olivermuty

Lol! Wat? Got more details?(On cerebras)

Did you have a bad day outside of the internet today? These three thread posts come off as you being a bit of an asshole for no reason. Why wouldnt the guy put financial context on your claim he should go for a three times more expensive sub?

They have cache, but it costs the same indeed, no idea what the point of the cache is

Cerebras the tech is awesome, cerebras the company is a trainwreck

Yes, but this is sadly a lesson most founders (especially american ones) need to self teach.

I certainly did and struggle with both sleep and stress induced anxiety, 9 years after my founder burnout.

Brave as well?

The americanism of politics being taboo is so stupid it aches.

My xpeng g9 goes about 570km in summer. Less in winter, like 480 maybe. Longest range ICE i had was a mercedes wagon that went 1050km on one tank of gas.

Filling the wagon today would cost me like 170 euro. Filling my xpeng happens overnight and is about 7-9 euro depending on grid pricing.

Do you really believe this? Lol

Could you please elaborate what labour law drives the labour out of france?

It is (source, old employee of me now works for them)

Yeah christians would never go crusading like that!

Hmmm, I wonder where the word «crusade» comes from though

I’ll be darned. Too bad it will become abandonware most likely given the authors history lol

Is this a satire thing or will it really be remix3? Didnt they just rebrand to react router FROM remix??

At 40gb and a rumoured 5 to 7 TB size of the proprietary flagships you are looking at several megawatts to run one single model instance. Cerebras is insanely power hungry. It is funny how they are essentially a parallell happenstance (chips being made for other compute stuff also works for LLMs) to gaming processors accidentally being good for LLMs.

The world will be much more interesting when real bespoke hardware built for actual LLM usage comes to market. This means silicon of the SIMD flavour or other variants, but using DRAM so you can pack more tightly.

I would like to be added so I can benchmark you, in process of picking a EU vendor for data storage.

You’re either very gullible or you are the answer to the question you just asked.

Filler pr jippo fluffer article aside, anyone tried to self host ubicloud lately? A year and a half ago it was super cumbersome, wondering if I should give it a new try now.

None of this infra is worth anything more than five years from now.

could you elaborate on this? we're looking at moving off cloudflare r2 in the somewhat near future and garage is on our short-list

By «this» I meant the docx parts, not the rest ;)

would be cool if they published this as oss

Scaleway, ovh and a bunch of other inference as a service providers dont count?

I have been logging in via ssso on business non enterprise plan for a year. Am I a part of an a/b test or what?

This is only a problem if an agent is made in a lazy way (all of them).

Chat completion sends the full prompt history on every call.

I am working on my own coding agent and seeing massive improvements by rewriting history using either a smaller model or a freestanding call to the main one.

It really mitigates context poisoning.

Well RTO mandate means you lose ONLY the good people with other options or make the people with no options have animosity since the deal was changed

Its funny that the current state of vibomania makes me very unsure if this comment is (good) satire or not lol

This tool is for working with established codebases and phoenix.new is more about bootstrapping (for now), so they are not overlapping at all imo.

No such thing needez, just alias ssh to ssh.exe and ssh-add to ssh-add.exe and it works out of the box.

Incredibly weak to not seize the opportunity to name this «Mangentic» :D

What a weird outlook.

For me its more «I am happy that the police removes dangerous uninsurable vehicles from the road»