Incredible artificial intelligence Easy Diffusion 3.0. The Middle and High School Guide
Incredible artificial intelligence Easy Diffusion 3.0. The Middle and High School Guide
Alexander Chesalov
This book is written for middle and high school students.With the help of it and the Easy Diffusion 3.0 artificial intelligence system, you will learn how to create unique and inimitable images that you can use in your studies or for entertainment.The book contains educational information about existing types of artificial intelligence and a wonderful album of more than a hundred illustrations.
Incredible artificial intelligence Easy Diffusion 3.0
The Middle and High School Guide
Alexander Chesalov
Illustrator Vsevolod Chesalov
Illustrator Diana Chesalova
Cover designer Alexander Chesalov
Illustrator Chesalov Alexander
© Alexander Chesalov, 2024
© Vsevolod Chesalov, illustrations, 2024
© Diana Chesalova, illustrations, 2024
© Alexander Chesalov, cover design, 2024
© Chesalov Alexander, illustrations, 2024
ISBN 978-5-0062-7082-4
Created with Ridero smart publishing system
Gratitude
I thank all the people who are involved in the development of the Easy Diffusion generative artificial intelligence system. Your work helps people understand how artificial intelligence works, as well as discover new creative abilities in themselves.
Introduction
The wonderful world of Artificial Intelligence can be fully revealed to us only when you and I can see the positive results of its work, especially those created with our direct participation. These results must be understandable and explainable to everyone, and they must also be ethical, unbiased and not violate the law.
Today, Artificial Intelligence can do a lot, for example: write the text of a new poem or even a whole story, reproduce it in different voices of famous actors or musicians, analyze a large amount of numerical data and make a forecast for the future, play computer games with us or with a thousand people at once games. Perhaps the most impressive thing, in my opinion, that Artificial Intelligence can do today is create unique and incredible images. These images can be the embodiment of a three-dimensional world of a fantastic future in a computer game or be a virtual simulation of the world around us. The images can be three-dimensional or two-dimensional, and can also be made in various painting styles by famous artists from different time periods. But the most interesting thing is that these images may contain creatures or objects that do not exist in our world.
In this book, you will acquire new knowledge about how diverse Artificial Intelligence can be, and you will become familiar with its most popular terms and definitions. You will learn what is «artificial superintelligence», «narrow Artificial Intelligence», «friendly Artificial Intelligence», «human-centered Artificial Intelligence», «distributed Artificial Intelligence», «trusted Artificial Intelligence», «composite Artificial Intelligence», «explainable» Artificial Intelligence» and much more.
I will tell you how you can transform and visualize the knowledge you have acquired using an Artificial Intelligence system, turning the text definitions of various terms from this book into fantastic drawings.
In order to bring all this to life, we will study the work of a free program – the Easy Diffusion generative Artificial Intelligence system. This program creates photorealistic images from text queries and graphic images.
Easy Diffusion is an easy-to-install assembly (distribution) of the Stable Diffusion Artificial Intelligence system. Easy Diffusion will allow us to create unique images based on the most popular neural network models today, such as Stable Diffusion, Midjourney, OpenJourney, DreamShaper, SDXL Yamer’s Realism!, ProtoVision XL, Art Universe, Realistic Vision and many others.
With the help of this book, you will learn not only how to use Easy Diffusion Artificial Intelligence as a creative tool, but you will also be able to independently install, configure and uninstall this program on your computer, connect various additional modules to it, which will significantly expand its capabilities for creating and processing images.
With the help of Artificial Intelligence, you will learn to create unique and inimitable images that you can use in your studies or for entertainment.
In order to be objective in my work and make the book as accessible as possible to a wide range of readers, I asked my son Vsevolod (5th grade student) and my daughter Diana (8th grade student) to help me work on the textbook and create various images that reflect the essence of the definitions of Artificial Intelligence presented in the book. It was interesting for me to involve them in a new activity for them, and it was also very important to understand for myself how quickly they would learn to work with the Easy Diffusion Artificial Intelligence system and be able to create their own high-quality images.
It turned out that fifteen minutes of instruction was enough to begin their first acquaintance with Easy Diffusion Artificial Intelligence and become familiar with the work of the Realistic Vision model. Their first image was a steampunk style hamster. Steampunk is a science fiction movement incorporating technology and arts and crafts inspired by technology and 19th century science fiction[1 - Steampunk. [Electronic resource] // ru.wikipedia.org URL: https://ru.wikipedia.org/wiki/Стимпанк (date: 26.07.2023).]. Nevertheless, they gained more or less serious experience working with Easy Diffusion after some time.
It took us more than six weeks of work to fill this book with illustrations. From more than five hundred drawings, we have selected for you the most, in our opinion, impressive and interesting. The total computer time we spent working with the Easy Diffusion Artificial Intelligence is approximately three hundred and sixty hours. What can we say about the fact that writing the book itself, correcting and editing the text, as well as selecting the necessary pictures required a lot of time and effort.
Almost six months have passed since the first edition of this book was published. During this time, more than seventeen thousand readers became acquainted with the first edition of this book. Since then, the version of the Easy Diffusion program has also changed.
This book is written as a textbook for middle and high school students. It contains detailed instructions for working with the Easy Diffusion 3.0 Artificial Intelligence system, which allows you to transform the text you write or your image into a new unique image. It also contains educational information about existing types of Artificial Intelligence. The book features a beautiful set of over one hundred illustrations demonstrating the AI capabilities of Easy Diffusion 3.0.
We really hope that our positive and positive experience that we received in the process of studying the Easy Diffusion Artificial Intelligence system will be useful to you in your studies and creativity.
Enjoy your reading and productive work!
Artificial intelligence Easy Diffusion: first steps
Step 1. Most popular programs
Today, there are quite a large number of different programs that, based on Artificial Intelligence technologies, create incredibly beautiful images.
In my opinion, the following programs are the most popular:
– Midjourney is a very popular Artificial Intelligence system created by the company of the same name, Midjourney. Artificial Intelligence Midjourney is a web service on the Internet that allows you to create very beautiful images without using the computing resources of your computer. To use it, you must have Discord installed – a convenient messenger for group work, which is often used by players around the world. The main drawback of the program is that not so long ago Midjourney became completely paid.
Program website:
https://www.midjourney.com
– Leonardo.Ai is a colorful web service on the Internet that provides access to the Leonardo.Ai neural network of the same name. With it, you can create images from text and other images. The advantages of Leonardo.Ai include an easy-to-use interface. The disadvantage is the limitation on the number of images created per day. Images are created very quickly and with high quality, but it will take a lot of time until you learn how to make something high-quality. If you want to spend more time working with this program, be prepared to pay for its services. Very often Leonardo.Ai is compared to Midjourney, but which one is better is up to you to decide.
Program website:
https://leonardo.ai
– Dreamstudio.ai is a web service created by Stability AI, which has developed one of the most famous neural networks or, as they say, «models» called Stable Diffusion. The web service is very convenient and allows you to create any images very quickly. This is because Dreamstudio.ai uses very powerful computers provided by Stability AI. In other words, your computer is not involved in the process of creating images, which is a significant advantage of the program. At the same time, the disadvantage of using Dreamstudio.ai is the limitation on the number of images created.
Program website:
https://beta.dreamstudio.ai/dream
– StableDiffusionWeb UI is a web-based program that you can install directly on your computer. Its advantages include the fact that it is a completely free option for using the Stable Diffusion model. The program allows you to choose flexible settings to improve the quality of the image and convert it to a new version or create another image based on it. Disadvantages include some difficulties associated with installation and the need for you to have additional knowledge of the programs necessary for its operation (for example, such as Python and Git). Stable Diffusion can be recommended to all those users who have already gained extensive experience with Easy Diffusion.
Program websites:
https://github.com/AUTOMATIC1111/Stable-diffusion-webui
https://www.python.org/downloads/windows/
https://git-scm.com/download/win
https://Stablediffusionweb.com/#demo
– Stablecog is a very easy to use, but at the same time very interesting web service. Unfortunately, it’s paid. However, you can create twenty free images per day. I liked it because out of ten models you can choose two created by our compatriots. These models are called Kandinsky and Kandinsky 2.2. In addition, you can increase the resolution of the created image four times (for example, from 1248x832 to 4992x3328).
