add support of nanollava model

[eaidova]

What does this PR do?

added support of nanollava model family (qnguyen3/nanoLLaVA).

example of usage:

from PIL import Image
from optimum.intel.openvino import OVModelForVisualCausalLM
from transformer import AutoTokenizer, AutoProcessor, TextStreamer

model_id = "qnguyen3/nanoLLaVA"

model = OVModelForVisualCausalLM.from_pretrained(model_id, trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(model_id, trsut_remote_code=True)
processor = AutoProcessor.from_pretrained(model.config.mm_vision_tower)

prompt = "Describe this image in detail"

def process_text_input(text, tokenizer):
    text_chunks = [tokenizer(chunk).input_ids for chunk in text.split("<image>")]
    input_ids = torch.tensor(text_chunks[0] + [-200] + text_chunks[1], dtype=torch.long).unsqueeze(0)
    attention_mask = torch.ones_like(input_ids, dtype=torch.int64)
    return input_ids, attention_mask

messages = [{"role": "user", "content": f"<image>\n{prompt}"}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
url = "https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/8bf7d9f2-018a-4498-bec4-55f17c273ecc"
image = Image.open(requests.get(url, stream=True).raw)
image_tensor = processor(images=image, return_tensors="pt")["pixel_values"][0]
input_ids, attention_mask = process_text_input(text, tokenizer)

streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
output_ids = model.generate(input_ids, attention_mask=attention_mask, images=image_tensor, max_new_tokens=128, use_cache=True, streamer=streamer)

Before submitting

• [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• [ ] Did you make sure to update the documentation with your changes?
• [x] Did you write any new necessary tests?

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.