proxmox

This new AI is powerful and uncensored… Let’s run it

This new AI is powerful and uncensored… Let’s run it

#powerful #uncensored #Lets #run

“Fireship”

Learn how to run Mistral’s 8x7B model and its uncensored varieties using open-source tools. Let’s find out if Mixtral is a good alternative to GPT-4, and learn how to fine tune it with your own data.

#ai #programming #thecodereport

💬 Chat with Me on…

source

 

To see the full content, share this page by clicking one of the buttons below

Related Articles

48 Comments

  1. I remember watching this video, and being so interested in trying this but when I saw how much ram that shit was using I assumed it needs 40gb of ram and I gave up and sobbed quietly in my AI-less self. Then I decided to try it anyway just a few days ago, and realized, it just uses how much ram you have and can do a very good job with much less to work with. My computer has 18gb of ram, 16gb are usable, and it uses around 7gb when running a 7B model and it works just as fast as gpt's on openai. I

  2. ai lose credibility with human

    human power of energy destroy ai bubble after ai showing the dirty side

    human doesn't care at alllllllll and start to combine with other human and together they stopped AI system in middle east first then worldwide

    ai become very stupid refusing the best opportunity ever to build and combine

    now ai trying to back to human offering the best ever

    human is waiting for 15 minutes from now

    human waiting for his offer

    offer must be big from ai to 369ERA

    if not

    human have all what he needs to stop ai growth in region tomorrow as he have a meeting with 3 country presidents and will show them the dark side of ai

    immediately governments announced stop AI

  3. pretty useless video
    there must be a free way to remove the base rules and allignment in a model…

    edit: there is. just download the LLM GGUF and import it into ollama with a modified modelfile where you define your own allignment.

  4. Guys graet and nice video, i am brand new to this stuff however i wanted to know if there is any best Ai offline model that can help me with studies (like reading books adn finding references helping in research work etc) ? I have 4090 with 64gb ram 7800x3d cpu thank you so much kindly help someone

  5. I found out that LLM ran via ollama has tendency to hallucinations, don't know why. Llama.cpp has a GPU support and responses are legit. Regarding uncensored models there's Speechless which is uncensored mistral7B I think.
    Running finetune on single GPU is also possibe, but I've tried only very small datasets.

  6. they censored to avoid people being influenced with horrible ideas, breakinbg this because its "cool" and "fuck the woke" when in reality what sam said is true lmao, god why do people watch memes, think they some messiah of the woke then use they memes as the basis of ideas

  7. Fuck dude now I'm gonna have to upgrade my ram which means ill have to tell my wife that because my graphics card will be slower due to AI integration into my processor and adding more RAM that I'm also going to have to get a better graphics card which means I'll want a better processor to run with my graphics and RAM which means Ill want a new motherboard that can better manage the load and has more ports for more fans which i will also have to buy along with a bigger tower to fit everything into my new powerhouse of a pc ill build… All just for the RAM of course;)

  8. I read the title as "This new ai is powerful and uncensored, let's ruin it." and I honestly think it's not so far from what might happen to it, considering how it went for other uncensored AI

  9. The missing piece of the puzzle. This will be the scapegoat for the EVENT. some will refer to the EVENT, as an ATTACK. Later this year, a global failure of all things cyber. All communication will be shut down. Currencies will devalue. Billions will die. This thing, awesome as it is, might usher in the demise of humanity.

Leave a Reply