Add-ons

Generates a response to the Text Prompt. By default, it is generated by accessing “gpt-4-1106-preview MODEL” (can be changed to any MODEL). Text prompts for starting a conversation are defined separately for system messages and user messages. Generally, the system message contains the answer rules and the personality of the answering character, and the user message contains the question text.

Creates an image given a PROMPT text by the “DALL-E 3 MODEL” via OpenAI API. By default a 1024×1024 image is generated, but 1024×1792 or 1792×1024 are also available. Note that PROMPT must be defined within 4000 characters.

Creates a response for a prompt with images. MODEL (GPT-4 with Vision) takes in images and answers questions related to them, such as what the image represents, what is in the image, and more (eg. ideas for dinner based on what is in fridge).

Communicates with Azure OpenAI Service API (ChatGPT running on Microsoft infrastructure). Supports the assignment of advanced parameters. Specifically, sampling temperature, upper % sampling, recurrence prohibition, frequency prohibition and Logit bias can be set.

Lists status updates for the fine-tuning job. The number of all events and the TSV of all event details can be obtained; created time, event level, event message. (upto 20)

Immediately cancels the Fine Tuning Job (FtJob). FtJob status will be “cancelled”. If logging of fine-tuning events is required, perform a separate operation to obtain a list of events.

Gets info about the Fine-Tune Job (FtJob). FineTune status can be obtained. If the job status is “succeeded”, the MODEL name (ID) can also be obtained. It is typically placed in a flow that is called repeatedly every hour.

Creates a job (FtJob) that fine-tunes a specified MODEL. Stores the job ID. The job will take minutes or hours. The Fine-Tuned MODEL (FtMODEL) will be available after the job is completed. Note that the training file (dataset) must be uploaded in advance.