Azure OpenAI Part 2: Show me the code!

前一篇 介紹了申請以及使用 Azure OpenAI Studio (https://oai.azure.com/portal) 快速部署 Model 以及驗證. 這篇介紹如何使用 Python Code 整合進原本的應用程式.

1. 在前一篇最後, 利用 Playground 測試 OpenAI 時, 不知道有沒有發現右上方有 Code View 這個按鈕?
   

2. 比對一下 Azure OpenAI 和 OpenAI 的範例.

   • Azure OpenAI 的範例

     import os
     import openai
     openai.api_type = "azure"
     openai.api_base = "https://<YOUR-ENDPOINT-NAME>.openai.azure.com/"
     openai.api_version = "<OPENAI-API-VERSION>"
     openai.api_key = os.getenv("OPENAI_API_KEY")
     
     response = openai.Completion.create(
       engine="<MODEL-NAME>",
       prompt="..."
       temperature=0.7,
       max_tokens=60,
       top_p=1,
       frequency_penalty=0,
       presence_penalty=0,
       stop=None)    
     

   • OpenAI 官方網站的範例

     import os
     import openai
     
     openai.api_key = os.getenv("OPENAI_API_KEY")
     
     start_sequence = "\nA:"
     restart_sequence = "\n\nQ: "
     
     response = openai.Completion.create(
       model="<MODEL-NAME>",
       prompt="...",
       temperature=0,
       max_tokens=100,
       top_p=1,
       frequency_penalty=0,
       presence_penalty=0,
       stop=["\n"]
     )
     

   看起來只需要 OPENAI_API_KEY 就可以了.

3. 回到 Azure Cognitive service 畫面, 點選 OpenAI 設定主畫面左方的選單 Keys and Endpoint. 將 Key1 的值複製作為 OPENAI_API_KEY

4. 這次來試試 code-davinci-002 吧!

   import os
   import openai
   openai.api_type = "azure"
   openai.api_base = "https://<ENDPOIND-NAME>.openai.azure.com/"
   openai.api_version = "2022-06-01-preview"
   openai.api_key = os.getenv("OPENAI_API_KEY")
   
   prompt = """Answer the question as truthfully as possible using the provided text, and if the answer is not contained within the text below, say "I don't know"
   
   Q: 幫我產生 ssh tunnel 的 bash script
   A:"""
   
   openai.Completion.create(
       engine="code-davinci-002",
       prompt=prompt,
       temperature=0,
       max_tokens=300,
       top_p=1,
       frequency_penalty=0,
       presence_penalty=0,
       model=COMPLETIONS_MODEL
   )["choices"][0]["text"].strip(" \n")    
   

   某種程度也算對…