TDApiClient.API_Request | OpenAI or Azure OpenAI teradataml Extension | Teradata - TDApiClient.API_Request - Teradata Vantage

TDApiClient.API_Request is a static helper function to invoke the in-database function API_Request.

This function returns a teradataml DataFrame.

Required Arguments:

• dataframe: Specifies an input teradataml DataFrame which acts as the input query for the API_Request in-database function.
• See list of additional required arguments and corresponding parameters in the in-database function API_Request in the following table.

Optional Arguments:

• options: Specifies key-value arguments to be passed to the in-database function API_Request. See list of these key-value arguments in the following table.

This table shows the mapping between TDApiClient.API_Request arguments to the corresponding parameters in the in-database function API_Request. For details of these parameters, see API_Request In-database Function Syntax Elements.

Example Prerequisites

To run the following examples, load necessary package first.

from tdapiclient import TDApiClient
from teradataml import DataFrame

Example 1: Use TDApiClient.API_Request with OpenAI

auth_info_fmt_str = ('{{ "key": "{}" }}')
auth_info = auth_info_fmt_str.format("open-ai-key")

embeddings_df = TDApiClient.API_Request(DataFrame("product_reviews"),
                                        "open-ai-embedding",
                                        authorization=auth_info,
                                        num_embeddings='1576',
                                        model_name='text-embedding-ada-002',
                                        text_column="comment")

Example 2: Use TDApiClient.API_Request with Azure OpenAI using endpoint URL

auth_info_fmt_str = ('{{ "Key": "{}" }}')
auth_info = auth_info_fmt_str.format("az-ai-key")

az_embeddings_df = TDApiClient.API_Request(DataFrame("product_reviews"),
                                           "az-ai-embedding",
                                           authorization=auth_info,
                                           endpoint='https://test-azure-open-ai-instance.openai.azure.com/openai/deployments/embedding-ada/embeddings?api-version=2023-05-15',
                                           num_embeddings='1576',
                                           model_name='text-embedding-ada-002',
                                           text_column="comment")

Example 3: Use TDApiClient.API_Request with Azure OpenAI using resource and deployment parameters

auth_info_fmt_str = ('{{ "Key": "{}", "Resource" : "{}", "Deployment" : "{}" }}')
auth_info = auth_info_fmt_str.format("az-ai-key", "az_resource_name", "az_deployment_name")

az_embeddings_df = TDApiClient.API_Request(DataFrame("product_reviews"),
                                           "az-ai-embedding",
                                           authorization=auth_info,
                                           num_embeddings='1576',
                                           model_name='text-embedding-ada-002',
                                           text_column="comment")