This script deftly enlists the Groq client to choreograph a dance with the llama3-70b-8192 model, playing out a skit where it conjures up an email about the thrilling world of government contract bidding in NY, sans the fluffy pleasantries.

groq_llama3.py

from groq import Groq

# Initialize Groq client with API key
client = Groq(api_key="gsk_123")

completion = client.chat.completions.create(
    model="llama3-70b-8192",
    messages=[
        {
            "role": "system",
            "content": "Wise Assistant"
        },
        {
            "role": "user",
            "content": "Context:\nCompose a draft email to my client about how to submit bids for government contracts in NY. Omit any initial greetings or pleasantries, such as 'I hope this message finds you well."
        }
    ],
    temperature=0.1,
    max_tokens=8192,
    top_p=1,
    stream=True,
    stop=None,
)

for chunk in completion:
    print(chunk.choices[0].delta.content or "", end="")

This script deftly enlists the Groq client to choreograph a dance with the llama3-70b-8192 model, playing out a skit where it conjures up an email about the thrilling world of government contract bidding in NY, sans the fluffy pleasantries. Picture it as a digital concierge, coolly crafting a no-nonsense guide sans the "Hope this finds you well" snooze fest. The parameters set the stage for a low-drama, precise output, ensuring that the email stays as engaging as a tax form—and streams the fun in bite-sized chunks.