OpenAI
App Vendor: OpenAI
App Category: IT Services
Connector Version: 1.1.0
API Version: 1.0.0
About App
OpenAI is an AI research and deployment company. OpenAI provides various artificial intelligence models such as GPT-3, Codex, and more that enable natural interfaces with large-scale models.
The OpenAI app is configured with Orchestrate to perform the following actions:
Action Name | Description |
|---|---|
Create Completion | This action creates a text completion from a model. |
Create Image | This action creates an image. |
Edit Prompt | This action edits a prompt. |
Get All Models | This action retrieves a list of models that the user has access to. |
Get Model Details | This action retrieves the details of a model. |
Create Chat Completion | This action creates a model response for a chat conversation. |
Generic Action | This is a generic action to perform any additional use case on Open AI. |
Configuration Parameters
The following configuration parameters are required for the Open AI app to communicate with the Open AI enterprise application. The parameters can be configured by creating instances in the app.
Parameter | Description | Field Type | Required/Optional | Comments |
|---|---|---|---|---|
API Token | Enter the API token to authenticate the client. | Password | Required |
Action: Create Completion
This action creates a text completion from a model.
Action Input Parameters
Parameter | Description | Field Type | Required/Optional | Comments |
|---|---|---|---|---|
Model Identifier | Enter the model identifier. Example: "text-davinci-003" | Text | Required | You can retrieve the model identifiers using the Get All Models action. |
Prompt | Enter the prompt. Example: "write a paragraph on trends in cybersecurity over the last decade" | Text | Required | |
Max Tokens | Enter the maximum number of tokens to return. | Integer | Optional | Default value: 256 |
Temperature | Enter the temperature to use for sampling. | Float | Optional | Default value: 0.7 It is recommended to alter either temperature or top_p but not both. |
Top P | Enter the top_p to use for sampling. | Float | Optional | Default value: 1 It is recommended to alter either temperature or top_p but not both. |
Frequency Penalty | Enter the frequency penalty to use for sampling. | Float | Optional | Allowed values: -2.0 to 2.0 |
Presence Penalty | Enter the presence penalty to use for sampling. | Float | Optional | Allowed values: -2.0 to 2.0 Default value: 0 |
Stop | Enter the stop sequence to use for sampling. | Text | Optional | Default value: Null |
Number of Samples | Enter the number of samples to retrieve. | Integer | Optional | Default value: 1 Use this parameter carefully and ensure that you have reasonable settings for max_tokens and stop parameters. |
Action: Create Image
This action creates an image.
Action Input Parameters
Parameter | Description | Field Type | Required/Optional | Comments |
|---|---|---|---|---|
Prompt | Enter the text description of the image. | Text | Required | The maximum allowed length is 1000 characters. |
Number of Samples | Enter the number of images to generate. | Integer | Optional | Default value: 1 Allowed values: Between 1 to 10 |
Size | Enter the size of the image to be generated. | Text | Optional | Default value: 1024x1024 Allowed values:
|
Action: Edit Prompt
This action edits a prompt
Action Input Parameters
Parameter | Description | Field Type | Required/Optional | Comments |
|---|---|---|---|---|
Model Identifier | Enter the identifier of the model. Example: "text-davinci-003" | Text | Required | You can retrieve the model identifiers using the Get All Models action. |
Input | Enter the input text to use as a starting point for the edit. Example: "write a paragraph on trends in cybersecurity over the last decade" | Text | Required | |
Instruction | Enter the instruction to edit the prompt. | Text | Required | |
Number of Samples | Enter the number of edits to be generated for the input and instruction. | Integer | Optional | Default value: 1 |
Temperature | Enter the temperature to use for sampling. | Integer | Optional | Default value: 1 It is recommended to alter either temperature or top_p but not both. |
Top P | Enter the top_p to use for sampling. | Integer | Optional | Default value: 1 It is recommended to alter either temperature or top_p but not both. |
Action: Get All Models
This action retrieves a list of models that the user has access to.
Action Input Parameters
This action does not require any action input parameter.
Action: Get Model Details
This action retrieves the details of a model.
Action Input Parameters
Parameter | Description | Field Type | Required/Optional | Comments |
|---|---|---|---|---|
Model Identifier | Enter the identifier of the model to retrieve details. Example: "text-davinci-003" | Text | Required | You can retrieve the model identifiers using the Get All Models action. |
Action: Create Chat Completion
This action creates a model response for the given chat conversation.
Action Input Parameters
Parameter | Description | Field Type | Required/Optional | Comments |
Model ID | Enter the model ID of the model to use. Example: "text-davinci-003" | Text | Required | You can retrieve the model ID using the action Get All Models. |
Messages | Enter a list of messages describing the conversation so far. Example: $LIST[{"role": "customer", "text": "Can you help me find a pen?"}] | List | Required |
|
Extra Params | Enter the extra parameters.
| Key Value | Optional | Allowed keys:
|
Action: Generic Action
This is a generic action to perform any additional use case on Open AI.
Action Input Parameters
Parameter | Description | Field Type | Required/Optional | Comments |
|---|---|---|---|---|
Endpoint | Enter the endpoint. Example: https://api.openai.com/v1/engines/davinci/completions | Text | Required |
|
Method | Enter the HTTP method to use. Example:
| Text | Required |
|
Payload | Enter the payload to pass with the request. Example: {"prompt": "Once upon a time"} | Key Value | Optional |
|
Query Params | Enter the query parameters to filter results. Example: {"limit": 100} | Key Value | Optional |
|