Hello, nice to meet you. Today, I will explain what is Lora in Stable Diffusion. Lora is one of the essential elements in Stable Diffusion. And you can learn our shadow and use it as Lora. Lora can be applied in a variety of checkpoints in an infinite number of ways. You can mix animation-like Lora with a real-life checkpoint. You can mix Lora with animation-like Lora with a real-life checkpoint. You can create a completely new image. I've already explained the concept of Lora in a video before. I've already explained the basic explanation of Stable Diffusion in a video. I've already explained the basic explanation of Stable Diffusion in a video. I think it would be good to refer to that video. Lora has a wide variety of uses. Today, I'm going to explain where to download Lora and how to apply it. Today, I'm going to explain where to download Lora and how to apply it. If you're new to Stable Diffusion, you might not be familiar with the concept of Lora. But if you learn this step by step with me, you can apply it easily. If you follow my explanation, you'll be able to create an amazing and fun image. Downloading Lora is the same as downloading other models. First, let's move to CBTie. I'm going to move to CBTie and apply this Lora. It's a fun Lora. It's a Lora that teaches you how to smile in animation. I'm going to apply this smile in real life. I'm going to check if this smile works well in animation. I'll explain how to download this Lora and apply it. It's the same way to download it. You can click the download button on the top or click the file button on the left. So when you click it, it's usually just a download. But CBTie has been posting a lot of these things lately. If you have an ID, you can download it. But if you don't, you can download it only if you have an ID. So if you see a message saying, Log in, I recommend you sign up and download it. I downloaded it like this. A file called FaceMemeSafeTensor has been downloaded. So where do I put this file? It's simple. If you put it in the folder I'm moving to, you can apply it to other Loras as well. Now, I'm going to cut this file and go into the Stable Diffusion Web UI folder. Go in and there's a folder called Models. Come into the Models folder. Now, there are several more. You can come into the Lora folder here. That's it. If you paste it here, it's over. Now we're ready to use Lora. Then let's go back to Stable Diffusion. Stable Diffusion Now, I'm back in Stable Diffusion. It's simple to apply Lora here. I'm explaining it in 1.5.2 version. If you're using 1.6.0 version, you can downgrade it for a while and apply it to my version. Actually, it doesn't matter much, so I think you should try it. The reason I don't use the latest version is because I keep telling you, There's a lot of expansion error, so I prefer the stable 1.5.2 version. I've told you several times in advance. Now, let's apply Lora. Now, I have to enter Lora in the positive prompt here. What should I do? Now, it's just like a boy is smiling. The way to apply it here is to click on the shape of the flower on the right. Please click once. Now, here's the texture inversion, hyper network, checkpoint, Lora. After clicking the Lora tab, If this Lora doesn't appear yet, If it doesn't appear, press the refresh button. Then Lora will probably be applied like this. So, click on the mouse cursor of this positive prompt. If you click on the Lora in this Lora tab, it will be applied like this. Now, there's no comma next to Smile Smile, so press the comma once. If you write it like this, Lora is applied. It's really easy, right? We don't have to open this bracket and enter Lora. You can apply Lora like this with a click. Then I'll have to make it. I'll take the shape of the flower on the right and create an image of how it's made. A boy is smiling. If you look at this Lora in CBTie, it's a little funny. You need to create an image that makes you laugh with a comic feeling. You can see that this Lora is well applied. I'll make it once. Now, I've made it, and I can see that Lora is well applied. Of course, the resolution is not perfect right now, but I won't explain it. It can get too long and complicated, so I'll just look at the Lora applied. He's smiling. If you compare it with the smiling face of Lora, which I checked in CBTie, You can see that a similar feeling is applied. In this way, Lora mixes a little bit of the checkpoint I'm using right now. I think it would be easy to think that way. Now, then, no, just because you're lucky, you can make this face. You might say that. Then I'll remove this applied Lora and make it. The way to remove it is simple. You can just erase this Lora. I'll erase it and create it. I'll keep the rest of the prompts the same. If you create it, it's completely different from the image that was created a while ago. You can see a smiling boy with only the checkpoint applied. So depending on what Lora you use, the image may change in a thousand ways. But when you use these Loras, sometimes Lora doesn't work well. What you need to try in that case is called a trigger word. What is a trigger word? Now, when you download Lora, If you look to the right, there's a tab called Trigger Word in CBTie. So I'm applying this Lora. If this Lora doesn't work properly, Put these words in this positive prompt. It's like this. So if I just applied this Lora, It's good, but when it doesn't work well, Try copying and pasting these words. It's like this. So when there's a trigger word, There are Loras that apply Lora a little better. So if you apply Lora while applying this, This checkpoint and this Lora don't have a big collision. It looks like it's good to use a trigger word and it's good to not use a trigger word. But later on, if you mix up a lot of checkpoints and Loras, Sometimes it doesn't work well. I recommend you to check and use it to see if the trigger word is missing. The last thing I'm going to explain is that I applied Lora like this. But it's still too weak. There are a lot of cases where it doesn't work properly. In that case, I still want this Lora to be a little stronger. If you want to, you can apply a positive prompt. I'll explain this later. You can give a positive prompt to a word. You can also give a positive prompt to Lora. It usually gives a value between 1 and 2. You don't have to enter the number here. If you look here, you can put the cursor next to the number 1. Press Ctrl and raise the direction key up. Then you can see that the number goes up automatically. On the contrary, if you press Ctrl and lower the direction key, You can see that the value is lowered. This is a value. I want this Lora to be more developed. You can put this in a positive prompt and ask AI. So the higher the value, the higher the probability of Lora being developed. But if you raise the number like this, everything will be solved. You can think of it now. But if the value is too high, it will break. Because the probability of the image breaking is higher, I think it's better to give a value of 1.5 or 1.6. Then I'll give you a value of 1.6 and create an image again with the same conditions. If you create an image, you can see that Lora has been applied more. When you look at the overall image, you feel a little broken. So you have to touch it several times and realize the feeling you want. At the same time, you have to find a combination of how to make the image less broken. The explanation of Lora is not over today. I'll explain the deepened content next time. I'll introduce you again. That's it for today's video. I'll see you next time with better and more informative content. Thank you.
