61–70 of 92 posts
For anyone wondering “how slow is this?” IIUC, Kimi K3 on RTX 6000 Ada (48GB) takes 292 s/token https://github.com/lyogavin/airllm/releases/tag/v3.1.0
I do think these “run a bigger model than will fit in VRAM” projects are necessary steps, but are they functionally useful or helpful to anyone currently? For example, is anyone out there running a big Qwen for coding on a 16-32GB machine with these techniques?
For anyone wondering “how slow is this?” IIUC, Kimi K3 on RTX 6000 Ada (48GB) takes 292 s/token https://github.com/lyogavin/airllm/releases/tag/v3.1.0
For anyone wondering “how slow is this?” IIUC, Kimi K3 on RTX 6000 Ada (48GB) takes 292 s/token https://github.com/lyogavin/airllm/releases/tag/v3.1.0
I do think these “run a bigger model than will fit in VRAM” projects are necessary steps, but are they functionally useful or helpful to anyone currently? For example, is anyone out there running a big Qwen for coding on a 16-32GB machine with these techniques?
I've been building a SaaS that deals with data that can't be distributed to third parties. Some of the useful AI stuff I can add is not time sensitive and can run overnight. Things like this allow me to use higher quality models without selling my house for GPUs.
Seeing a lot of these “run 1TB models with 1GB RAM” projects recently. Most seem vibe coded and probably won’t be maintained. Hoping a winner emerges with some real momentum behind it.
Earlier quoted context omitted.
You don't really need a maintainer when codex or claude code can set it up for you; thats how I got Trellis2 working on windows and tiny VRAM despite Microsoft recommending you have 24GB VRAM and Linux. Models are pretty disposable now.
So, instead of having one or two maintained projects that work super well, now we have hundreds or thousands of people half-assing it locally each time they need it? Doesn't sound efficient.
now we'll just always be vulnerable from different angles in different and unique ways across the globe.
better? well, no.. but it is different, and biology has offered tons of wisdom about why it's a good thing to achieve things different ways across the world, maybe some far-future version of this weird fragmentation of work that is happening now will contribute to some kind of herd effect that reduces the severity or magnitude of some bug or work of malice.
who knows though, i'm probably just considering it too positively. The disintegration of consolidated effort is a pita, I agree.
Earlier quoted context omitted.
If I could justify wear and tear and electricity, I was willing to do something like this for batch processing. The batches would be a bunch of prompts whose outputs I'd look at the next day. Maybe common operations, like QA or refactoring, on whatever software I wrote. If so, I could use a larger model than I have real-time hardware for. The largest, well-trained models can often get the output mostly right in one t…
at 5 minutes per token, you could look at the results next week
If you had a deadly condition, and no diagnosis worked, and a specific model had the answer... past that I wouldn't use it.
Earlier quoted context omitted.
MoE has nothing to do with multimodality. MoE is a concept proposed in 1991, before the deep learning era (which is before what I call the transformers era). You can think of it like sharing. Contrary to popular belief; 'experts' in MoE LLMs do not specialize. There's no expert trained to be good at maths, or python, or writing, or whatever. It's an inference optimization. As for reasoning in non-text modalities, you…
Wait I thought the router ends up specializing the experts? Like there is no explicit goal aside from each 'expert' getting roughly equal weight? And it happens that when you train the router you do end up passing certain classes of problem to each expert - just as a training result nothing as clean as a python expert. But math vs creative writing will tend to rely on different experts over the majority of the infere…
Earlier quoted context omitted.
MoE has nothing to do with multimodality. MoE is a concept proposed in 1991, before the deep learning era (which is before what I call the transformers era). You can think of it like sharing. Contrary to popular belief; 'experts' in MoE LLMs do not specialize. There's no expert trained to be good at maths, or python, or writing, or whatever. It's an inference optimization. As for reasoning in non-text modalities, you…
Wait I thought the router ends up specializing the experts? Like there is no explicit goal aside from each 'expert' getting roughly equal weight? And it happens that when you train the router you do end up passing certain classes of problem to each expert - just as a training result nothing as clean as a python expert. But math vs creative writing will tend to rely on different experts over the majority of the infere…
Because the natural continuation of this is to say like, "Ok I want to load the bird detection expert and the navigation expert but leave the medieval European history expert behind", and my understanding is that this is not really how it works. At least at the moment.