- Remove optional 'modes' parameter from aclear_cache() and clear_cache() methods
- Replace deprecated drop_cache_by_modes() with drop() method for complete cache clearing
- Update API endpoint to ignore mode-specific parameters and clear all cache
- Simplify frontend clearCache() function to send empty request body
This change ensures all LLM cache is cleared together.
The `_migrate_entity_relation_data` function previously processed directed edges from `get_all_edges`, which could lead to duplicates (e.g., (A,B) and (B,A)) and an incorrect relation count.
This commit normalizes edges by sorting their source and target nodes before adding them to the relation set. This ensures all edges are treated as undirected and are properly deduplicated.
- The `initialize_storages` method must be explicitly called after LightRAG creation.
The `finalize_storages` method should be called before LightRAG destyoyed.
- Added explicit data migration check
- Introduces an index mapping documents to their corresponding entities and relations. This significantly speeds up `adelete_by_doc_id` by replacing slow graph traversal with a fast key-value lookup.
- Refactors the ingestion pipeline (`merge_nodes_and_edges`) to populate this new index. Adds a one-time data migration script to backfill the index for existing data.
Replace regex-based JSON extraction with json-repair for better handling of malformed LLM responses. Remove deprecated JSON parsing utilities and clean up keyword_extraction parameter across LLM providers.
- Remove locate_json_string_body_from_string() and convert_response_to_json()
- Use json-repair.loads() in extract_keywords_only() for robust parsing
- Clean up LLM interfaces and remove unused parameters
- Add json-repair dependency
- Add metadata field to DocProcessingStatus with start_time and end_time tracking
- Record processing timestamps using Unix time format (seconds precision)
- Update all storage backends (JSON, MongoDB, Redis, PostgreSQL) for new field support
- Maintain backward compatibility with default values for existing data
- Add error_msg field for better error tracking during document processing
- Add ollama_server_infos attribute to LightRAG class with default initialization
- Move default values to constants.py for centralized configuration
- Refactor OllamaServerInfos class with property accessors and CLI support
- Update OllamaAPI to get configuration through rag object instead of direct import
- Add command line arguments for simulated model name and tag
- Fix type imports to avoid circular dependencies
This commit renames the parameter 'llm_model_max_token_size' to 'summary_max_tokens' for better clarity, as it specifically controls the token limit for entity relation summaries.
- Add nano-vectordb and networkx to pyproject.toml dependencies
- Replace dynamic imports with direct imports for 4 default storage implementations
- Improve startup performance while maintaining backward compatibility
This commit refactors query parameter management by consolidating settings like `top_k`, token limits, and thresholds into the `LightRAG` class, and consistently sourcing parameters from a single location.
- Remove MAX_TOKEN_SUMMARY parameter and related configurations
- Eliminate forced token-based truncation in entity/relationship descriptions
- Switch to fragment-count based summarization logic using FORCE_LLM_SUMMARY_ON_MERGE
- Update FORCE_LLM_SUMMARY_ON_MERGE default from 6 to 4 for better summarization
- Clean up documentation, environment examples, and API display code
- Preserve backward compatibility by graceful parameter removal
This change resolves issues where LLMs were forcibly truncating entity relationship
descriptions mid-sentence, leading to incomplete and potentially inaccurate knowledge
graph content. The new approach allows LLMs to generate complete descriptions while
still providing summarization when multiple fragments need to be merged.
Breaking Change: None - parameter removal is backward compatible
Fixes: Entity relationship description truncation issues
- Refactor the trigger condition for LLM-based summarization of entities and relations. Instead of relying on character length, the summary is now triggered when the number of merged description fragments exceeds a configured threshold. This provides a more robust and logical condition for consolidation.
- Introduce the `OLLAMA_NUM_CTX` environment variable to explicitly configure the context window size (`num_ctx`) for Ollama models. This decouples the model's context length from the `MAX_TOKENS` parameter, which is now specifically used to limit input for summary generation, making the configuration clearer and more flexible.
- Updated `README` files, `env.example`, and default values to reflect these changes.
The document processing pipeline would crash with a TypeError when a document was submitted as raw text via the API, as the file_path attribute would be None. This change adds a check to handle the None case gracefully, preventing the crash and allowing text-based documents to be indexed correctly.
