Basalt

Interact with Basalt API to manage prompts, monitoring, and datasets
Join our community

Actions9

Overview

This node interacts with the Basalt API to create a new item (row) in a specified dataset. It is useful for automating data entry into datasets managed by Basalt, such as adding new records with specific values, names, ideal outputs, and metadata. For example, it can be used to programmatically add training data rows to a machine learning dataset or to update datasets with new information in a workflow.

Use Case Examples

  1. Adding a new row to a dataset with specific column values and metadata.
  2. Automating the creation of dataset items with ideal outputs for training AI models.

Properties

Name Meaning
Slug The slug identifier of the dataset where the item will be created.
Values A JSON object representing the values for each column in the dataset row.
Item Options Additional optional properties for the dataset row, including name, ideal output, and metadata.
Request Options Settings to control request behavior such as batching, SSL certificate validation, proxy usage, and timeout.

Output

JSON

  • id - The unique identifier of the created dataset item.
  • slug - The slug of the dataset where the item was created.
  • values - The values assigned to each column in the created dataset row.
  • name - The name of the created dataset row, if provided.
  • idealOutput - The ideal output for the created dataset row, if provided.
  • metadata - Additional metadata associated with the created dataset row.

Dependencies

  • Basalt API with an API key credential for authentication

Troubleshooting

  • Ensure the dataset slug is correct and exists in Basalt to avoid 'dataset not found' errors.
  • Validate the JSON structure of the 'Values' and 'Metadata' fields to prevent request failures due to malformed JSON.
  • Check network connectivity and proxy settings if requests time out or fail to connect.
  • If SSL certificate validation errors occur, consider enabling the 'Ignore SSL Issues' option cautiously.

Links

Discussion