Nessun risultato trovato

La tua ricerca non ha prodotto risultati.

È consigliabile provare quanto segue per riuscire a trovare quello che stai cercando:

  • Controlla l'ortografia della ricerca per parola chiave.
  • Utilizza sinonimi per la parola chiave digitata, ad esempio prova “applicazione” anziché “software”.
  • Prova una delle ricerche popolari mostrate di seguito.
  • Inizia una nuova ricerca.
Domande di tendenza

Oracle Cloud Infrastructure Virtual Machines for Data Science

With the explosion of business data—ranging from customer data to the Internet of Things—data scientists need the flexibility to explore and build models quickly. But purchasing new hardware to meet temporary or peak demand can involve significant capital expense as well as a considerable amount of time.

Oracle Cloud Infrastructure Virtual Machines (VMs) for Data Science are preconfigured environments that enable you to build models and deliver business value faster. Built on Oracle Cloud Infrastructure, these VMs offer exceptional performance, security, and control. You can expand your compute resources as needed using compute autoscaling and keep costs under control by stopping compute instances when they are not needed.

Compute options suitable for this VM image include a virtual machine with an NVIDIA GPU that can be up and running in under 15 minutes with preinstalled common IDEs, notebooks, and frameworks. Oracle Cloud Infrastructure VMs for Data Science include basic sample data and code for you to test and explore.

Virtual Machines for Data Science
Major wireless carrier achieves faster performance with AI solution built on Oracle Cloud Infrastructure

Major wireless carrier achieves faster performance with AI solution built on Oracle Cloud Infrastructure

A large mobile network operator delivers an AI-powered virtual voice assistant in multiple languages to millions of users. The environment uses a cluster with 2 nodes of 8 GPUs each, connected as a cluster with 16 GPUs and 768GB of memory in each node, significantly reducing the training time of the model.

The solution uses 100 million trainable parameters optimized in each iteration. Results include a speech-to-text performance increase of 2.4x and text-to-speech handled 30 to 50 percent faster, along with faster training of models.

Virtual Machines for Data Science

Benefits

Built on Oracle Cloud Infrastructure, our solution for data science provides exceptional performance, security, and control and enables you to build models and deliver business value faster.

Fast

Get up and running quickly. Just deploy the preconfigured image and start working. When you’re finished, teardown is just as easy.

Easy to Use

Launch these images yourself in the cloud, quickly and easily—without the assistance or intervention of your IT organization.

Everything You Need

The all-in-one image includes a complete set of preinstalled tools. You can easily add and customize, either before deployment with the Terraform script or manually after the system is running.

Flexible

Add additional compute resources in the cloud quickly and easily, by autoscaling or using Oracle Cloud Infrastructure Resource Manager.

Customizable

Use a GPU shape for deep-learning model training and inference or CPU-based compute for machine learning, according to your needs.

Low Cost

Reduce your IT costs. For about US$30, you can run one model for a day on a Tesla P100 GPU in the cloud.

Use Cases

Oracle’s preconfigured environment for deep learning is useful in many industries across a wide range of applications.

 

Natural language processing

 

Image recognition and classification

 

Fraud detection for financial services

 

Recommendation engines for online retailers

 

Risk management

Virtual Machine for Data Science Image Content

Operating System

  • Image Family: Oracle Linux 7.x
  • Operating System: Oracle Linux
  • Kernel Version: kernel-uek-4.14.35-1902.8.4.el7uek.x86_64
  • CUDA Version: 10-1-10.1.168-1
  • cuDNN Version: 7.3.1
  • Release Date: Dec. 19, 2019

Machine Learning Frameworks (Python-based)

  • TensorFlow
  • Keras
  • Theano
  • scikit-learn
  • PyTorch
  • NumPy
  • Pandas
  • Seaborn

Integrated Development Environments (IDEs) and Notebooks

  • Anaconda Open Source Distribution
  • Spyder
  • PyCharm
  • Atom
  • Jupyter Notebook
  • Sublime Text

Included Labs

If you are looking to test the environment or learn more about deep learning and data science, Jupyter Notebooks that provide self-guided instruction are included. Just open the readme.md file in the Jupyter Notebook in the virtual machine.

  • Lab 1: Introduction to Machine Learning Packages: scikit-learn
    Walks through the scikit-learn tutorial, and covers how to build and tune models on scikit-learn. Includes exercises.
  • Lab 2: Introduction to ML Packages: PyTorch
    Covers the PyTorch tutorial and how to build and tune models on neural networks in PyTorch for vision tasks, natural language processing, and related uses.
  • Lab 3: Advanced Neural Networks and Transfer Learning for Natural Language Processing
    Provides a tutorial on convolutional and recurrent neural networks.
  • Lab 4: Advanced Neural Networks and Transfer Learning for Vision
    Explains how to implement custom CNNs and use pretrained, state-of-the-art CNNs.