Program website:
https://stablecog.com/generate
– Easy Diffusion is a completely free Artificial Intelligence system that combines the capabilities of using Midjourney and Stable Diffusion. Easy Diffusion offers you a convenient and simple web interface that allows you not only to create very beautiful images, but also to connect various additional modules and models to its work. By default, the program already has the well-known Stable Diffusion model preinstalled. Additionally, you can download many other models from www.civitai.com, including the «Openjourney» model, which will allow you to replace the use of Midjourney. Another advantage of the program is that it is very easy to install on a computer or laptop and works with not very powerful video cards. The only drawback of the program is the time it takes you to create images. For example, if an image in “Dreamstudio.ai” is created in 20 seconds, then in Easy Diffusion it can take up to 120 seconds to create.
Once again, I would like to draw your attention to the fact that Easy Diffusion is an easy-to-install assembly of software packages (or, in other words, a distribution kit) of the Stable Diffusion Artificial Intelligence system for various operating systems. Here and further, I will not talk about Stable Diffusion, but rather Easy Diffusion due to the fact that it is with this distribution and program that we will work.
Program website (program distribution):
https://github.com/easydiffusion/easydiffusion
Step 2. Install Easy Diffusion
In order for us to install the Easy Diffusion program, follow the following link and download the required distribution for your operating system (MS Windows, Linux or MacOS):
https://github.com/easydiffusion/easydiffusion
There are three distributions on the program page:
• For MS Windows operating system.
• For the Linux operating system.
• For the MacOS operating system.
Before installing the program, please pay attention to the minimum requirements for your computer. You must have:
– Video card from NVIDIA with 2 GB of RAM (better with 4 GB). A laptop video card will do, but keep in mind that you must have the latest drivers installed not only for the video card, but also for all components of the MS Windows system.
– The minimum amount of computer RAM is 8 Gigabytes.
– The minimum amount of disk space (either on the «C:\” drive or on any other drive) is 40 Gigabytes (25 for Easy Diffusion and 15 for models). I recommend installing the program on a computer flash drive (not an external drive, but an internal drive of a computer or laptop). It will make image creation faster.
Next, we will consider installing Easy Diffusion on a computer with the MS Windows operating system.
Before installing, you need to make sure that everything is in order on your MS Windows system. To do this, you need to right-click on the «My Computer» icon and select the «Manage» command from the context menu. The Computer Management program will open, and in the Device Manager section you will see not only all the devices on your computer, but also their operating status. If there is a malfunction of any of the devices, you need to install new drivers for this device. You can find them on the Internet.
Next, download the Easy-Diffusion-Windows program.
After downloading the file named Easy-Diffusion-Windows, launch the installation and click the «Next» button:
We agree to the license agreement and click the «I Agree» button:
Click the «I Agree» button again:
Select the disk on which the program will be installed. The main recommendation is that the program should be located in the root of the disk, that is, it should be installed on the disk, but not inside other folders, but only in its own.
For example, on the drive C:\EasyDiffusion\, as shown in the image below:
Or on drive D:\EasyDiffusion\, as shown in the image below:
Select the disk you need and click the «Install» button.
The process of installing and unpacking the files on your computer has begun:
The installation process takes some time.
During this process, the program will install all the necessary files and models for work, including installing the Stable Diffusion model (version 1.5), with which you will create your first works of computer art:
The program will install other auxiliary models (for example, GFPGAN model, Real-ESRGAN model, etc.), which are used by Easy Diffusion to improve the images you create.
At the end of the process, a CLIP model will be established, which allows you to determine how well the generated image matches the query you wrote.
After installation is complete, the following window will appear:
But don’t rush to click the «Finish» button.
Be sure to check the «Create Desktop Shortcut» box and uncheck the «Run Easy Diffusion 3.0» box.
And only now click the «Finish» button.
After completing all the steps, the program will be completely installed on your computer.
Why didn’t we launch the program immediately after installation?
All this is necessary because before we start working with the Easy Diffusion generative Artificial Intelligence system, we need to download and install a few more additional model files that will allow us to create incredibly beautiful images.
What is a model?
A model is a file containing a set of parameters of a trained neural network. The file size can be small or it can be several GB in size. It depends on what a particular model is created and used for.
Each model is trained on a large number of thematic images. A set of images or data is called a «dataset».
There are a large number of different models for creating photorealistic images in various artistic styles. Some models work better for creating anime images, while others work better for imitating famous artists.
To continue working, we need to follow the following links and download the latest versions of the model files with which we will work:
– Realistic Vision версия 5.1
(https://civitai.com/models/4201/realistic-vision-v51);
– DreamShaper версия 8.0
(https://civitai.com/models/4384/dreamshaper);
– Color Fusion версия 1.0
(https://civitai.com/models/20617/color-fusion);
– Midjourney версия 4.0
(https://huggingface.co/prompthero/openjourney/blob/main/mdjrny-v4.ckpt);
– Openjourney версия 4.0 (модель обучена на рисунках Midjourney)
(https://civitai.com/models/86?modelVersionId=27392);
– SDXL Yamer’s Realism! 2.0
(https://civitai.com/models/136669/sdxl-yamers-realism-realisticanime3d).
For example, we need to download the Realistic Vision model file version 5.1.
Follow the link above to the page where there is a model file:
https://civitai.com/models/4201/realistic-vision-v51
Then we find the file of the latest version of the model for downloading:
And download the file by clicking the «Download» button:
The realisticVisionV51_v51VAE.safetensors file, 1.9 GB in size, will be downloaded to your computer.
If you do not have an electronic version of the book, you can enter the names of the model files in any search engine and find them on the Internet. For example, upon request «download model Realistic Vision» you will be offered a list of sites for downloading, including the site civitai.com, from which I downloaded most of the files I needed.
After you download all or most of the models offered to you for download, you need to transfer all the files (using the «cut» and «put» commands) to a folder along one of the paths: «C:\EasyDiffusion\models\Stable-diffusion» or «D:\EasyDiffusion\models\Stable-diffusion», depending on where you installed your program.
For this:
– Open the folder on the disk where Easy Diffusion is installed (for example, «C:\EasyDiffusion\” or «D:\EasyDiffusion\»).
– Then find and open the «models» folder.
– In it you will see a folder called «Stable-diffusion». Open it and transfer the model files you downloaded into it.
Congratulations!
You have become the proud owner of your own Easy Diffusion Artificial Intelligence system, installed on your personal computer and supplemented with new models.
Now we can proceed to the first launch.
Step 3. Launch Easy Diffusion for the first time
If immediately after installing the program you selected the «Run Easy Diffusion 3.0» command, as shown in the figure below:
A black window will appear in front of you with running lines of incomprehensible text:
You shouldn’t be afraid of this window, since this is the window where the application itself works, or, more precisely, its server part. This window will always appear when the program starts working. There is no need to close it until you decide to finish your work. Just roll it up.
You need to wait a minute or two (depending on the performance (power) of your computer), after which another window will open in the Internet browser (installed by default). This will be the same Easy Diffusion user interface (or it can also be called the client application of the program), with which we will work throughout the entire time.
Easy Diffusion’s interface is dark, but that’s fixable. In the next chapter we will learn how to make our work more comfortable using program settings.
Thus, when we first launched the program, two windows opened: a window for the server part of the program and a window for the interface.
But don’t rush to get started.
I would recommend that you close both windows and start the program again. This is primarily due to the fact that when you restart the Easy Diffusion program will automatically check for updates, download and install all the necessary new files on its own.
Wait a little longer and your default Internet browser will open again.
After which we can begin with you the next step of work – setting up comfortable and stable operation of Easy Diffusion.
Step 4. Setting up Easy Diffusion
Setting up Easy Diffusion is actually very simple.
By default, the program has options that provide a high level of comfortable work and do not require any changes. However, there are a few that require our attention.
The figure below shows the user interface of the program:
So, let’s get started.
Easy Diffusion has only five main application work window tabs, which are placed one after the other at the top of the user interface work area:
– «Generate».
– «Settings».
– «Help & Community».
