cerved

10 ms·

91.
by cerved 2y ago
Same. I used to have a 56" OLED but it was a tad large. 48 is perfect. The hardest part was buying a good desk mount. It doesn't work great with an all white screen but I use dark mode for most things
92.
by cerved 2y ago
isn't backtracking a complete search strategy?
93.
by cerved 2y ago
Who on earth doesn't want faster builds? It's make interpolation if it's using Make syntax and shell if it's not. It's not the same syntax
94.
by cerved 2y ago
Is the US government not already negotiating the prices for cars and food it buys?
95.
by cerved 2y ago
What bash but not really bash quirks are you referring to? Only quirk I can think of is that each line in a recipe is a separate shell and it makes inline comments wonky The beauty of Make is the automatic parallelization and incremental bu
96.
by cerved 2y ago
same problem in Java
97.
by cerved 2y ago
what's preventing you from doing it in one pass?
98.
by cerved 2y ago
It's an analogy. He's making the point that even though something can scale at an exponential rate, it doesn't mean there is a business need for such scaling
99.
by cerved 2y ago
but the OP was talking about the size of the context window, not the size of the training corpus
100.
by cerved 2y ago
Why should the response be better just because there is "more data"? Should I be adding extra random tokens to my prompts to make the LLM "smarter"?
101.
by cerved 2y ago
I'm not sure I would draw the same conclusion, stricter requirements on typing is less cumbersome these days with better auto complete and LLMs
102.
by cerved 2y ago
depends what you consider interesting and/or acceptable solution I suppose
103.
by cerved 2y ago
they do have a Gurobi license
104.
by cerved 2y ago
How do you come to this conclusion? They way I read it rather sounds like they use the CP-SAT solver with LNS and LS workers, parallelized over multiple machines
105.
by cerved 2y ago
Not at all. Plenty of people are researching finding satisfiable solution that are quasi-optimal at scale. Once you reach a certain scale of optimization problem, proven optimality in a reasonable time-frame is often infeasible
106.
by cerved 2y ago
tbf, it's a constraint optimization problem
107.
by cerved 2y ago
MiniZinc is a high-level CP language and tool chain, not a solver. Google has been developing their suite of OR tools for 15+ years, it's not like they built something from scratch just for this
108.
by cerved 2y ago
please do, it was very annoying
109.
by cerved 2y ago
Depends how good you want the application to be
110.
by cerved 3y ago
seems like Vim does too
111.
by cerved 3y ago
but serializing/deserializing decimal using the json module is futile
112.
by cerved 3y ago
Another thing you can do when you have two versions, one which fails and one which doesn't, but where the commits are of the messy work in progress nature and not atomic changes, is gradually stage the changes from the bad version onto
113.
by cerved 3y ago
In `sed` it's end of string. String is usually end of line, but not if you use stuff like `N`, to manipulate multi-line strings
114.
by cerved 3y ago
I didn't which is why I asked
115.
by cerved 3y ago
How? Pipes are not synchronous
116.
by cerved 3y ago
It's asynchronous, which is pretty neat
117.
by cerved 3y ago
can you elaborate the difference, besides the different names of the runtime?
118.
by cerved 3y ago
maybe the borrow checker determined their lifetime was up
119.
by cerved 3y ago
In widows, you get bonus points if your program needs administrative rights
120.
by cerved 3y ago
Doesn't the compiler complain if you don't use using or try? I forget

More ›