> Hugging Face is the bottleneck, not your link.<p>README could clearly make use of a cleanup, seems to be more like a session log dump now than a good introduction to the project for a new user. Maybe try something like "Remove anything from the README.md that wouldn't be helpful to someone who sees this project with zero context, for the first time. Rewrite all paragraphs and sections to be concise and remove all fluff, leave only important details new users must know before using the project".
As someone who is just looking at the theoretical benchmarks of each of these models I'm curious if anyone could share what are the problems (maybe around code) that flash-next was able to solve which 27b was not able to
This is the best I could find: <a href="https://huggingface.co/Qwen/Qwen3.8-Flash-Next?utm_source=chatgpt.com#language" rel="nofollow">https://huggingface.co/Qwen/Qwen3.8-Flash-Next?utm_source=ch...</a><p>About the specifics, I have only anecdotal evidence, but I guess this info can be found somewhere
I'm hoping to see progress in this space.<p>Folks talking about how 32G is not enough for local use, but then there's been work like this to empower it.<p>My hope is that the new 32G M6 will be "useful" locally, possibly because of work like this.
It's hard to believe 16GB unified memory will give you 5 tok/sec unless you are ignoring the thermal warnings. I am running Qwen3.6-35B-A3B on my 16GB M3 and get 7-8 tokens/sec with all the optimizations while keeping the peak memory and thermal warnings at check. <a href="https://github.com/deepanwadhwa/samosa-chat" rel="nofollow">https://github.com/deepanwadhwa/samosa-chat</a>
Now I'm feeling pretty good about getting 10-11 tokens/sec running Qwopus 3.6-35B-A3B Q6_K on an old Mac Pro 2013 (trashcan) with 128GB RAM (DDR3), 12 core Xeon, dual D700s. Arch Linux and llama.cpp.
interesting! Yes, thermal is important. Pretty cool project man! Starred and checking it out!
how much energy does it consume?
It seems we could use a new kind of memory that streams the weight data in, like GDDR in reverse.
High Bandwidth Flash? <a href="https://www.sandisk.com/company/newsroom/blogs/2025/scaling-beyond-the-wall-inside-sandisks-high-bandwidth-flash-for-ai" rel="nofollow">https://www.sandisk.com/company/newsroom/blogs/2025/scaling-...</a>
yes! I guess future hardware designs will have something like that!
Is this going to destroy my SSD?
"Disk is the gate that bites first"<p>AI;DR
There are already a handful of repos doing essentially exactly this: `mlx-moe-offload`, `streamlx`, `mlx-moe`, `mlx-flash`, and `deepseek-v4-flash-mlx` - i.e. keep the resident parts of an MoE in unified memory and page/stream routed experts from SSD on Apple Silicon.<p>At this point I'd much rather see people collaborate on one of these implementations, benchmark against them, or upstream the useful bits into MLX/MLX-LM instead of producing yet another near-identical repo.<p>The local-LLM ecosystem really does not need every implementation idea rediscovered five times and wrapped in a new README. AI-assisted coding makes producing a new repo cheap; maintaining, benchmarking, and integrating one is the actually valuable part.
This is one of the aspects of this year that I've been finding very grating and wasteful. Collaboration still happens among people with the ability to do so and the technical skills, but everyone else is taking their own helicopter to the top of the mountain, "putting it out there", and there's just a ton of redundant projects that do the same thing.
I see your point. As an oss defender myself, I agree, however, the spirit of this is to see how fast I can make it. I'm sharing this with the community, which I think is aligned with the original oss spirit.<p>It's an experiment for myself but I am committing to maintain it. I've been an oss person for a loooong time, way before AI was a thing. Think about it as a new, from-scratch take at it, not as a re-reproduction.
Hey carloslfu, kudos from the other side of the internet, don't get down on people nitpicking everything here, experimenting and discovering is part of learning so keep going!, remember this is the place that said dropbox was dumb and could be replaced by a script.
Vouched especially since OP might have a perspective on this. And readers may want to look up those other repos and compare for themselves.
Why should they do that? For you? You could merge those projects and see if they get traction.
And there are `Mference` and `SwiftLM` too, I think they are doing the same use case.
AI NIH
> every implementation idea rediscovered five times and wrapped in a new README<p>That's open source since forever, unfortunately.
I agree with the sentiment, but have you seen those videos in which all men say other men are gay? This feels like the same, so much AI paranoia!<p>I genuinely want to contribute. And hey! I was doing oss this since 2014 so waay before AI was cool.
It's what happens when you don't do market research.
I'm sorry this makes it seem like I didn't do my research. I did a TON. To fix it I'll add a benchmark/comparison table. Also, I wouldn't call it market research since this is not commercial AT ALL.
Does a painter check to make sure that a portrait hasn't been painted? What a dismissive comment.
Half the people on here are using Ollama. No one is doing market research.
[dead]
[dead]
[dead]