I've just created c/Ollama!
from catty@lemmy.world to selfhosted@lemmy.world on 24 Jun 21:56
https://lemmy.world/post/31938331

I’ve just re-discovered ollama and it’s come on a long way and has reduced the very difficult task of locally hosting your own LLM (and getting it running on a GPU) to simply installing a deb! It also works for Windows and Mac, so can help everyone.

I’d like to see Lemmy become useful for specific technical sub branches instead of trying to find the best existing community which can be subjective making information difficult to find, so I created !Ollama@lemmy.world for everyone to discuss, ask questions, and help each other out with ollama!

So, please, join, subscribe and feel free to post, ask questions, post tips / projects, and help out where you can!

Thanks!

#selfhosted

threaded - newest

gashead76@lemmy.world on 24 Jun 22:07 next collapse

Cool! I’ll subscribe. I’ve got about a dozen projects I’d like to build with Ollama, if I’ll get the motivation and free time who knows?

catty@lemmy.world on 24 Jun 22:09 collapse

Start now! Install it, get a python environment up and running if you haven’t already, and get that first play-around project working which you work outwards from!

gashead76@lemmy.world on 24 Jun 22:16 collapse

Already setup! I think the first thing I want to do is setup retrieval augmented generation. Several of my hobby ideas will require it I think. Started trying to read up on it a couple days ago and I had a serious lack of focus going on.

I’ve been kind of hoping to come across a super simple way to implement it, but haven’t exactly looked much yet.

catty@lemmy.world on 24 Jun 22:35 collapse

Sounds like a great first question! Go for it!

infeeeee@lemmy.zip on 24 Jun 22:14 next collapse

Instance independent link: !Ollama@lemmy.world

Share links to communities this way, so everyone can subscribe easily.

You should also post about this in !newcommunities@lemmy.world and !communitypromo@lemmy.ca for better discoverability!

catty@lemmy.world on 24 Jun 22:20 collapse

Thanks will do all that!

brucethemoose@lemmy.world on 24 Jun 22:47 next collapse

TBH you should fold this into localllama? Or open source AI?

I have very mixed (mostly bad) feelings on ollama. In a nutshell, they’re kinda Twitter attention grabbers that give zero credit/contribution to the underlying framework (llama.cpp). And that’s just the tip of the iceberg, they’ve made lots of controversial moves, and it seems like they’re headed for commercial enshittification.

They’re… slimy.

They like to pretend they’re the only way to run local LLMs and blot out any other discussion, which is why I feel kinda bad about a dedicated ollama community.

It’s also a highly suboptimal way for most people to run LLMs, especially if you’re willing to tweak.

I would always recommend Kobold.cpp, tabbyAPI, ik_llama.cpp, Aphrodite, LM Studio, the llama.cpp server, sglang, the AMD lemonade server, any number of backends over them. Literally anything but ollama.

…TL;DR I don’t the the idea of focusing on ollama at the expense of other backends. Running LLMs locally should be the community, not ollama specifically.

southernbeaver@lemmy.world on 24 Jun 23:30 next collapse

What would you recommend to hook to my home assistant?

brucethemoose@lemmy.world on 24 Jun 23:40 next collapse

Totally depends on your hardware, and what you tend to ask it. What are you running? What do you use it for? Do you prefer speed over accuracy?

EncryptKeeper@lemmy.world on 25 Jun 01:18 next collapse

I’m going to go out on a limb and say they probably just want a comparable solution to Ollama.

brucethemoose@lemmy.world on 25 Jun 01:57 collapse

OK.

Then LM Studio. With Qwen3 30B IQ4_XS, low temperature MinP sampling.

That’s what I’m trying to say though, there is no one click solution, that’s kind of a lie. LLMs work a bajillion times better with just a little personal configuration. They are not magic boxes, they are specialized tools.

Random example: on a Mac? Grab an MLX distillation, it’ll be way faster and better.

Nvidia gaming PC? TabbyAPI with an exl3. Small GPU laptop? ik_llama.cpp APU? Lemonade. Raspberry Pi? That’s important to know!

What do you ask it to do? Set timers? Look at pictures? Cooking recipes? Search the web? Look at documents? Do you need stuff faster or accurate?

This is one reason why ollama is so suboptimal, with the other being just bad defaults (Q4_0 quants, 2048 context, no imatrix or anything outside GGUF, bad sampling last I checked, chat template errors, bugs with certain models, I can go on). A lot of people just try “ollama run” I guess, then assume local LLMs are bad when it doesn’t work right.

