Skip to main content

Prompt Builder Introduction

The Prompt Builder is a core tool within the Digisquares platform, designed to enable users to create, manage, and customize AI-driven prompts efficiently. These prompts can be tailored for various AI models, allowing users to automate and streamline text and image generation tasks based on the specific capabilities of each model. Whether you're working on simple text-based tasks or building complex AI-driven workflows, the Prompt Builder offers flexibility and scalability.

Upon entering the Prompt Builder, users are greeted by the Modal Gallery, which presents an organized catalog of available AI models. These models vary in capability, size, and purpose, ensuring users can find the right model for their specific needs.

Here are the key details for each model displayed in the Modal Gallery:

  • Model Name: This represents the unique identifier for the AI model, such as llama-2-70b-chat. Each model name reflects its intended purpose or architecture, helping users distinguish between models quickly.

  • Model Type: This indicates the AI model's main functionality, such as Text Generation, Image Generation, or Code Generation. Knowing the type ensures that users select a model aligned with their desired task.

  • Description: A short description explains the model’s capabilities, providing users with insight into what the model is designed to achieve. For instance, it could specify that the model is optimized for deep conversational understanding or complex text generation tasks.

  • Model Size: The size of the model (in GB) indicates its computational requirements. Larger models typically offer more advanced capabilities but may require more processing power. This information helps users gauge whether their system can handle the selected model.

  • Update Information: The Last Updated field helps users ensure they are working with the latest version of a model. Regular updates can introduce performance improvements, bug fixes, or new features.

The Model Gallery is designed to be highly interactive, allowing users to filter and sort models based on different criteria, such as model size, popularity, or update recency.

Key Features of the Prompt Builder​

Once inside the Prompt Builder, several key features help guide users through creating and configuring their prompts:

1. Search and Filter​

  • Search by Name: At the top of the Model Gallery, users can search for a specific model by typing its name. This feature is particularly helpful when the user already knows the model they want to use, making the selection process quicker and more efficient.

  • Sort by Popularity or Task Type: The Prompt Builder allows users to sort models based on predefined criteria like popularity (most-used models) or their primary task (text generation, code generation, etc.). This ensures users can focus on models that align closely with their project requirements.

2. Dynamic Configuration​

Once a model is selected, users are taken to the Prompt Configuration screen. Here, the prompt can be fully customized to meet specific project requirements:

  • Input Types: Users can define what kind of input the prompt will accept, such as Text, JSON, or File. This is crucial in shaping how the model interacts with the provided data and what it can output.

  • Output Types: After processing, the prompt can generate outputs in formats like Text, JSON, or File. This ensures that the AI model’s output is usable in a variety of contexts, from chatbots to document generation.

  • Dynamic Input: Enabling dynamic inputs allows the model to accept varied and flexible inputs. For example, users can set up prompts that take multiple forms of input or handle inputs that change based on context.

  • Function Call: Users can enable a Function Call during the prompt’s execution, allowing it to interact with external systems or perform predefined actions (such as fetching data from APIs). This is particularly useful in more complex workflows where prompts need to trigger additional processes.

Once all configurations are set, users can Save the prompt, making it ready for use or further refinement.

Practical Applications of Prompt Builder​

The Prompt Builder is designed to accommodate a wide range of use cases, offering extensive flexibility to match different project needs:

  • Text Generation: Create prompts that generate natural language text based on a variety of input types. These can be used for automated content creation, conversational agents, or summarization tasks.

  • Image Generation: For models that support image generation, prompts can be customized to generate images from text descriptions or other inputs, enabling use cases like graphic design, product visualization, or creative artwork.

  • Code Generation: With models optimized for code, users can build prompts that generate, complete, or debug code snippets. This is particularly useful for automating development tasks or providing coding assistance.

  • Data Processing: The Prompt Builder also supports prompts for structured data processing, including JSON input/output, making it possible to build workflows that integrate AI with databases, APIs, and business systems.

Best Practices for Prompt Builder​

Here are some best practices to maximize your experience with the Prompt Builder:

  • Model Selection: Always ensure that the chosen model aligns with your project’s requirements. For instance, use a model designed for text generation if your primary goal is creating written content.

  • Optimize for Performance: Larger models may offer more advanced features but come with higher computational requirements. Make sure to choose a model that your system can handle efficiently.

  • Utilize Dynamic Inputs: If your prompt needs to handle varying data types or inputs, leverage dynamic inputs to allow for flexibility and adaptability.

  • Test and Iterate: Always test your prompts before finalizing them. Depending on the complexity of the prompt, it may require multiple iterations to optimize performance and accuracy.

tip

To start creating your custom AI-driven prompts, select a model from the Model Gallery. Leverage dynamic inputs and function calls to build powerful, adaptable prompts that streamline tasks and improve workflow efficiency.