BojAI

BojAI is an open-source framework for building, training, evaluating, and deploying machine learning pipelines with zero boilerplate and total flexibility.

Whether you are creating a text classifier, training a vision model, or building your own architecture from scratch, BojAI gives you the tools to go from raw data to deployed model using clean CLI or UI.

BojAI does not impose any restrictions on the model, training, or evaluation loops. You can customize them to run using any type of library including PyTorch, Transformers, or even call an API that runs direct GPU code.

Most ML frameworks make you choose between power or simplicity, pre-built tools or full control, UI-based tools or code-based customization. BojAI gives you control over all of that:

• Use and customize pre-built pipelines for common tasks (text, number, or image data classification, text-to-text and image-to-text generation).

• Create, test, and use your own custom processor, model, trainer, and deployer with no restrictions.

• Do all this using either a command-line interface or a user visual interface.

To do all this, you just need to run simple commands and start coding. When you are done with your processor, you can test it using our CLI or UI interface. Then, code the model and the trainer and test and finally code the deployment logic and test. Finally, you have a fully-tested pipeline which runs on a UI or CLI.

Where to start

To get started, see install BojAI Vexor and get started page.

Research and Development

This open-source framework is an active research and development project. We are constantly producing new knowledge and improving its functionality. If you would like to contribute, read our Contribution guidelines or visit our research page to view current research projects and papers.

License and Use

BojAI Vexor is released under the Apache 2.0 License, and is free for academic and personal use. While we cannot enforce usage beyond the license’s terms, we strongly advocate for responsible and ethical application of machine learning technologies.

We recognize the power of machine learning and aim to make it more accessible through BojAI. With this accessibility comes the responsibility to use such tools for constructive and ethical purposes. We explicitly discourage the use of BojAI in any context related to:

• Military or warfare applications

• Violations of human rights

• Activities that contribute negatively to climate change

• Animal exploitation or abuse

In short: do not use BojAI to harm people, animals, or the planet.

Users found to be engaging in harmful activities may be restricted from accessing future updates or contributions.