1000 Prompts in 1 Click - Dynamic Prompt Wildcards for Automatic 1111

Olivio Sarikas
1 May 202312:21

TLDRThis video tutorial introduces viewers to the SD Dynamic Prompts extension, which simplifies the creation of numerous prompt variations with a single click. The host, with the help of Don Bumby, guides users through installing the extension and utilizing it to generate keyword lists with AI, such as Chat GPT, and then incorporating these into prompts for AI-generated images. The video covers various methods to control and randomize outputs, including using wildcard lists, batch sizes, and combination generation. It also explores advanced techniques like using ControlNet for pose and composition control, and in-painting for targeted modifications. The host emphasizes the importance of saving prompt information with images for future reference and encourages viewers to experiment with the extension's features to discover unique image ideas.

Takeaways

  • 📚 Install the SD Dynamic Prompts extension by Don Bumby for Automatic 1111 to create dynamic prompts.
  • 🔍 Use AI, like chat GPT, to generate keyword lists and save them as text files for use in prompts.
  • 📝 Organize keywords into individual lines in a text file for clear separation and easy use in prompts.
  • 🌐 Place the text file containing keywords into the appropriate folder within the Automatic 1111 extensions directory.
  • 🔄 Restart Automatic 1111 to ensure the new keyword list is recognized and available for use.
  • 🎨 Create prompts with wildcards by using double underscore __ around the filename (without .txt) to insert randomized keywords.
  • 📈 Set a batch size to generate multiple variations of an image based on the keywords in the list.
  • 📊 Use angle brackets <> to manually input lists of keywords directly into the prompt for on-the-fly variations.
  • 🧩 Select multiple keywords from a list by specifying the number within the angle brackets and using $$ to denote selection.
  • 🔄 Utilize combination generation to create all possible combinations of keywords within brackets for extensive variations.
  • ✂️ Employ in-painting techniques in Automatic 1111 to change specific aspects of an image, like hair color, while keeping others consistent.
  • 🎉 Discover new creative ideas by using the prompt magic feature, which adds new prompt ideas based on your initial prompt.

Q & A

  • What is the name of the extension that is used to create dynamic prompts?

    -The extension used to create dynamic prompts is called SD Dynamic Prompts.

  • How can you install the SD Dynamic Prompts extension?

    -To install the SD Dynamic Prompts extension, you need to go to the extensions section in Stable Diffusion, click on 'Available', use the 'Load from' button to search for SD Dynamic Prompts, click 'Install', and wait for the installation to finish.

  • What is the purpose of using a text file to store keyword lists?

    -A text file is used to store keyword lists to organize and separate keywords for use in dynamic prompts. Each keyword should be on an individual line, allowing for easy management and modification.

  • How can you randomize keywords in a prompt using the extension?

    -To randomize keywords in a prompt, you use the double underscore syntax followed by the name of the text file (without the .txt extension) and again double underscores. This tells the extension to choose randomly from the list of keywords in the specified file.

  • What is the benefit of setting a batch size when using randomized keywords?

    -Setting a batch size allows the system to generate multiple variations of the prompt at once, each with a different random selection from the keyword list, which can save time and provide a variety of results.

  • How can you manually include a list of keywords in a prompt without using a text file?

    -You can manually include a list of keywords in a prompt by using angle brackets (< >) and separating each keyword with a straight line. This allows you to create a list directly within the prompt without the need for an external file.

  • What is the purpose of the 'combination generation' feature?

    -The 'combination generation' feature allows the system to create all possible combinations of the keywords listed within the angle brackets, providing a wide range of variations based on the keywords you've chosen.

  • How can you control the output of the image to ensure consistency in certain aspects like clothing style?

    -To control the output, you can use the 'In-Pane' feature in the extension, which allows you to mask out the area you want to replace and then generate a new image with the desired changes while keeping other aspects consistent.

  • What is the 'Prompt Magic' feature and how does it help in creating prompts?

    -The 'Prompt Magic' feature automatically adds new prompt ideas to the end of your existing prompt, based on what you have written so far. It can help in discovering new and creative ideas for image generation that you might not have thought of.

  • How can you ensure that the output image retains a specific pose or composition?

    -To ensure a specific pose or composition, you can use the ControlNet feature with the 'open pose full' option, which generates a stick figure for the body, a point mask for the face, and tracks hand and finger positions, giving you high control over the output.

  • How can you view the prompt and settings used to create an image in Automatic 1111?

    -You can view the prompt and settings used to create an image by going to the 'PNG info' tab in Automatic 1111, loading the image, and then viewing the information on the right side, which includes the prompts, settings, and model information.

Outlines

00:00

😀 Introduction to AI Prompt Variations

The video begins with an introduction to creating numerous prompt variations using AI. The presenter shares how to utilize the SD Dynamic prompts extension, developed with the help of Don Bumby, to generate keyword lists and randomize keywords for stable output. The audience is guided through the process of installing the extension on stable diffusion, using it for text-to-image and image-to-image tasks, and exploring its settings. The presenter also suggests using chat GPT to create keyword lists, saving them as text files, and placing them in specific folders within the automatic 1111 directory. The video demonstrates how to integrate these lists into prompts to generate varied outputs, such as different hair colors and cuts, by using wildcards and setting batch sizes for random selection.

