gmakstutis/LightRAG
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
5664 commits 5 branches 72 tags 705 MiB
Commit graph

3619 commits

Author SHA1 Message Date
yangdx
9262f66d13 Bump API version to 0255 2025-11-17 17:07:18 +08:00
yangdx
90f52acf0c Fix linting 2025-11-17 12:28:53 +08:00
yangdx
c13f9116d9 Add embedding dimension validation to EmbeddingFunc wrapper 
• Validate total elements divisibility
• Check vector count matches input count
• Raise clear error messages on mismatch
• Ensure embedding output correctness
• Add docstring for EmbeddingFunc class
 2025-11-17 12:26:54 +08:00
yangdx
b5589ce4d5 Merge branch 'main' into embedding-limit 2025-11-15 01:10:34 +08:00
yangdx
4343db753a Add macOS fork safety check for Gunicorn multi-worker mode 
• Check OBJC_DISABLE_INITIALIZE_FORK_SAFETY
• Prevent NumPy/Accelerate crashes
• Show detailed error message
• Provide multiple fix options
• Exit early if misconfigured
 2025-11-15 00:58:23 +08:00
yangdx
5dec4deac7 Improve embedding config priority and add debug logging 
• Fix embedding_dim priority logic
• Add final config logging
 2025-11-14 23:22:44 +08:00
yangdx
de4412dd40 Fix embedding token limit initialization order 
* Capture max_token_size before decorator
* Apply wrapper after capturing attribute
* Prevent decorator from stripping dataclass
* Ensure token limit is properly set
 2025-11-14 22:56:03 +08:00
yangdx
963a0a5db1 Refactor embedding function creation with proper attribute inheritance 
- Extract max_token_size from providers
- Avoid double-wrapping EmbeddingFunc
- Improve configuration priority logic
- Add comprehensive debug logging
- Return complete EmbeddingFunc instance
 2025-11-14 22:29:08 +08:00
yangdx
39b49e92ff Convert embedding_token_limit from property to field with __post_init__ 
• Remove property decorator
• Add field with init=False
• Set value in __post_init__ method
• embedding_token_limit is now in config dictionary
 2025-11-14 20:58:41 +08:00
yangdx
ab4d7ac2b0 Add configurable embedding token limit with validation 
- Add EMBEDDING_TOKEN_LIMIT env var
- Set max_token_size on embedding func
- Add token limit property to LightRAG
- Validate summary length vs limit
- Log warning when limit exceeded
 2025-11-14 19:28:36 +08:00
yangdx
680e36c6eb Improve Bedrock error handling with retry logic and custom exceptions 
• Add specific exception types
• Implement proper retry mechanism
• Better error classification
• Enhanced logging and validation
• Enable embedding retry decorator
 2025-11-14 18:51:41 +08:00
yangdx
05852e1ab2 Add max_token_size parameter to embedding function decorators 
- Add max_token_size=8192 to all embed funcs
- Move siliconcloud to deprecated folder
- Import wrap_embedding_func_with_attrs
- Update EmbeddingFunc docstring
- Fix langfuse import type annotation
 2025-11-14 18:41:43 +08:00
yangdx
4401f86f07 Refactor exception handling in MemgraphStorage label methods 2025-11-14 11:01:26 +08:00
yangdx
1ccef2b932 Fix null reference errors in graph database error handling 
- Initialize result vars to None
- Add null checks before consume calls
- Prevent crashes in except blocks
- Apply fix to both Neo4J and Memgraph
 2025-11-14 10:39:04 +08:00
yangdx
c164c8f631 Merge branch 'main' of github.com:HKUDS/LightRAG 2025-11-13 20:42:47 +08:00
yangdx
1889301597 Merge branch 'feat/add_cloud_ollama_support' 2025-11-13 20:41:58 +08:00
yangdx
77ad906d3a Improve error handling and logging in cloud model detection 2025-11-13 20:41:44 +08:00
yangdx
cc031a3db9 Add macOS compatibility check for DOCLING with multi-worker Gunicorn 2025-11-13 19:18:04 +08:00
LacombeLouis
844537e378 Add a better regex 2025-11-13 12:17:51 +01:00
yangdx
a24d8181c2 Improve docling integration with macOS compatibility and CLI flag 
- Add --docling CLI flag for easier setup
- Add numpy version constraints
- Exclude docling on macOS (fork-safety)
 2025-11-13 18:58:09 +08:00
yangdx
746c069ab0 Implement lazy configuration initialization for API server 
• Add lazy config initialization
• Maintain backward compatibility
• Support programmatic usage
• Add gunicorn dependency
• Explicit config in entry points
 2025-11-13 15:28:05 +08:00
yangdx
4b31942e2a refactor: move document deps to api group, remove dynamic imports 
- Merge offline-docs into api extras
- Remove pipmaster dynamic installs
- Add async document processing
- Pre-check docling availability
- Update offline deployment docs
 2025-11-13 13:34:09 +08:00
yangdx
c230d1a28d Replace asyncio.iscoroutine with inspect.isawaitable for better detection 2025-11-13 12:56:01 +08:00
yangdx
297e460740 Merge branch 'main' into tongda/main 2025-11-13 12:37:37 +08:00
yangdx
940bec0b31 Support async chunking functions in LightRAG processing pipeline 
- Add Awaitable and Union type imports
- Update chunking_func type annotation
- Handle coroutine results with await
- Add return type validation
- Update docstring for async support
 2025-11-13 12:37:15 +08:00
Louis Lacombe
f7432a260e Add support for environment variable fallback for API key and default host for cloud models 2025-11-12 16:11:05 +00:00
yangdx
