gunalx

Its happy to see a new open image model from qwen. But the license is a let down. And it dosent even beat their closed qwen3 image wich is already a bit old.

Not at all in mine. Deepseek has some of the worst prose of the close to frontier models in my opinion.

Wait I don't realise ram prices where this bad. Bought my kit for about a tenth of that a while ago. Ouch.

If it can do innside out tracking its way more data efficent to have compute onboard. And if you already have a decent amount of compute you can scale it up just a little more to have. a fully standalone unit.

I would say this is correct behaviour though. Any touchup by ai directly means it is strictly ai generated. Even if it looks almost identical to a nontampered original. This is exactly the kind of things you do want to catch. I also would just assume they simply look for the ai "watermark" for this.

Look at the nanokvm pro. 4k@45 with pass through.

Bummer to not see the atx control support. Its the main use i have from my ipkvm to do fde decryption and boot the machine once wake on lan fails.

There is too many ads to report all of them. But a don't show me this, that also affected on report as a negative signal to reduce the possibility of that specific ad to you.

Can't wait till a customer rep is impossible to get to because all support is outsourced to ai.

Right. I got the feeling of it being saturated because all the top 5 fully completed it.

More often than not just knowing the contact has OK connection is good enough.

As far as I know the marker chips is mostly to tell the devices to not abuse the cable.

Wouldn't work. A dynamic but verifiable problem. Is just a perfect target for a RL environment. If you don't have the verifiable part the benchmark is useless, or really expensive with human review. (Or just open ended)

In fact we should though.

Ignoring the obvious ai slop webpage. I don't really trust the benchmark. It seems either pretty saturated, or inconsistent just based on the results.

A casual look around the office. Not a good representative selection for the entire population but counting to 10 and asking, everyone used ai in some fashion, extrapolate the data and you get 100%. (Not a ai first office even)

Fact is most people don't really care. And they don't really know why they should care either. All the problems feels like something far away, while they are being exploited and sold without fully realising.

I dont think anyone knowing full and well would be comfortable with it. At least that would dumbfound me.

I mean. Fundamentally they are just autoregressive next token predictors. Fundamently fou can simplify to f(x) -> x+y where x is input tokens, y is the next token and f() is the model function. Yes the model function is complex but still.

The problem I see with vue is what vue patterns are you using. And the llms may be trained on more outdated ones more than current practices.

God that had a lot of llm'isms. Couldnt get through it with all the repeats and info bloat.

This is so funny. Would love a more readable font option in the same style though.

Just as expensive unobtanium as nvidia. Maybe even more so.

Its probably the huawei ascend 910, they are using.

They are high end really expensive Huawei ascend GPUs. It is kinda bruteforcing the performance on a older semiconductor processing tech, so total production is pretty low.

Early glm models gave off that wibe. But now its more inspired by. With a bit of Claude in there. But I do think they actually do RL otherwise glm5.3 shouldn't have been able to beat fable on the few tests it did.

Yes, the terms are dubious. But they are also reasonably lenient with enforcement. They also don't require persona id verification, witch is wat turned me away from openai.

I have had the opposite where I felt 5.3 as a stronger model than 5.2. Its feels way more in tune with my code, and does more nuanced edits. Though I do handhold my models a lot, so might fall outside the agentic term.

glm 5.3 gets awfully slow during peak hours. But you might not hit them to frequently.

Cute. But I don't see the real uppsides here vs any other minipc.

Yes true. But also way to often neglected.

Single page applications are one of the most anoying web patterns i know of. Please just let me have one page for each ting to do so i actually can bookmark it properly.

Well that would not work as the metadata is also part of the sha, and if you have a sha predictably contain or hash to itself you have a bad hashing algorithm. (Unless of course you mean sha of the weigths and not the gguf itself)