How to install Stable Diffusion WebUI Colab Alternative (free)

29 Sept 202308:12

TLDRThis video provides a step-by-step guide on how to install and use a Stable Diffusion WebUI alternative for free, with a 4-hour daily quota. The process is simpler and more convenient than using Google Colab. Users are guided through signing up for an account on ngroc, accessing SageMaker Studio, and setting up a GPU environment. The tutorial includes instructions for installing necessary requirements, downloading a base model, and running Stable Diffusion using an ngroc token. The video emphasizes the convenience of Amazon storage for saving files and models, and offers tips for managing storage and cleaning up the environment. The presenter also mentions an ultimate version of the notebook for downloading additional models and extensions, available on Patreon, and compares the convenience of this method to Google Colab.


  • ๐ŸŒ Access any stable diffusion model with a user-friendly interface, free of charge, with a daily 4-hour quota for GPU usage.
  • ๐Ÿ“š Follow the ngroc website to sign up or log in, and use your Google account for convenience.
  • ๐Ÿ“ง Create a new account on SageMaker Studio, which might require an email request and verification, sometimes taking up to a day.
  • ๐Ÿ“ฑ To utilize GPU on Kaggle, provide your phone number for verification.
  • โš™๏ธ Use CPU hours for initial setup to save GPU time, and choose GPU runtime when ready to use the model.
  • ๐Ÿš€ A ready-to-use notebook is provided, which includes steps for requirements, downloading the base model, and running stable diffusion.
  • ๐Ÿ’พ Files are saved in Amazon storage, so there's no need to reinstall requirements or redownload models each time.
  • ๐Ÿ”„ The process includes downloading the model (e.g., RV5), which is restricted by Amazon SageMaker's available storage.
  • ๐Ÿ”— After setting up, you can generate images using the ngroc token and the provided URL in the Automatic 1111 interface.
  • ๐Ÿงน Maintenance options are available to clean up virtual environments, remove cache, and check storage, with the ability to delete models to free space.
  • โน Remember to stop the runtime when finished to avoid unnecessary usage, and consider supporting the video for further assistance and updates.
  • ๐Ÿ” For those unfamiliar with coding, there's an ultimate version of the notebook available on Patreon, which simplifies downloading additional models.

Q & A

  • What is the daily quota for using the stable diffusion model with the latest user interface?

    -The daily quota for using the stable diffusion model is 4 hours per day.

  • How can one sign up for an account on the ngroc website?

    -You can sign up for an account on the ngroc website using your Google account by pressing 'sign up' and then 'create account'.

  • What is the process to get access to GPU on SageMaker Studio?

    -To get access to GPU on SageMaker Studio, you need to create a new account, verify your email, and specify your phone number.

  • How long does it take to get access to GPU if there's no availability?

    -If there's no immediate availability of GPU on SageMaker, you might need to wait, which could take about 20 minutes.

  • What is the advantage of using CPU hours for tuning the notebook in SageMaker?

    -Using CPU hours for tuning the notebook is advantageous because it can be done easily without consuming the valuable GPU time.

  • How many hours of CPU and GPU runtime are available in SageMaker Studio?

    -In SageMaker Studio, there are eight hours available for CPU and four hours for GPU runtime.

  • What is the benefit of using Amazon storage for saving files?

    -The benefit of using Amazon storage is that your files are constantly saved, eliminating the need to reinstall requirements or re-download models every time.

  • How long does it take to install all the requirements for the stable diffusion model?

    -It takes about three minutes to install all the required dependencies for the stable diffusion model.

  • What is the limitation regarding the storage of models on Amazon SageMaker?

    -There is a heavy restriction on the available storage on Amazon SageMaker, so users need to be cautious about how they use their storage space.

  • How can one download additional models like LoRA and controlnet models?

    -Users can download LoRA and controlnet models by themselves or use the creator's ultimate version of the notebook available on his Patreon page, which has the ability to download these models.

  • What should one do to clean up space for new models on Amazon SageMaker?

    -To clean up space, one can navigate to the 'models' folder in Amazon SageMaker and delete the stable diffusion or other models that are no longer needed.

  • What is the final step one should take after finishing work on SageMaker Studio?

    -The final step is to stop the runtime by pressing the designated button to ensure no unnecessary usage of resources.



๐Ÿš€ Accessing Stable Diffusion Models with Amazon SageMaker Studio

The video provides a step-by-step guide on how to access stable diffusion models using Amazon SageMaker Studio. It starts with signing up or logging into the ngroc website and then moving on to SageMaker Studio app. The process includes creating a new account, verifying email, and specifying a phone number for GPU access. The video emphasizes the importance of careful operation to avoid destroying the notebook or environment. It also mentions the availability of CPU and GPU hours, and the convenience of using a pre-prepared notebook for GPU runtime. The viewer is guided on how to upload files, hide interface bars for clarity, and run through the notebook's steps, which include installing requirements, downloading the base model, and running Stable Diffusion. The video also advises on the efficient use of Amazon storage and concludes with a reminder to stop the runtime when finished.


๐ŸŒŸ Using ngroc for Stable Diffusion and Supporting the Video

This paragraph explains how to use an ngroc auth token to connect and run Stable Diffusion, leading to a faster image generation process. The video demonstrates obtaining the token from the ngroc account and pasting it into the notebook to establish a connection. It then shows the URL for the latest Automatic 1111 user interface with the RV5 model. The presenter requests viewer support by watching the video to the end and possibly multiple times for promotion. The video highlights the importance of cleaning up virtual environments and checking available storage, and it provides instructions on how to do so within the notebook. It also touches on the limitations of storage on Amazon SageMaker and advises on deleting models to free up space. The video concludes with information on downloading additional models, using an ultimate version of the notebook available on Patreon, and the availability of extensions and UI versions for different platforms. The presenter thanks the viewers for their support and emphasizes the convenience of the method over Google Colab.



