I installed centaur tabs and accepted I can’t see two tabs but I can see my whole buffer :)

Yup, it definitely will help speed up inference on models you can get working.

My personal recommendation is to start with something like PyTorch Mobile or Tensorflow Lite (whichever you prefer). The main benefit is that you can take a model in PyTorch and compile it down to a representation that will use the NN API

You can pretty quickly use the examples in this repo to try out running a language model like BERT. It will also show you the process of converting a model and running it in your phone.

https://github.com/pytorch/android-demo-app

If you’re going after maximum performance on a particular model then it might make more sense to learn the NN API directly try to build it yourself. Personally I would probably try to work with the open source community to add an NN API backend in llama.cpp

https://github.com/ggerganov/llama.cpp/issues/2687

Pretty cool! Currently I’m using Orgzly (well… Orgzly revived since development stopped https://github.com/orgzly-revived/orgzly-android-revived )

Web might be enough for me but do you plan on making an Android version?

Right now I use git to track all of my org documents but syncing and capturing from my phone is a pain. I need git because I run actions to update calendar services when things change. Would love something that I could seamlessly integrate with capture that also is a useful note viewer.

First if you’re coming from vim save yourself a lot of time and use Doom Emacs. Learn on Doom and someday you can decide if you want to build your own config from scratch after you know what you’re doing.

Out of the box this is going to feel the most similar to vim and it’s very easy to add packages that have already been integrated.

You should now be able to edit a file how you’re used to. Next is to learn the basics that will help you learn more. So all the help menus (SPC-h), fuzzy function finder (M-x or SPC-:), Eval (SPC-;), file finder (SPC-SPC).

The doom documentation does a pretty good job of walking you through this:

• https://github.com/doomemacs/doomemacs/blob/master/docs/getting_started.org
• https://noelwelsh.com/posts/doom-emacs/

It’s going to take you more than 3 hours.

Only issue I currently have is the title bar is huge now. I haven’t tracked down why yet but if I toggle tool-bar-mode on and off then it goes back to how it was.

That was fast, nice work! I posted in your last thread. Now that you support GPT-4 I can probably use this.

Couple Q’s before I set some time aside to try it out:

1. Do you have a public API if I want to define some custom prompt commands? I’d like to do some simple things like make functions that are region specific with a custom prompt string. I could hack it up but it would be helpful to know if there’s an API surface that I should use to keep things stable between versions.
2. Where do you want to go from here?