Update `WatsonxLLM.metadata` property to avoid validation error when `model_limits` field isn't present (#17839)

* build: support `python3=3.9,<4.0` for `llama-index-vector-stores-databricks`

Signed-off-by: themattmorris <matt.morris114@gmail.com>

* build: bump `llama-index-vector-stores-databricks` minor version

Signed-off-by: themattmorris <matt.morris114@gmail.com>

---------

Signed-off-by: themattmorris <matt.morris114@gmail.com>

Fix "Invalid value for 'content': expected a string, got null." openai error in case of empty assistant messages (#17921)

* update mocked payload

* fix outlook import paths

* reduce flakyness

* update for redisvl 0.4.0

* update toml

* bump version

* simplify creation

* fix: Convert Embedding Functions to Async

This PR updates the embedding functions to run asynchronously. By wrapping the synchronous implementations with asyncio.to_thread, both _aget_query_embedding and _aget_text_embedding now execute without blocking the event loop.

* Update pyproject.toml

Add support for anthropic.claude-3-7-sonnet-20250219-v1:0 model with 200k token context window

* attach more metadata to retrieved image node

* remove print statement

* bump version

vbump

* include some metadata when returning image node from llamacloud retriever

* version bump

* add tests + sync implementation