Support for prompts text file input for multi-video-generation

[glmcdona]

Suggest support allowing for generation of multiple videos from a single hv_generate_video.py call. This saves on time when you want to generate a bunch of videos at once by being able to avoid all the repeated model loads and unloads for each video.

Attaching PR that adds this feature in a minute.

[glmcdona]

#88 PR.