https://huggingface.co/settings/tokens
First, create one read token and add it to the Secrets of your space as HF_TOKEN. If you add tokens to an environment variable or write it directly in your code, it will be exposed to the whole world.
Any name for the HF_TOKEN part.

import os
hf_token = os.getenv("HF_TOKEN")

from transformers import AutoConfig
config = AutoConfig.from_pretrained("meta-llama/Meta-Llama-3.1-70B", revision="main", token=hf_token)
config.rope_scaling = {"type": "llama3", "factor": 8.0}
model = AutoModelForCausalLM.from_pretrained("meta-llama/Meta-Llama-3.1-70B", config=config, use_auth_token=True, token=hf_token)

I'm pretty sure this model was only for people on a Pro subscription, but if you're posting, you must be on a subscription, so I'm sure you'll be fine.