๐Ÿ’กStable Diffusion

Stable Diffusion refers to a type of machine learning model used for generating images from textual descriptions. It is a part of the broader field of artificial intelligence known as generative models. In the context of the video, it is the core technology that enables users to create images using textual prompts without the need for manual drawing or design skills. The video provides a tutorial on how to access and use a Stable Diffusion model through an alternative platform to Google Colab.


WebUI stands for Web User Interface, which is the graphical interface that users interact with through a web browser. In the video, the WebUI is mentioned in relation to the Stable Diffusion model, indicating that there is a user-friendly interface available for accessing and using the Stable Diffusion technology over the web.


Google Colab, often referred to simply as 'Colab,' is a cloud-based platform offered by Google that allows users to write and execute Python code in a collaborative environment. It is popular among data scientists and machine learning practitioners. The video discusses an alternative to Google Colab for accessing Stable Diffusion models, suggesting that the alternative is more convenient and offers a daily usage quota.


ngrok is a tool that creates a secure tunnel from a public DNS name to a local machine. It is often used to expose local development servers to the internet. In the video, ngrok is mentioned as a part of the process to set up the Stable Diffusion WebUI, where users are instructed to sign up for an account to obtain an auth token necessary for the connection.

๐Ÿ’กSageMaker Studio

Amazon SageMaker Studio is an integrated development environment (IDE) provided by Amazon Web Services (AWS) for building, training, and deploying machine learning models. The video script describes the process of signing up for and using SageMaker Studio as an alternative platform to run Stable Diffusion models with GPU support.


GPU stands for Graphics Processing Unit, which is a specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display device. In the context of the video, the GPU is essential for the high-performance computation required by Stable Diffusion models to generate images quickly.

๐Ÿ’กVirtual Environment

A virtual environment in computing is an isolated working copy of a set of software tools, applications, or data. In the video, the speaker cautions viewers about the potential to accidentally destroy their notebook or virtual environment by running incorrect code cells, emphasizing the importance of careful operation within the provided notebook interface.


In the context of software development, 'requirements' refer to the specific needs that a system or solution must meet. The video mentions installing 'requirements' as the first step in setting up the Stable Diffusion environment, indicating that certain software packages and dependencies must be installed before the model can be used.

๐Ÿ’กModel Download

The process of downloading a machine learning model, such as a Stable Diffusion model, involves obtaining the pre-trained files that the model uses to generate outputs. The video script outlines the steps for downloading a base model onto the Amazon SageMaker platform, which is a prerequisite for running Stable Diffusion.

๐Ÿ’กStorage Restrictions

Storage restrictions refer to the limitations placed on the amount of data that can be stored on a given platform or system. The video discusses the limited storage available on Amazon SageMaker and advises users to manage their storage carefully, especially when downloading and managing multiple models.


In computing, 'runtime' refers to the period during which a program or process is executing. The video mentions starting the runtime to initiate the use of the GPU for the Stable Diffusion model. It also talks about stopping the runtime when the work is completed to conserve resources and avoid unnecessary charges.


Patreon is a membership platform that allows creators to receive financial support from their fans or patrons. In the video, the speaker mentions that an 'ultimate version' of their notebook and additional extensions for the Stable Diffusion WebUI are available on their Patreon page, indicating that there is a tiered system where supporters gain access to more advanced features or tools.


Access any stable diffusion model with the latest user interface for free, with a 4 hours per day quota.

The process is easier and more convenient than using Google Colab.

Sign up or log in to the ngroc website using a Google account.

Create a new account on SageMaker Studio or request access if you don't have one.

You may need to wait for a day to get the link for creating a new account on SageMaker.

Verify your email and specify your phone number to access GPU on Kaggle.

Watch the video carefully to avoid destroying your notebook or environment.

You have eight hours for CPU and four hours for GPU on SageMaker Studio.

Use CPU hours for tuning your notebook to save valuable GPU time.

A ready-to-use notebook is provided for easy setup.

SageMaker may sometimes not have available GPUs, requiring a short wait.

Once you have access to GPU, you can open the product and start using Jupyter Lab.

Files are constantly saved in your Amazon storage, eliminating the need to reinstall requirements or download models every time.

Install all necessary requirements in the first step of the notebook.

Downloading the base model is the second step, with a limit of up to five models due to storage restrictions.

Run Stable Diffusion in the third step by specifying the ngroc token and using the xformer for faster image generation.

The ngroc token can be found in your ngroc account under 'get started'.

After running the Stable Diffusion cell, you can access the latest Automatic 1111 user interface with the RV5 model.

Support the video by watching it until the end and potentially re-watching for better promotion.

Clean up virtual environments, cache, and files in case of errors using the maintenance cell.

Check available storage and manage your folders and models carefully due to storage restrictions.

Stop your runtime when finished to conserve resources.

Download LoRA and controlnet models on your own or use the creator's ultimate notebook version with these capabilities on Patreon.

SwarmUI and ConfUI versions for Amazon SageMaker Studio Lab and Kaggle Automatic1111 are available on Patreon.

This method is considered more convenient than Google Colab as it doesn't require downloading the model or installing requirements repeatedly.