OpenAI API example

This document provides a guide on how to use the OpenAI API within the Valohai platform.

Prerequisites

1. OpenAI Account: Ensure you have an OpenAI account. similar services.
2. Funding: Add funds to your OpenAI account to enable API access.
3. API Token: Obtain an API token from OpenAI.

Configuration

Set Up Environment Variable: The API token should be set as an environment variable within Valohai as OPENAI_API_KEY.

This can be done in two ways:

• Execution Environment Variable: Add the token when creating an execution from the UI under “Create Execution -> Environment Variables.” This variable will be available only for the execution where it was created.
• Project Environment Variable: For broader access, add it under “Project Settings -> Environment Variables” tab and check the “Secret” checkbox. This makes the variable available across all project executions.

Project Setup

Create a New Project: Initialize a new project on Valohai and define your execution step in the valohai.yaml configuration file.

``` yaml
- step:
    name: openai-test
    image: python:3.9
    command:
      - pip install --upgrade openai
      - python run-openai.py
```

Executing Your First API Call

Run the API Call: Use the following Python script run-openai.py to make a call to the OpenAI API.

``` python
from openai import OpenAI
client = OpenAI()

completion = client.chat.completions.create(
  model="gpt-3.5-turbo",
  messages=[
    {"role": "system", "content": "You are a poetic assistant, skilled in explaining complex programming concepts with creative flair."},
    {"role": "user", "content": "Compose a poem that explains the concept of recursion in programming."}
  ]
)

print(completion.choices[0].message)
```

You can get more examples of API calls in the OpenAI documentation.