Generating creative copy
You're developing an AI-powered content assistant for a SaaS marketing team. The team needs to automate social media posts about their latest software updates, and you need to adjust response diversity so that multiple calls to the model result in different variations.
You have been provided the Llama
class instance in the llm
variable and the code to call the completion. You are also given a sample prompt to test with.
This exercise is part of the course
Working with Llama 3
Exercise instructions
- Adjust the top-p parameter to a value in the upper half of its range so that it generates more varied responses.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
output = llm(
"Write a tweet announcing a new analytics dashboard feature for enterprise users.",
max_tokens=15,
# Set top-p to a value in the upper range for more varied responses
top_p=____
)
print(output['choices'][0]['text'])