikit – Naučte se a mnoho dalšíhoOpen-source frameworks such as PyTorch, TensorFlow, and scikit-learn and many more
Posílení učení s využitím Ray RLlibReinforcement learning with Ray RLlib
Můžete dokonce použít MLflow ke sledování metrik a nasazení modelů nebo Kubeflow k sestavení koncových kanálů pracovního postupu.You can even use MLflow to track metrics and deploy models or Kubeflow to build end-to-end workflow pipelines.
Vytváření modelů ML v Pythonu nebo RBuild ML models in Python or R
Spusťte školení na místním počítači pomocí Azure Machine Learning Python SDK nebo R SDK.Start training on your local machine using the Azure Machine Learning Python SDK or R SDK. Pak můžete horizontální navýšení kapacity na Cloud.Then, you can scale out to the cloud.
Díky mnoha dostupným výpočetním cílům, jako je Azure Machine Learning compute a Azure Databricks, a s pokročilými službami ladění pomocí parametrůmůžete vytvářet lepší modely rychleji pomocí výkonného cloudu.With many available compute targets, like Azure Machine Learning Compute and Azure Databricks, and with advanced hyperparameter tuning services, you can build better models faster by using the power of the cloud.
Můžete také automatizovat školicí a optimalizační model pomocí sady SDK.You can also automate model training and tuning using the SDK.
Vytváření modelů ML v studiuBuild ML models in the studio
Azure Machine Learning Studio je webový portál v Azure Machine Learning pro možnosti s nízkým kódem a bez kódu pro školení modelů, nasazení a správu prostředků.Azure Machine Learning studio is a web portal in Azure Machine Learning for low-code and no-code options for model training, deployment, and asset management. Sada Studio se integruje se sadou Azure Machine Learning SDK, aby se zajistilo bezproblémové prostředí.The studio integrates with the Azure Machine Learning SDK for a seamless experience. Další informace najdete v tématu co je Azure Machine Learning Studio.For more information, see What is Azure Machine Learning studio.
Návrhář služby Azure Machine LearningAzure Machine Learning designer
Použijte Návrháře ke školení a nasazení modelů strojového učení bez psaní kódu.Use the designer to train and deploy machine learning models without writing any code. Vyzkoušejte si kurz pro návrháře a začněte.Try the designer tutorial to get started.
Sledovat experimentyTrack experiments
Naučte se sledovat a vizualizovat experimenty v oblasti datových věd v studiu.Learn how to track and visualize data science experiments in the studio.
A mnohem víc...And much more...
Navštivte Azure Machine Learning studia na adrese ml.Azure.com.Visit Azure Machine Learning studio at ml.azure.com.
MLOps: nasazení správy životního cyklu &MLOps: Deploy & lifecycle management
Pokud máte správný model, můžete ho snadno použít ve webové službě, v zařízení IoT nebo v Power BI.When you have the right model, you can easily use it in a web service, on an IoT device, or from Power BI. Další informace najdete v článku věnovaném nasazení a umístění.For more information, see the article on how to deploy and where.
Pak můžete nasadit nasazené modely pomocí Azure Machine Learning SDK pro Python, Azure Machine Learning Studionebo rozhraní příkazového řádku Machine Learning.Then you can manage your deployed models by using the Azure Machine Learning SDK for Python, Azure Machine Learning studio, or the machine learning CLI.
Tyto modely lze spotřebovat a vracet předpovědi v reálném čase nebo asynchronně u velkých objemů dat.These models can be consumed and return predictions in real time or asynchronously on large quantities of data.
A díky pokročilým kanálům strojového učenímůžete v rámci přípravy dat, školení a vyhodnocení modelu spolupracovat na všech krocích prostřednictvím nasazení.And with advanced machine learning pipelines, you can collaborate on each step from data preparation, model training and evaluation, through deployment. Kanály umožňují:Pipelines allow you to:
- Automatizace kompletního procesu strojového učení v clouduAutomate the end-to-end machine learning process in the cloud
- Znovu použít komponenty a v případě potřeby znovu spustit krokyReuse components and only rerun steps when needed
- Použití různých výpočetních prostředků v každém krokuUse different compute resources in each step
- Spustit úlohy dávkového vyhodnocováníRun batch scoring tasks
Pokud chcete používat skripty pro automatizaci pracovního postupu machine learningu, rozhraní příkazového řádku Machine Learning poskytuje nástroje příkazového řádku, které provádějí běžné úlohy, jako je například odeslání školicího běhu nebo nasazení modelu.If you want to use scripts to automate your machine learning workflow, the machine learning CLI provides command-line tools that perform common tasks, such as submitting a training run or deploying a model.
