Zuck’s LLaMA 3 Is HERE and SMASHES Benchmarks (Open-Source)

[@1MinuteFlipDoc]

sweet! it's available on Ollama already! hahaha

[@CognitiveComputations]

And yes uncensored

[@BoostedFA]

Music to my ears.

[@spidey-y3n]

uncensored? which version is it? i tried the instruct version and it's censored

[@stickmanland]

@@spidey-y3n Dolphin LLaMa 3

[@Oscar_delta13]

​@@spidey-y3n uncensored version is on Open Web UI community models.
Or, you can use a specific system prompt.
With patience... llama 3 in my experience is very heavily catered to providing objective truths so logical reasoning can be used to easily tune it.

[@aladinmovies]

Amazing model

[@emiandere]

Keep posting man love you ❤

[@incapablehulk3538]

why they had to name this technology after the most random animal

[@BoostedFA]

I was thinking of making my own user interface for ollama since youve sparked my interest into running other models instead of GPT. Got an open source one you could recommend?

[@ayushmishra5861]

Where are you testing Llama with UI ? Do they have a website?

[@sennetor]

yes, via the facebook/metaccount - you need to have one unless oyu have hardware to run

[@flameguy3416]

Local

[@stickmanland]

Yes, meta chat or whatsapp

[@Oscar_delta13]

Its OpenWeb UI + Ollama

[@borntodoit8744]

how to install LL3
#goto ollama library to see models to download
#install ollama
#install llama3 8b (4.7GB)
ollama run llama

[@marcusk7855]

Is there anyone that benchmarks models on home gpus. To compare like a RTX 4070 16G to a 3090 etc..? I want to buy a GPU for my PC just to run local models but don't know how much I need to spend.

[@wayne30047]

How can I chat with it online?

[@renderartist]

Sweet! 👏🏼👏🏼

[@MrGaborKukucska]

Awesome!!! 🎉

[@MrKarlyboy]

Sorry but jeff minter will have something to say about this lol

[@stanTrX]

Llama3 or Mistral?

[@stickmanland]

LLaMa 3

[@JohnMartindale-f9y]

Meta! 🤔😢🤷‍♂️
Like Mark Zuckerberg Meta?😬

[@dollarking9641]

3d2y