Skip to content

Latest commit

 

History

History
 
 

prompt_chatgpt

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 

🤖 Python Prompt ChatGPT Function

Ask question, and let OpenAI GPT-3.5-turbo answer.

🧰 Usage

GET /

HTML form for interacting with the function.

POST /

Query the model for a completion.

Parameters

Name Description Location Type Sample Value
Content-Type The content type of the request body Header application/json N/A
prompt Text to prompt the model Body String Write a haiku about Mondays

Sample 200 Response:

Response from the model.

{
  "ok": true,
  "completion": "Monday's heavy weight, Dawning with a sigh of grey, Hopeful hearts await."
}

Sample 400 Response:

Response when the request body is missing.

{
  "ok": false,
  "error": "Missing body with a prompt."
}

Sample 500 Response:

Response when the model fails to respond.

{
  "ok": false,
  "error": "Failed to query model."
}

⚙️ Configuration

Setting Value
Runtime Python (3.9)
Entrypoint src/main.py
Build Commands pip install -r requirements.txt
Permissions any
Timeout (Seconds) 15

🔒 Environment Variables

OPENAI_API_KEY

A unique key used to authenticate with the OpenAI API. Please note that this is a paid service and you will be charged for each request made to the API. For more information, see the OpenAI pricing page.

Question Answer
Required Yes
Sample Value sk-wzG...vcy
Documentation OpenAI Docs

OPENAI_MAX_TOKENS

The maximum number of tokens that the OpenAI response should contain. Be aware that OpenAI models read and write a maximum number of tokens per API call, which varies depending on the model. For GPT-3.5-turbo, the limit is 4096 tokens.

Question Answer
Required No
Sample Value 512
Documentation OpenAI: What are tokens?