sgt101

22 ms·

301.
by sgt101 2y ago
you made my day!
302.
by sgt101 2y ago
I find 100tps unlikely, I see 13tps on an 80GB A100 for a 70B 4bit quantized model. Can you link - I am answering an email on monday where this info would be very useful!
303.
by sgt101 2y ago
Well, Deepseek trained them?
304.
by sgt101 2y ago
I'll see your architectural innovation and raise you a loss function revolution. https://arxiv.org/pdf/2412.21149
305.
by sgt101 2y ago
in every contingency - this is correct I think. It's been very important to the Nato stance to have multiple nuclear armed states, it made Soviet calculations extremely difficult. With the USA leaving the alliance I think that either a
306.
by sgt101 2y ago
But will the UK and France bail out Germany? Yes: if Germany acts as a serious and honest partner in defence. No: if Germany refuses to be a responsible security player and swings back to Russia as soon as it can. If Germany reopens the gas
307.
by sgt101 2y ago
More likely "reparations" - €500bn a year or we're coming to Berlin ! Genuinely - this is a one way bet for the Russians, smash up Estonia and extort the Germans, spend the money on more guns and then kill all the Poles. Comi
308.
by sgt101 2y ago
It's super interesting. There are two levels... The pZombie type level where we look at the LLM as if it were a black box and simply account for its behavior. At this level LLM's claim to have knowledge and also claim knowledge of
309.
by sgt101 2y ago
How about "How to thrive in a ChatGPT world"?
310.
by sgt101 2y ago
Yes. This is the issue. "not on board", "anti-innovation", "not a team player", "disruptive", "unhelpful", "negative". bye bye bye bye.... I see a lot of devs and IC's taking
311.
by sgt101 2y ago
You try raising Merino sheep in the Hebrides! Blackfaces can live all year on the hill, if you get clothes made from their wool, so can you :)
312.
by sgt101 2y ago
They don't have any concept of falsehood either, so this is very different from a human making things up with the knowledge that they may be wrong.
313.
by sgt101 2y ago
I wish the title wasn't so aggressively anti-tech though. The problem is that I would like to push this course at work, but doing so would be suicidal in career terms because I would be seen as negative and disruptive. So the good mess
314.
by sgt101 2y ago
nahh - when a distribution is very tight but with very shallow tails - I think low variance.
315.
by sgt101 2y ago
- high wages are not evenly spread geographically in the USA - high wages are not evenly spread demographically in the USA - the wage distribution seems to me to be unusual in that it has a very long fat tail whereas the UK's is very c
316.
by sgt101 2y ago
SE is a good example - I get a lot of help from LLM tools and I think we're learning how to use them better across realistic SDLC processes as well, but we're not replacing lots of people at the moment. On the other hand I saw a b
317.
by sgt101 2y ago
hmm that's a hard one - let's ask Deepseek!
318.
by sgt101 2y ago
There's a big difference. The membership of these classes is determined in the worst case - so if there is no polynomial time solution in the worst case then it's NP. For this problem we don't care if it's possible that
319.
by sgt101 2y ago
I wonder if that's really set in stone or just a stance that suited Apples marketing for a few years. Maybe there's a business case for some bigger memory Macs now!
320.
by sgt101 2y ago
Hmm - if only macpro's had memory that could be upgraded like in the olden days.
321.
by sgt101 2y ago
"My king, it's amazing these things can contain and regurgitate knowledge on demand. The next step is an army of Gollems that will destroy your enemies, all I need is three bags of gold." - Sargon the scribe, near Baghdad fir
322.
by sgt101 2y ago
Compute has been more than doubling because people have been spending silly money on it. How long ago would a proposal for a $10m cluster for ML have been thought surreal by any funding agency? Certainly less than 10 years ago. Now people a
323.
by sgt101 2y ago
hodl?
324.
by sgt101 2y ago
My wife can't wear a bitcoin. She can wear a gold necklace and will be pleased to do it even if the price goes to 10%
325.
by sgt101 2y ago
I'm no neuroscientist, but I hung out with a few, so here are some bits I picked up from them. In biological brians: - neurons have signal surpression (-ive activation) as well as propagation (+activation) - neurons appear to process a
326.
by sgt101 2y ago
We've had "reasoning" machines for a long time - I learned chess playing against a computer in the 1980's. But we don't have reasoning that can be applied generally in the open world yet. Or at least I haven't
327.
by sgt101 2y ago
Given the observations of the workings of brains though (slow, low energy) we're back to some magic to make this happen?
328.
by sgt101 2y ago
It would be startling if humans could reason this way with two kilos of meat consuming 40w. Even more surprising given that we get answers in ~1s with a cross brain comms time of about 0.25s To me it's clear that human reasoning is dif
329.
by sgt101 2y ago
Options: Finetune the models to be better Optimise the prompts to be better Train better models
330.
by sgt101 2y ago
One thing we did with Zeus which remained proprietary was to build a graphplan planner to replace the STRIPS planner. This was better, but unfortunately it only shifted right by one step - very quickly we found that there were planning issu

More ›