Parameters for product descriptions
You're working with Amazon Bedrock to generate product descriptions. You need to understand how different parameter settings affect the model's output. Using the provided generate_description function, you can experiment with different combinations of temperature, top_p, and max_tokens to see how they influence the style and consistency of generated product descriptions.
The boto3 and json libraries, and generate_description() function, have been pre-imported.
This exercise is part of the course
Introduction to Amazon Bedrock
Exercise instructions
- Set parameters for three configurations, using different values of the parameters for a conservative, balanced, and creative output.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
bedrock = boto3.client('bedrock-runtime', region_name='us-east-1')
product = "Wireless Noise-Cancelling Headphones"
# Configuration 1: Conservative/Consistent
result1 = generate_description(bedrock, product, temperature=____, top_p=0.85, max_tokens=50)
# Configuration 2: Balanced
result2 = generate_description(bedrock, product, temperature=0.5, top_p=0.92, max_tokens=____)
# Configuration 3: Creative
result3 = generate_description(bedrock, product, temperature=0.8, top_p=____, max_tokens=200)
print("Conservative Result:", result1)
print("Balanced Result:", result2)
print("Creative Result:", result3)