Skip to content

3.1.2.3 OpenAI LLM Use

OpenAI Usage

1. Create the relevant file.

Create a YAML file, for example, user_openai.yaml Paste the following content into your user_openai.yaml file.

YAML
name: 'user_openai_llm'
description: 'user define openai llm'
model_name: 'gpt-3.5-turbo'
max_tokens: 1000
metadata:
  type: 'LLM'
  module: 'agentuniverse.llm.default.default_openai_llm'
  class: 'DefaultOpenAILLM'

2. Environment Setup

Must be configured: OPENAI_API_KEY
Optional: OPENAI_PROXY, OPENAI_API_BASE

2.1 Configure through Python code

Python
import os
os.environ['OPENAI_API_KEY'] = 'sk-***'
os.environ['OPENAI_PROXY'] = 'https://xxxxxx'
os.environ['OPENAI_API_BASE'] = 'http://xxxx/v1'

2.2 Configure through the configuration file

In the custom_key.toml file under the config directory of the project, add the configuration:

TOML
OPENAI_API_KEY="sk-******"
OPENAI_PROXY="https://xxxxxx"
OPENAI_API_BASE="http://xxxx/v1"

3. Obtaining the OpenAI API KEY

Reference

4. Tips

In agentuniverse, we have already created an llm with the name default_openai_llm. After configuring the OPENAI_API_KEY, users can directly use it.