Adding --max_length parameter for embeddings

[przepeck]

🛠 Summary

CVS-187555
Adding --max_length parameter for embeddings to speed up model loading and inference on NPU.

Model loading time:
unset: ~13min
200: 18s

🧪 Checklist

• Unit tests added.
• The documentation updated.
• Change follows security best practices.
  ``

[Copilot]

Pull request overview

Adds support for configuring embeddings max_length (via CLI/graph export and MediaPipe node options) so model context length can be fixed explicitly—intended to improve loading/inference behavior on NPU—along with test coverage for the new setting.

Changes:

• Add --max_length to embeddings graph-export CLI and propagate it into generated graph pbtxt (max_length node option).
• Extend embeddings node initialization/servable flow to carry an optional configured max length and apply it during preprocessing.
• Add/update unit and integration tests plus test graphs/configs covering max-length behavior.

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
src/test/ovmsconfig_test.cpp	Extends config parsing tests to cover --max_length for embeddings graph settings.
src/test/graph_export_test.cpp	Updates graph export expectations and settings to include max_length.
src/test/embeddingsnode_test.cpp	Adds HTTP-level tests verifying requests within/exceeding configured max_length.
src/test/embeddings/graph_ov_max_length.pbtxt	Adds a new embeddings graph fixture with max_length set.
src/test/embeddings/config_embeddings.json	Registers the new test model/graph in embeddings test config.
src/graph_export/graph_export.cpp	Emits max_length into generated embeddings graph templates when configured.
src/graph_export/embeddings_graph_cli_parser.cpp	Adds CLI option parsing for --max_length into embeddings graph settings.
src/embeddings/embeddings_servable.hpp	Adds optional configured max-length member/plumbing into the servable.
src/embeddings/embeddings_servable.cpp	Applies configured max length as an override during preprocessing.
src/embeddings/embeddings_node_initializer.cpp	Reads max_length from node options and passes it into the servable.
src/embeddings/embeddings_calculator_ov.proto	Extends calculator options proto with optional uint32 max_length.
src/capi_frontend/server_settings.hpp	Extends embeddings graph settings struct with maxLength.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[dtrawins]

looks good. add also documentation and include new parameter in the embeddings demo with NPU