Come back later, this is a big WIP, a big fat mud ball of WIPs.
A big playground for multiple purposes, discovering different kind of technology.
A place to test and adjust for benchmarking and comparing different approaches with different models and techniques. Later on a subset of this will be used for the Rawwen REST Api (Mostly MIT ones).
Rawen itself is a super flexible command framework, allowing devs to create their own commands by writing 3rd party scripts with js,ts,py scripts and injecting them into defined commands.
The goal is to have the best subset of techniques for a basic Rawen instance that can :
- Summarize
- QA docs/content
- SQL QA
- Transcribe
- Speak
- Translate
- Gen. Image
- UI context awareness / understanding
- Auto mapping, UI application
- Dynamic UI generation
- Auto UI pathing
- Fully Local
- Self hosted
- API consumed
The question is when to use which one, depending on the use case, the hosting variant will be different. For personal use just cosume APIs lazylily you cant keep up with the requirements of most of the models.
Self hosted reccomendation :