Show HN: Getting GLM 5.2 running on my slow computer

Show HN: Getting GLM 5.2 running on my slow computer(github.com)

140 points by vforno 14 hours ago | 27 comments

A few days ago I found myself trying out GLM 5.2 and was really positively impressed. The capabilities and security I was getting from this LLM are similar to those I've gotten from models like Claude or GPT, and this really surprised me.

But then I thought, "I wonder how it would work on a normal computer like mine," and above all, "I wonder if it would work without going into OOM on a computer like mine." So I started working with the help of agents to test this possibility.

I started converting the model to int4, understanding MTP usage, and if possible implementing DSA for long context. How it responds in int4 and whether the quality is maintained or not. Until I got to the point, on my computer with 32GB of RAM, I was able to communicate with GLM 5.2 with times that, of course, aren't high in cold start, but even then, we're talking about 0.1 tok/s, but that wasn't important to me. The important thing was the journey to reach this goal. I just wanted it to work at all costs, even slowly.

So I created Colibrì, which was born from a very simple idea, to be honest, but tested in every way, where a 744B Mixture-of-Experts model activates only ~40B parameters per token—and only ~11 GB of those change from token to token (the routed experts). So:

The dense part (attention, shared experts, embeddings—~17B params) stays resident in RAM at int4 (~9.9 GB); The 21,504 routed experts (75 MoE layers × 256 experts + the MTP head, ~19 MB each at int4) live on disk (~370 GB) and are streamed on demand, with a per-layer LRU cache, an optional pinned hot-store, and the OS page cache as a free L2.

The engine is a single C file (c/glm.c, ~1,300 lines) plus small headers. No BLAS, no Python at runtime, no GPU.No GPU or serious hardware because I don't have that hardware so I can't test it on hardware that is more powerful than my computer.Colibrì is a one-person project, written and tested entirely on a 12-core laptop with 25 GB of RAM — the numbers above are the ceiling of what I can measure at home.

Any feedback is welcome! (and if anyone wanted to participate in the project I would be delighted)

Repo: https://github.com/JustVugg/colibri

walrus01 1 hour ago |

My main question is whether when put into practical use, this can be measured in tokens/second, or more like 1 token per minute... I have seen locally hosted LLM that are as slow as 1 tok/second still be very useful if you give it a project to do something overnight and metaphorically walk away from it, check back with what it has done in 6 or 8 hours.

0.05 to 0.1 tok/s on the other hand, as reported in the URL for the lowest class of hardware, isn't really usable for much.

edit: I think this is a fantastic project in general concept, and look forward to seeing more efforts towards the general idea of being able to run a 350B to 900B size model locally, even if as slow as 1 tok/s, on hardware that ordinary people can afford. Anything along the general concept of "we have fast read NVME SSD storage, we have a big ass model on local disk, we'll read it at 11GB/tok as we need it, not try to load the whole thing".

vforno 1 hour ago | |

In the readme you can see benchmark which everyone with different hardware is running Colibrì, and I have to say I've seen great times! I'm always doing more to improve!

walrus01 1 hour ago | | |

I have a 16-core system with 256GB RAM here I could try it with but regretfully it's so old the CPUs aren't AVX2 capable. Otherwise it makes a fairly good llama-server test system for CPU only stuff. Oh well. Time to upgrade (painful to the wallet these days).

Cieric 31 minutes ago |

I was actually just working on the same thing as this, but I went down the route of mmapping the entire model into memory to avoid the extra ram usage. I also had Claude implement Medusa[1] on the model to try and avoid loading an additional model into memory but still get the benefits of MTP. Currently at a stop light so I can't list everything and I didn't get to read your full post either yet.

To expand since I just got home, I'm making all of my modifications to llama.cpp, the goal was to eventually put this on a SBC of some kind with an nvme to handle the mmapped files. I think the theoretical limit of my current setup is about 1.8 tok/s based on prior testing but that is also with the additional medusa heads not fully trained (I honestly don't know if the counting it's generated tokens or not.)

In the end it seems like the idea we had is similar, I just don't know how to write an llm parser/runner from scratch yet and instead of specifying what needed to stay in memory I just let the linux kernel handle it.

