marat_ai

The Best AI Avatar Tool

This service is really great, looks like canva for ai video. Great features and price

🔥 Ultimate Stable Diffusion in Google Colab

A basic version of a new and simple UI for Stable Diffusion, which works in free Colab and is optimized for phones. The ultimate version includes:Prompt translationMultiple image generationBatch image savingCustom models and samplers.youtube tutorial: https://www.youtube.com/watch?v=JWtRkXvwA6c🔗 notebook link:https://colab.research.google.com/drive/1JyCUg5BvN4NX1zc5c7r7ctJGUKsDlBsQ

Base Stable Diffusion in Google Colab

A basic version of a new and simple UI for Stable Diffusion, which works in free Colab and is optimized for phones. The ultimate version includes:Prompt translationMultiple image generationBatch image savingCustom models and samplers.link to ultimate version: https://www.patreon.com/posts/114078538youtube tutorial: https://youtu.be/JWtRkXvwA6c🔗 notebook link:https://colab.research.google.com/drive/1VUvKrrBIt3FshJyt_PtKMxtYOOPyfkR2

AI Avatar with Open-Source Tools | Ultimate Guide

AI Text-to-Video: Easy, No GPU Required

CogVideo Notebook for Lightning AI: Text-to-Video AI Generatortutorial:https://youtu.be/HqomwK8swwQnotebook:https://drive.google.com/file/d/18qgsU0j_BS6bdB-ScCjIkSKG-UQdVyzK

🔥 Run Latest AI FLUX in One Click on Colab, Kaggle, and Lightning.ai

I've made it as simple as possible for your convenience. Just create a notebook and paste the four lines of code below.P.S.Don’t forget to enable GPU runtime!!git clone https://github.com/ai-marat/flux_wui!pip install -r flux_wui/requirements.txtfrom flux_wui.main import setup_pipeline_and_widgetssetup_pipeline_and_widgets()upd1: added guidance_scale and uuid like a filename

FLUX in free tier Google colab

FLUX is the best open model for generating images. More info can be found here.You can try it out in Colab, and I’m going to create a notebook for Lightning.It doesn’t work so quickly as it used a Tesla T4, which is already quite outdated, but it’s still a nice alternative for private image generation. It works like a charm in Lightning.ai, thanks to the fast Tesla L4.https://colab.research.google.com/drive/1pIxUeslU79gpUrgPlUzvN9vWSMZYY0BO

[Early Access] The best deep fake instrument

A powerful tool for creating deepfakes, much better than the previous version (Roop), with more features and settings.This version allows to use it for NSFW. In educational purposes only.Instructions:For cloud lightning.ai users:Simply run the 'NSFW liberator' cell.For local users:Add the content_analyser.py file to facefusion/facefusion/ and substitute files.Update: The CUDA issue for the lightning.ai notebook has been fixed, and the lin...

The best deep fake instrument base version

A powerful tool for creating deepfakes, much better than the previous version (Roop), with more features and settings.update1: CUDA isn't working in lightning.ai. I'm trying to fix it; in the meantime, I've added a link to a Colab notebook.🔥 Update2 : The CUDA issue for the lightning.ai notebook has been fixed, and the link has been updated.facefusion github: https://github.com/facefusion/facefusionpinokio install: https://program....

Automatic1111 for Stable diffusion base version

Now, we have a working and stable alternative to Google Colab. To use this file, please watch the video tutorial.Notebook link: https://drive.google.com/file/d/1xlAmvIUQdWBoIoNzb2go79JqmD5Ny3jATutorial link: https://www.youtube.com/watch?v=sIXfW8r60OQ

[Early Access] Automatic1111 for Stable diffusion Ultimate version

Now, we have a working and stable alternative to Google Colab. To use this file, please watch the video tutorial.The ultimate version includes:Extensions: Control Net, Animate Diff, Reactor, 3D Open Pose, etc.15+ ControlNet modelsPlayground WebUI model + base SDXLCivitAI and Hugging-face extensions which allow downloading any model inside WebUINotebook link: https://drive.google.com/file/d/1eaN0V0GSkHP7n1dgcQHA7qLw_J6ENglLTutorial link: https://w...

