Renato Lui Geh

> Yesh the new version is less complicated and much cleaner, will add some more docs there :) Good to know! Looking forward to this new version. > I think...

For me, as a user, I usually know beforehand if I have a decomposable, deterministic, etc circuit. So (1) seems like a good feature to have with minimum effort for...

Sure. Feel free to make this happen. Any contributions are more than welcome. :) Would this require us to move GoSPN to its own organization?

You raise some interesting points. In fact, the TensorFlow integration issue (#19) fits quite well with your arguments. As mentioned in #19, TF still lacks several key API features for...

Not sure if relevant to this issue exactly, but one thing that might be helpful for end-users (it would be helpful to me at least) would be to explicitly state...

I have managed to pinpoint exactly which flag causes this error. Here's a reproducible minimal example: ```python import lm_eval, torch mgr = lm_eval.tasks.TaskManager() lm_eval.simple_evaluate(model=lm_eval.models.huggingface.HFLM(pretrained="google/gemma-2b", dtype=torch.float32), device="cuda", tasks=["ai2_arc"], limit=10, bootstrap_iters=0) ```...

You are correct, we didn't in fact `chmod +x AppRun` our AppImages. It's curious though how we had previously done tests on other machines and this error hadn't come up....

Instead of writing his own replies, it'd be better if the player were given a set of dialogue options to choose from. This way we prevent the player from writing...

Suggestion: For each `name`, follow the convention below: - If `name` names a `function`, then use `camelCase`; else - If `name` names a `class`, then use `CapitalCase`; else - If...

I personally don't think it'd be worth the effort. Every line of code always starts with an action with arguments being either registers or directions. Coloring all these keywords this...