Vedat Baday

> But, I've seen few examples of successful sampling using Gpus with 24GB or smaller memory. Why is that? They use smaller lower decoding frames. I would suggest you start...

It is sad that docs do not mention it but you can access it using the return value of the component ```py result = st_ace('Type here', key='mytextarea') print("Provided code:", result)...

totally agree with having the logic in fal itself. `deploy` is on https://github.com/fal-ai/fal/pull/173 atm, reg. of that having a simple command like `fal deploy` for deploying everything inside the app...