[llm_bench] Add video generation pipeline to benchmark

[sbalandi]

Description

example:
hf model:

1. donwload Lightricks/LTX-Video
   python ./tools/llm_bench/benchmark.py -m /./models/ltx_hf/ -pf ./tools/llm_bencts/video_generation_ltx_video.jsonl -f pt --task text-to-video -n 3

Optimum model:

optimum-cli export openvino -m Lightricks/LTX-Video ./models/ltx_video
python ./tools/llm_bench/benchmark.py -m /./models/ltx_video/ -pf ./tools/llm_bencts/video_generation_ltx_video.jsonl --optimum --task text-to-video -n 3

Note: --static_reshape have added but don't work well, need to investigate

CVS-176897

Fixes #(issue)

Checklist:

• Tests have been updated or added to cover the new code.
• This patch fully addresses the ticket.
• I have made corresponding changes to the documentation.