switched cumproduct to cumprod in base_diffusion for numpy compliance + added utf8 compliance in update_model_index

[manaspgandhi]

Motivation

I am using mmgeneration for a stable diffusion model. I noticed some issues in the codebase that I changed to make it work, and wanted to make sure everyone could use the codebase without issues.

Modification

• In /mmgeneration/mmgen/models/diffusions/base_diffusion.py I changed np.cumproduct to np.cumprod. This is because np.cumproduct has been deprecated as of NumPy 1.25.0 and entirely removed in NumPy 2.0, being replaced with np.cumprod. This change will ensure compliance with the updated numpy.
• In /mmgeneration/.dev_scripts/github/update_model_index.py I changed the with open(md_file, 'r') in get_task_dict(md_file) to with open(md_file, 'r', encoding='utf-8'), because on windows machines, it does not always automatically read the file as utf08, and causes errors for certain situations, such as pre-commit. This change will ensure all windows users don't face an issue when trying to open that md file in python.