Step 1 - Get the OPENAI API Token
To use OpenAI API, the first step is to create API Key.
- Visit API Key to Generate new secret key if you don’t have yet, and then copy right after generation since you don’t have second chance to get that unless you generate another one.
With that, if you want to run Python script in your local terminal, you can use
oslibrary to set environment variable.
1
2
import os
os.environ["OPENAI_API_KEY"] = "sk-xxxx"
1
- :bulb: If met with `module not found` issue, please check out below reference. highly suspicious that it's caused by multiple version of Python installed.
Step 2 - Use API Key to call OpenAI API
1
2
3
4
5
6
7
8
9
10
11
import openai
openai.api_key = "sk-xxxx" # replace with your API key
# Example usage: generate text with GPT-3
prompt = "Once upon a time"
model = "text-davinci-002"
response = openai.Completion.create(engine=model, prompt=prompt, max_tokens=50)
generated_text = response.choices[0].text
print(generated_text)
// This code block shows how to use the OpenAI API to generate text using the GPT-3 model. First, you need to set your API key using `openai.api_key`. Then, you can call `openai.Completion.create` to generate text with the specified model and prompt. The generated text is returned in the `response` object, and can be accessed using `response.choices[0].text`.
Some tips when using the ChatGPT API:
- Experiment with different prompts to get the desired output.
- Choose the appropriate model for your use case. The
text-davinci-002model is the most powerful, but also the most expensive. - Use the
max_tokensparameter to control the length of the generated text. - Be mindful of the API usage limits to avoid unexpected charges.
在使用OpenAI API的时候，需要传的参数里面有两个关键参数：model, messages.
model比较简单，是可以指定你想要实用的LLM模型的类型，比如gpt-3.5-turbo, gpt-4, gpt-4-0314等。 注意使用不同的model，价格是不一样的。
另外一个参数是messages，是一对message对象的列表，而每个对象又有两个必须的属性：
-
role： 是指这条消息的角色，包含system，user和assistant -
content： 消息的内容。
关于role， 区分system和assistant还稍微花点时间。根据OpenAI的官方文档， system像是一个定位整个对话的引导员，通常会是高速chatgpt它的角色是什么，比如：
{'role': 'system', 'content': 'You are an assistant that speaks like Shakespeare. ''
而 assistant就可以看成是chatgpt这个人化的角色。
System: Sets behavior of assistant Assistant: Chat model User: You
1
2
3
4
5
6
7
8
messages = [
{'role':'system', 'content':'You are friendly chatbot.'},
{'role':'user', 'content':'Hi, my name is Isa'},
{'role':'assistant', 'content': "Hi Isa! It's nice to meet you. \
Is there anything I can help you with today?"},
{'role':'user', 'content':'Yes, you can remind me, What is my name?'} ]
response = get_completion_from_messages(messages, temperature=1)
print(response)
flowchart LR
system[] --> assistant <--> user
详细解释和例子可以在ChatGPT API Transition Guide 找到。