What does that actually do? Force the "effort" to be static to what I set?

How does it do with the "car wash" benchmark? :D

What about Qwen? Does it get that right?

I've run several local models that get this right. Qwen 3.5 122B-A10B gets this right, as does Gemma 4 31B. These are local models I'm running on my laptop GPU (Strix Halo, 128 GiB of unified RAM).

And I've been using this commonly as a test when changing various parameters, so I've run it several times, these models get it consistently right. Amazing that Opus 4.7 whiffs it, these models are a couple of orders of magnitude smaller, at least if the rumors of the size of Opus are true.

Does Gemma 4 31B run full res on Strix or are you running a quantized one? How much context can you get?

I'm running an 8 bit quant right now, mostly for speed as memory bandwidth is the limiting factor and 8 bit quants generally lose very little compared to the full res, but also to save RAM.

I'm still working on tweaking the settings; I'm hitting OOM fairly often right now, it turns out that the sliding window attention context is huge and llama.cpp wants to keep lots of context snapshots.

I had a whole bunch of trouble getting Gemma 4 working properly. Mostly because there aren't many people running it yet, so there aren't many docs on how to set it up correctly.

It is a fantastic model when it works, though! Good luck :)

Americans are for the most part incredibly out of touch and still think they are at the center of the world.

They voted in the orange turd twice, a third of their population didn't even vote.

The US has become a laughing stock worldwide and internally looks like a prelude to Idiocracy.

Well, at least they know how to hire smart people, from China/Russia/Germany/India.

Bringing it back to your message, they love this! To be the center of attention, show off how strong and powerful of a nation they are.

Except, they messed with the wrong region. Same mistake Russia made.

Sounds to me like you have a lot of envy and jealousy. That's normally what tantrums and name calling result from; that and a need for attention. Congrats.

Thank you for the diagnosis

Bots too, vanderBOT!

I used to work in robotics, and can't remember the password for my usual username so I pulled this one out of thin air years ago

I had a problem with a Realtek wifi card, where it would become slow for a few seconds every couple of minutes, had to disable a setting , maybe it helps you: https://wiki.archlinux.org/title/Network_configuration/Wirel...

I would have looked at this if I hadn't already jumped ship. It sounds like the problem I was having.

That's a bargain Johnny boy! My company gives me $250 in AI tokens to use every day!

It's great, this is the kind of content I look for in HackerNews, I'm so fucking tired of LLMs and AI posts

I would appreciate some extra markings on the keycaps for chords and numbers, and a sans-serif font, but this looks awesome as it is.

I love your take <3

Following your analogy, if there is a way for the lawyer to never lose a case due to misinterpreting the law...