ULTIMATE Llama 3 UI: Dive into Open WebUI & Ollama!

okay let's get started setting up open web UI on our computers which is what you see up on the screen right now there are a couple prerequisits that I do expect you to have already taken care of one is make sure you already have olama set up we will not be going over that setup if you don't know how to do that check out the description below there are two videos one for mac and one for Windows for getting olama set up on your computer and the next prerequisite is having Docker set up so we'll cover that in this video jump over to the docker website okay I'm back and I've got the docker website pulled up here now we're going to need to download Docker if you don't already have it installed if you already have it installed go ahead and Skip for the chapters are below in the description all right so if you don't have Docker installed we're going to go up here to the top and we're going to go up to products and then we're going to go to Docker desktop now it's going to detect the OS that you're currently running I'm running a Mac and it's running the Apple chip now if you're running Windows you will obviously select the windows installer and if you're on Linux will select the Linux installer so I've select Mac here and next I am going to go to my downloads folder where I have that install and run the installer so I'll be back once this is done downloading okay I'm back and I'm in my downloads folder and you can see I have Docker downloaded here I'm on a Mac so I'm going to just double click on this here and I am going to drag this to my application folder now if you're in Windows just go ahead and walk through the installer it's pretty straightforward and run through that and come back after you're done installing that all right so I'm going to jump over to my applications folder and get Docker up and running on my computer okay I'm back and I'm in my applications folder and I have Docker selected here I am going to double click on this and what it's going to do it's going to start the docker engine on the back end for me and then this pops up here so we haven't pulled down any Docker images yet so that's what we will do next and if you're on a Windows computer you should be able to go down to your task bar and see that Docker is running in your taskbar once you go find the docker program so now that we've taken care of prerequisites we have Ama running on our computer and we have Docker install now let's jump over to the open UI website so that we can get this installed onto our computer via Docker okay I've got the open web UI website pulled up here and this is what you'll be greeted with it may be different by the time you watch this video so what we're going to do next is go up to the very top here and click on docs and we are going to go to the get started section and we're going to scroll down here feel free to reference the link in the description for the open UI website but we're going to go roll down here and we're going to find the docker command that we need to run here and the command that we're going to run is the first command here because we already have Ama installed locally on our laptop so I'm going to go ahead and copy this command here and next I am going to open up terminal but before we jump over into the terminal I do want to let you know this is by no means meant to be a deep dive into how Docker operates and runs it's just meant to give you enough knowledge on how to get Docker up and running so you can get open web UI running within a Docker container all right let's jump over into the terminal okay I've got terminal pulled up here if you're on a Windows it'll be your command line and go ahead and open that up and I am going to now paste the command that I copied from the open web UI documentation here and I am going to hit enter now what you'll notice it didn't find the image locally on my computer so it's pulling that image down onto my computer and installing that for me in Docker so I'll come back after this is done okay I'm back and my open web UI image has been pulled down into Docker so you should see a screen similar to this now we're going to open up the docker application and just briefly go over what has been set up for us so I'll see you back in the docker application okay I'm back and I have my Docker desktop application open up here on my computer and for those who are not familiar with how to open that up if you're on a Mac you're going to go up to the top here and you're going to click go to dashboard and it'll open it up for you in fill the windows go to where you have your programs installed and click on Docker desktop and that should open up for you so the first thing that you can see here is you can see I have a container that was created with that last command that we ran in terminal and this is where our actual open web UI application exists so we'll be able to click this and open that up but before we do that let's jump over to the images section here and the image represent the copy of the application that is stored in our Docker desktop so we can create multiple containers based on this image so that's what our container is based on and next we have the volume section here now this is where our persistent disc exist and this is attached to our container so if we were to go up to the top here and delete delak this container that volume would persist so just think of it as a persistent disc on your computer that stores data for you so I wanted to go over that briefly with you before we dove into the open web UI application itself so let's go back up to Containers here and now let's click on the section where you see 3000 colon 8080 and we're going to click on that and it's going to open up Local Host 3000 for us okay we can see that the browser just opened up a signin page for us but we need to create an actual account now I'm going to click sign up down there and we have this screen appear so go ahead and walk through these steps here and then come back and go ahead and log into your open web UI account now one thing to call out these credentials are stored locally within your container so this is not sent back over to open web UI this is stored locally within this container okay so after you've created your open web UI account and you have logged into your open web console here you'll be presented with this interface here and you're probably already familiar with this if you've used chat GPT is essentially a clone with some extra features built in here now I'm only going to do a highle overview of the most necessary features that you need to know about in order to get the most value as you initially start using the open web UI application now you can see up at the top here that we have the option to select a model here now I've only downloaded the Llama 3 instruct model now the models you see here up at the top drop down here are the actual models that we've pulled into our AMA instance on our computer so if you're not familiar with olama again go ahead and check out the two videos down below for Windows and Mac on how to set up o llama on your computer all right so what I'm going to do is select the Llama 3 Model here and I am going to say write me some sample python code I always just like to use this as an example here to see what kind of output we get so we can see it's printing us some sample python code here next we can see that we have features over here just the same as chat GPT we can go over here and share this we can