app.py

import os
from huggingface_hub import InferenceClient

client = InferenceClient(
    provider="auto",
    api_key=os.environ["HF_API_TOKEN"],
)

stream = client.chat.completions.create(
    model="meta-llama/Llama-3.2-1B-Instruct",
    messages=[],
    temperature=0.5,
    top_p=0.7,
    stream=True,
)

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