[Neural Speed] Improvements to run.py script

[aahouzi]

Type of Change

• Handles models that require a HF token access ID (llama, llama2, etc..)

Description

• Same as above

Expected Behavior & Potential Risk

• N/A

How has this PR been tested?

• Tested with same commands provided in the README file, on diverse models requiring token access ID: llama2-7b, llama2-13b, llama2-70b. The script completes its executions as expected.

Dependency Change?

• huggingface_hub package, but I guess this is already a dependency of transformers ?