– «Models tools».
– «What’s new.»
All our work will take place in the main «Generate» window, which we will devote most of our time to in the next chapter.
In this chapter you will look at the contents of the following bookmarks:
– «Settings».
– «Help & Community».
– «Models tools».
– «What’s new».
«Settings»
To get started, we first need the Settings window.
For your convenience, I will use the «Light» theme of the program:
The Settings window provides the following options:
– «Theme». Allows you to change the current theme to the one you like. I’m using the «Light» theme.
– «Auto-Save Images». Let’s activate the option. It allows us to save all our images and descriptions for them in a folder we specify.
– «Save Location». Leave the default value or change it to the path where our files will be stored. For example, C:\EasyDiffusion\models\Stable Diffusion UI.
– «Metadata format». We change to «txt» – the text data format in which our requests will be stored. They can be read using any text editor.
– «Block NSFW images.» Let’s activate the option. Prohibits working with photos of certain categories.
– «Enable Sound». Let’s activate the option. We will always receive a sound notification that our image is ready.
– «Process newest jobs first.» Determines the order in which images are processed. By default, jobs are processed in order. We don’t change.
– «Extract LoRA tags from the prompt.» LoRA stands for Low-rank adaptation and is translated as «Low-rank adaptation». LoRA speeds up the process of training models. We do not change the parameter.
– «Open browser on startup.» The option is responsible for launching the Internet browser after the server part has started. We don’t change.
– «GPU Memory Usage». The option transfers all program work from the video card to the central processor. If you select «High», the program may run very slowly or stop working altogether. We need to select «Law».
If your video card is not NVIDIA GeForce RTX, then set it to «Low».
– «Use CPU (not GPU).» The option shifts all the work of the program not to the video card, but to the central processor. If you select a central processor, the program can run very, very slowly. We don’t change.
– «Auto-Save Settings». Click the «Configure» button. All options must be selected:
To do this, change the «Enter Prompt» value.
And close the window.
– «Confirm dangerous actions.» The option is responsible for the security of the application and must be active.
– «Profile Name». Leave the default value «default».
– «Beta channel». The option provides access to not yet very tested and unstable functions that may appear in the next version of the application. We don’t change.
– «Use the new v3 engine (diffusers).» A very important function for us, which in the left panel of Text-to-image AI software increases the list of parameters for customizing the created image. Let’s activate the option.
After you have made the changes, click the «Save» button to save all the changes made.
Below are a few more options that you do not need to change:
Three options control the network settings of your Easy Diffusion application. We don’t change them.
Please note that under the settings there is system information that allows us to see the data of the CPU and video card we are using.
My computer details are:
– Intel64 Family 6 Model 60 Stepping 3, GenuineIntel.
– NVIDIA GeForce GTX 980 (cuda:0) (3.5Gb free / 4.3 Gb total).
My laptop details:
– Intel64 Family 6 Model 158 Stepping 10, GenuineIntel.
– GeForce GTX 1650 (cuda:0) (2.2Gb free / 4.3 Gb total).
And below is the address where we can access our server. In order to open the Easy Diffusion user interface in the Internet browser, write: localhost:9000/ or 127.0.0.1:9000/. After which the program’s working window will open.
«Help & Community»
Following the Settings window is the Help & Community window.
In this window, by clicking on one of the links, you can independently study additional information about working with the Easy Diffusion program, as well as learn about installing additional models and modules (plugins).
You can also join the system user community.
From the latest link «Source code on GitHub» you can download the Easy Diffusion source codes from GitHub.
«Models tools» / «Text-to-image AI model software»
Behind the «Help & Community» window is the «Models tools» window, which has two tabs:
– «LoRA Keywords».
– «Merge models».
In the LoRA Keywords window, you can create sets of keywords for yourself to speed up the image creation process. We will dwell on this option in more detail as we work.
Using the «Merge models» window, you can combine different models with each other. For example, those that we downloaded after installation from the Internet.
If you like a couple of models that you work with most often, you can combine them into one model and give it your name. But there is one small limitation: the models must be of the same type.
From my experience I will say that there is no particular need to combine models. But, if you use two dozen different models, you may need this functionality of the Easy Diffusion application to combine according to the desired topics or styles.
«What’s new»
Following the «Models tools» window is the «What’s now» window. This window provides a list of all updates and changes that have been made to the current and previous versions of the Easy Diffusion application.
It will also be useful for us to know that our Easy Diffusion 3.0 program is 40% faster than Easy Diffusion v2.5, which has a serious impact on the speed of creating our images.
Step 5. Let’s get started
As we have already seen, the Easy Diffusion Artificial Intelligence system is a fairly simple program both to install and configure.
Once again, I would like to draw your attention to the fact that the main purpose of Easy Diffusion is to create a new unique image from a text description or from a combination of text and image based on the most popular neural network models today (which we downloaded and installed earlier).
In order for us to create our first image, we launch the program and go to its work area (the «Generate» window):
In the «Generate» window on the left you will see a toolbar that allows you to enter a text description and define all the necessary parameters for the new image you are creating:
For our first image, we will only add a description and select a model to base this image on. We will not change any parameters now.
Please also note that above the Toolbar there are three additional buttons that allow us to add a description from a text or structured file, as well as use additional image modifiers:
– «Load from a file». This button allows you to load text describing a previously created image from a text file or a JSON file. It is very useful when you are working with a large number of images. Previously, in the Settings window, we enabled the «Auto-Save Images» option and defined the «Metadata format» as the text «txt». This way our system saves all your images and descriptions in text format automatically.
• «Embedding». Configured through the «Models tools» tab of the application’s working window.
• «Image Modifiers» / «Image modifiers». A very useful button, especially at first, when you have not yet learned how to compose complex descriptions of new images. In the modifiers you can choose several different filters for your images:
Below the buttons there are two windows:
– «Enter Prompt». This is the main window, which is intended to contain a text description (we will call it briefly a «query») of the image that we want to create. We will work with him throughout our reading of the book.
• «Negative Prompt». This is an additional window in which everything that should not be displayed in the created picture is described in short phrases.
Below the request windows there are two buttons, united by one heading «Initial Image (img2img)»:
– «Browse». This button allows us to add our image, which the Easy Diffusion system uses to create a new image similar in design style and presentation form from it and the request we wrote.
– «Draw». The button opens the image editing window. In this window we can edit the image we added as a source. For example, clear the background of an image.
Below is the «Make image» button, which is very important for us, with which we will start the process of creating our new unique images.
Under the «Make image» button there is a large list of parameters for the image we want to create.
Let’s take a closer look at all the «Image Settings» functions:
– «Seed». This is the starting number of the random number generator that is used to create a unique image. By default, the program itself creates a unique number. If we know the Seed and other things being equal, we can create the same picture again. We do not change this option and leave it as it is.
– «Number of Images». This option determines the number of images that are created after clicking the «Make image» button. We leave the default value (equal to «1»). If you change the number, it will affect the overall performance of the system. We do not change this option and leave it as it is.
– «Model». This window displays a list of models that we downloaded for you earlier. Models are used to create unique images. We will look at their work later in this and subsequent chapters.
– «Clip Skip». This option allows the system to more accurately analyze and interpret our request. When the option is disabled, then the new image created matches our request as closely as possible. We do not change this option and leave it disabled.
– «ControlNet Image». This is a new and very interesting option of the Easy Diffusion Artificial Intelligence system version 3.0. With ControlNet, you and I can not only create a unique image, but we can also make it look like a celebrity. ControlNet will be covered in more detail in a separate chapter.
– «Custom VAE». This option allows us to use a special autoencoder, which helps speed up the creation process and improve image quality. The Easy Diffusion manual suggests downloading and using the file vae-ft-mse-840000-ema-pruned.ckpt. I must say that it worked great with Easy Diffusion 2.5, but does not work correctly with our version of Easy Diffusion version 3.0. We do not change this option and leave it disabled.
– «Sampler». These are auxiliary Text-to-image AI software for improving the created image. The interesting thing is that some samplers are suitable for certain types of images, while other samplers are suitable for other types of images. You will have to determine which one is better during your work.
