Building a conversation
The Llama model you've been working with can accept prompts in the form of messages from different roles. In this exercise, you'll create messages to prompt a customer support chatbot for an internet service provider.
The Llama model is available as llm
.
This exercise is part of the course
Working with Llama 3
Exercise instructions
- Complete the conversation in the
conv
list so it includes the"user"
message provided asprompt
. - Pass the conversation to the Llama model,
llm
, and print the result.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
prompt = "Give me four short steps to troubleshoot my internet connection."
conv = [
# Complete the user message
{
"role": "____",
"content": ____
}
]
# Pass the conversation to the model
result = llm.create_chat_completion(messages=____, max_tokens=20)
print(result)