GPT4All HUGE Upgrade 🚀 Commercial License, 1-Click Install, New UI, New Base Model

hey welcome back in today's video we're going to talk about GPT for all two and this is the brand new version of GPT for all which was one of the original local llms that you can run on any machine just four gigabytes is all you needed and any consumer computer mostly could use it but now we have a new and improved version ready to go the company behind GPT for all nomic AI has made a number of improvements to it so let's go over those improvements then I'm going to show you how to install it and then we're going to do a demo let's get into it so the previous version of GPT for all was based on the Llama model so a lot of people didn't know that the license for llama did not allow for using it commercially even though a lot of people did this new version is based on a model called gptj from a company called a Luther Ai and now you can use it commercially which is awesome so the team from nomec AI completely retrained GPT for all based on the GPT TJ model and now you can use it commercially GPT for all was already one of the best if not the best local models it was fast it was smart and it was pretty comparable to GPT 3.5 by open Ai and the gptj model is just as good if not better and they've made the process of installation a thousand times easier wrapping it in an installation package making it essentially one click not only that but they have a really beautiful interface that you can use now and one click updates to the app so they've really made it end user friendly for people who aren't super technical which I really appreciate the last thing that I'll mention before we actually get into the installation is that you can soon provide nomic AI with your data as you're using GPT for all this new version and what that'll do is that'll give them a ton of training data for subsequent mod levels and this is going to be really important for having GPT for all and other local llms get on par with open ai's models so I'm really excited about that I'm definitely going to be sharing my data with them and with that let's get into the installation so the installation really couldn't be easier now it's one link you download the installation package and once you open it up it's going to download the model install all the files the interface and everything you need let me show you that now so let's download the installation package I'm using a Mac so I'm going to use the Mac installation and we're going to download it now you may get a warning from chrome that this file may be malicious feel free to send it to Google for them to scan it before actually opening it now I've already done that so I'm going to go ahead and open it now here's the install file go ahead and double click now if you try to just double click it and open it you're going to get a warning on a Mac that this is an unverified developer so you can go ahead and click cancel now to get past this you're going to right click on and click open and then open so this is it this is the installer we're going to click next and then we're going to choose our installation directory I'm going to put it in applications click next again and select components the only thing in there is the chat so I'm going to click next accept the license next install and so now it's going to download the models it's going to download all the other files necessary to run this so let's let it run for a few minutes and I'll come back okay so we're finished so we click finish and now we're going to open up the applications folder and go to the GPT for all chat folder within there open up the bin folder and there's this little file called chat double click on it and there it is we're all done here's the interface it's beautiful it's simple it looks mobile friendly and it's based on GPT J Riot python code outputs all prime numbers from one to one thousand so it's interesting at the beginning it says as an AI language model I don't have the capability to write python code and then it goes on to write python code so I've actually tested this before and it didn't say that and it just output the code so that's very interesting now I haven't tested this code but it does from looking at it seem correct and then it explains what the code does at the end so let's try something else who is Michael Jordan and it seems to be really fast I'm on a MacBook Air M2 so it's a pretty modern computer but yeah it seems very very fast let's try another one what are Transformers and why are they important important to AI Transformers and AI are essential components that allow a neural network to learn and improve they are used to adjust the weights and biases of the network during training etc etc so that's it congratulations to the nomic AI team for shipping this this will make adopting local models that you can install on any Edge device really really easy I've seen examples of installations on really really interesting devices like TI-84 graphing calculators and Nintendo switches so this is just going to make everything even easier congratulations to the nomic AI team if you like this video please consider giving a like And subscribe and I'll see you in the next one [Music] foreign

Channel: Matthew Berman

Views: 35,875

Keywords: chatgpt, gpt4all, gpt4all install, local llm, ai, artificial intelligence, chatgpt explained, what is chatgpt, chatgpt tutorial, chat gpt, llama, alpaca, gptj, gpt-j, gpt j, eleuther ai, install llm, install ai locally, llama model, alpaca model

Id: scEMax2r4ts

Length: 6min 0sec (360 seconds)

Published: Thu Apr 13 2023