rename this we can delete this chat and we can also archive this chat here so I'm going to go ahead and archive this and we can look at that in the settings later on so I'm am going to go up to the top you still have the ability to create new chats so just like chat GPT we can select new chat here or new chat here here and when you create a new chat you'll obviously have to select a model from the dropdown but that's not an option obviously within chat GPT if you hang around to the end of the video I'll show you how to pull new models directly within the open web UI interface okay we've covered the basics nothing here really new to see if you're familiar with using chat GPT so I'm going to go over some other features that I think it may be important for you to know as you initially start using the open web UI I'm going to go here where my name is and select that and then we have the admin panel now I'm going to click the admin panel here and you can see it has the users I have signed up here so I've just have one user but you can create multiple users to access this instance of open web UI so nothing really special here going on of course you can delete a user and edit a user all right and if you have any questions about anything maybe you want to see something in a little more detail again this is just going to be a basic overview of open web UI just put a comment down below and maybe create another video on that all right and now I'm going to go over here and I'm going to click playground so similar to the playground on the open ai's website I can add a system prom here and I can just test the chats out here so I'm going to select the Llama 3 Model here and I'm just going to say what is your name and click submit here and of course he says I'm just a AI I don't have a personal name so that's how you use the playground nothing really fancy there there going on now if we go back here and we go to Archive chats we can click there and we can see that chat that I archived earlier when we initially started using the open web UI interface here and of course we can share this can actually unarchive this chat here and we can delete this archive chat here so I'm going to go ahead and click delete and I no longer have archive chats it's basically the same feature that's in chat GPT all right I'm going to go back over here and I am next going to go to the settings SE section now I'm not going to dive super deep in the settings section I'm only going to dive into the areas I think you really need to know right out of the gate and again I can make a more advanced video on open web UI at a later date depending on the comments that I get so in the general section we can set our theme do we want it dark do we want it system my system set to dark so it's showing up as dark and I can change my language here also we can set a custom system prompt here for our underlying models I'm just going to leave that blank for now so that really covers the general section and now I'm only going to cover two more sections here I'm going to cover the connection section here just in case you have a different configuration or you want to change your configuration for your underlying olama instance now here this is pointing to my base olama URL now this here this URL here is a Docker specific URL and it's just saying point back to the host so whatever host is hosting this Docker instance point back to it at this particular Port here so essentially this here is equal equal to this here now if I copy that and I open up a new tab and I paste it you can just see it says AMA is up and running now I won't go too deep into this here but this is again specific to Docker because if it had Local Host here it would be the container pointing back to itself and the container itself doesn't have olama hosted on it we're looking at our local instance of AMA so I'm on a Mac if I go up here to the top we can see ama's running here directly on my computer here so hopefully you understand why that's there if you wanted to say you wanted to host AMA on a completely separate computer within your lab setup or something like that and you had the IP you could change this here to point to an IP of a separate computer where olama is running at so hopefully you understand that feel free to comment below if I wasn't super clear on that all right so in the the last section that I am going to cover in the settings is the models and again like I said if you hung around to this point in the video this is where we can actually go and pull different olama models directly from ama.com right here so for example say I want 53 and I want to pull that down I know the name of that model there so I can enter it in here and it will download that model for me I also can delete a model here so I am going to go ahead and click download here and show you this downloading directly within the open web UI without me having to open up AMA on the termal so I'm going to click download here and then you'll eventually see progress bar start to tick here so we can see that it's now gathering this model downloading it onto my computer this is a fairly small model here so the 53 I believe it was three gigs and some change for the size of that model at least for the the dis space that it takes up on your computer and so for the time that I'm been sitting here it downloaded pretty quickly depending on your internet connection it may take a little bit longer so we'll let this finish downloading here I will actually go ahead and well it's done already now so we can see it's successfully been downloaded now so I am going to click X here on this settings tab and again I'm not going to go into these are not the most important settings to get up and running again I may make a separate video on these other settings but if you understand the general the connections in the model section it's good enough to get up and running so we're going to exit out of this here and let's go back and create a new chat here now so I'm click new chat here I'm going to scroll up to the top here where it says select model and we can now see we have our new model here downloaded onto our computer without actually having to go into the terminal or the command line and run the AMA pole command so I'm going to select that and tell it to write me a short story and see what kind of output we get so it's just writing us a a short story here within the output now we didn't cover everything that was available to us in the open web UI such as model files and prompts and documents again if you would like to see a separate video on those things feel free to comment down below and let me know and then I will create a separate video dedicated to those sections here but think we've learned enough here on how to get open web UI up and running and if you'd like to create your very own personalized olama models feel free to check out the video up on the screen where I walk through that in a step-by-step Manner and if you like the content hit like And subscribe I do videos like this on a weekly basis thanks for hanging out to the end of this video and see you in the next one

Channel: AI DevBytes

Views: 12,131

Keywords: DevTechBytes, generative ai, ai, ollama, ollama windows, llama 3, llama 3 tutorial, llama, meta ai, llama test, llm, artificial intelligence, openai, chatpgt, models, open source, run ollama locally, llama 2, large language model, ollama on macos, how to install ollama

Id: D4H5hMMoZ28

Length: 14min 10sec (850 seconds)

Published: Tue Apr 30 2024