Regolo AI icon

Regolo AI

Use Regolo AI (OpenAI-compatible)
Join our community

Actions3

Overview

The node provides functionality to generate text embeddings using a specified embedding model. It is designed to convert short sentences or paragraphs into vector representations (embeddings) that can be used for various natural language processing tasks such as semantic search, clustering, recommendation systems, or similarity comparisons.

Typical use cases include:

  • Creating embeddings of user reviews or comments to find similar content.
  • Transforming textual data into numerical vectors for machine learning models.
  • Generating embeddings for chatbot inputs to improve intent recognition.

For example, you might input a product review sentence and receive an embedding vector that captures its semantic meaning, which can then be compared with other embeddings to find related reviews.

Properties

Name Meaning
Model The embedding model to use. Options include dynamically loaded models filtered to those supporting embeddings, plus a "Custom" option to manually specify a model ID.
Custom Model If "Custom" is selected for Model, this field allows typing a custom embedding model ID (e.g., "gte-qwen2").
Input The text string to embed. Typically a short sentence or paragraph (e.g., "The food was delicious and the waiter...").
Simplify Boolean flag indicating whether to return a simplified array of embeddings instead of the raw API response. Defaults to true.

Output

The output JSON contains the embedding data generated by the selected model for the provided input text.

  • If Simplify is enabled (default), the output will be a simplified structure containing only the data property from the API response, which holds the embeddings array.
  • If Simplify is disabled, the full raw response from the embedding API is returned.

No binary data output is involved in this operation.

Dependencies

  • Requires an API key credential for authentication with the Regolo AI service (OpenAI-compatible).
  • The base URL for API requests is configured via credentials.
  • The node dynamically fetches available embedding models from the /model/info endpoint and filters them to those supporting embeddings.

Troubleshooting

  • Common issues:

    • Invalid or missing API key credential will cause authentication failures.
    • Specifying an unsupported or incorrect embedding model ID may result in errors or empty responses.
    • Providing very long input text might exceed model limits or cause truncation.

  • Error messages:

    • Authentication errors typically indicate invalid or missing API credentials; verify your API key and credential setup.
    • Model not found or unsupported model errors suggest checking the model selection or entering a valid custom model ID.
    • Input validation errors may occur if the input text is empty or too long; ensure input meets model requirements.

Links and References

Discussion