[FEA] Support configurable table-fusion indexing and NVE layer types for exportable inference embeddings

[geoffreyQiu]

Introduce configurable table fusion for exportable inference embeddings:

• Add an exportable table-fusion indexer interface, decoupled from embedding storage.
• Support identity mapping, per-table hash lookup, and a fused hash namespace using bit-concatenated table and feature IDs.
• Support configurable NVE layer types:LinearUVM, GPU-resident, Hierarchical
• Add export/AOTI coverage for the supported indexer and NVE layer combinations.

---

By submitting this issue, you agree to follow our code of conduct and our contributing guidelines.