marauding_gibberish142

joined 2 weeks ago
[–] [email protected] 1 points 6 hours ago

I have an alternative for you if your power bills are cheap: X99 motherboard + CPU combos from China

[–] [email protected] 1 points 9 hours ago (2 children)

In general how much VRAM do I need for 14B and 24B models?

[–] [email protected] 3 points 11 hours ago

I didn't know that. I thought just one ROCM binary to install, run Ollama and that's it. Thanks for the explanation

[–] [email protected] 1 points 11 hours ago (4 children)

Do you have any recommendations for running the Mistral small model? I'm very interested in it alongside CodeLlama, OogaBooga and others

[–] [email protected] 1 points 11 hours ago (1 children)

Wait how does that work? How is 24GB enough for a 38B model?

[–] [email protected] 1 points 11 hours ago

The 7900XTX was $1000 when it launched, I wouldn't mind it used either.

[–] [email protected] 2 points 11 hours ago (1 children)

I don't mind multiple GPUs but my motherboard doesn't have 2+ electrically connected X16 slots. I could build a new homeserver (I've been thinking about it) but consumer platforms simply don't have the PCIE lanes for 2 actual x16 slots. I'd have to go back to Broadwell Xeons for that, which are really power hungry. Oh well, I don't think it matters considering how power hungry GPUs are now.

[–] [email protected] 1 points 11 hours ago (1 children)

I am OK with either Nvidia or AMD especially if Ollama supports it. With that said I have heard that AMD takes some manual effort whilst Nvidia is easier. Depends on how difficult ROCM is

[–] [email protected] 1 points 11 hours ago (6 children)

Thank you. Are 14B models the biggest you can run comfortably?

[–] [email protected] 1 points 11 hours ago (2 children)

Do you have 2 PCIE X16 slots on your motherboard (speaking in terms of electrical connections)?

44
Consumer GPUs to run LLMs (lemmy.dbzer0.com)
submitted 1 day ago* (last edited 1 day ago) by [email protected] to c/[email protected]
 

Not sure if this is the right place, if not please let me know.

GPU prices in the US have been a horrific bloodbath with the scalpers recently. So for this discussion, let's keep it to MSRP and the lucky people who actually managed to afford those insane MSRPs + managed to actually find the GPU they wanted.

Which GPU are you using to run what LLMs? How is the performance of the LLMs you have selected? On an average, what size of LLMs are you able to run smoothly on your GPU (7B, 14B, 20-24B etc).

What GPU do you recommend for decent amount of VRAM vs price (MSRP)? If you're using the TOTL RX 7900XTX/4090/5090 with 24+ GB of RAM, comment below with some performance estimations too.

My use-case: code assistants for Terraform + general shell and YAML, plain chat, some image generation. And to be able to still pay rent after spending all my savings on a GPU with a pathetic amount of VRAM (LOOKING AT BOTH OF YOU, BUT ESPECIALLY YOU NVIDIA YOU JERK). I would prefer to have GPUs for under $600 if possible, but I want to also run models like Mistral small so I suppose I don't have a choice but spend a huge sum of money.

Thanks


You can probably tell that I'm not very happy with the current PC consumer market but I decided to post in case we find any gems in the wild.

[–] [email protected] 9 points 1 day ago

Seedboxes go from €2 to €100+ a month depending on how much you will torrent and how much space you need on the box alongside other factors. My personal choices are Gigarapid and Ultra but there are others

[–] [email protected] 0 points 3 days ago

What are you talking about mate? The VPS will be a wireguard server, your device will connect to it and use the VPS' IP address to connect to your game

 

I've been thinking about this for a bit but I couldn't come up with anything.

The idea is that you have a VOIP number and some self-hosted VOIP infrastructure connected to a landline phone. WhatsApp, Signal and voice traffic from other apps would be redirected to this landline phone instead of your mobile phone.

Is there a way to do this? How do I get started?

Reasoning: I can now keep my phone isolated, wrapped in a thick towel and inside a solid box to prevent it from eavesdropping on me inside my own house.

Please do not respond with messages like "you're too paranoid", it doesn't help.

Thanks

 

Hi,

The general consensus amongst the Android community is that rooting is detrimental to privacy. In a sense, I agree with them since privilege escalation because of human error becomes a much bigger threat if the user has root access.

Android has a big privacy problem encapsulated in one word: "baseband". Your modem and other hardware running in your device don't run FOSS firmware and are likely actively malicious towards your privacy.

I am a Linux user, and I understand that concepts do not necessarily transfer well between the two. With that in mind:

  1. If I wanted to be absolutely certain that sensistive hardware like Camera, Microphone and Modem were truly off, would shutting them off as root hold any real significance?
    • I do not know what the equivalent of Intel ME is called in the Android space, but I doubt that a highly complex OS is running beneath general Android as we know it. I think it's just the firmware of the individual device that we need to worry about.
  2. Is it possible to replace the bootloader on some Android devices/prevent it from loading unwanted firmware?

With Google taking Android behind closed doors, I suspect we will start seeing some suspicious snippets of code here and there with questionable purpose, but which might be missed by FOSS volunteers because of the sheer volume of work that is. I'm thinking of ways we can try to evade this blatant grab of our personal data.

 

I'm looking at quad port 2.5Gbe Intel PCIe cards. These cards seem to be mostly x4 physically (usually PCIe gen 3) whilst I have a PCIe Gen4 X1 slot, which is more the theoretical bandwidth that the card can support. The card needs at the most PCIE Gen 3 X2 == PCIE Gen 4 X1 in terms of bandwidth.

How do I fit the card into a PCIe x1 slot? Won't it lose performance if all the pins are not connected to the physical PCIe connector? Is there a PCIe x1 riser that the community likes that is somewhat affordable?

Thanks

view more: next ›