Blab Embed
Generate embeddings using Upstage Solar embedding models. Supports up to 100 strings per request with max 204,800 total tokens. Each text should be under 4000 tokens (optimal: under 512 tokens).
Overview
This node generates text embeddings using Upstage Solar embedding models, optimized for either search queries or document passages. It supports processing either a single text input or multiple texts at once, making it useful for applications like semantic search, text similarity, or natural language understanding tasks. For example, it can convert user queries or document contents into vector representations for downstream machine learning or search indexing.
Use Case Examples
- Generating an embedding vector for a single user query to improve search relevance.
- Processing multiple document passages to create embeddings for semantic document retrieval.
Properties
| Name | Meaning |
|---|---|
| Model | The Solar embedding model to use, optimized either for search queries/questions or for documents/passages. |
| Input Type | Determines whether to process a single text input or multiple texts at once. |
| Text | The single text string to generate embeddings for, used when Input Type is 'Single Text' and no text field is specified. |
| Texts | Multiple texts to generate embeddings for, one per line, used when Input Type is 'Array of Texts'. |
| Text Field | Field name from input data containing the text to embed; if empty, uses the 'Text' parameter instead. |
Output
JSON
text- The input text string for which the embedding was generated (only for single text input).embedding- The generated embedding vector for the input text.model- The embedding model used for generating the embeddings.usage- Information about token usage or API usage statistics.dimension- The dimensionality of the embedding vector (only for single text input).full_response- The full API response from the embedding service.embeddings- Array of objects containing text, embedding vector, and index for each input text (only for multiple texts input).
Dependencies
- Upstage Solar embedding API accessed via HTTP with authentication using an API key credential named 'blabApi'.
Troubleshooting
- Common issues include providing no input text or empty input, which results in an error 'No input text provided'. Ensure that the input text or texts are correctly specified either via the 'Text' parameter, 'Texts' parameter, or the specified 'Text Field' in the input data.
- API request failures due to invalid or missing API credentials can cause errors. Verify that the 'blabApi' credential is correctly configured.
- Input texts exceeding the token limits (max 100 texts, total 204,800 tokens, each text max 4000 tokens) may cause errors or unexpected behavior. Ensure inputs respect these limits.