OpenAI Serverless

A Stack for managing OpenAI API access using Serverless Framework against AWS. This is useful for exposing your custom OpenAI API usage to the public or run experiments.

Features

• JWT Authentication (multi-user support)
  • Usage counting and segregation for each JWT user (In Progress)
• Improved Answers Endpoint (/better_answers)
• File uploads via S3
  • Queued file uploads (Planned)

Prerequisites

• NodeJS LTS
• yarn
• pyenv
• Pipenv
• make
• Docker (for building dependencies during deployment)

Install

# install serverless dependencies
yarn install
# install python dependencies
pipenv install

Env Vars

Please check .env.example. Copy to .env and add the OpenAI API Key.

Run make create-jwt to create the JWK for signing. Put that in to .env.

Run make sign-jwt user="<username>" to get a signed JWT for requests.

How to test locally

make local args="--stage dev"

How to deploy

First make sure we have the right AWS profile set up, it is best to inject keys via AWS_PROFILE env var. (Put into .env is fine).

make deploy

How to make requests

For example, with /better_answers:

curl --request POST \
  --url http://localhost:3000/dev/better_answers \
  --header 'Authorization: jwt <jwt_token>' \
  --header 'Content-Type: application/json' \
  --data '{
	"file": "file-xxxx",
	"question": "Can I update my User ID?",
	"temperature": 0,
	"search_model": "babbage",
	"model": "davinci-instruct-beta-v3",
	"examples_context": "In 2017, U.S. life expectancy was 78.6 years.",
	"examples": [["What is human life expectancy in the United States?","78 years."]],
	"max_tokens": 32,
	"max_rerank": 50,
	"z_threshold": 1,
	"stop": ["\n", "===", "<|endoftext|>"]
}'

Current Methods

TODO

License

MIT