Fix default max_length computation to address issue #10461 (#12255)

.../llama_index/embeddings/huggingface_optimum_intel/base.py

...index-embeddings-huggingface-optimum-intel/pyproject.toml

...ptimum/llama_index/embeddings/huggingface_optimum/base.py

...llama-index-embeddings-huggingface-optimum/pyproject.toml

...gs-huggingface/llama_index/embeddings/huggingface/base.py

...eddings/llama-index-embeddings-huggingface/pyproject.toml