Automatic DreamBooth Installer For RunPod & Upgrade Auto1111 & Libraries To Latest Including cuDNN

[FurkanGozukara]

Automatic DreamBooth Installer For RunPod & Upgrade Auto1111 & Libraries To Latest Including cuDNN

Full tutorial: https://www.youtube.com/watch?v=MOVEcvn5h-0

With the script I shared on Patreon, you can automatically install the latest version of Automatic1111, along with the DreamBooth extension, on RunPod. Moreover, my script will automatically upgrade the cuDNN, Cuda, and Torch libraries to their latest versions, along with the Automatic1111 version.

Download Scripts ⤵️

https://www.patreon.com/posts/84716845

GitHub Readme File ⤵️

https://github.com/FurkanGozukara/Stable-Diffusion/blob/main/Tutorials/How-To-Install-DreamBooth-Extension-On-RunPod.md

Our Discord server ⤵️

https://bit.ly/SECoursesDiscord

If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on 🥰 ⤵️

https://www.patreon.com/SECourses

Technology & Science: News, Tips, Tutorials, Tricks, Best Applications, Guides, Reviews ⤵️

https://www.youtube.com/playlist?list=PL_pbwdIyffsnkay6X91BWb9rrfLATUMr3

Playlist of StableDiffusion Tutorials, Automatic1111 and Google Colab Guides, DreamBooth, Textual Inversion / Embedding, LoRA, AI Upscaling, Pix2Pix, Img2Img ⤵️

https://www.youtube.com/playlist?list=PL_pbwdIyffsmclLl0O144nQRnezKlNdx3

00:00:00 Introduction to Automatic RunPod Dreambooth extension installer

00:00:12 How to register and open your RunPod account

00:00:23 Which RunPod machine selection for DreamBooth training

00:00:35 Which template to select for Automatic1111 and DreamBooth training

00:01:09 How to download installation scripts and start installation

00:01:19 How to share your RunPod JupyterLab links with others

00:03:30 How to connect Automatic1111 Web UI from 3000 port

00:04:18 How to restart Web UI in RunPod and watch CMD

00:06:01 Speed of RTX 3090 GPU on RunPod

Video Transcription

• 00:00:00 Greetings everyone. This is a special Patreon exclusive video. 

• 00:00:03 In this video, I will show how to install  latest Auto1111, latest libraries and DreamBooth 

• 00:00:10 extension on RunPod.io pods. So let's open our RunPod account. 

• 00:00:14 You can also register from this link. Let's open it, login. 

• 00:00:20 Let's select our pod. 

• 00:00:22 I prefer community cloud and I  prefer over 20GB VRAM having GPUs. 

• 00:00:27 For example, RTX 3090, which  is very decent, only 29 cents. 

• 00:00:32 Select your template from here. Make sure that you have  

• 00:00:35 selected Web Automatic 8.0.3. It can be higher version when you are  

• 00:00:40 watching this, but it still should work. You can also customize your RunPod. 

• 00:00:46 For example, let's make it  110GB, set override, continue. 

• 00:00:49 You can also use advanced and bidding, but  I don't suggest that for training because 

• 00:00:55 it can just turn it off if  someone else overbids you. 

• 00:00:59 So let's wait until the  Connect button appears here. 

• 00:01:03 OK, Connect button appeared. Let's connect to our JupyterLab. 

• 00:01:08 The installation is so easy. Just download these two scripts  

• 00:01:12 by clicking on them. OK. 

• 00:01:14 Not loaded yet. Let's refresh. 

• 00:01:16 By the way, if you want to share your RunPod  with someone else, right click here, copy 

• 00:01:20 link and share that link. With that way someone else can also connect your JupyterLab. 

• 00:01:25 OK, still not started. Sometimes it is taking time. 

• 00:01:30 Sometimes some machines are also being broken  and you are being need to start a new machine. 

• 00:01:36 OK, it is started. The first thing we need to do is  

• 00:01:40 changing python relauncher.py. So I prepared this GitHub readme file  

• 00:01:45 because it is easier here to copy. So I copied this. 

• 00:01:50 Got your JupyterLab, got your python relauncher.py  here, you see, just change this part, save 

• 00:01:59 and restart your pod. This is only one time. 

• 00:02:01 You don't need to do this again and again. Let's just wait until it is getting restarted. 

