fix: check model is loaded before starting

[vansangpfiev]

Describe Your Changes

This pull request introduces several updates to improve logging, enhance functionality, and fix redundant or unnecessary code in the engine's configuration, extensions, and services. Key changes include refining logging levels, adding new parameters for model handling, and addressing redundant variable declarations.

Logging Improvements:

• Updated logging level from CTL_DBG to CTL_TRC in YamlHandler::ReadYamlFile to reduce verbosity for debug-level logs and align with trace-level logging. (engine/config/yaml_config.cc, engine/config/yaml_config.ccL51-R51)

New Features:

• Added support for the ignore_eos parameter in ConvertJsonToParamsVector, allowing the engine to ignore end-of-sequence tokens during processing. (engine/extensions/local-engine/local_engine.cc, engine/extensions/local-engine/local_engine.ccR83-R87)
• Enhanced LoadModel to check if a model is already loaded. If so, it logs a warning, returns a 409 error status, and avoids redundant loading. (engine/extensions/local-engine/local_engine.cc, engine/extensions/local-engine/local_engine.ccR510-R526)

Code Simplification:

• Removed redundant re-declaration of model_author in ModelSourceService::AddCortexsoRepo to streamline the logic and improve readability. (engine/services/model_source_service.cc, engine/services/model_source_service.ccL436-R436)

Parameter Updates:

• Added the --jinja parameter to the LoadModel method, enabling additional functionality for Jinja template processing. (engine/extensions/local-engine/local_engine.cc, engine/extensions/local-engine/local_engine.ccR567-R568)

Fixes Issues

• Closes #
• Closes #

Self Checklist

• Added relevant comments, esp in complex areas
• Updated docs (for bug fixes / features)
• Created issues for follow-up changes or refactoring needed