05:03

🎨 Advanced Prompting Techniques

The second paragraph delves into advanced prompting techniques. It explains how to manually write lists within the prompt using angle brackets for various attributes like clothing, hair color, and glasses. The method to select multiple terms from a list is also covered, using a numerical indicator followed by double dollar signs. The presenter then discusses combination generation, which creates all possible combinations of terms within brackets, and how to control the number of generations and batches. Additionally, tips for stabilizing output results are provided, such as using open pose and control net 1.1 for higher control over the generated images. The video also touches on using in-painting to change specific attributes like hair color without affecting the rest of the image.

10:03

🌟 Exploring Prompt Magic and Output Control

The final paragraph introduces prompt magic, a feature that adds new prompt ideas based on the existing prompt. The presenter explains how to use prompt magic with different methods, such as 'I feel lucky' for loading fitting prompts from lexicon art, and 'attention grabber' for emphasizing parts of the prompt. It's highlighted that all prompt information and settings are saved with the image, allowing users to review how an image was created. The video concludes with a reminder to check the PNG info tab for saved details and a call to action for viewers to like the video and engage with the channel.

Mindmap

Keywords

💡SD Dynamic Prompts

SD Dynamic Prompts is an extension mentioned in the video that allows users to create multiple variations of prompts with just one click. It is used to enhance the functionality of the AI tool Stable Diffusion, enabling the creation of keyword lists and randomization of these keywords to stabilize the output. The extension is a key tool for generating diverse and creative outputs in the video's context.

💡Stable Diffusion

Stable Diffusion is an AI model used for generating images from textual descriptions, known as prompts. In the video, it is the platform where the SD Dynamic Prompts extension is installed and utilized. It serves as the foundation for creating various image outputs based on the dynamic prompts generated.

💡Wildcards

In the context of the video, wildcards refer to the placeholders used in prompts to represent a list of keywords. These wildcards are replaced by the AI with different keywords from the provided lists, allowing for the generation of numerous variations of an image based on a single prompt. They are crucial for the dynamic nature of the prompts created.

💡Batch Size

Batch size in the video refers to the number of images generated at one time when using dynamic prompts. It is an important parameter for controlling the output of the AI, allowing users to choose how many variations they want to generate from a single prompt.

💡ControlNet 1.1

ControlNet 1.1 is a feature within the AI tool that provides additional control over the generated images. It uses an input image to guide the generation process, ensuring that certain aspects of the output, such as pose or composition, match the input. It is used in the video to fix the output result and maintain consistency in the generated images.

💡In-Paint

In-Paint is a feature that allows users to make specific changes to generated images without affecting the rest of the image. In the video, it is used to change the hair color while keeping the clothing style and background consistent. It represents a more targeted approach to modifying the output of the AI.

💡Prompt Magic

Prompt Magic is a feature of the SD Dynamic Prompts extension that automatically adds new prompt ideas to the end of a user's existing prompt. It is designed to help users discover new and creative image concepts that they might not have thought of on their own. In the video, it is presented as a way to enhance the variety and creativity of the generated prompts.

💡Combination Generation

Combination Generation is a process described in the video where the AI creates every possible combination of the keywords listed within the brackets. This allows for a wide range of variations to be generated from a single prompt, offering a high level of diversity in the output.

💡Negative Prompts

Negative prompts are used in the context of the video to specify elements that the user does not want to include in the generated images. They are an important part of refining the output of the AI to match the user's vision more closely.

💡PNG Info Tab

The PNG Info tab is a feature within the AI tool that allows users to view and manage the metadata associated with a generated image. In the video, it is mentioned as a way to recall the prompts and settings used to create an image, which is useful for reproducing or refining similar outputs in the future.

💡Automatic 1111

Automatic 1111, as mentioned in the video, is likely a version or a specific instance of the AI tool being used. It is the environment where the SD Dynamic Prompts extension is installed and where the image generation process takes place. The name suggests a focus on automation and ease of use.

Highlights

The video demonstrates how to use AI to create keyword lists and randomize them for use in Stable Diffusion's SD Dynamic Prompts extension.

SD Dynamic Prompts is an extension that allows for automatic variations in prompts with just one click.

The extension can be installed through the Stable Diffusion extensions menu by searching for 'SD Dynamic Prompts'.

After installation, users can access the Dynamic Prompts feature in the text-to-image and image-to-image sections.

Keyword lists can be created using AI, such as Chat GPT, and saved as text files to be used in the extension.

The video provides a tutorial link for detailed instructions on how to use the SD Dynamic Prompts extension.

Wildcard lists can be used in prompts to select random elements from a predefined list of keywords.

The batch size can be adjusted to control how many variations of a prompt are generated.

The video shows how to manually write lists within a prompt using angle brackets for inline keyword variations.

Combination generation allows for the creation of all possible combinations of keywords within brackets.

ControlNet 1.1 can be used to fix the output result, ensuring consistency in pose and composition.

In-Pane editing allows for targeted changes, such as altering hair color without affecting clothing style.

Prompt Magic is a feature that adds new prompt ideas to the end of a prompt based on what has already been written.

The 'I feel lucky' button loads prompts from Lexicon Art that fit the user's desired creation.

Attention Grabber emphasizes different parts of the prompt using adjustable weights for minimum and maximum attention.

Automatic 1111 saves prompt information and settings inside the image file for easy reference.

The PNG info tab in Automatic 1111 allows users to view the prompts and settings used to create an image.

The video encourages viewers to like and subscribe for more helpful content on using AI and Stable Diffusion.