– «Image Size». This option allows you to set the size of the resulting image. I highly recommend that you use the 512x512 size. After you select the one you like from the many images you have created, you can use the «Upscale» / «Scale» command to enlarge the image to a size of 2048x2048 and download it to your computer.
– «Inference Steps». The picture we create at the first stage consists of so-called random noise. The program, analyzing our request, reduces this noise step by step and gives us the final version of the image. This option determines the number of steps after which we obtain a sufficiently detailed image. You can try changing it by setting a value in the range from 25 to 50. Increasing the value affects system performance. The optimal ratio of speed and quality of Easy Diffusion is achieved at a value of 25.
– «Guidance Scale». This option allows us to tell Easy Diffusion to analyze our request more closely so that the final image it produces more accurately matches its description. The optimal value is 7.5. The higher the value, the less creativity Easy Diffusion has.
– «LoRA» (Low-rank adaptation). LoRA models are used to fine-tune and speed up the large models that we previously downloaded and installed and will use to create our images. There are various LoRA models that are suitable for enhancing some part of the image (for example, eyes) or style (for example, anime). By default, they are not installed on the system. You can download them from the website civitai.com by typing «LoRA» in its search bar. We do not change this option and leave it disabled (empty). However, I recommend that after training, you go to civitai.com and download the models that suit you for further experiments.
– Seamless Tiling. This option allows you to make a double image seamlessly stitched together from one of your images, that is, make a panoramic image. We do not change this option and leave it disabled (empty).
– «Output Format». This option specifies the final image format. You can choose three: jpeg, png and webp. I recommend leaving jpeg.
– «Image Quality». The default value is 75. If you want a higher quality image but need to wait a little longer to create it, set the value to 95.
We will look at the «Render Settings» functions later in the exercise. For now we leave them unchanged.
Let’s start creating our first image.
In order to create our first image from text, in the «Enter Prompt» window we need to write our first text request («Prompt»).
Let’s do the following:
– In the «Enter Prompt» window, write the following request: «Artificial Intelligence that will change the world, beautiful cyborg face.» Please note that Easy Diffusion only understands queries in English. If you are not very good at it, please use any online translator available to you.
– In the «Image Settings» section, select the «sdxlYamersRealism_version2» algorithm.
– Image Size: 512 x 512.
Leave all other parameters unchanged and click the «Make image» button.
If you open the application backend window, you will see how the Easy Diffusion system works:
Easy Diffusion may need to download additional files to create our image. After which, after a few minutes, she will begin to create a new image based on our request.
Switch to the Easy Diffusion graphical interface window.
Hooray!
You have created your first unique image using an Artificial Intelligence system. It’s located to the right of the Toolbar in the workspace of the user interface:
If you don’t get the same picture as me, don’t be alarmed and try to create it again (I didn’t succeed the first time either). Very often in Easy Diffusion you can get the desired image the second, third or fifth time. This is a feature of the program. This often happens with simple queries, as in our example. We’ll look at complex queries in the next chapter.
Our next step will be to explore the enhancement and manipulation options that most often need to be done after we create any image. To do this, hover your mouse over the new image and you will see the following commands:
– «Upscale». If you like the new image and feel there is no need to improve it further, run the Upscale command. After which the image will be enlarged to a maximum resolution of 2040 x 2048 and you can save it on your computer.
– «Fix faces». The team corrects faces in images. Most often, fingers and facial details are affected in images.
– «Draw another 25 steps.» If you think the image needs to be improved, run the «Draw another 25 steps» command. It often happens that there will be more details in the picture and its quality will become better. It also happens that no significant changes are noticeable. This requires patience and perseverance.
– «Make Similar Images.» With this command we tell the system to make five more similar versions of the images and the first received version.
– «Download Image». The command saves your image in the location you want.
– «JSON». The command saves the description and all parameters of the image you created in a marked file. You can load it the next time you want to create a similar image by clicking the «Load from a file» button.
– «Use as Input». This is a very useful command that allows you to improve the image from time to time. When using it, the Artificial Intelligence system takes as a basis not only our request, but also your image, on the basis of which a new image is created. We will look at the work of this command in more detail in the next chapter.
Now let’s take another 25 steps to improve our image by running the command «Draw another 25 steps»:
As a result, we will get a similar image, but with some changes:
Now let’s go back to our first image of «Artificial Intelligence» below.
Hover your mouse over the image and select the «Make Similar Image» command. This command will allow us to create five more new image variations from the original one.
Five new options will be placed in one program window (for clarity, I left four):
You can use this function an unlimited number of times until you get the perfect image. This is exactly what I did when choosing pictures for the book cover.
Once you select the perfect one, you can delete all other images. Hover your mouse over the image you want to delete and click the cross in the back right corner:
As a result, only those images that you need for further processing will remain in the window.
As I noted earlier, all the models that we additionally installed in Easy Diffusion, without exception, have one significant drawback – they do not draw eyes or fingers well, and very often various strange artifacts appear in the pictures. To correct the eyes, in the image we like, we need to move the mouse over the picture again and execute the «Fix Faces» command:
As an example, I’ll leave one AI face and run the Fix Faces command on that image.
It is necessary to make a small remark and clarify that the «Fix Faces» command quite often does not improve the image, and sometimes spoils it. However, it’s always worth trying out what might come out of your experiments. Be bolder and experiment more often.
In my case, the faces turned out a little blurry or blurry:
The final result (which we like best) needs to be converted to a higher resolution and saved on our hard drive in order to later use it for further work. To do this, execute the «Upscale» command by hovering the mouse cursor over the picture.
After this, execute the «Download Image» command. The downloaded image will have a resolution of 2040 x 2048 pixels per inch. This is quite enough to design a book cover or make a beautiful postcard.
Congratulations!
You’ve just created your first unique image using the Easy Diffusion Artificial Intelligence system, which can easily be called a small piece of digital art.
In subsequent chapters of the book, we will learn a lot about Artificial Intelligence and increase our level of knowledge when working with the Easy Diffusion program.
Step 6. Create a beautiful image
In this chapter, we will look at the most interesting option for working with the Easy Diffusion system – creating a new image from a combination of text and another image. We will also pay special attention to writing a more complex query.
Please note that in the «Enter Prompt» window we are not writing a literary description of the picture we want to receive, but we are writing a short request:
The Artificial Intelligence we work with, unfortunately, does not yet read our thoughts, and it is important for it that we speak to it in a language it understands. The only thing that can be considered a great achievement for developers is that the query is written as a set of key phrases (consisting of words that are easy for both humans and machines to understand), separated by a comma.
So, let’s begin.
Let’s look at a specific example.
We will create with you in a few steps an image of a beautiful cyborg based on the description from the previous chapter.
To do this, perform the following steps:
• In the «Enter Prompt» window, write the following phrase: «Artificial Intelligence that will change the world, beautiful cyborg face.»
• «Model»: «sdxlYamersRealism_version2».
• «Sampler»: «Euler Ancestral».
• Image Size: 512 x 512.
Leave all other parameters unchanged and click the «Make image» button.
Let’s add a small clarification about the color of our cyborg’s eyes in the «Enter Prompt» window: «Artificial Intelligence that will change the world, beautiful cyborg face, green eyes», and also give the Easy Diffusion command to use our first image as the source (parent image) to create a new (child image): «Use as Input».
Конец ознакомительного фрагмента.
Текст предоставлен ООО «Литрес».
Прочитайте эту книгу целиком, купив полную легальную версию (https://www.litres.ru/chitat-onlayn/?art=70541059?lfrom=390579938) на Литрес.
Безопасно оплатить книгу можно банковской картой Visa, MasterCard, Maestro, со счета мобильного телефона, с платежного терминала, в салоне МТС или Связной, через PayPal, WebMoney, Яндекс.Деньги, QIWI Кошелек, бонусными картами или другим удобным Вам способом.
notes
Примечания
1
Steampunk. [Electronic resource] // ru.wikipedia.org URL: https://ru.wikipedia.org/wiki/Стимпанк (date: 26.07.2023).
Alexander Chesalov
This book is written for middle and high school students.With the help of it and the Easy Diffusion 3.0 artificial intelligence system, you will learn how to create unique and inimitable images that you can use in your studies or for entertainment.The book contains educational information about existing types of artificial intelligence and a wonderful album of more than a hundred illustrations.
