Moderating the model response
Even though the chatbot has generated a tailored response, it's important to verify that the content is safe to send to the customer. In this step, you'll moderate the response to ensure all category scores fall below the customer safety threshold.
The model's reply from the previous exercise is available as chatbot_reply.
This exercise is part of the course
Multi-Modal Systems with the OpenAI API
Exercise instructions
- Send a moderation request using the
text-moderation-latestmodel withchatbot_replyas input. - Extract the response's category scores, convert them to a dictionary using
.model_dump(), and check if any score exceeds0.7.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
client = OpenAI(api_key="")
# Send the moderation request
response = ____
# Extract scores and convert to dictionary
scores = ____
if all(____ for score in scores.values()):
print("AI Response flagged for moderation!")
chatbot_reply = """I'm sorry, but I can't provide a response to that request. Please contact [email protected] for further assistance."""
else:
print("AI Response is safe.")