feat: retry logic on chunks stream requests

[TordAreStromsnes]

Bulk Data Streaming Reliability:

• Added a stream_retries parameter to get_bulk_data_stream and the underlying _request_binary_stream method, enabling automatic retries for transient failures during streaming. Retries resume after the bytes already yielded, assuming deterministic endpoint behavior.
• Updated internal streaming logic to track progress and handle resuming streams after failures, with incremental backoff between retry attempts.

Documentation and Example Improvements:

• Updated docstrings, guides, and API documentation to describe the new stream_retries parameter, explain its usage, and clarify that streaming timeouts apply to connection setup and waiting for new bytes, not the total download duration.
• Expanded usage examples in README.md, guides, and API docs to demonstrate how to use stream_retries and set appropriate timeouts when streaming large binary datasets.

Configuration and Example Consistency:

• Updated DSIS configuration examples across documentation and code samples to consistently include the dsis_site parameter and to use the correct environment (e.g., "qa" or "prod") as appropriate for the context.

Other Documentation Cleanups:

• Removed redundant model_name and model_version from some guide examples to prevent confusion.
• Clarified explanations for null value handling and memory management in binary data guides.