I keep getting a connection error when trying to connect to OpenAI’s API in my Python app. I’m building a chatbot using streamlit and langchain but can’t get past this issue.
I’ve checked the OpenAI docs but nothing seems to work. The error happens every time I try to make a request.
Here’s my code:
def enhance_user_input(chat_history, user_message):
api_response = openai_client.completions.create(
model="gpt-3.5-turbo",
prompt=f"""Based on this chat history and user message, create a better question for searching our database.\n\nCHAT HISTORY: \n{chat_history}\n\nUser Message: {user_message}\n\nImproved Question:""",
temperature=0.5,
max_tokens=200,
top_p=1,
frequency_penalty=0,
presence_penalty=0
)
return api_response['choices'][0]['text']
def build_chat_history():
full_conversation = ""
for index in range(len(st.session_state['bot_replies'])-1):
full_conversation += "User: " + st.session_state['user_messages'][index] + "\n"
full_conversation += "Assistant: " + st.session_state['bot_replies'][index+1] + "\n"
return full_conversation
# Main execution
with st.spinner("Processing..."):
chat_context = build_chat_history()
improved_question = enhance_user_input(chat_context, user_input)
The error shows up when calling the completions.create method. Has anyone faced this before?