Viewing profile — searealist

Sous vide is fool proof, you cant mess it up. All you have to worry about is the sear, and that's easy too.

The internet loves using 137 for Ribeyes for the reason you said. I've had good luck going with 135 for ~4 hours.

What this article wants you to believe: Police are actively using flock to identify people traveling over state borders to buy marijuana and search their cars. What actually happen…

I read that Anglo as in Anglo-Saxon, and I was confused why they would throw that in a article. It turns out Anglo is a diamond company.

NEM 3.0 is more fair. The grid is expensive and just as expensive if it is needed to be fully utilized 1% of the time or 100% of the time. The problem is California forcing new con…

There is a real tradeoff: - The musl allocator is only slow with multi-threading. - Almost all other allocators have trouble reclaiming memory when using multi-threading. This ofte…

Because it's probably still more expensive in electricity costs compared to openrouter. Certainly it's not 98% cheaper.

My answer doesn't change if it is M5s. Where is the math showing a 98% discount over K3 on openrouter. Heck, where is the math showing it is any % cheaper? How much electricity wil…

I'm not aware of any service that gives you a 98% discount and is served off of M1s. Did you do the math for what this would cost vs K3 on openrouter?

No one will ever derive any utility from running models at this speed. Please prove me wrong. Give me the number of tokens input and output (and dont forget about reasoning) and ac…

> For even the sparsest MoE open models, having more than a handful of inferences in the batch is enough to make it more likely than not that you'll get some MoE weight reuse withi…

Fair point. I guess it depends on if you would be using ToU otherwise. It looks like about 50% of Californians use ToU plans, but the number is only 10% nation-wide.

> It's worse than that: a typical MoE model routes a separate set of experts at every layer, not just every token! But in practice, RAM offload (for systems with non-unified VRAM) …

It's easy to have your EV only charge off-peak, though. It's just a setting.

Thank god this is written in modern Zig and not Zig88.

Holding your cache in VRAM for 5 minutes or 3 hours have very different costs to them. "They already charge me to park my car, why can't I leave it there for a year for the same pr…

They could always just extend the cache timeout beyond 5 minutes themselves. They don't do that because it is an expensive resource and there is a trade-off bewteen saving computat…

1) To be clear, it costs you 10x more for uncached input tokens _for the next call_, which are still 5x cheaper than output tokens. 2) Now imagine Anthropic or OpenAI now charge yo…

How does an ASIC manage to have 60x the memory bandwidth needed to achieve that speedup?

The models themselves support up to 1M. You are just charged more for all context over 400k. The 272k limit is just a client limitation to ensure you can never go over 400k since t…

Have you ever said "I'm starving"? Do you think that undermines the experience of people actually starving in war or famine?

Please don't act as the hyperbole police. People exaggerate all the time (I'm starving, etc). It's normal, and you are being a jerk to call them out.