Present features for Nebula as well - How should we do it?

[mjovanc]

How could we present features for both Delta and Nebula on the "Features" section? Do you have an idea how we could present that?

[SvMak]

Yep I have some idea. Could you please provide a list of features for the Nebula project?

[DarthB]

I'm not certain on how much we should promise as we are quite early, so I decided to give to you two versions.
What are you thoughts?

Humble List:

• A command-line tool as a package manager for your favorite datasets and models
• Access to ML datasets with metadata and raw-data
• Access to pretrained models
• Access to the public nebula-registry with common datasets and models maintained by the community
• Self host your private nebula-registry for your confidential work

Let's sell this:

• A command-line tool as a package manager for your favorite datasets and models
• Virtual environment support (run multiple ML projects on the same machine)
• Access ML datasets based on metadata, manage their versions, variants, dependencies and lifecycles
• Access to pretrained models, manage their versions, adaptations and relationships
• Access to template code/projects based on the delta-ml framework
• Access to the public nebula-registry with common datasets and models maintained by the community
• Self host your private nebula-registry for your confidential work

[mjovanc]

@DarthB I like these a lot! Should we do something like this to highlight on the website:

Command-line tool

Manage datasets and models directly from a powerful CLI.

Virtual environments

Run multiple ML projects on the same machine without conflicts.

Dataset management

Organize datasets by metadata, versions, variants, dependencies, and lifecycles.

Pretrained models

Access and manage pretrained models with versioning and adaptations.

Template projects

Use prebuilt templates based on the Delta framework for faster setup.

Public registry

Browse datasets and models shared by the community in the Nebula registry.

Private registry

Host your own Nebula registry for secure and confidential work.

[mjovanc]

This is what we could use for Delta:

Also can we make a dark-reddish badge called "future" similar to what we do in the "hero" boxes @SvMak? I would like to highlight things that we plan for. Maybe you could add "future" to all of Nebula and we can remove them once they become ready. Would that be OK with you @DarthB ?

Fast

It is designed to be fast, ensuring minimal overhead in model training and deployment workflows.

Usability

APIs are designed for simplicity, making it easy for beginners to get started while providing advanced customization options for experienced users.

Extensibility

The framework is modular, allowing users to plug in custom layers, optimizers, or preprocessing pipelines tailored to their unique needs.

Efficient and scalable tools

It provides highly efficient and scalable tools for building and training neural networks, supporting both small-scale experiments and large-scale production systems.

Distributed and Parallel Training (future)

Native support for distributed and parallel training ensures that Delta scales effortlessly across multi-core systems and cloud environments.

Classical ML (future)

Includes support for classical ML algorithms such as decision trees, random forests, SVMs and more.

Integration to Nebula (future)

Direct access to datasets and models managed by the Nebula registry, public or private.

[SvMak]

@mjovanc, @DarthB

[mjovanc]

I think that looks good, can we maybe center the whole container also? Maybe that will be nicer?

[SvMak]

I think that looks good, can we maybe center the whole container also? Maybe that will be nicer?

The text in the description of the Delta points is longer. The container is currently centered, it can be moved to the right by limiting the width, but this will make the left side even more elongated. I would leave the current view for now.

[SvMak]

If you have any questions, please reopen this issue.