add video-llava model support

[kaixuanliu]

This PR enables model LanguageBind/Video-LLaVA-7B-hf. Throughput comparison of Gaudi2 and A100 is(single card):

datatype	A100	Gaudi2
FP32	6.22	23.27
BF16	26.74	29.22

[kaixuanliu]

@regisss @mandy-li @libinta , pls help review

[kaixuanliu]

cmd line:
python3 run_example.py --model_name_or_path "LanguageBind/Video-LLaVA-7B-hf" --warmup 3 --n_iterations 5 --batch_size 1 --use_hpu_graphs --bf16

[regisss]

@kaixuanliu A couple of comments:

• Could you please add a test to test both this new example and the model. You can do the same as what was done for other examples with a dedicated test file as you can see here: https://github.com/huggingface/optimum-habana/tree/main/tests ?
• Could you also add this model to the table in the README and in the doc ?

[kaixuanliu]

@regisss Done per your suggestions