Open WebUI & OpenAI DALL-E 3: Effortless Text to Image Generation

okay by the end of this video you'll be able to easily generate images using open web UI and Dolly 3 but before we get started if you haven't already set up open web UI on your computer check out part one in this video series down in the description below which will walk you how to set up open web UI on your computer step by step all right let's get started I've got open web UI pulled up in my browser so if you haven't already go ahead and log into your instance of open web UI and next navigate down to where your name is and we're next going to go to the settings section so I'm going to click settings and after we go to settings since we're going to set up image generation with an open web UI we're going to next navigate over to the images section here once we're at the images section here we're not going to use the default setting here we're going to use the open AI Dolly setting here now what we're going to need to do is get our API key within open AI so I'm going to assume you've got an account with open AI I'm going to hop back over there and get my API key and that's where I'll meet you next okay once you've logged into your open AI account you're going to next want to navigate over to your API key section here we're now going to click API keys I don't have any API Keys created so I'm going to go ahead and click create new secret key now you'll be presented with this screen here I am going to name my key open web UI demo and just following security best practices here I am going to select restrict it I am only going to give this key access to the model capabilities it doesn't need any other access all right so I'm going to select right and next I'm going to scroll down here and click see secret key okay so you can see my key has been created I'll delete this one before I post this video so we're going to go ahead and copy this key and we're going to hop back over to the open web UI interface okay I've got my open web UI instance pulled back up here again I'm at the images settings section here now be sure you copy your API key or open AI we don't need to change this here we'll leave that the same and next what I'm going to do is paste in my key here and then one other thing we need to do is turn on experi Al here so I'm going to turn that on and I am going to use the dolly 3 there's Dolly 2 you can save some cost if you want to use the lower quality model I like to use Dolly 3 it generates better images for you so I'm going to leave that setting there and the other thing I'll call out here is we need to set the image size now we're only able to enter three different image sizes here for Dolly 3 now I've select 1024x 1024 that is the smallest size image that you can create using the dolly 3 Model now I'm going to jump over to the documentation really quickly to show you where you can go and get these image sizes so you know okay I've got the dolly 3 documentation pulled up here and we can scroll further down here to get an idea of the different image sizes that we can use and they call out that it was trained on certain types of images so therefore it can only generate 1024 x 1024 1024x 1792 or 1792 by 1024 size images so I just wanted to call that out to you so you know where to go get these image sizes all right let's hop back over to open web UI okay we're now ready to save our image settings so I'm going to go here and click save here and we can see settings were successfully saved so I'm going to close that out now we're ready to start actually generating images all right so let's do that next okay now that we set up our image settings let's start generating some images here now I'll walk you through this process step by step now one thing I do want to call out here I am using llama 3 instruct at the top to create my image prompt so what I'm going to do is go down to the bottom here and I have an image prompt already put together here and it says create a one- sentence image prompt of a line and tiger facing off us your creativity now the reason I put one sentence image prompt here is these models tend to get quite wordy when you start asking them things so I want to limit it in its output all right so let's go ahead and send this over to the model and see what we get back here and this is the output that the model gave us so we can use this output or we can go here and tweak this output here so if I wanted to change some of the verbiage in this this output that the model gave to me I can go here and click edit if I wanted to but I'm not going to do that I'm going to see what it gives me based on this prompt that is created for me now in order to create the image we're going to go here and we're going to hover over the image icon down there and we're just going to click that to generate image now this may take a few minutes to create that image for you so I'll be back once this image is done being created okay I'm back and you can see the image that's been created for us so it did a pretty good job at creating the image based on the description of the prompt down below now I will call out one thing this is not entirely free cuz we are using our open AI API key now you may be curious on what the cost would be on that so if you hang out to the end of the video I will show you how you can see how much it's going to cost for creating certain types of images using Dolly okay let's try creating one more image here now one thing to call out during the editing of this video I did change the dimensions of the image this image Dimension is 1792 by 1024 so let's go over to the settings again here and go to our image section and let's change this back to 1024x 1024 and click save and I'm going to exit out of that and next let's start a new chat I'm going to go up and click new chat and I'm going to paste another prompt that I have here and I am going to enter that in there and I said hey create a 1 to two sentence image prompt of a metallic blue Ferrari 458 Italia racing through a mountain path and for those who aren't big into cars that's just an exotic sports car so we can see that the output that we got here for our image promp so let's go down here and click generate image and see what type of image it generates for so I'll be back after this image is done being created okay I'm back and the image is now being generated and it did a pretty good job at creating the image based on the prompt that we have here down below now I know I mentioned if you hang out to the end of the video I'll show you how to check out the prices for generating images using Dolly 3 so let's jump over to look at how we can go about seeing how much this is going to cost us for generating images okay I'm back and I've got the open aai pricing page pulled up here now it covers all the different model cost here what I'm going to do is just run run a find on the page and type in images here to take me right down to the image model section here now we can see the different breakdowns by model here so we can see Dolly 3 standard Dolly 3 HD and Dolly 2 so what we're using here is the dolly 3 HD and so it's costing roughly 8 cents per image when you using the 1024 x 1024 and then when you use the other two Dimensions it's costing 12 cents per image so I do want to make you aware of that and be transparent that this is going to cost you something if if you're using open AI as your backend model if you watch my channel you know I really like open source models but I did get a comment down below someone wanted to know how to set this up in open web UI using the dolly model so that's why I'm creating this video here but I really do like focusing on primarily open source models but once there's other image generation models that are open- source that can run on Lower tier Hardware I will start covering those right now you really need a beefy GPU to do really good image generation on your computer with the current models all right thanks for hanging out to the end of the video hopefully you found it helpful if you did hit like subscribe turn on the notification Bell I create videos like this on a weekly basis so again thanks for hanging around and see you in the next video

Channel: AI DevBytes

Views: 1,636

Keywords: AI DevBytes, AIDevBytes, openai, open webui, openwebui, generative ai, Dall-E 3, dalle 3, openai dalle 3, openai dall·e 2, artificial intelligence, chatgpt, ai robots, ai, open web ui, ai image generator

Id: DHVQ1UBaYMQ

Length: 7min 12sec (432 seconds)

Published: Wed May 15 2024