Convenience Namespace for Using/Obtaining Models

[jaketothepast]

Hello!

I saw this in the Roadmap, and I've somewhat implemented this myself for a side project using your library. Are you open to contributions with this item?

My approach:

I have a .edn file that defines the filepath to drop the models, in and if I load up my application and call into my namespace with a model URL that hasn't yet been downloaded, my code will download that model and return a promise and resolve the llama-context lazily in the application code. I can contribute this model loading namespace!