WhirlpoolBrewer@lemmings.world on 25 Jun 16:01 next collapse

I have a MacBook 2 pro (Apple silicon) and would kind of like to replace Google’s Gemini as my go-to LLM. I think I’d like to run something like Mistral, probably. Currently I do have Ollama and some version of Mistral running, but I almost never used it as it’s on my laptop, not my phone.

I’m not big on LLMs and if I can find an LLM that I run locally and helps me get off of using Google Search and Gimini, that could be awesome. Currently I use a combo of Firefox, Qwant, Google Search, and Gemini for my daily needs. I’m not big into the direction Firefox is headed, I’ve heard there are arguments against Qwant, and using Gemini feels like the wrong answer for my beliefs and opinions.

I’m looking for something better without too much time being sunk into something I may only sort of like. Tall order, I know, but I figured I’d give you as much info as I can.

brucethemoose@lemmy.world on 25 Jun 16:22 next collapse

Honestly perplexity, the online service, is pretty good.

As for local running, one question first: how much RAM does your Mac have? This is basically the factor for what model you can and should run.

WhirlpoolBrewer@lemmings.world on 25 Jun 16:35 collapse

8GB

brucethemoose@lemmy.world on 25 Jun 16:43 collapse

8GB?

You might be able to run Qwen3 4B: huggingface.co/mlx-community/…/main

But honestly you don’t have enough RAM to spare, and even a small model might bog things down. I’d run Open Web UI or LM Studio with a free LLM API, like Gemini Flash, or pay a few bucks for something off openrouter. Or maybe Cerebras API.

…Unfortunely, LLMs are very RAM intensive, and >4GB (more realistically like 2GB) is not going to be a good experience :(

WhirlpoolBrewer@lemmings.world on 25 Jun 17:11 collapse

Good to know. I’d hate to buy a new machine strictly for running an LLM. Could be an excuse to pickup something like a Framework 16, but realistically, I don’t see myself doing that. I think you might be right about using something like Open Web UI or LM Studio.

brucethemoose@lemmy.world on 25 Jun 18:47 collapse

Yeah, just paying for LLM APIs is dirt cheap, and they (supposedly) don’t scrape data. Again I’d recommend Openrouter and Cerebras! And you get your pick of models to try from them.

Even a framework 16 is not good for LLMs TBH. The Framework desktop is (as it uses a special AMD chip), but it’s very expensive. Honestly the whole hardware market is so screwed up, hence most ‘local LLM enthusiasts’ buy a used RTX 3090 and stick them in desktops or servers, as no one wants to produce something affordable apparently :/

psudojo@ioc.exchange on 25 Jun 18:49 collapse

@brucethemoose @WhirlpoolBrewer

*1650 and it works like a charm 🤌🏾

brucethemoose@lemmy.world on 25 Jun 21:17 collapse

1650

You mean GPU? Yeah, it’s good, I was strictly talking about purchasing a laptop for LLM usage, as most are less than ideal for the money. Laptop vram pools are relatively small and SO-DIMMS are usually very slow.

Things will get much better once the “Max” AMD SKUs proliferate.

brucethemoose@lemmy.world on 25 Jun 16:32 collapse

Actually, to go ahead and answer, the “fastest” path would be LM Studio (which supports MLX quants natively and is not time intensive to install), and a DWQ quantization (which is a newer, higher quality variant of MLX models).

Hopefully one of these models, depending on how much RAM you have:

huggingface.co/…/Qwen3-14B-4bit-DWQ-053125

huggingface.co/…/Magistral-Small-2506-4bit-DWQ

huggingface.co/…/Qwen3-30B-A3B-4bit-DWQ-0508

huggingface.co/…/GLM-4-32B-0414-4bit-DWQ

With a bit more time invested, you could try to set up Open Web UI as an alterantive interface (which has its own built in web search like Gemini): openwebui.com

And then use LM Studio (or some other MLX backend, or even free online API models) as the ‘engine’

Alternatively, especially if you have a small RAM pool, Gemma 12B QAT Q4_0 is quite good, and you can run it with LM Studio or anything else that supports a GGUF. Not sure about 12B-ish thinking models off the top of my head, I’d have to look around.

WhirlpoolBrewer@lemmings.world on 25 Jun 16:41 collapse

This is all new to me, so I’ll have to do a bit of homework on this. Thanks for the detailed and linked reply!

brucethemoose@lemmy.world on 25 Jun 17:10 collapse

I was a bit mistaken, these are the models you should consider:

huggingface.co/mlx-community/Qwen3-4B-4bit-DWQ

huggingface.co/…/gemma-3-4b-it-qat-q4_0-gguf

huggingface.co/unsloth/Jan-nano-GGUF (specifically the UD-Q4 or UD-Q5 file)

they are state-of-the-art at this size, as far as I know.

