Thomas Antony

@duraki Were you able to get this to work with D4M ?

I have found that the only way to actually shut down the call is to call `shutdown_background()` on the tokio runtime used to create the client after dropping the client...

I was able to get past this by forcing `protobuf` and `protoc-rust` to version 2.17.0 in my Cargo.toml. The problem is that later versions of these crates upgrade the version...

Have you seen this? https://www.fyears.org/2015/06/electron-as-gui-of-python-apps.html

Yes. I am sending the whole virtual DOM tree with each render command. I know it's not exactly the most efficient way to do this. I considered doing the diff-ing...

I haven't looked at this in a while. However, I am open to pull-requests and suggestions. I also created another demo based on this same concept, but using Tkinter instead...

@peyer There already exists a C++ version here: https://github.com/corsix/amx

> I see that llama.cpp has added a [C-style API](https://github.com/ggerganov/llama.cpp/pull/370), exciting stuff! Yea. My bindings were based on my own C++ API ([#77](https://github.com/ggerganov/llama.cpp/pull/77) which is now closed). Georgi decided that...

> > > I see that llama.cpp has added a [C-style API](https://github.com/ggerganov/llama.cpp/pull/370), exciting stuff! > > > > > > Yea. My bindings were based on my own C++ API...

Thanks for the changes. I just released v0.1.11 - this includes the new memory mapped I/O feature and requires updating the weight files. But it makes loading the models a...