Introduces a 'fail_safe_mode' option to the Embedding Model and OpenSearch (Multi-Model Multi-Embedding) components, allowing errors to be logged and None returned instead of raising exceptions. Refactors embedding model fetching logic for better error handling and updates component metadata, field order, and dependencies. Also adds 'className' fields and updates frontend node folder IDs for improved UI consistency.
Introduces a 'fail_safe_mode' option to the Embedding Model component, allowing errors to be logged and None returned instead of raising exceptions. Refactors embedding initialization logic for OpenAI, Ollama, and IBM watsonx.ai providers to support this mode, and updates UI configuration and metadata accordingly.
Replaces all references to 'OpenSearchHybrid-Ve6bS' with 'OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4' in main.py, processors, and file service. Adds a utility for injecting provider credentials into Langflow request headers and integrates it into chat and file services for improved credential handling.
Replaced imports from config_manager with settings in chat_service.py and langflow_file_service.py to use get_openrag_config from config.settings. This change ensures consistency with the updated configuration structure.
Replaces OpenSearchHybrid with OpenSearchVectorStoreComponentMultimodalMultiEmbedding in ingestion_flow.json, updating all relevant edges and embedding connections. Updates docker-compose.yml to use local builds for backend, frontend, and langflow, and improves environment variable handling for API keys. This refactor enables multi-model and multimodal embedding support for document ingestion and search.
Introduces SELECTED_EMBEDDING_MODEL as a global environment variable in docker-compose files and ensures it is passed in API headers for Langflow-related services. Updates settings and onboarding logic to set this variable without triggering flow updates, improving embedding model configuration consistency across services.