Incredible artificial intelligence Easy Diffusion 3.0
The Middle and High School Guide
Alexander Chesalov
Illustrator Vsevolod Chesalov
Illustrator Diana Chesalova
Cover designer Alexander Chesalov
Illustrator Chesalov Alexander
© Alexander Chesalov, 2024
© Vsevolod Chesalov, illustrations, 2024
© Diana Chesalova, illustrations, 2024
© Alexander Chesalov, cover design, 2024
© Chesalov Alexander, illustrations, 2024
ISBN 978-5-0062-7082-4
Created with Ridero smart publishing system
Gratitude
I thank all the people who are involved in the development of the Easy Diffusion generative artificial intelligence system. Your work helps people understand how artificial intelligence works, as well as discover new creative abilities in themselves.
Introduction
The wonderful world of Artificial Intelligence can be fully revealed to us only when you and I can see the positive results of its work, especially those created with our direct participation. These results must be understandable and explainable to everyone, and they must also be ethical, unbiased and not violate the law.
Today, Artificial Intelligence can do a lot, for example: write the text of a new poem or even a whole story, reproduce it in different voices of famous actors or musicians, analyze a large amount of numerical data and make a forecast for the future, play computer games with us or with a thousand people at once games. Perhaps the most impressive thing, in my opinion, that Artificial Intelligence can do today is create unique and incredible images. These images can be the embodiment of a three-dimensional world of a fantastic future in a computer game or be a virtual simulation of the world around us. The images can be three-dimensional or two-dimensional, and can also be made in various painting styles by famous artists from different time periods. But the most interesting thing is that these images may contain creatures or objects that do not exist in our world.
In this book, you will acquire new knowledge about how diverse Artificial Intelligence can be, and you will become familiar with its most popular terms and definitions. You will learn what is «artificial superintelligence», «narrow Artificial Intelligence», «friendly Artificial Intelligence», «human-centered Artificial Intelligence», «distributed Artificial Intelligence», «trusted Artificial Intelligence», «composite Artificial Intelligence», «explainable» Artificial Intelligence» and much more.
I will tell you how you can transform and visualize the knowledge you have acquired using an Artificial Intelligence system, turning the text definitions of various terms from this book into fantastic drawings.
In order to bring all this to life, we will study the work of a free program – the Easy Diffusion generative Artificial Intelligence system. This program creates photorealistic images from text queries and graphic images.
Easy Diffusion is an easy-to-install assembly (distribution) of the Stable Diffusion Artificial Intelligence system. Easy Diffusion will allow us to create unique images based on the most popular neural network models today, such as Stable Diffusion, Midjourney, OpenJourney, DreamShaper, SDXL Yamer’s Realism!, ProtoVision XL, Art Universe, Realistic Vision and many others.
With the help of this book, you will learn not only how to use Easy Diffusion Artificial Intelligence as a creative tool, but you will also be able to independently install, configure and uninstall this program on your computer, connect various additional modules to it, which will significantly expand its capabilities for creating and processing images.
With the help of Artificial Intelligence, you will learn to create unique and inimitable images that you can use in your studies or for entertainment.
In order to be objective in my work and make the book as accessible as possible to a wide range of readers, I asked my son Vsevolod (5th grade student) and my daughter Diana (8th grade student) to help me work on the textbook and create various images that reflect the essence of the definitions of Artificial Intelligence presented in the book. It was interesting for me to involve them in a new activity for them, and it was also very important to understand for myself how quickly they would learn to work with the Easy Diffusion Artificial Intelligence system and be able to create their own high-quality images.
It turned out that fifteen minutes of instruction was enough to begin their first acquaintance with Easy Diffusion Artificial Intelligence and become familiar with the work of the Realistic Vision model. Their first image was a steampunk style hamster. Steampunk is a science fiction movement incorporating technology and arts and crafts inspired by technology and 19th century science fiction[1 - Steampunk. [Electronic resource] // ru.wikipedia.org URL: https://ru.wikipedia.org/wiki/Стимпанк (date: 26.07.2023).]. Nevertheless, they gained more or less serious experience working with Easy Diffusion after some time.
It took us more than six weeks of work to fill this book with illustrations. From more than five hundred drawings, we have selected for you the most, in our opinion, impressive and interesting. The total computer time we spent working with the Easy Diffusion Artificial Intelligence is approximately three hundred and sixty hours. What can we say about the fact that writing the book itself, correcting and editing the text, as well as selecting the necessary pictures required a lot of time and effort.
Almost six months have passed since the first edition of this book was published. During this time, more than seventeen thousand readers became acquainted with the first edition of this book. Since then, the version of the Easy Diffusion program has also changed.
This book is written as a textbook for middle and high school students. It contains detailed instructions for working with the Easy Diffusion 3.0 Artificial Intelligence system, which allows you to transform the text you write or your image into a new unique image. It also contains educational information about existing types of Artificial Intelligence. The book features a beautiful set of over one hundred illustrations demonstrating the AI capabilities of Easy Diffusion 3.0.
We really hope that our positive and positive experience that we received in the process of studying the Easy Diffusion Artificial Intelligence system will be useful to you in your studies and creativity.
Enjoy your reading and productive work!
Artificial intelligence Easy Diffusion: first steps
Step 1. Most popular programs
Today, there are quite a large number of different programs that, based on Artificial Intelligence technologies, create incredibly beautiful images.
In my opinion, the following programs are the most popular:
– Midjourney is a very popular Artificial Intelligence system created by the company of the same name, Midjourney. Artificial Intelligence Midjourney is a web service on the Internet that allows you to create very beautiful images without using the computing resources of your computer. To use it, you must have Discord installed – a convenient messenger for group work, which is often used by players around the world. The main drawback of the program is that not so long ago Midjourney became completely paid.
Program website:
https://www.midjourney.com
– Leonardo.Ai is a colorful web service on the Internet that provides access to the Leonardo.Ai neural network of the same name. With it, you can create images from text and other images. The advantages of Leonardo.Ai include an easy-to-use interface. The disadvantage is the limitation on the number of images created per day. Images are created very quickly and with high quality, but it will take a lot of time until you learn how to make something high-quality. If you want to spend more time working with this program, be prepared to pay for its services. Very often Leonardo.Ai is compared to Midjourney, but which one is better is up to you to decide.
Program website:
https://leonardo.ai
– Dreamstudio.ai is a web service created by Stability AI, which has developed one of the most famous neural networks or, as they say, «models» called Stable Diffusion. The web service is very convenient and allows you to create any images very quickly. This is because Dreamstudio.ai uses very powerful computers provided by Stability AI. In other words, your computer is not involved in the process of creating images, which is a significant advantage of the program. At the same time, the disadvantage of using Dreamstudio.ai is the limitation on the number of images created.
Program website:
https://beta.dreamstudio.ai/dream
– StableDiffusionWeb UI is a web-based program that you can install directly on your computer. Its advantages include the fact that it is a completely free option for using the Stable Diffusion model. The program allows you to choose flexible settings to improve the quality of the image and convert it to a new version or create another image based on it. Disadvantages include some difficulties associated with installation and the need for you to have additional knowledge of the programs necessary for its operation (for example, such as Python and Git). Stable Diffusion can be recommended to all those users who have already gained extensive experience with Easy Diffusion.
Program websites:
https://github.com/AUTOMATIC1111/Stable-diffusion-webui
https://www.python.org/downloads/windows/
https://git-scm.com/download/win
https://Stablediffusionweb.com/#demo
– Stablecog is a very easy to use, but at the same time very interesting web service. Unfortunately, it’s paid. However, you can create twenty free images per day. I liked it because out of ten models you can choose two created by our compatriots. These models are called Kandinsky and Kandinsky 2.2. In addition, you can increase the resolution of the created image four times (for example, from 1248x832 to 4992x3328).
