-
Notifications
You must be signed in to change notification settings - Fork 0
/
policygeneration.py
37 lines (25 loc) · 1.45 KB
/
policygeneration.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
import openai
import config
# create policy cards
# Replace with your OpenAI API key
openai.api_key = config.OPENAI_API_KEY
def generate_policy_suggestions(user_need, effect_on_need, n_suggestions=3):
model_engine = "text-davinci-003"
prompt = f"Generate {n_suggestions} potential policy suggestions based on the following user need and its effect: \nUser need: {user_need}\nEffect on need: {effect_on_need}\n\nPolicy suggestions:"
response = openai.Completion.create(
engine=model_engine,
prompt=prompt,
max_tokens=150,
n=n_suggestions,
stop=None,
temperature=0.7,
)
policy_suggestions = [choice.text.strip() for choice in response.choices]
return policy_suggestions
# example use case
user_need = "Affordable housing for low-income families"
effect_on_need = "Increased availability of affordable housing options and reduced homelessness"
policy_suggestions = generate_policy_suggestions(user_need, effect_on_need)
for i, suggestion in enumerate(policy_suggestions, start=1):
print(f"Policy suggestion {i}: {suggestion}")
# This code will generate potential policy suggestions using the OpenAI API based on the user need and its effect. You can customize the number of suggestions by changing the n_suggestions parameter. To integrate this into your chatbot application, you can include the generate_policy_suggestions function in your Flask app and create an API endpoint for policy generation.