Live data from Hacker News

YaLM-100B: Pretrained language model with 100B parameters

github.com

301–310 of 666 posts

Re: YaLM-100B: Pretrained language model with 100B parameters

#301
post #13

I have to wonder if 10 years down the line, everyone will be able to run models like this on their own computers. Have to wonder what the knock-on effects of that will be, especially if the models improve drastically. With so much of our social lives being moved online, if we have the easy ability to create fake lives of fake people one has to wonder what's real and what isn't. Maybe the dead internet theory will rea…

> I have to wonder if 10 years down the line, everyone will be able to run models like this on their own computers. Isn’t that already the case? Sure, it costs $60K, but that is accessible to a surprisingly large minority, considering the potency of this software.

What kind of computer would they be?

Can you spec it out roughly?

Re: YaLM-100B: Pretrained language model with 100B parameters

#302
post #253

Earlier quoted context omitted.

Source for all the claims?

Those are trivial to verify. If you can't do your 5 minutes of research, OP should not feel obliged to humor your attempt to create busy work for them.

Plenty of people cite their work and arguments in contentious discussions.

Re: YaLM-100B: Pretrained language model with 100B parameters

#304

Side note: Yandex search is awesome, and I really hope they stay alive forever. It's the only functional image search nowadays, after our Google overlords neutered their own product out of fear over lawyers/regulation and a disdain for power users. You can't even search for images "before:date" in Google anymore.

I regularly use it for a sample of what Google and Bing are intentionally omitting.

Re: YaLM-100B: Pretrained language model with 100B parameters

#305

Earlier quoted context omitted.

Is there a reason why it is required to fill the swapfile with zeroes here? Normally you'd see something like "dd of=/swapfile bs=1G seek=3 count=0", creating a file of size 3G but with no space allocated (yet). It's much quicker to complete the setup this way.

I assume if you force the file system to allocate inodes you are likely to have a less fragmented file than if you create a sparse file that gets inodes assigned over time when each part is used.

Which won't matter on SSDs

Re: YaLM-100B: Pretrained language model with 100B parameters

#307

Earlier quoted context omitted.

The achievements cannot be tainted. Kolmogorov complexity is (I hope) untainted. Also, Hilbert’s problems are not untainted (and he never flew Nazi Germany!).

I think the correct word might be fled -> he never fled.

Yes, sorry.

Re: YaLM-100B: Pretrained language model with 100B parameters

#308
post #13

I have to wonder if 10 years down the line, everyone will be able to run models like this on their own computers. Have to wonder what the knock-on effects of that will be, especially if the models improve drastically. With so much of our social lives being moved online, if we have the easy ability to create fake lives of fake people one has to wonder what's real and what isn't. Maybe the dead internet theory will rea…

Running the models like this on own computer is already possible with DeepSpeed. I think it even supports training albeit it would be extremely slow.

https://www.deepspeed.ai/

Re: YaLM-100B: Pretrained language model with 100B parameters

#309
post #167

First of all regardless for political situation this is great step in making ML research actually open. So huge thanks for those developers who pushed to make it public. Still... Yandex is in fact share responsibility for Russian government actions. While it impossible to fight censorship they could certainly shut down their News service completely. Yandex could also certainly move more of their company and staff out…

How much responsibility does Google share for US wrecking Afghanistan, Iraq, Libya and Syria?

Re: YaLM-100B: Pretrained language model with 100B parameters

#310

It's just crazy how much it costs to train such models. As I undestand 800 A100 cards would cost about 25.000.000 without considering the energy costs for 61 days of training.

Lambda labs will rent you an 8xA100 instance for 3 months for $21,900. That would put it at around $2m
Post reply on HN