Turn Yourself Into Art - Fast & Easy DREAMBOOTH Collab (No GPU Required)

In this video, I'm going to show how you can use a very easy and fast dream booth collab, to train your own model. A model that will know the way you look.
Then you can use this model to generate any art and include yourself in it.
Because we are using a collab, you don't need a strong GPU, and you can basically run this on any PC.

Link to this Collab:

Birme - Cropping Tool:

How to install Automatic1111 on your PC:

The Complete Beginners Tutorial for Stable Diffusion:

💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾
Between my full-time job and my family life,
I try to find free time to create content for this channel.
You can support me and help this channel keep growing:
💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾💾

💻 This Is My Development Setup (Affiliate): 💻
============
Main Monitor:
Secondary Monitor:
Graphics Card:
CPU:
RAM:
Keyboard:
Mouse:
Headphones:
Microphone:
=============

AI-generated art
Dreambooth
Stable diffusion 1.5
Midjourney
Artificial Inteligance
#stablediffusion #aiart #art

[gustavogus]

hello friend. Please, could you elaborate more on the captions?
Did you put the same caption in all pictures, even the ones where youre far away from the camera?
The ones that goes up to your chest?
The ones where your arms appear?
The ones where it shows up to your waist?
And the ones where the legs appear?
Did you change the caption in these cases?
What did you put?

Thanks!

[doingnothingatwork]

This was a very helpful and clearly presented and explained video! Thank you so much!! (subscribed)

[artndollsnstuff]

if i'm an artist and I want my art style to be publicly available, how can it be done?

[fabriziogamboapalacios]

Great video, well explained. New sub!

[randomscandinavian]

I see that you don’t check the «use external captions» option, so as far as I know Dreambooth will not consider all of those captions you wrote but rather make up its own «mind» about what it sees. Having said that, I did always get better results before captions were added as a feature so perhaps it is best not to use them at all for faces?

Also the number of training steps will vary depending on the number of input images. I will try your settings of 1000 at 5e-6 since I also happen to have 28 images in my current set.

I didn’t notice that it said «smart crop» so thanks for that! I have misread it as simply «crop» so since my images are already very carefully composed, I will now uncheck that. 👍

[matteomorganti]

Thank you. this video really helped me. If i want to load more images in my already trained model, how may i add them to it?

[kkshooter]

i just started getting this error today.

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
torchaudio 2.0.2+cu118 requires torch==2.0.1, but you have torch 2.0.0 which is incompatible.
torchdata 0.6.1 requires torch==2.0.1, but you have torch 2.0.0 which is incompatible.
torchtext 0.15.2 requires torch==2.0.1, but you have torch 2.0.0 which is incompatible.
torchvision 0.15.2+cu118 requires torch==2.0.1, but you have torch 2.0.0 which is incompatible.

[hartiym]

why are all colab's different from those in the video? What do I need to do now change the parts

[Showbiz_CH]

I am trying to follow this tutorial. Every time I attempt to write captions for my images, Google Colab disconnects. It's frustrating because I have 110 images and I have to start over each time.

Is there a way to prevent Google Colab from disconnecting, or should I simply reduce the number of images to 20?

[WillKMB]

When you say stable diffusion folder, where do you download that? Sorry if this is a stupid question

[hartiym]

No module named 'wget'...what is this error, how can i fix it

[monstaman]

For anyone trying this, this video has several mistakes:
1. Use max 20 images for best results
2. Keep steps somewhere between 1500 and 2000

[ThemisPapaioannou]

Not working for free users anymore. Went Pro.