Classification with Python and OpenAI
def get_completion_from_messages(messages,
model="gpt-3.5-turbo",
temperature=0,
max_tokens=500):
response = openai.ChatCompletion.create(
model=model,
messages=messages,
temperature=temperature,
max_tokens=max_tokens,
)
return response.choices[0].message["content"]
When classifying data with AI, it can be useful to have primary and secondary categories. The categories are fixed, and the instructions are hard coded for each category.
This prompts returns JSON.
delimiter = "####"
system_message = f"""
You will be provided with customer service queries. The customer service query will be delimited with {delimiter} characters.
Classify each query into a primary category and a secondary category.
Provide your output in json format with the keys: primary and secondary.
Primary categories: Billing, Technical Support, Account Management, or General Inquiry.
Billing secondary categories:
Unsubscribe or upgrade
Add a payment method
Explanation for charge
Dispute a charge
Technical Support secondary categories:
General troubleshooting
Device compatibility
Software updates
Account Management secondary categories:
Password reset
Update personal information
Close account
Account security
General Inquiry secondary categories:
Product information
Pricing
Feedback
Speak to a human
"""
user_message = f"""I want you to delete my profile and all of my user data"""
messages = [
{'role':'system',
'content': system_message},
{'role':'user',
'content': f"{delimiter}{user_message}{delimiter}"},
]
response = get_completion_from_messages(messages)
print(response)
user_message = f"""\
Why is my monthly payment so high?"""
messages = [
{'role':'system',
'content': system_message},
{'role':'user',
'content': f"{delimiter}{user_message}{delimiter}"},
]
response = get_completion_from_messages(messages)
print(response)