Get fully-managed notebooks up and running in seconds.
Give your projects and models a quicker start with notebooks.
Are you a data scientist or developer looking to launch a notebook in less than 15 minutes? Use our AI Notebooks solution for quick access to Jupyter or VS Code, and launch your notebooks quickly with the resources you need. You also get secure user access, simplified use of your data, and the most popular artificial intelligence frameworks (TensorFlow, PyTorch, Hugging Face and Scikit-learn) to develop your business.
For developers and data scientists
We offer fully-managed artificial intelligence (AI) work environments. This way, you can focus on your projects rather than your infrastructure. Launch notebooks in just a few clicks, or from the command line — then add your data. Your project is now operational.
Clear and flexible pricing
Allocate dedicated CPU and GPU resources to your notebook when you need them. You are charged for resources according to your usage. We offer simple, per-minute pricing, so you can manage your budget.
The most popular AI frameworks
With our AI Notebooks solution, you can use the most widely-used frameworks natively: TensorFlow, PyTorch, Scikit-learn, MXNet and Hugging Face. You can also import many additional libraries.
Why choose AI Notebooks?
Simplify your transition to data science with OVHcloud’s portfolio of AI data and training
Simplicity and security to accelerate your AI projects
OVHcloud AI Notebooks improves productivity for your data scientists. It simplifies their day-to-day work by taking away the hassle of time-consuming engineering tasks. We take care of your infrastructure’s management and security, as well as your notebook instances. With no commitment and pay-as-you-go usage, our Public Cloud solution can be adapted to suit your needs for maximum flexibility and power.
A European cloud that respects data privacy
Artificial intelligence and machine learning offer many ways of adding value to your data. It is important to opt for a cloud service that respects data confidentiality. As a European cloud provider, OVHcloud ensures sovereignty for your data. We follow GDPR regulations, and manage all of our infrastructures along with their security. This way, we can avoid any potential competitive breaches caused by legislations like the CLOUD Act in the US.
Built on open and open-source standards
AI Notebooks offers the most popular code editing applications: Jupyter and VS Code. By default, each notebook is compatible with industry-leading applications and frameworks such as Scikit-learn, TensorFlow, PyTorch and Hugging Face. This means you get total reversibility for your projects.
Once you have created your Public Cloud project, you can launch your notebooks directly on CPU and GPU computing resources. You can do this via the OVHcloud Control Panel, an API, or in the command line. In just a few clicks, you can deploy Jupyter or VS Code notebooks linked to your preferred framework (TensorFlow, PyTorch, MXNet or Hugging Face).
Via the OVHcloud Control Panel
Our web interface offers many features, including a dashboard to monitor your notebooks, the ability to launch new instances, manage your data sets, and much more.
Via the command line
Example: You want to start a notebook with Jupyter and the PyTorch framework, as well as your data, in two command lines.
Import your local data to an Object Storage container
ovhai data upload GRA myBucketa my_dataset.zip
Run a Jupyter notebook with PyTorch pre-installed and mount your data in a read-write folder/data with 2 GPUs
ovhai notebook run \
--gpu 2 \
--volume myBucket@GRA:/data:RW \
No installation required
The AI Notebooks service is fully managed by OVHcloud. It does not require any technical skills to be used. This is useful for data scientists who want to launch a work environment quickly and easily.
Available in seconds
It only takes a few seconds to launch a notebook with a full machine learning environment.
CPU and GPU resources
In AI Notebooks, you can select the type and volume of resources you want to allocate to your notebook, so that it suits your needs. You can also use both CPU and GPU, which is very useful for machine learning.
A range of frameworks to choose from
Choose from a wide range of frameworks, including the most popular ones: TensorFlow, PyTorch, Scikit-learn, MXNet and Hugging Face.
Jupyter and VS Code
AI Notebooks offers the two most popular open-source live code editors. This means you can improve productivity by selecting an environment users are familiar with, and you also get guaranteed reversibility.
Data management made simple
AI Notebooks offers a simple and efficient way to work with your own data, stored in our Object Storage solution. Your data is synced by leveraging CPU and GPU resources for optimal performance. This data is visible in your notebook, so you can work quickly and efficiently without time-consuming import constraints.
Increased user security
With AI Notebooks, you can control access to your notebooks. You are free to choose from specific user roles and applications, or allow an open connection for everyone.
Compatibility with your current notebooks
Easily import your existing notebooks, created on similar platforms (Colab, Jupyter.org), or hosting platforms like GitHub or GitLab.
With AI Notebooks, you can start and manage your notebooks via the OVHcloud Control Panel, the API, or in the command line. This gives you the freedom to manage your pipelines.
At no extra cost, AI Notebooks natively offers tools like Grafana to monitor your projects. Monitor the load of your CPU and GPU resources in real time, or even ongoing network traffic.
Simple and pay-as-you-go rates
You can opt for pay-per-use billing with AI Notebooks. You only pay for notebooks that are currently running, depending on the CPU and GPU resources you have selected. This pricing method matches your resource usage, and allows you to start a notebook for a few pennies a day.
OVHcloud AI Notebooks billing
For each training task you launch, you pay per minute for the computing resources used, depending on their lifespan and the number of GPUs allocated.
No commitment is required to use this service.
Train your AI, machine learning and deep learning models efficiently and easily, and optimise your GPU usage.
Easily deploy machine learning models and applications to production, create your API access points effortlessly, and make effective predictions.
Artificial intelligence accelerated by cloud solutions
Artificial intelligence technology is based on the use of computing units to learn, just as humans would. With the rise of cloud computing and digital transformation, projects involving AI are taking off now more than ever. Data storage and the computing power required to process data are decisive factors in achieving success. Infrastructure-as-a-Service (IaaS) solutions are an asset because you can use them to adapt your configuration as required, and choose the computing power you need. Some of the applications closely linked to AI are big data architectures. They process massive volumes of data with their algorithms in record time. Software applications like Apache Spark, Apache Hadoop and MongoDB databases are widely used in big data solutions. The data analyst — who manages data processing — uses certain programming languages like Python, Scala, R and Java. The quality of the data collected is important for achieving maximum accuracy in results. Other uses for AI include predictive analysis, business intelligence, machine learning, data management and much more.
What is a notebook for machine learning?
Notebooks revolutionise collaborative work within machine learning projects. A notebook is an application you can use to create code, and share it within a team. On top of helping teams work collaboratively on your machine learning project, it can also be used to view data set rendering.
What is Jupyter in machine learning?
Jupyter is an open-source web application that enables users to work in up to 40 different programming languages. These include Python, Julia, Ruby and R. Jupyter — all programming languages that are closely associated with machine learning. Various tools have been developed, such as Jupyter Notebook, which facilitates collaborative work within scientific projects. There are multiple fields of application for machine learning, as well as data analytics, statistics, and data visualisation. Jupyter Notebook for machine learning is much more than just an IDE (Integrated Development Environment), and is suitable for any type of data science project. With our solution, get Jupyter Notebook online through the cloud.
What is an AI framework?
An AI framework is a software infrastructure dedicated to artificial intelligence. It is a set of software tools that form the basis of an AI project’s architecture. Here are some examples of AI frameworks: Scikit-learn, TensorFlow, Jupyter Notebook and PyTorch.