Program website:
https://stablecog.com/generate
– Easy Diffusion is a completely free Artificial Intelligence system that combines the capabilities of using Midjourney and Stable Diffusion. Easy Diffusion offers you a convenient and simple web interface that allows you not only to create very beautiful images, but also to connect various additional modules and models to its work. By default, the program already has the well-known Stable Diffusion model preinstalled. Additionally, you can download many other models from www.civitai.com, including the «Openjourney» model, which will allow you to replace the use of Midjourney. Another advantage of the program is that it is very easy to install on a computer or laptop and works with not very powerful video cards. The only drawback of the program is the time it takes you to create images. For example, if an image in “Dreamstudio.ai” is created in 20 seconds, then in Easy Diffusion it can take up to 120 seconds to create.
Once again, I would like to draw your attention to the fact that Easy Diffusion is an easy-to-install assembly of software packages (or, in other words, a distribution kit) of the Stable Diffusion Artificial Intelligence system for various operating systems. Here and further, I will not talk about Stable Diffusion, but rather Easy Diffusion due to the fact that it is with this distribution and program that we will work.
Program website (program distribution):
https://github.com/easydiffusion/easydiffusion
Step 2. Install Easy Diffusion
In order for us to install the Easy Diffusion program, follow the following link and download the required distribution for your operating system (MS Windows, Linux or MacOS):
https://github.com/easydiffusion/easydiffusion
There are three distributions on the program page:
• For MS Windows operating system.
• For the Linux operating system.
• For the MacOS operating system.
Before installing the program, please pay attention to the minimum requirements for your computer. You must have:
– Video card from NVIDIA with 2 GB of RAM (better with 4 GB). A laptop video card will do, but keep in mind that you must have the latest drivers installed not only for the video card, but also for all components of the MS Windows system.
– The minimum amount of computer RAM is 8 Gigabytes.
– The minimum amount of disk space (either on the «C:\” drive or on any other drive) is 40 Gigabytes (25 for Easy Diffusion and 15 for models). I recommend installing the program on a computer flash drive (not an external drive, but an internal drive of a computer or laptop). It will make image creation faster.
Next, we will consider installing Easy Diffusion on a computer with the MS Windows operating system.
Before installing, you need to make sure that everything is in order on your MS Windows system. To do this, you need to right-click on the «My Computer» icon and select the «Manage» command from the context menu. The Computer Management program will open, and in the Device Manager section you will see not only all the devices on your computer, but also their operating status. If there is a malfunction of any of the devices, you need to install new drivers for this device. You can find them on the Internet.
Next, download the Easy-Diffusion-Windows program.
After downloading the file named Easy-Diffusion-Windows, launch the installation and click the «Next» button:
We agree to the license agreement and click the «I Agree» button:
Click the «I Agree» button again:
Select the disk on which the program will be installed. The main recommendation is that the program should be located in the root of the disk, that is, it should be installed on the disk, but not inside other folders, but only in its own.
For example, on the drive C:\EasyDiffusion\, as shown in the image below:
Or on drive D:\EasyDiffusion\, as shown in the image below:
Select the disk you need and click the «Install» button.
The process of installing and unpacking the files on your computer has begun:
The installation process takes some time.
During this process, the program will install all the necessary files and models for work, including installing the Stable Diffusion model (version 1.5), with which you will create your first works of computer art:
The program will install other auxiliary models (for example, GFPGAN model, Real-ESRGAN model, etc.), which are used by Easy Diffusion to improve the images you create.
At the end of the process, a CLIP model will be established, which allows you to determine how well the generated image matches the query you wrote.
After installation is complete, the following window will appear:
But don’t rush to click the «Finish» button.
Be sure to check the «Create Desktop Shortcut» box and uncheck the «Run Easy Diffusion 3.0» box.
And only now click the «Finish» button.
After completing all the steps, the program will be completely installed on your computer.
Why didn’t we launch the program immediately after installation?
All this is necessary because before we start working with the Easy Diffusion generative Artificial Intelligence system, we need to download and install a few more additional model files that will allow us to create incredibly beautiful images.
What is a model?
A model is a file containing a set of parameters of a trained neural network. The file size can be small or it can be several GB in size. It depends on what a particular model is created and used for.
Each model is trained on a large number of thematic images. A set of images or data is called a «dataset».
There are a large number of different models for creating photorealistic images in various artistic styles. Some models work better for creating anime images, while others work better for imitating famous artists.
To continue working, we need to follow the following links and download the latest versions of the model files with which we will work:
– Realistic Vision версия 5.1
(https://civitai.com/models/4201/realistic-vision-v51);
– DreamShaper версия 8.0
(https://civitai.com/models/4384/dreamshaper);
– Color Fusion версия 1.0
(https://civitai.com/models/20617/color-fusion);
– Midjourney версия 4.0
(https://huggingface.co/prompthero/openjourney/blob/main/mdjrny-v4.ckpt);
– Openjourney версия 4.0 (модель обучена на рисунках Midjourney)
(https://civitai.com/models/86?modelVersionId=27392);
– SDXL Yamer’s Realism! 2.0
(https://civitai.com/models/136669/sdxl-yamers-realism-realisticanime3d).
For example, we need to download the Realistic Vision model file version 5.1.
Follow the link above to the page where there is a model file:
https://civitai.com/models/4201/realistic-vision-v51
Then we find the file of the latest version of the model for downloading:
And download the file by clicking the «Download» button:
The realisticVisionV51_v51VAE.safetensors file, 1.9 GB in size, will be downloaded to your computer.
If you do not have an electronic version of the book, you can enter the names of the model files in any search engine and find them on the Internet. For example, upon request «download model Realistic Vision» you will be offered a list of sites for downloading, including the site civitai.com, from which I downloaded most of the files I needed.
After you download all or most of the models offered to you for download, you need to transfer all the files (using the «cut» and «put» commands) to a folder along one of the paths: «C:\EasyDiffusion\models\Stable-diffusion» or «D:\EasyDiffusion\models\Stable-diffusion», depending on where you installed your program.
For this:
– Open the folder on the disk where Easy Diffusion is installed (for example, «C:\EasyDiffusion\” or «D:\EasyDiffusion\»).
– Then find and open the «models» folder.
– In it you will see a folder called «Stable-diffusion». Open it and transfer the model files you downloaded into it.
Congratulations!
You have become the proud owner of your own Easy Diffusion Artificial Intelligence system, installed on your personal computer and supplemented with new models.
Now we can proceed to the first launch.
Step 3. Launch Easy Diffusion for the first time
If immediately after installing the program you selected the «Run Easy Diffusion 3.0» command, as shown in the figure below:
A black window will appear in front of you with running lines of incomprehensible text:
You shouldn’t be afraid of this window, since this is the window where the application itself works, or, more precisely, its server part. This window will always appear when the program starts working. There is no need to close it until you decide to finish your work. Just roll it up.
You need to wait a minute or two (depending on the performance (power) of your computer), after which another window will open in the Internet browser (installed by default). This will be the same Easy Diffusion user interface (or it can also be called the client application of the program), with which we will work throughout the entire time.
Easy Diffusion’s interface is dark, but that’s fixable. In the next chapter we will learn how to make our work more comfortable using program settings.
Thus, when we first launched the program, two windows opened: a window for the server part of the program and a window for the interface.
But don’t rush to get started.
I would recommend that you close both windows and start the program again. This is primarily due to the fact that when you restart the Easy Diffusion program will automatically check for updates, download and install all the necessary new files on its own.
Wait a little longer and your default Internet browser will open again.
After which we can begin with you the next step of work – setting up comfortable and stable operation of Easy Diffusion.
Step 4. Setting up Easy Diffusion
Setting up Easy Diffusion is actually very simple.
By default, the program has options that provide a high level of comfortable work and do not require any changes. However, there are a few that require our attention.
The figure below shows the user interface of the program:
So, let’s get started.
Easy Diffusion has only five main application work window tabs, which are placed one after the other at the top of the user interface work area:
– «Generate».
– «Settings».
– «Help & Community».
– «Models tools».
– «What’s new.»
All our work will take place in the main «Generate» window, which we will devote most of our time to in the next chapter.
In this chapter you will look at the contents of the following bookmarks:
– «Settings».
– «Help & Community».
– «Models tools».
– «What’s new».
«Settings»
To get started, we first need the Settings window.
