[Task]: Support embeddings using ML models in `MLTransform`

[AnandInguva]

What needs to happen?

doc - https://docs.google.com/document/d/1En4bfbTu4rvu7LWJIKV3G33jO-xJfTdbaSFSURmQw_s/edit#heading=h.wskna8eurvjv discusses on supporting generation of embeddings in MLTransform.

This would be an umbrella issue to track different features/tasks required for embeddings in MLTransform.

Issue Priority

Priority: 2 (default / most normal work should be filed as P2)

Issue Components

• Component: Python SDK
• Component: Java SDK
• Component: Go SDK
• Component: Typescript SDK
• Component: IO connector
• Component: Beam YAML
• Component: Beam examples
• Component: Beam playground
• Component: Beam katas
• Component: Website
• Component: Spark Runner
• Component: Flink Runner
• Component: Samza Runner
• Component: Twister2 Runner
• Component: Hazelcast Jet Runner
• Component: Google Cloud Dataflow Runner

[AnandInguva]

Tasks needs to be done after PR #29564 is merged.

• Add set_model_handler method
• Support Dead letter queue for RunInference and asses its compatibility with tft transforms.
• Add inference_fn to the embedding config
• Add support for artifact locations other than local system and GCS.
• Integrate ArtifactFetcher with MLTransform.