AI Music: Stable Audio Colab in one click

1. Install requirements.2. Enter the prompt to generate music!Unfortunately, it's only 47 seconds long. However, this weights might be fine-tuned with your music, which is extremely interesting to test. https://colab.research.google.com/drive/1PYZ_O7fi6nndmVcK20Yw9oN0gFycy4HK?usp=sharing

The New Stable Diffusion model is out! Test it now with my new Colab notebook!

I haven't tested the model yet, but at least I made a notebook.At this moment, Cascade is very memory inefficient, but I optimized it a bit, and it works.P.S. Guys, thanks for your support.more details: https://huggingface.co/stabilityai/stable-cascade🔥 notebook link:https://colab.research.google.com/drive/1yVUnHn6Wtb3k8ay4o0DSiELiGpO2krwZ

WUI (widget based UI) Stable Diffusion in Google Colab

It's a very nice notebook for basic Stable Diffusion usage.This notebooks supports: custom LoRA, and only official SDXL with Refiner.Update: If you are experiencing issues with loading LoRA, it's because Civit.ai requires authorization for loading certain models. This requirement seems unnecessary, but to ensure proper functionality, I must implement Civit.ai authorization. The solution is simple: load your LoRA model to any other platfor...

[Early Access] Ultimate WUI (widget based UI) Stable Diffusion in Google Colab

It's a very nice notebook for basic Stable Diffusion usage.This notebooks supports: custom LoRA, custom checkpoints, and official SDXL with Refiner.This notebook probably will never be banned because this form of 'UI' doesn't violate any Google Colab rules.I could add an img2img tab with ControlNet in the future. Surprisingly, even inpainting is possible, but it might all be implemented in an ugly manner. We'll see.🔥 notebo...

Kaggle started blocking civit.ai. Here is how to bypass it 🤫

Blocking won't let you download from civit.ai, including custom checkpoints, LoRA, etc.However, I found a simple solution: at the moment, just use any link shortening service like http://tinyurl.com/. Fortunately, Kaggle doesn't detect the readdress.p.s. Keep in mind that such services are often blocked in countries that are not very free; just find one that works for you or use vpn.

MidJourney inspired UI (Focus). Advanced features in Kaggle

If you have been using Focuus in free tier Colab, you may have noticed that sometimes it doesn't work, and the session might disconnect. This issue occurs when using upscale, faceswap, and other advanced features due to shortages of RAM. I resolved this problem by creating a notebook on Kaggle, which is also free like Colab but comes with much more RAM. It functions similarly to Automatic1111 with an ngrok token. To use this, you just need to...

[Early Access] The best deep fake instrument

A powerful tool for creating deepfakes, much better than the previous version (Roop), with more features and settings. It's regularly blocked by Google, but at this moment, it works like a charm. I think I'll create a version without an interface but with the ability to work without Google bans. Sagemaker and Kaggle versions are also possible.🔥 notebook link: https://colab.research.google.com/drive/1ZHPqtu6jtvNNEip2eG10FSNasbu0cFHK

[Public] DeepFake in one click for the free tier in Google Colab

It's a simple notebook that allows you to make easy and quick face swaps, even in free Google Colab. Previous notebooks with GradioUI were banned, so this is a really nice alternative.🚨 THE NOTEBOOK HAS JUST BEEN UPDATEDPrevious problems have been resolved; they were due to issues in the new version of Colab libraries. I have fixed them, and now it works with GPU and renders videos much faster.🔥 notebook link: https://colab.research.goo...

DeepFake in one click for the free tier in Google Colab

It's a simple notebook that allows you to make easy and quick face swaps, even in free Google Colab. Previous notebooks with GradioUI were banned, so this is a really nice alternative.🔥 upd1: Now it also works with CPU (3 seconds 1920x1080 ~ 50 minutes with the best quality)🔥 notebook link: https://colab.research.google.com/drive/1ygXdO7SE-PsYnvnRwZbaDaz3nLd-KOh5🔗 tutorial link: coming soon

🔥Stable Video Diffusion🔥 Colab notebook

Now we can easily create a solid video from an image with just one click and of course FREE! That is mind-blowing!To avoid problems with the notebook, please, check the tutorial first🔗 tutorial link:https://youtu.be/k8z3Bn_Wkbg🔥 notebook link: https://colab.research.google.com/drive/1cR2WVcFqIugmvIiL67ScktaAWvwn1U3K#scrollTo=aaimSFWfLPgb

[Early Access] 🔥Stable Video Diffusion🔥 one-click install!