Oh last note, I also capped llama.cpp usage to 16GB of my 32GB, so it might be possible to get it down even lower.

[1] https://arxiv.org/abs/2401.10774

kodablah 29 minutes ago |

I've taken a similar strategy w/ image/video gen at https://github.com/cretz/thinfer (see video branch for a ton of work).

Basically I kept needing an inference engine that could stream weights in and out as needed in an LRU manner. So I ended up vibe coding this thing that accepts a `--vram-budget` and stays under it (mostly). It turns out moving mmap'd bytes in and out of VRAM is way cheap compared to compute. Coupled with some pipelining/double-buffering, I almost always end up compute bound not memory bound. Granted I use way smaller models heh.

shrinks99 1 hour ago |

Pretty cool! I've also been playing around with GLM 5.2 this week and was equally impressed. At work we're running it locally on some crazy expensive hardware as a test before starting another project so it's great to see people taking this massive FOSS model release and running it on an average machine, even if it's not terribly practical at this point.

Nice work!

vforno 1 hour ago | |

Really thanks!!

miohtama 1 hour ago |

This is the hacker spirit

vforno 1 hour ago | |

Thank you so much, it's true! It all started with this spirit!

bobim 48 minutes ago |

I'm not fully understanding this business of MoE so please forgive me if this is a dumb question, but would it be possible to use MPI with a small cluster to distribute the load?

vforno 44 minutes ago | |

It’s a good question.

In theory MPI could distribute experts across nodes. In practice, for small clusters the added network latency usually hurts more than it helps.

Better suited for big clusters with fast interconnects. For now we're focusing on single-machine speed (caching, GPU hybrid, etc.).

khalic 1 hour ago |

I love seeing that kind of tinkering

vforno 1 hour ago | |

Really thanks!

mariopt 1 hour ago |

I wonder if you could replicate this in a Colourful GeForce RTX 50-series GPU, they ship it with 2 NVMe drive slots.

vforno 1 hour ago | |

I'd love to! Right now I only have a very consumer-grade computer that I've had fun with! We'll see!

xfalcox 1 hour ago |

Question to the OP, have you tested this on a machine where the entire model and context fit in RAM ?

walrus01 52 minutes ago | |

I think if you had something like a theoretical used/refurb 2U rackmount server with two older multi core CPUs, 768GB of RAM, you would see faster performance loading a Q6 or Q8 GGUF of GLM5.2 into a freshly-compiled latest copy of llama-server, with the "no-mmap" option turned on to intentionally load the whole thing into RAM at the time the llama-server daemon launches.

If you want a CPU-only machine with 512GB to 1024GB of RAM, despite extreme cost rises, there are still some great options out there from companies selling ex-lease stuff that's 3, 4, 5 years old. It'll be loud as hell under full CPU load when running inference, so if you plan to use it at home, put it in your garage or basement or laundry room or somewhere similar on the far end of a network cable.

The software that OP has published appears to be specifically designed to hold only the active parameters in RAM (<100GB) and read content off local NVME SSD as needed on the fly. All that NVME SSD read wouldn't be necessary if you can hold the model in RAM, even in the absence of any GPUs.

vforno 1 hour ago | |

No because I have only 32gb of ram too low

stavros 17 minutes ago |

This is great, well done! I love seeing people run things where they weren't meant to be run.

kzrdude 1 hour ago |

Your coding style is halfway to IOCCC. I'm just jealous though :)

Pragmata 1 hour ago |

Would this cause issues with SSD lifespan?

vforno 1 hour ago | |

What causes problems is the rewriting in this case are only read while writing is the cache! However, I'm working to improve more and more and make some parts lighter!

Pragmata 27 minutes ago | | |

Is it possible to run this into an agent? pi, claude code, etc..? I've only tried it with LM studio, but i'm guessing this is a bit different

xfalcox 1 hour ago | |

README covers that

https://github.com/JustVugg/colibri#ssd-wear-warning

nerder92 1 hour ago |

Is this inspired by antirez work on ds4?

Amazing job!

vforno 1 hour ago | |

Antirez is the number one!thanks really thanks!