C2D

True, covering all of the decorator APIs with a near identical duplication of the code doesn't seem like a good idea. This is the 'simplest' solution, but it should be...

I 'resolved' this issue by implementing the extending the c++ class in javascript; I believe you can also implement a `set(index, value)` function to work around it and expose it,...

Looking into implementing `SHA`

Is this something that is still under consideration? It would quite cool for embedded, since you'd be able to do 'dynamic' allocations on the stack without the need for a...

Bug report when testing: When provided with an invalid grammar (for the llama.cpp backend) it will stop stop responding to new requests to generate. It seems like ollama does not...