[OpenVINO] Add support for Granite-4.0 family (incl. models with MoE and Mamba2)

[rkazants]

What does this PR do?

Example of conversion cmd-line for ibm-granite/granite-4.0-micro:

optimum-cli export openvino -m ibm-granite/granite-4.0-micro granite-4.0-micro

Example of inference for ibm-granite/granite-4.0-micro using OpenVINO backend:

from transformers import AutoTokenizer
from optimum.intel.openvino import OVModelForCausalLM

model_path = "ibm-granite/granite-4.0-micro"

tokenizer = AutoTokenizer.from_pretrained(model_path)
model = OVModelForCausalLM.from_pretrained(model_path)

# change input text as desired
input_text = "The capital of France is"
# tokenize the text
input_tokens = tokenizer(input_text, return_tensors="pt").to(device)
# generate output tokens
output = model.generate(**input_tokens, max_length=10)
# decode output tokens into text
output = tokenizer.batch_decode(output)
print(output[0])

Before submitting

• [N/A] 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?
• 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.