I’d be interested in setting up the highest quality models to run locally, and I don’t have the budget for a GPU with anywhere near enough VRAM, but my main server PC has a 7900x and I could afford to upgrade its RAM - is it possible, and if so how difficult, to get this stuff running on CPU? Inference speed isn’t a sticking point as long as it’s not unusably slow, but I do have access to an OpenAI subscription so there just wouldn’t be much point with lower quality models except as a toy.

I’m only an armchair physicist, but I believe this isn’t possible due to relativity. I know that, at least, there are cases where two observers can disagree on whether an event occurred simultaneously. Besides all the other relativity weirdness, that alone seems to preclude a truly universal time standard. I would love for someone smarter than me to explain more and/or correct me though!

The issue is that, in the function passed to reduce, you’re adding each object directly to the accumulator rather than to its intended parent. These are the problem lines:

if (index == array.length - 1) {
	accumulator[val] = value;
} else if (!accumulator.hasOwnProperty(val)) {
	accumulator[val] = {}; // update the accumulator object
}

There’s no pretty way (that I can think of at least) to do what you want using methods like reduce in vanilla JS, so I’d suggest using a for loop instead - especially if you’re new to programming. Something along these lines (not written to be actual code, just to give you an idea):

let curr = settings;
const split = url.split("/");
for (let i = 0; i < split.length: i++) {
    const val = split[i];
    if (i != split.length-1) {
        //add a check to see if curr[val] exists
        let next = {};
        curr[val] = next;
        curr = next;
    }
    //add else branch
}

It’s missing some things, but the important part is there - every time we move one level deeper in the URL, we update curr so that we keep our place instead of always adding to the top level.

If anybody’s curious, I tried it with GPT4 and it got it right.

Yeah personally I haven’t needed jQuery in years.

Currying is converting a function with n parameters to n functions that each have one parameter. This is done automatically in most primarily functional languages. Then, partial application is when you supply less than n arguments to a curried function. In short, currying happens at the function definition and partial application happens at the function call.

Currently the type of test_increment is (int, int) -> unit -> unit. What we want is int -> int -> unit -> unit. The more idiomatic way would have this function definition:

let test_increment new_value original_value () =

Which would require this change in the callers:

test_case "blah" `Quick (test_increment 1 0);

See, in most primarily functional languages you don’t put parentheses around function parameters/arguments, nor commas between them - in this case, only around and between members of tuples.

I’m not an OCaml person but I do know other functional languages. I looked into Alcotest and it looks like the function after “`Quick” has to be unit -> unit. Because OCaml has currying, and I think test_increment already returns unit, all you should have to do is add an extra parameter of type unit. I believe that would be done like this:

let test_increment (new_value, original_value) () =

Now the expression test_increment (1, 0) returns a function that must be passed a unit to run its body. That means you can change the lambdas to e.g. this:

test_case "blah" `Quick (test_increment (1, 0))

I don’t know OCaml precedence rules so the enclosing parentheses here may not be necessary.

I’d also note that taking new_value and original_value as a tuple would probably be considered not idiomatic unless it makes sense for the structure of the rest of your code, especially because it limits currying like we did with the unit being able to be passed later. Partial application/currying is a big part of the flexibility of functional languages.

Edit: if you’re getting into functional programming you may also consider calling increment_by_one “succ” or “successor” which is the typical terminology in functional land.

I think the point is that since it's open source, there's not as much worry of Microsoft ruining a good thing.

Excuse me but I have it from a very reputable source that the saying goes "fool me once…

Shame on… Shame on you.

Fool me-- can't get fooled again"

Eh, I'll take assembly over like Perl any day. I may be biased since I like to reverse engineer though.

Pretty sure Jython is, though not a Python 3 version. My exposure is as the scripting language, behind Java as the primary language, for Ghidra. And I hate it and installed Eclipse rather than deal with J/Python’s shit.