Update benchmark.py for specifying models to use

[theodufort]

Added the --models parameter to allow specific models to be ran, if you dont want to run some models

[Omega-Centauri-21]

• Improvement: This change provides more flexibility by allowing users to specify which models to benchmark.
• Code Quality: The changes are well-structured and maintain the readability of the script.
• Impact: It simplifies the process of benchmarking specific models and ensures users are informed if they specify non-existent models.

Overall, the changes are beneficial and should be approved.