Chris Bates

"Transformations (not using numpy) are like 80% of the run time" Are those demos in pure python?

That kind of thing (and some of the stuff in PyRR) could be accelerated very easily via use of the Numba JIT. Numba has a lot of nice hooks for...

From memory I found OpenCV hard to install via pip in a virtualenv. I ended up installing it via conda so I could keep the venv and python version isolated,...

> > After searching all the available Flutter video players, the best on I found is https://pub.dev/packages/flutter_meedu_videoplayer which is compatible on all platforms except desktop MacOS. It's fairly easy to...

I did a regular install and ran into the same issue. I am using WSL2 / Ubuntu. The issue is that Watchdog relies on inotify and if you are using...

Migrating from discord to here as to facilitate discussion on possible design: Another option would be have a chat template prompt take another prompt as input ``` answer = outlines.generate.text(model,sampler)...

In practice I tend to compose all these things (model, system prompt, chat template) and create a function f that simply takes the chat history and next prompt and returns...

From experimenting with many models I have a bunch of chat templates like the following ``` @outlines.prompt def chatml_template(system,query,history=[]): '''system {{ system }} {% for example in history %} user...

One idea would be to have a generate.chat method that is separate to generate.text that supports a template in the constructor?

Another thing I sometimes do (which might influence the design) is send in a chat history as a few shot prompt example using the chat template and then ask for...