70cc2419f2 Fix empty dict handling after JSON sanitization 
• Replace truthy checks with `is not None`
• Handle empty dict edge case properly
• Prevent data reload failures
• Add comprehensive test coverage
• Fix JsonKVStorage and DocStatusStorage
 2025-11-12 16:40:57 +08:00
yangdx
dcf1d28681 Fix migration to reload sanitized data and prevent memory corruption 
• Reload cleaned data after sanitization
• Update shared memory with clean data
• Add specific surrogate char tests
• Test migration sanitization flow
• Prevent dirty data in memory
 2025-11-12 16:16:28 +08:00
yangdx
6de4123f74 Optimize JSON string sanitization with precompiled regex and zero-copy 
- Precompile regex pattern at module level
- Zero-copy path for clean strings
- Use C-level regex for performance
- Remove deprecated _sanitize_json_data
- Fast detection for common case
 2025-11-12 15:42:07 +08:00
yangdx
777c987371 Optimize JSON write with fast/slow path to reduce memory usage 
- Fast path for clean data (no sanitization)
- Slow path sanitizes during encoding
- Reload shared memory after sanitization
- Custom encoder avoids deep copies
- Comprehensive test coverage
 2025-11-12 13:48:56 +08:00
yangdx
8c07c91833 Remove deprecated response_type parameter from query settings 
- Bump API version to 0254
- Remove response format UI controls
- Hard-code response_type in query params
- Add migration for version 19
- Clean up settings store structure
 2025-11-12 12:19:30 +08:00
yangdx
f28a0c25b1 Improve JSON data sanitization to handle tuples and dict keys 
- Sanitize dictionary keys
- Preserve tuple types
- Handle nested structures better
 2025-11-12 00:50:18 +08:00
yangdx
6918a88f92 Add specialized JSON string sanitizer to prevent UTF-8 encoding errors 
• Remove surrogate characters (U+D800-DFFF)
• Filter Unicode non-characters
• Direct char-by-char filtering
 2025-11-12 00:38:47 +08:00
yangdx
d1f4b6e515 Add data sanitization to JSON writing to prevent UTF-8 encoding errors 
• Add _sanitize_json_data helper function
• Recursively clean strings in data
• Sanitize before JSON serialization
• Prevent encoding-related crashes
• Use existing sanitize_text_for_encoding
 2025-11-12 00:11:13 +08:00
yangdx
fdcb4d0b6d Replace PyPDF2 with pypdf for PDF processing 
- Update import from PyPDF2 to pypdf
- Change dependency to pypdf>=6.1.0
- Update all requirements files
- Remove PyPDF2 from lock file
- Use modern pypdf library
 2025-11-11 01:38:09 +08:00
Tong Da
245df75d9c easier version: detect chunking_func result is coroutine or not 2025-11-10 20:49:50 +08:00
yangdx
913fa1e415 Add concurrency warning for JsonKVStorage in cleanup tool 2025-11-09 23:04:04 +08:00
Tong Da
d137ba5843 support async chunking func to improve processing performance when a heavy chunking_func is passed in by user 2025-11-09 14:52:42 +08:00
yangdx
1f9d0735c3 Bump API version to 0253 2025-11-09 14:42:22 +08:00
yangdx
37b7118901 Fix table alignment and add validation for empty cleanup selections 2025-11-09 14:17:56 +08:00
yangdx
1485cb82e9 Add LLM query cache cleanup tool for KV storage backends 
- Interactive cleanup workflow
- Supports all KV storage types
- Batch deletion with progress
- Comprehensive error reporting
- Preserves workspace isolation
 2025-11-09 13:37:33 +08:00
yangdx
2f16065256 Refactor keyword_extraction from kwargs to explicit parameter 
• Add keyword_extraction param to functions
• Remove kwargs.pop() calls
• Update function signatures
• Improve parameter documentation
• Make parameter handling consistent
 2025-11-09 12:02:17 +08:00
yangdx
88ab73f6ae HotFix: Restore streaming response in OpenAI LLM 
The stream and timeout parameters were moved from **kwargs to explicit
parameters in a previous commit, but were not being passed to the OpenAI
API, causing streaming responses to fail and fall back to non-streaming
mode.Fixes the issue where stream=True was being silently ignored, resulting
in unexpected non-streaming behavior.
 2025-11-09 11:52:26 +08:00
yangdx
7bc6ccea19 Add uv package manager support to installation docs 2025-11-09 04:31:07 +08:00
yangdx
754d2ad297 Add documentation for LLM cache migration between storage types 2025-11-09 00:41:07 +08:00
yangdx
a75efb06dc Fix: prevent source data corruption by target upsert function 
• Prevent mutations bugs by using copy() when storing cache values
• Protect filtered cache data and ensure batch data isolation
 2025-11-09 00:02:19 +08:00
yangdx
987bc09cab Update LLM cache migration docs and improve UX prompts 2025-11-08 23:48:19 +08:00
yangdx
1a91bcdb5f Improve storage config validation and add config.ini fallback support 
• Add MongoDB env requirements
• Support config.ini fallback
• Warn on missing env vars
• Check available storage count
• Show config source info
 2025-11-08 22:48:49 +08:00
yangdx
57ee7d5ac8 Merge branch 'main' into llm-cache-migrate 2025-11-08 22:15:46 +08:00
yangdx
3d9de5ed03 feat: improve Gemini client error handling and retry logic 
• Add google-api-core dependency
• Add specific exception handling
• Create InvalidResponseError class
• Update retry decorators
• Fix empty response handling
 2025-11-08 22:10:09 +08:00