Skip to main content

Parallel Processing Node

The Parallel Processing Node processes system prompts using various AI models simultaneously, generating multiple responses with adjustable temperature settings. This powerful node enables efficient parallel execution of AI tasks, allowing you to run the same prompt across different models or configurations at the same time.

Parallel Processing Node

Basic Usage​

Use the Text, Text Input, Display Text, and Parallel Processing Node for your process.

Inputs​

The Parallel Processing Node accepts the following inputs:

  • Prompt Input: The main user prompt or content to be processed by all AI models simultaneously.
  • Overwrite System Prompt: Custom system instructions that guide the AI's behavior across all parallel processes (optional).

Outputs​

  • Output: Multiple AI-generated responses, one from each configured process, executed in parallel.

Configuration​

Process Types​

Configure multiple AI models to run simultaneously. Each process can be independently configured with:

  • Process Type Selector: Choose from available AI models:
    • GPT-4o-mini: Fast and cost-effective option for most general tasks.
    • GPT-4o: Advanced reasoning and more sophisticated responses.
    • GPT-3.5: Efficient model for simpler tasks.
    • Gemini 1.5 Pro: Google's advanced AI model.
    • Claude 3.5 Sonnet: Anthropic's balanced model for various tasks.
    • Other supported models: Additional AI models available in your LearningFlow instance.

Temperature Slider Bar​

Each parallel process has its own temperature control:

  • Lower values (0.0-0.3): More focused, deterministic, and consistent responses.
  • Medium values (0.4-0.7): Balanced creativity and coherence.
  • Higher values (0.8-1.0): More creative, diverse, and unpredictable responses.

Adding Multiple Processes​

  • Click "+ Add New" to add additional parallel processes.
  • Each process runs independently and simultaneously.
  • Configure each process with different models or temperature settings as needed.

Example Workflows​

Multi-Language Translation​

Scenario: Translate user input into three different languages simultaneously using parallel processing.

Parallel Processing Translation Example

Steps to Create the Flow:

  1. Start with the Start Node.

  2. Add and connect a Text Input for user text submission.

    • Configure the Text Input with Medium selected text box height for comfortable input.

  3. Create three Text nodes with translation instructions:

    Text Node 1 - Korean Translation:

    Translate the user's input as Korean language.

    Text Node 2 - Japanese Translation:

    Translate the user's input as Japanese language.

    Text Node 3 - Thai Translation:

    Translate the user's input as Thai language.

  4. Add and connect a Parallel Processing Node:

    i. Configure Process 1 (Korean):

    • Choose the Process Type: GPT-4o-mini
    • Connect Text Input's Output to Prompt Input
    • Connect Text Node 1 (Korean instructions) to Overwrite System Prompt
    • Set Temperature: 0.3 (for consistent translation)

    ii. Configure Process 2 (Japanese):

    • Click "+ Add New" to add a second process
    • Choose the Process Type: GPT-4o-mini
    • Connect Text Input's Output to Prompt Input
    • Connect Text Node 2 (Japanese instructions) to Overwrite System Prompt
    • Set Temperature: 0.3 (for consistent translation)

    iii. Configure Process 3 (Thai):

    • Click "+ Add New" to add a third process
    • Choose the Process Type: GPT-4o-mini
    • Connect Text Input's Output to Prompt Input
    • Connect Text Node 3 (Thai instructions) to Overwrite System Prompt
    • Set Temperature: 0.3 (for consistent translation)

  5. Add three Display Text nodes to show each translation:

    • Connect Process 1 Output to Display Text 1 (Korean translation)
    • Connect Process 2 Output to Display Text 2 (Japanese translation)
    • Connect Process 3 Output to Display Text 3 (Thai translation)

Example Input:

Hello, how are you today? I hope you're having a wonderful day!

Result: Users receive three simultaneous translations in Korean, Japanese, and Thai, all processed at the same time for maximum efficiency.

Benefits of Parallel Processing:

  • All three translations execute simultaneously, not sequentially
  • Faster overall processing time compared to running translations one after another
  • Consistent temperature settings ensure reliable translation quality
  • Easy to add or remove languages by adjusting the number of parallel processes