secret

app.py

@@ -1,10 +1,9 @@
 import gradio as gr
 # from huggingface_hub import InferenceClient
 from openai import OpenAI
-
-# Modify OpenAI's API key and API base to use vLLM's API server.
-openai_api_key = "12345"
-openai_api_base = "https://a502-131-112-63-87.ngrok-free.app/v1"
+import os
+openai_api_key = os.getenv('api_key')
+openai_api_base = os.getenv('url')
 model_name = "cyberagent/calm3-22b-chat"
 """
 For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
@@ -36,17 +35,6 @@ def respond(
 
     response = ""
 
-    """
-    #streaming無し: gradio側が対応してない
-    completion = client.chat.completions.create(model=model_name,
-                                                messages=messages,
-                                                temperature=temperature,
-                                                max_tokens=max_tokens,
-                                                top_p=top_p,
-                                                )
-    text = completion.choices[0].message.content.strip()
-    return text
-    """
     for message in client.chat.completions.create(
         model=model_name,
         messages=messages,