• 00:02:08 You can refresh this page and  see if it is restarted or not. 

• 00:02:12 Not still yet. OK, let's try to connect again. 

• 00:02:15 OK, restart has been completed. Go to your workspace folder. 

• 00:02:21 I am inside workspace. Upload your files. 

• 00:02:24 Go to download, select the files downloaded,  upload them, open a new terminal in here inside 

• 00:02:31 workspace. You see currently I am inside workspace. 

• 00:02:34 Copy the first command. Paste and hit enter. 

• 00:02:38 Be patient until the Auto1111 restarts. 

• 00:02:44 This is the script. It deletes the venv. 

• 00:02:47 It downloads the latest files. It install the latest CUDA DLL. 

• 00:02:51 It is also downloading the best  VAE file automatically for you. 

• 00:02:55 You see, it is automatically installing  the latest cuDNN files. 

• 00:03:00 This is significantly improving the speed. When you see this running on local URL,  

• 00:03:06 it means that the first part of installation has been completed. 

• 00:03:09 Let's open a new terminal. Let's copy the second command. 

• 00:03:14 Paste it. Hit enter. 

• 00:03:16 Then it will install the  second part of the script. 

• 00:03:20 And it will automatically  launch the newest web UI. 

• 00:03:26 You see, it is installing  everything automatically. 

• 00:03:28 You don't need to do anything. The installation is completed  

• 00:03:31 and the web UI started. Let's connect with 3000 port. 

• 00:03:36 OK, it is started. You see the DreamBooth extension is here. 

• 00:03:41 Moreover, we have now Torch 2.0.1 version  with CUDA 11.8 and xFormers 0.0.20. 

• 00:03:50 Let's check out the speed. Moreover, I also have installed  

• 00:03:55 automatically this VAE file. So let's also apply that. 

• 00:03:59 OK, let's go to the speed fast car. Let's make it like that. 

• 00:04:05 And let's see the speed. You see we are reaching over 19 IT per second  

• 00:04:11 with RTX 3090 GPU. This is amazing speed for this GPU. 

• 00:04:18 Then let's say you wanted to  restart your web UI for any reason. 

• 00:04:23 You could be have turned off your Pod. You want to refresh your web UI. 

• 00:04:28 So copy this command. Open a new terminal anywhere here. 

• 00:04:34 Copy paste it. And it will kill the previous  

• 00:04:37 instance and it will restart your web UI. You see it has killed the previous instance  

• 00:04:42 and it is restarting again. I will also demonstrate after  

• 00:04:46 turning off Pod and turning it on. So the web UI instance restarted. 

• 00:04:51 Let me also show you. I will stop this Pod. 

• 00:04:54 By the way, if you delete the pod,  then it will get permanently deleted. 

• 00:04:58 Otherwise, it won't be. I deleted this pod. 

• 00:05:01 Then stop pod. Start it again. 

• 00:05:04 So after you stop your pod and start it again  or after you restart it, it will automatically 

• 00:05:09 start a new web UI instance, but you won't  be able to see its command line arguments. 

• 00:05:16 Its command line interface. So what we are going to do is we will kill  

• 00:05:20 the initial started web UI instance and we will start it again. 

• 00:05:24 I will now demonstrate it after  the connect button appears. 

• 00:05:27 OK, connect button appeared. Let's connect to JupyterLab. 

• 00:05:30 You see, I already have an instance of web UI. It is started, but I won't see what is happening. 

• 00:05:37 So I open a new terminal, copy  the command here, paste it here. 

• 00:05:43 Then it will start a new instance and  I will be able to see everything here. 

• 00:05:48 OK, it is getting started. OK, it is started. 

• 00:05:51 Let's refresh. Now we will see what is happening here. 

• 00:05:54 This is very useful when you  are doing training, especially. 

• 00:05:58 Fast car generate. 

• 00:06:03 And it has started and I am  seeing whatever is happening. 

• 00:06:06 And you see, it is almost  reaching 20 it per second. 

• 00:06:10 This is very, very good 19.5 it per second. This is all for today. 

• 00:06:17 I hope you have enjoyed. 

• 00:06:18 Please keep supporting me and  ask me any questions you want. 

• 00:06:21 You can also leave a comment here  and hopefully I will reply you back.