Pokud chcete začít používat Azure Machine Learning, přečtěte si Další kroky.To get started using Azure Machine Learning, see Next steps.
Integrace s dalšími službamiIntegration with other services
Azure Machine Learning funguje s ostatními službami na platformě Azure a integruje se s open source nástroji, jako je třeba Git a MLFlow.Azure Machine Learning works with other services on the Azure platform, and also integrates with open source tools such as Git and MLFlow.
- Výpočetní cíle, jako je Služba Azure Kubernetes, Azure Container Instances, Azure Databricks, Azure Data Lake Analytics a Azure HDInsight.Compute targets such as Azure Kubernetes Service, Azure Container Instances, Azure Databricks, Azure Data Lake Analytics, and Azure HDInsight. Další informace o výpočetních cílech najdete v tématu co jsou výpočetní cíle?.For more information on compute targets, see What are compute targets?.
- Azure Event Grid.Azure Event Grid. Další informace najdete v tématu věnovaném využívání Azure Machine Learningch událostí.For more information, see Consume Azure Machine Learning events.
- Azure monitor.Azure Monitor. Další informace najdete v tématu monitorování Azure Machine Learning.For more information, see Monitoring Azure Machine Learning.
- Úložiště dat, jako jsou účty Azure Storage, Azure Data Lake Storage, Azure SQL Database, Azure Database for PostgreSQL a Azure Open DataSet.Data stores such as Azure Storage accounts, Azure Data Lake Storage, Azure SQL Database, Azure Database for PostgreSQL, and Azure Open Datasets. Další informace najdete v tématech přístup k datům ve službě Azure Storage Services a vytváření datových sad pomocí Azure Open DataSets.For more information, see Access data in Azure storage services and Create datasets with Azure Open Datasets.
- Virtuální sítě Azure.Azure Virtual Networks. Další informace najdete v tématu Přehled izolace a ochrany osobních údajů ve virtuální síti.For more information, see Virtual network isolation and privacy overview.
- Azure Pipelines.Azure Pipelines. Další informace najdete v tématu výuka a nasazení modelů strojového učení.For more information, see Train and deploy machine learning models.
- Protokoly úložiště GitGit repository logs. Další informace najdete v tématu integrace Gitu.For more information, see Git integration.
- MLFlow.MLFlow. Další informace najdete v tématu MLflow ke sledování metrik a nasazení modelů MLflow jako webové služby .For more information, see MLflow to track metrics and Deploy Mlflow models as a web service
- Kubeflow.Kubeflow. Další informace najdete v tématu vytváření koncových kanálů pracovního postupu.For more information, see build end-to-end workflow pipelines.
Zabezpečená komunikaceSecure communications
Váš Azure Storage účet, výpočetní cíle a další prostředky se dají v rámci virtuální sítě bezpečně použít k učení modelů a provádění odvození.Your Azure Storage account, compute targets, and other resources can be used securely inside a virtual network to train models and perform inference. Další informace najdete v tématu Přehled izolace a ochrany osobních údajů ve virtuální síti.For more information, see Virtual network isolation and privacy overview.
Další krokyNext steps
Vytvořte svůj první experiment s upřednostňovanou metodou:Create your first experiment with your preferred method:
-
- Začněte ve vlastním vývojovém prostředíGet started in your own development environment
- Použití poznámkových bloků Jupyter na výpočetní instanci ke školení & nasazení modelů MLUse Jupyter notebooks on a compute instance to train & deploy ML models
- Použití automatizovaného strojového učení ke studiu & nasazení modelů MLUse automated machine learning to train & deploy ML models
- Použití možností přetažení & designeru pro výuku & nasazeníUse the designer's drag & drop capabilities to train & deploy
- Použití rozhraní příkazového učení pro strojové učení ke školení a nasazení modeluUse the machine learning CLI to train and deploy a model
Seznamte se s kanály strojového učení, které umožňují vytvářet, optimalizovat a spravovat scénáře strojového učení.Learn about machine learning pipelines to build, optimize, and manage your machine learning scenarios.
Přečtěte si podrobnou architekturu Azure Machine Learning a koncepty .Read the in-depth Azure Machine Learning architecture and concepts article.