Viewing profile — Iolaum

TBH Taalas was a company I was existed about as a consumer. A dense model like gemma4-31b or qwen3.6-27b running at 10k t/s sounds like an awesome thing to have. Would be willing t…

There's an old joke that goes about like: How do you solve Global Warming? Nuclear Winter ... I really hope this is not how we end up geo-engineering our way out of Climate change.…

It's not even at society scale IMO. How many times do people at a company deprioritize small-to-medium fixes on things that may/will become a problem later? Sometimes you have to w…

Not sure this is a meaningful factor for normal circumstances (for example below 1km altitude and for water that's safe to drink). According to CGPT at 1km altitude the boiling tem…

Do think about b2b. Companies are already paying much more for AI. a new K3 (or similar model) every 6 months for a monthly rate of ~100$ per month is something MANY businesses wou…

Do they? With Fable/Opus duo they have a better product and their customers are paying for quality. They want to be the premium LLM provider letting others compete for the commodit…

A really fast qwen-3.6-27B type of model could be useful. With a specialized harness and this speed I 'd expect it to find many applications. Implementing a coding plan is the mini…

QWEN-3.6-27B P.S. Couldn't resist :p

That's like saying a sling is the same as an assault rifle. Yes both are weapons but scale and capabilities matter.

Personalized psyops against voters ...

Was the temperature 0? Cause unless I don't understand it right, any non-zero temperature implies probabilistic next token prediction. You did mention, seed, which I haven't seen a…

While I agree this is the case, I think this is a consequence of low education for the general public and low quality of the public discourse. If the quality of those two goes up, …

Since they did this with their own harness I m not sure it's apples to apples comparison.

Hey if you can externalize that cost, more money left for you and the investors !!!

Own a framework desktop 128gb. GPU bandwidth limits the amount of tokens/sec that you get. I 've mostly been running QWEN-3.6-35B-A3B at Q8 and QWEN-3.5-122B-A10B at Q4 with Q8 kv-…

It's not even IP. Model outputs are not protected (to the best of my understanding).

Maybe because a non public agreement was in place?

Open Source models decelerate growth of closed AI. For people who think (or want) AI = closed_AI then that argument has weight. Good luck getting them to update their priors.

OpenCode has an autoretry functionality (that progressively waits more after each failed request). I m surprised other harnesses don't have that.

There's an emerging practice of using Q4 quants and Q8 KV cache for local inference. At that point you can run both Qwen3.5-122B-A10B (my personal choice on Framework Desktop 128gb…

As long as they are transparent about what quant they serve the model and any other optimization they do that also affects performance of inferred tokens.

I m not sure about that. It raises the cost of doing business, but US tech giants still get to dominate the EU tech landscape.

TSLA and SPCX issues are not short term.

Likewise here, my setup costs 2k+ more eur as well. Given that memory bandwidth remains the same I don't think it's worth buying if you have the 1st gen framework desktop. Usecase …

Model Looks amazing! Even more important, subjectively, is that this model will run very well on Strix Halo (e.g. Framework Desktop), DGX Spark kinds of devices. Looking forward to…