Improve provider-specific errors in vectorizer _set_model_dims()

Summary

Replace generic fallback exceptions in vectorizer _set_model_dims() with provider-specific, actionable errors.

Level: Beginner

Current State

• Several vectorizer implementations catch broad Exception with TODO comments.
• Resulting messages are vague and slow down troubleshooting.

Proposed Change

• Update _set_model_dims() error handling in:
  • redisvl/utils/vectorize/text/openai.py
  • redisvl/utils/vectorize/text/cohere.py
  • redisvl/utils/vectorize/text/mistral.py
  • redisvl/utils/vectorize/text/azureopenai.py
  • redisvl/utils/vectorize/text/huggingface.py
  • redisvl/utils/vectorize/bedrock.py
  • redisvl/utils/vectorize/vertexai.py
  • redisvl/utils/vectorize/voyageai.py
• Catch known provider/network/auth cases and include remediation hints.

Definition of Done

• TODO fallback comments are removed.
• Errors include provider/model context and useful next-step guidance.
• Existing tests pass; add targeted tests where needed.

Out of Scope

• Refactoring vectorizer architecture or retry policy behavior.

Read the _set_model_dims() implementations and TODO fallback paths in the eight listed provider files, then inspect the existing vectorizer tests before running them. Done means removing those TODO fallbacks, adding provider/model context and actionable guidance for known provider, network, and authentication cases, and keeping existing tests passing.