Stability AI recently released their new generative model, Stable Video Diffusion. I'm making a video about it, but you, my lovely subscribers, can already play around with this notebook. Enjoy!

🔥Updated🔥 MidJourney inspired UI for Stable Diffusion. Ultimate Google Drive Version

MidJourney inspired UI for Stable Diffusion. The main advantage is that it still works on the free tier in Google Colab. All data and models are saved in your Google Drive.🔗 tutorial link:https://youtu.be/teHOM5lca3c🔥 notebook link:https://colab.research.google.com/drive/1WzrHwVSU1Mc3IMhdDhRELf3MmJCBqsjH🚨 upd. The notebook updated, custom base models and LoRA have been added.

[Public] MidJourney inspired UI for Stable Diffusion

MidJourney inspired UI for Stable Diffusion. The main advantage is that it still works on the free tier in Google Colab.🔥 tutorial link:https://youtu.be/teHOM5lca3c🔗 notebook link: https://colab.research.google.com/drive/1OG-t4snVisfZoMcmRCPmgJjjtwdJCIIjThere might be some minor errors; just restart the cell or notebook in this case. Sometimes, Google indicates that the notebook is banned, but that's not the case. After a restart, every...

[Early Access]One-click super AI upscaler

Guys, I have a gem for you!I was following a project that provided super upscaling, but unfortunately it required a lot of RAM, so I was waiting for it to be optimized.But the guys released a demo on the Chinese equivalent of Huggingface, and it works flawlessly. With minimal settings, you can achieve excellent upscaling. I managed to upscale a cat with a resolution of 128x128 to 768x768!I think I can create a notebook for Kaggle, as they have re...

[Early Access] SDXL + Refiner in Google Colab

It turns out that the Stable Diffusion developers have closed free access to the ClipDrop service with the SDXl model. I used it quite often when I needed to quickly generate something.I have decided to create a Colab notebook that performs this exact task without a user interface. You simply input a prompt and a negative prompt, and Google Colab will generate an image for you without any problems or restrictions. The most important thing is that...

[Public] StableDiffusion InvokeAI Base Cloud version

This Google Colab notebook will allow you to run any Stable Diffusion models without paid subs and any disconnect issues.It can download LoRAs, ControlNet, Checkpoints.Please, watch the tutorial for Invoke UI before using it:https://youtu.be/lmpAopZ41K4Updatesupd1. Latest issue with image showing has been resolvedupd2. Latest issue with upscales has been resolvedupd3: Latest issue with 'No model' message has been resolved 🔥

[Early Access] 🔥Colab🔥 Stable Diffusion InvokeAI Ultimate Google Drive Version

If you are a newcomer, you may be overwhelmed by the vast number of notebooks and updates. You can simply use the latest version____________________DescriptionI've created a Google Colab notebook that will allow you to run Stable Diffusion without a paid subscription and any disconnection issues. As usual, this Colab notebook has practically no limitations.You can download LoRAs, Checkpoints, Embeddings, ControlNet, VAEs, and more. Despite a ...

[Public] Base notebook with Automatic1111 for SDXL, free Colab Alternative (SageMaker Lab)

Now, we have a working and stable alternative to Google Colab. To use this file, please watch the video tutorial.NB: It is not a Colab Notebook! You need to download the file.Tutorial link: https://www.youtube.com/watch?v=bUd8T0Oe_e4https://drive.google.com/file/d/1Jek76pvY6x9fcLoVYzrhcGbl_NVqP2Gz

Ultimate notebook with Automatic1111 (LoRA, ControlNet, VAE, embedding) updated

Now, we have a working and stable alternative to Google Colab. To use this file, please watch the video tutorial.Despite a free version, this notebook automatically downloads a batch of Lora and ControlNets models.NB: It is not a Colab Notebook! You need to download the file.upd: added VAE, embedinghttps://drive.google.com/file/d/1HYrOJkMsDvGdRUOTaA8ghj8g5TSVGBVU