For your convenience, I will use the «Light» theme of the program:
The Settings window provides the following options:
– «Theme». Allows you to change the current theme to the one you like. I’m using the «Light» theme.
– «Auto-Save Images». Let’s activate the option. It allows us to save all our images and descriptions for them in a folder we specify.
– «Save Location». Leave the default value or change it to the path where our files will be stored. For example, C:\EasyDiffusion\models\Stable Diffusion UI.
– «Metadata format». We change to «txt» – the text data format in which our requests will be stored. They can be read using any text editor.
– «Block NSFW images.» Let’s activate the option. Prohibits working with photos of certain categories.
– «Enable Sound». Let’s activate the option. We will always receive a sound notification that our image is ready.
– «Process newest jobs first.» Determines the order in which images are processed. By default, jobs are processed in order. We don’t change.
– «Extract LoRA tags from the prompt.» LoRA stands for Low-rank adaptation and is translated as «Low-rank adaptation». LoRA speeds up the process of training models. We do not change the parameter.
– «Open browser on startup.» The option is responsible for launching the Internet browser after the server part has started. We don’t change.
– «GPU Memory Usage». The option transfers all program work from the video card to the central processor. If you select «High», the program may run very slowly or stop working altogether. We need to select «Law».
If your video card is not NVIDIA GeForce RTX, then set it to «Low».
– «Use CPU (not GPU).» The option shifts all the work of the program not to the video card, but to the central processor. If you select a central processor, the program can run very, very slowly. We don’t change.
– «Auto-Save Settings». Click the «Configure» button. All options must be selected:
To do this, change the «Enter Prompt» value.
And close the window.
– «Confirm dangerous actions.» The option is responsible for the security of the application and must be active.
– «Profile Name». Leave the default value «default».
– «Beta channel». The option provides access to not yet very tested and unstable functions that may appear in the next version of the application. We don’t change.
– «Use the new v3 engine (diffusers).» A very important function for us, which in the left panel of Text-to-image AI software increases the list of parameters for customizing the created image. Let’s activate the option.
After you have made the changes, click the «Save» button to save all the changes made.
Below are a few more options that you do not need to change:
Three options control the network settings of your Easy Diffusion application. We don’t change them.
Please note that under the settings there is system information that allows us to see the data of the CPU and video card we are using.
My computer details are:
– Intel64 Family 6 Model 60 Stepping 3, GenuineIntel.
– NVIDIA GeForce GTX 980 (cuda:0) (3.5Gb free / 4.3 Gb total).
My laptop details:
– Intel64 Family 6 Model 158 Stepping 10, GenuineIntel.
– GeForce GTX 1650 (cuda:0) (2.2Gb free / 4.3 Gb total).
And below is the address where we can access our server. In order to open the Easy Diffusion user interface in the Internet browser, write: localhost:9000/ or 127.0.0.1:9000/. After which the program’s working window will open.
«Help & Community»
Following the Settings window is the Help & Community window.
In this window, by clicking on one of the links, you can independently study additional information about working with the Easy Diffusion program, as well as learn about installing additional models and modules (plugins).
You can also join the system user community.
From the latest link «Source code on GitHub» you can download the Easy Diffusion source codes from GitHub.
«Models tools» / «Text-to-image AI model software»
Behind the «Help & Community» window is the «Models tools» window, which has two tabs:
– «LoRA Keywords».
– «Merge models».
In the LoRA Keywords window, you can create sets of keywords for yourself to speed up the image creation process. We will dwell on this option in more detail as we work.
Using the «Merge models» window, you can combine different models with each other. For example, those that we downloaded after installation from the Internet.
If you like a couple of models that you work with most often, you can combine them into one model and give it your name. But there is one small limitation: the models must be of the same type.
From my experience I will say that there is no particular need to combine models. But, if you use two dozen different models, you may need this functionality of the Easy Diffusion application to combine according to the desired topics or styles.
«What’s new»
Following the «Models tools» window is the «What’s now» window. This window provides a list of all updates and changes that have been made to the current and previous versions of the Easy Diffusion application.
It will also be useful for us to know that our Easy Diffusion 3.0 program is 40% faster than Easy Diffusion v2.5, which has a serious impact on the speed of creating our images.
Step 5. Let’s get started
As we have already seen, the Easy Diffusion Artificial Intelligence system is a fairly simple program both to install and configure.
Once again, I would like to draw your attention to the fact that the main purpose of Easy Diffusion is to create a new unique image from a text description or from a combination of text and image based on the most popular neural network models today (which we downloaded and installed earlier).
In order for us to create our first image, we launch the program and go to its work area (the «Generate» window):
In the «Generate» window on the left you will see a toolbar that allows you to enter a text description and define all the necessary parameters for the new image you are creating:
For our first image, we will only add a description and select a model to base this image on. We will not change any parameters now.
Please also note that above the Toolbar there are three additional buttons that allow us to add a description from a text or structured file, as well as use additional image modifiers:
– «Load from a file». This button allows you to load text describing a previously created image from a text file or a JSON file. It is very useful when you are working with a large number of images. Previously, in the Settings window, we enabled the «Auto-Save Images» option and defined the «Metadata format» as the text «txt». This way our system saves all your images and descriptions in text format automatically.
• «Embedding». Configured through the «Models tools» tab of the application’s working window.
• «Image Modifiers» / «Image modifiers». A very useful button, especially at first, when you have not yet learned how to compose complex descriptions of new images. In the modifiers you can choose several different filters for your images:
Below the buttons there are two windows:
– «Enter Prompt». This is the main window, which is intended to contain a text description (we will call it briefly a «query») of the image that we want to create. We will work with him throughout our reading of the book.
• «Negative Prompt». This is an additional window in which everything that should not be displayed in the created picture is described in short phrases.
Below the request windows there are two buttons, united by one heading «Initial Image (img2img)»:
– «Browse». This button allows us to add our image, which the Easy Diffusion system uses to create a new image similar in design style and presentation form from it and the request we wrote.
– «Draw». The button opens the image editing window. In this window we can edit the image we added as a source. For example, clear the background of an image.
Below is the «Make image» button, which is very important for us, with which we will start the process of creating our new unique images.
Under the «Make image» button there is a large list of parameters for the image we want to create.
Let’s take a closer look at all the «Image Settings» functions:
– «Seed». This is the starting number of the random number generator that is used to create a unique image. By default, the program itself creates a unique number. If we know the Seed and other things being equal, we can create the same picture again. We do not change this option and leave it as it is.
– «Number of Images». This option determines the number of images that are created after clicking the «Make image» button. We leave the default value (equal to «1»). If you change the number, it will affect the overall performance of the system. We do not change this option and leave it as it is.
– «Model». This window displays a list of models that we downloaded for you earlier. Models are used to create unique images. We will look at their work later in this and subsequent chapters.
– «Clip Skip». This option allows the system to more accurately analyze and interpret our request. When the option is disabled, then the new image created matches our request as closely as possible. We do not change this option and leave it disabled.
– «ControlNet Image». This is a new and very interesting option of the Easy Diffusion Artificial Intelligence system version 3.0. With ControlNet, you and I can not only create a unique image, but we can also make it look like a celebrity. ControlNet will be covered in more detail in a separate chapter.
– «Custom VAE». This option allows us to use a special autoencoder, which helps speed up the creation process and improve image quality. The Easy Diffusion manual suggests downloading and using the file vae-ft-mse-840000-ema-pruned.ckpt. I must say that it worked great with Easy Diffusion 2.5, but does not work correctly with our version of Easy Diffusion version 3.0. We do not change this option and leave it disabled.
– «Sampler». These are auxiliary Text-to-image AI software for improving the created image. The interesting thing is that some samplers are suitable for certain types of images, while other samplers are suitable for other types of images. You will have to determine which one is better during your work.
– «Image Size». This option allows you to set the size of the resulting image. I highly recommend that you use the 512x512 size. After you select the one you like from the many images you have created, you can use the «Upscale» / «Scale» command to enlarge the image to a size of 2048x2048 and download it to your computer.
