jncfhnb

46 ms·

931.
by jncfhnb 2y ago
A well treated slave is still a slave
932.
by jncfhnb 2y ago
Don’t think that needed clarification from “pet”
933.
by jncfhnb 2y ago
I already responded to that two days ago in adjacent replies. You’re also completely wrong. Utilities do not get to pass on costs to specific customers. It is, at best, just a compelling point with regulators to build it into the rate case
934.
by jncfhnb 2y ago
Ok but a lot of the games I’m seeing here seem like they would have vastly different schedules. Like Risk of Rain 2. Like is it really feasible to run ocarina of time in parallel with risk of rain 2 given the difference in completion time?
935.
by jncfhnb 2y ago
What games actually work reasonably well with this? Particularly for someone who is not a speedrunning master of any particular title?
936.
by jncfhnb 2y ago
The climate and suitability for under grounding in Magdeburg is vastly different.
937.
by jncfhnb 2y ago
I’m only including distribution numbers in my figures here. Transmission does not make sense to underground.
938.
by jncfhnb 2y ago
The amortized per person cost is obviously not going to be constant when the demographic density and line mileage changes. That’s why it’s a foolish standard There’s NEVER 1M customers per mile in a distribution system. There’s only going t
939.
by jncfhnb 2y ago
Yes it is.
940.
by jncfhnb 2y ago
Seems hard to square spam being more of a problem in discord when visible spam is observably much less than in other forums
941.
by jncfhnb 2y ago
Per person per month for 10 years is a ridiculous unit to choose. Europe benefits from being very dense. The US grid is way more spread out. Undergrounding costs are around $1M per mile. The US has millions of miles of lines.
942.
by jncfhnb 2y ago
Idk if I have ever seen a spam account on a tech support forum. Like not a single spam post. Edit: on a discord*
943.
by jncfhnb 2y ago
You advertise yourself as Discord
944.
by jncfhnb 2y ago
If something is reliable, it is, by definition, consistently reliable. If something is unreliable, it is, by definition inconsistently unreliable.
945.
by jncfhnb 2y ago
Sometimes unreliable is a very funny qualifier. If something were consistently unreliable then unreliable wouldn’t be the right term.
946.
by jncfhnb 2y ago
The iskander and lancet drones are absolutely not used to nearly the same efficacy as atacms. Those things are regularly just hurled at low value civilian targets in large part because of poor targeting. They’re generally not the road mobil
947.
by jncfhnb 2y ago
The entire sequence is processed by the transformers to make the first predicted token. But that processing of the entire sequences is comprised of processing :n, and previously :n-1, and previously :n-2, etc cumulatively. If tokens 1:3000
948.
by jncfhnb 2y ago
I’m not saying they look at each token in isolation. I’m saying they look at the current token and previous tokens. it is not accurate that these things consider the entire sequence. They (generally) consider the entire sequence up until th
949.
by jncfhnb 2y ago
Tokens are generally consumed sequentially. The model does not care what token 3001 is when it’s creating the embedding representation of tokens 1-3000. You can use the cached state of the 3000th token rather than reprocessing it each time.
950.
by jncfhnb 2y ago
Not the brand new pro Russian account holder
951.
by jncfhnb 2y ago
The simplest example is contextual prompts. Let’s say you have a 3000 token prompt that you put in front of every query. You could A) have the LLM ingest and reprocess the same prompt every single time at needless cost Or B) process it one
952.
by jncfhnb 2y ago
Yes that would be the year that US lend lease began for the soviets
953.
by jncfhnb 2y ago
Ok buddy! We’re coming up on day 900 of the 3 day invasion. I’m sure you’ll get it soon.
954.
by jncfhnb 2y ago
It’s really quite reasonable. The basic capabilities of the ATACMS are not hugely changed and have been available for decades. Russia does not have an equivalent that is demonstrably deployed effectively. They’re also not the focal point of
955.
by jncfhnb 2y ago
Very little evidence of these being used to any effect comparable to ATACMs
956.
by jncfhnb 2y ago
Ok mr brand new account that loves russia
957.
by jncfhnb 2y ago
It absolutely has not kept up. ATACMs are 30 years old. Russia just doesn’t have anything like it.
958.
by jncfhnb 2y ago
It more or less is. This is something you can do with models that have open weights. People claiming those models are not “open source” really have no idea about the “preferred form” of “code” to change an LLM
959.
by jncfhnb 2y ago
Other answers here are mostly wrong. They are almost certainly NOT passing in the prompt as English and having the model reread the prompt again each time. What they’ll be doing is passing in the vector representation of the prompt at run t
960.
by jncfhnb 2y ago
There is. It’s just that we use LLMs like idiots. We should have more tooling to play with the vectors directly. But a lot of the literature is built around people playing with chat gpt, for which this is not an option as a business decisio

More ›