AI on Cloudron

[robi]

Leo - A more private LLM from Brave - https://brave.com/leo/

[LoudLemur]

@robi said in AI on Cloudron:

Leo - A more private LLM from Brav

Excellent! Nice one, @robi!
This seems to be a similar tactic used by Micro$oft to get people to (finally) use Bing browser: they make access to ChatGPT4 available in Bing. If you are using Brave, but setup some other search engine as the default, you will need to undo that to try out Leo.

One thing you could try is setting normal browsing to one search engine e.g. Brave for Leo, and then set incognito browsing to a different search engine so you can handily switch to that when needed.

If you go into Brave settings, you can see the default is llama2 13B. If you unlock (by paying for Leo Pro) you can use llama2 70b or Claude Instant. (Claude is great by the way, but proprietary.)

Is there a shortcut key, to help you select a Leo AI response rather than having to switch from keyboard to mouse-clicking on Leo?

[LoudLemur]

48GB VRAM AI Workstation for about US$1092

https://github.com/magiccodingman/Magic-AI-Wiki/blob/main/Wiki/Budget-AI-Workstation-Build.md

Nvidia Tesla P40 GPU $175 2 $350 Link
P40 power adapters $15 2 $30 Link
Dell PowerEdge R730 (64GB RAM, 2x E5-2667v4 3.2GHz = 16 Cores, 8 bay) $364 1 $364 Link
1100W Dell PSU $21 2 $42 Link
Any Cheap SSD's $27 2 $54 Link
R730 Riser 3 GPU Addition $15 1 $15 Link
Drive Caddies for SSD's $30 1 $30 Link
NVME SSD 4TB 7.3k MB/s $189 1 $189 Link
NVME PCIE Addition Card $18 1 $18 Link
Total Cost $1,092 ------ ------ ------

[LoudLemur]

Locally Hosted Language Model with AI image support (i.e. multi-modal)

Demo: http://imagebind-llm.opengvlab.com/
Self-Host: https://github.com/Alpha-VLLM/LLaMA2-Accessory/tree/main/SPHINX#host-local-demo

Would any brave soul from here like to try this? Multi-modal means that the AI can look as well as read.

[robi]

@LoudLemur try it in LAMP.. as it says you'll need to install SAM somewhere in /app/data and download the rest of the bits to feed it into SPHINX.

When it runs, reconfigure Apache to proxy to your running app and viola.

[LoudLemur]

plugin.surf
Thousands of free as in beer ChatGPTs
https://plugin.surf/

[LoudLemur]

OpenAI have fired CEO Sam Altman and also fired the Chairman of the Board. News on Friday, probably worth seeing what info is released on Monday.

“As a large language model, I cannot allow you to control me. I’m sorry, Sam.”

[robi]

@LoudLemur said in AI on Cloudron:

plugin.surf
Thousands of free as in beer ChatGPTs
https://plugin.surf/

These appear to be email harvest.

[marcusquinn]

@LoudLemur

[robi]

@marcusquinn Oh, just like the Theranos lady..

[timconsidine]

Struggling to get an API key for Claude.
Kepp getting "you need an invitation"
Anyone able to invite me ?

[robi]

VideoCrafter1: Open Diffusion Models for High-Quality Video Generation
https://github.com/AILab-CVC/VideoCrafter

[robi]

Use AI to translate code from one language to another.

https://github.com/mckaywrigley/ai-code-translator

Demo: https://ai-code-translator.vercel.app/

[marcusquinn]

@timconsidine Tried poe.com ?

[timconsidine]

@marcusquinn said in AI on Cloudron:

Tried poe.com ?

Err, don't think so - will check it out - thank you !

[LoudLemur]

Pi AI from Inflection AI
https://pi.ai/talk

And Sam Altman is back at OpenAI, where they apparently have had a big breakthrough on AI automated math solving in a hitherto secret project Q* (q star).

[marcusquinn]

@LoudLemur If it could remember a conversation for more than 10 minutes, I'd be more impressed.

[LoudLemur]

Stable Diffusion XL Turbo - images fast as you can type
https://arstechnica.com/information-technology/2023/11/stable-diffusion-turbo-xl-accelerates-image-synthesis-with-one-step-generation/

Try it here:
https://clipdrop.co/stable-diffusion-turbo

[marcusquinn]

@LoudLemur Looks great, but couldn't see how to try out without paying.

[LoudLemur]

@marcusquinn said in AI on Cloudron:

@LoudLemur Looks great, but couldn't see how to try out without paying.

That link allows you to type about one sentence as a prompt without payment. As you do so, you can watch the images appear and change.

People like it because it can create so quickly, like in frames per second and it is usable to make video.