– «Inference Steps». The picture we create at the first stage consists of so-called random noise. The program, analyzing our request, reduces this noise step by step and gives us the final version of the image. This option determines the number of steps after which we obtain a sufficiently detailed image. You can try changing it by setting a value in the range from 25 to 50. Increasing the value affects system performance. The optimal ratio of speed and quality of Easy Diffusion is achieved at a value of 25.
– «Guidance Scale». This option allows us to tell Easy Diffusion to analyze our request more closely so that the final image it produces more accurately matches its description. The optimal value is 7.5. The higher the value, the less creativity Easy Diffusion has.
– «LoRA» (Low-rank adaptation). LoRA models are used to fine-tune and speed up the large models that we previously downloaded and installed and will use to create our images. There are various LoRA models that are suitable for enhancing some part of the image (for example, eyes) or style (for example, anime). By default, they are not installed on the system. You can download them from the website civitai.com by typing «LoRA» in its search bar. We do not change this option and leave it disabled (empty). However, I recommend that after training, you go to civitai.com and download the models that suit you for further experiments.
– Seamless Tiling. This option allows you to make a double image seamlessly stitched together from one of your images, that is, make a panoramic image. We do not change this option and leave it disabled (empty).
– «Output Format». This option specifies the final image format. You can choose three: jpeg, png and webp. I recommend leaving jpeg.
– «Image Quality». The default value is 75. If you want a higher quality image but need to wait a little longer to create it, set the value to 95.
We will look at the «Render Settings» functions later in the exercise. For now we leave them unchanged.
Let’s start creating our first image.
In order to create our first image from text, in the «Enter Prompt» window we need to write our first text request («Prompt»).
Let’s do the following:
– In the «Enter Prompt» window, write the following request: «Artificial Intelligence that will change the world, beautiful cyborg face.» Please note that Easy Diffusion only understands queries in English. If you are not very good at it, please use any online translator available to you.
– In the «Image Settings» section, select the «sdxlYamersRealism_version2» algorithm.
– Image Size: 512 x 512.
Leave all other parameters unchanged and click the «Make image» button.
If you open the application backend window, you will see how the Easy Diffusion system works:
Easy Diffusion may need to download additional files to create our image. After which, after a few minutes, she will begin to create a new image based on our request.
Switch to the Easy Diffusion graphical interface window.
Hooray!
You have created your first unique image using an Artificial Intelligence system. It’s located to the right of the Toolbar in the workspace of the user interface:
If you don’t get the same picture as me, don’t be alarmed and try to create it again (I didn’t succeed the first time either). Very often in Easy Diffusion you can get the desired image the second, third or fifth time. This is a feature of the program. This often happens with simple queries, as in our example. We’ll look at complex queries in the next chapter.
Our next step will be to explore the enhancement and manipulation options that most often need to be done after we create any image. To do this, hover your mouse over the new image and you will see the following commands:
– «Upscale». If you like the new image and feel there is no need to improve it further, run the Upscale command. After which the image will be enlarged to a maximum resolution of 2040 x 2048 and you can save it on your computer.
– «Fix faces». The team corrects faces in images. Most often, fingers and facial details are affected in images.
– «Draw another 25 steps.» If you think the image needs to be improved, run the «Draw another 25 steps» command. It often happens that there will be more details in the picture and its quality will become better. It also happens that no significant changes are noticeable. This requires patience and perseverance.
– «Make Similar Images.» With this command we tell the system to make five more similar versions of the images and the first received version.
– «Download Image». The command saves your image in the location you want.
– «JSON». The command saves the description and all parameters of the image you created in a marked file. You can load it the next time you want to create a similar image by clicking the «Load from a file» button.
– «Use as Input». This is a very useful command that allows you to improve the image from time to time. When using it, the Artificial Intelligence system takes as a basis not only our request, but also your image, on the basis of which a new image is created. We will look at the work of this command in more detail in the next chapter.
Now let’s take another 25 steps to improve our image by running the command «Draw another 25 steps»:
As a result, we will get a similar image, but with some changes:
Now let’s go back to our first image of «Artificial Intelligence» below.
Hover your mouse over the image and select the «Make Similar Image» command. This command will allow us to create five more new image variations from the original one.
Five new options will be placed in one program window (for clarity, I left four):
You can use this function an unlimited number of times until you get the perfect image. This is exactly what I did when choosing pictures for the book cover.
Once you select the perfect one, you can delete all other images. Hover your mouse over the image you want to delete and click the cross in the back right corner:
As a result, only those images that you need for further processing will remain in the window.
As I noted earlier, all the models that we additionally installed in Easy Diffusion, without exception, have one significant drawback – they do not draw eyes or fingers well, and very often various strange artifacts appear in the pictures. To correct the eyes, in the image we like, we need to move the mouse over the picture again and execute the «Fix Faces» command:
As an example, I’ll leave one AI face and run the Fix Faces command on that image.
It is necessary to make a small remark and clarify that the «Fix Faces» command quite often does not improve the image, and sometimes spoils it. However, it’s always worth trying out what might come out of your experiments. Be bolder and experiment more often.
In my case, the faces turned out a little blurry or blurry:
The final result (which we like best) needs to be converted to a higher resolution and saved on our hard drive in order to later use it for further work. To do this, execute the «Upscale» command by hovering the mouse cursor over the picture.
After this, execute the «Download Image» command. The downloaded image will have a resolution of 2040 x 2048 pixels per inch. This is quite enough to design a book cover or make a beautiful postcard.
Congratulations!
You’ve just created your first unique image using the Easy Diffusion Artificial Intelligence system, which can easily be called a small piece of digital art.
In subsequent chapters of the book, we will learn a lot about Artificial Intelligence and increase our level of knowledge when working with the Easy Diffusion program.
Step 6. Create a beautiful image
In this chapter, we will look at the most interesting option for working with the Easy Diffusion system – creating a new image from a combination of text and another image. We will also pay special attention to writing a more complex query.
Please note that in the «Enter Prompt» window we are not writing a literary description of the picture we want to receive, but we are writing a short request:
The Artificial Intelligence we work with, unfortunately, does not yet read our thoughts, and it is important for it that we speak to it in a language it understands. The only thing that can be considered a great achievement for developers is that the query is written as a set of key phrases (consisting of words that are easy for both humans and machines to understand), separated by a comma.
So, let’s begin.
Let’s look at a specific example.
We will create with you in a few steps an image of a beautiful cyborg based on the description from the previous chapter.
To do this, perform the following steps:
• In the «Enter Prompt» window, write the following phrase: «Artificial Intelligence that will change the world, beautiful cyborg face.»
• «Model»: «sdxlYamersRealism_version2».
• «Sampler»: «Euler Ancestral».
• Image Size: 512 x 512.
Leave all other parameters unchanged and click the «Make image» button.
Let’s add a small clarification about the color of our cyborg’s eyes in the «Enter Prompt» window: «Artificial Intelligence that will change the world, beautiful cyborg face, green eyes», and also give the Easy Diffusion command to use our first image as the source (parent image) to create a new (child image): «Use as Input».
Конец ознакомительного фрагмента.
Текст предоставлен ООО «Литрес».
Прочитайте эту книгу целиком, купив полную легальную версию (https://www.litres.ru/chitat-onlayn/?art=70541059?lfrom=390579938) на Литрес.
Безопасно оплатить книгу можно банковской картой Visa, MasterCard, Maestro, со счета мобильного телефона, с платежного терминала, в салоне МТС или Связной, через PayPal, WebMoney, Яндекс.Деньги, QIWI Кошелек, бонусными картами или другим удобным Вам способом.
notes
Примечания
1
Steampunk. [Electronic resource] // ru.wikipedia.org URL: https://ru.wikipedia.org/wiki/Стимпанк (date: 26.07.2023).
Alexander Chesalov
Тип: электронная книга
Жанр: Техническая литература
Язык: на английском языке
Издательство: Издательские решения
Дата публикации: 30.10.2024
Отзывы: Пока нет Добавить отзыв
О книге: This book is written for middle and high school students.With the help of it and the Easy Diffusion 3.0 artificial intelligence system, you will learn how to create unique and inimitable images that you can use in your studies or for entertainment.The book contains educational information about existing types of artificial intelligence and a wonderful album of more than a hundred illustrations.