WhirlpoolBrewer@lemmings.world on 25 Jun 17:12 collapse

Awesome, I’ll give these a spin and see how it goes. Much appreciated!

southernbeaver@lemmy.world on 26 Jun 19:29 collapse

My HomeAssistant is running on Unraid but I have an old NVIDIA Quadro P5000. I really want to run a vision model so that it can describe who is at my doorbell.

TheHobbyist@lemmy.zip on 25 Jun 07:06 collapse

Perhaps give Ramalama a try?

github.com/containers/ramalama

tal@lemmy.today on 25 Jun 00:56 next collapse

While I don’t think that llama.cpp is specifically a special risk, I think that running generative AI software in a container is probably a good idea. It’s a rapidly-moving field with a lot of people contributing a lot of code that very quickly gets run on a lot of systems by a lot of people. There’s been malware that’s shown up in extensions for (for example) ComfyUI. And the software really doesn’t need to poke around at outside data.

Also, because the software has to touch the GPU, it needs a certain amount of outside access. Containerizing that takes some extra effort.

old.reddit.com/…/psa_please_secure_your_comfyui_i…

ComfyUI users has been hit time and time again with malware from custom nodes or their dependencies. If you’re just using the vanilla nodes, or nodes you’ve personally developed yourself or vet yourself every update, then you’re fine. But you’re probably using custom nodes. They’re the great thing about ComfyUI, but also its great security weakness.

Half a year ago the LLMVISION node was found to contain an info stealer. Just this month the ultralytics library, used in custom nodes like the Impact nodes, was compromised, and a cryptominer was shipped to thousands of users.

Granted, the developers have been doing their best to try to help all involved by spreading awareness of the malware and by setting up an automated scanner to inform users if they’ve been affected, but what’s better than knowing how to get rid of the malware is not getting the malware at all. ’

Why Containerization is a solution

So what can you do to secure ComfyUI, which has a main selling point of being able to use nodes with arbitrary code in them? I propose a band-aid solution that, I think, isn’t horribly difficult to implement that significantly reduces your attack surface for malicious nodes or their dependencies: containerization.

Ollama means sticking llama.cpp in a Docker container, and that is, I think, a positive thing.

If there were a close analog to ollama, like some software package that could take a given LLM model and run in podman or Docker or something, I think that that’d be great. But I think that putting the software in a container is probably a good move relative to running it uncontainerized.

brucethemoose@lemmy.world on 25 Jun 02:07 collapse

I don’t understand.

Ollama is not actually docker, right? It’s running the same llama.cpp engine, it’s just embedded inside the wrapper app, not containerized. It has a docker preset you can use, yeah.

And basically every LLM project ships a docker container. I know for a fact llama.cpp, TabbyAPI, Aphrodite, Lemonade, vllm and sglang do. It’s basically standard. There’s all sorts of wrappers around them too.

You are 100% right about security though, in fact there’s a huge concern with compromised Python packages. This one almost got me: pytorch.org/blog/compromised-nightly-dependency/

This is actually a huge advantage for llama.cpp, as it’s free of python and external dependencies by design. This is very unlike ComfyUI which pulls in a gazillian external repos. Theoretically the main llama.cpp git could be compromised, but it’s a single, very well monitored point of failure there, and literally every “outside” architecture and feature is implemented from scratch, making it harder to sneak stuff in.

tal@lemmy.today on 25 Jun 02:55 collapse

I’m sorry, you are correct. The syntax and interface mirrors docker, and one can run ollama in Docker, so I’d thought that it was a thin wrapper around Docker, but I just went to check, and you are right — it’s not running in Docker by default. Sorry, folks! Guess now I’ve got one more thing to look into getting inside a container myself.

hasnep@lemmy.ml on 25 Jun 11:01 collapse

Try ramalama, it’s designed to run models override oci containers

TheHobbyist@lemmy.zip on 25 Jun 07:06 collapse

Indeed, Ollama is going a shady route. github.com/ggml-org/llama.cpp/pull/11016#issuecom…

I started playing with Ramalama (the name is a mouthful) and it works great. There is one or two more steps in the setup but I’ve achieved great performance and the project is making good use of standards (OCI, jinja, unmodified llama.cpp, from what I understand).

Go and check it out, they are compatible with models from HF and Ollama too.

github.com/containers/ramalama

otter@lemmy.ca on 25 Jun 00:43 collapse

There is also !localllama@sh.itjust.works :)

crossposting between the communities can help grow both