Why is everyone in this thread constantly putting meaningless @ mentions at the top of their replies?

I was responding to your general statement that python is slow and so there is no point in making it faster, I agree that removing the GIL wont do much to improve the speed of code for programs making heavy use of numpy or things calling outside it.

That’s a bit suss too tbh. Did the C++ version use an existing library like Eigen too or did they implement everything from scratch?

It was written entirely from scratch which is kind of my point, a well writen python program can outperform a naive c implementation and is vastly simpler to create.

If you have the expertise and are will to put in the effort you likely can squeze that extra bit of performance out by dropping to a lower level language, but for certain workloads you can get good performance out of python if you know what you are doing so calling it extremely slow and saying you have to move to another language if you care about performance is missleading.

Numpy is written in C.

Python is written in C too, what’s your point? I’ve seen this argument a few times and I find it bizarre that “easily able to incorporate highly optimised Fortran and C numerical routines” is somehow portrayed as a point against python.

Numpy is a defacto extension to the python standard that adds first class support for single type multi-dimensional arrays and functions for working on them. It is implemented in a mixture of python and c (about 60% python according to github) , interfaces with python’s c-api and links in specialist libraries for operations. You could write the same statement for parts of the python std-lib, is that also not python?

Its hard to not understate just how much simpler development is in numpy compared to c++, in this example here the new python version was less than 50 lines and was developed in an afternoon, the c++ version was closing in on 1000 lines over 6 files.

Nope, if you’re working on large arrays of data you can get significant speed ups using well optimised BLAS functions that are vectorised (numpy) which beats out simply written c++ operating on each array element in turn. There’s also Numba which uses LLVM to jit compile a subset of python to get compiled performance, though I didnt go to that in this case.

You could link the BLAS libraries to c++ but its significantly more work than just importing numpy from python.

Python can be extremely slow, it doesn’t have to be. I recently re-wrote a stats program at work and got a ~500x speedup over the original python and a 10x speed up over the c++ rewrite of that. If you know how python works and avoid the performance foot-guns like nested loops you can often (though not always) get good performance.

The author is Danish, are you so certain that this isnt just slightly awkward usage of a second laguage that you are willing to throw transphobic at them as an insult?

And jesus, the writing in BE! There’s only so many times I can hear “haha look at these funny space people who think the dynamo is an invention from antiquity”

If storage space is important using uncompressed json is a bad choice, if you’re compressing the json it doesnt really matter if you have lots of exceptionCase: False fields as they will compress very well.

I dont have one, but I’m pretty sure you can drop to a desktop that you can do whatever you want with the files on the system just like any other linux distro.

Kinda, but its not black and white. For a start steam has a much longer track record of nearly 20 years not doing this, I’ve heard of them de-listing games and not allowing them to be sold any more but never of revoking games that have been sold. Secondly there are many games on steam that stream cant just revoke, games that use no DRM or DRM that isnt integrated into steamworks they cant just delete if you back it up.

But that being said there is the possibility of something like this happening on steam, which is why I’m glad there is still an active game piracy scene even if I dont use it any more.

Its not training the model, it’s the model using the context you provide it (in that instance). If you use an unfiltered LLM it will run with anything you say and go from there, for example you could tell it Mexico reclaimed Texas and it would carry on as if that’s true. But only until you close it down its not permanently changing the model it is just changing the context in which that instance is running.

The big tech companies are going to huge lengths to filter and censor their LLMs when used by the public both to prevent negative PR and because they dont want people to have unrestricted access to them.

Entirely depends what you are wanting to use it for. Unless you have a beast of a machine you cant run huge generalist models like chatGPT so you have to look for smaller models tuned to your use case. I’ve been liking mythomax for story telling and wizard coder for coding based tasks.

That’s only true with the corporate controlled ones, they filter all the results extensively to avoid it giving any answer that goes even slightly against American corporate norms. If you host your own LLM you get entirely unfiltered answers.

they do interact with matter, just incredibly weakly

You can easily get photoshop to reproduce one of Mondrain’s paintings. That’s on the user not the tool, i fail to see why the same doesnt apply to the tool of generative AI.