These properties are used to configure tOpenAIClient running in the Standard Job framework.
The Standard tOpenAIClient component belongs to the OpenAI family.
Basic settings
Property type |
Either Built-in or
Repository
.
|
Schema and Edit Schema | A schema is a row description, it defines the number of fields to be
processed and passed on to the next component. The schema is either
Built-in or stored remotely in the
Repository.
Click Edit schema to make changes to the schema. If the current schema is of the Repository type, three options are available:
|
Token | To enter the raw access token, click the [...] button
next to the corresponding field to bring out the Enter a new
password dialog box, enter the token in the text field, and click
OK. Note that you can enter either a JSON string in the pure password mode or a Java expression in the Java mode. You can toggle between the two modes by clicking switch to Java mode or switch to pure password mode on the top of the text field. |
Model | To select one of the available OpenAI models, click the [...] button next to the field. In the dialog box displayed, select the model that will be used or select the Use custom value check box and specify the model name in the Custom value field. |
Prompt | Enter your instructions in this field. For more information on prompt
writing best practices, read Prompt engineering. Example: Given the following addresses, guess the corresponding country names. |
Advanced settings
tStatCatcher Statistics |
Select this check box to gather the Job processing metadata at a Job level and at each component level. |
Base URL | Type in the base URL address of the OpenAI API server you want to access, by default it is https://api.openai.com. |
Endpoint | Enter the second part of the URL in order to complete the base URL filled in the Base URL field. Base URL and Endpoint values are joined with the character / when needed. |
Use raw JSON |
Select this option to manually enter the raw JSON body. |
Response body format | Select the response body format from the drop-down list. The right format
allows the connector to parse and apply operations to the response. Currently,
text, and JSON formats are supported.
|
Extract a sub-part of the response | Enter the path of a node to select a sub-element of the response. If the
element is an array, each element of the array will be looped over. For more
information about the syntax for entering the node name, see Data Shaping Selector Language syntax.
This field is optional and must remain empty to retrieve the entire JSON response. This field is only available if you select JSON from the Response body format drop-down list. Information noteNote: This field impacts the schema
of the output flow of this component. It is recommended to update the schema by
clicking the Update configuration button after setting
this field.
|
Customize output with DSSL | Select this option to customize the output by entering key/value pairs. Add
a row in the table by clicking the plus button under the table, enter the name of
the node in the Name field, and enter the value in the
Value field. For more information about the value syntax, see Data Shaping Selector Language syntax. This field is only available if you select JSON from the Response body format drop-down list. Information noteNote: This option impacts the schema
of the output flow of this component. It is recommended to update the schema by
clicking the Update configuration button after setting
the key/value pairs in the table under this option.
|
Global Variables
Global Variables |
ERROR_MESSAGE: the error message generated by the component when an error occurs. This is an After variable and it returns a string. This variable functions only if the Die on error check box is cleared, if the component has this check box. NB_LINE: the number of rows read by an input component or transferred to an output component. This is an After variable and it returns an integer. |