fix linting

examples/lightrag_api_ollama_demo.py

@@ -36,7 +36,10 @@ async def init():
         llm_model_name="gemma2:9b",
         llm_model_max_async=4,
         llm_model_max_token_size=8192,
-        llm_model_kwargs={"host": "http://localhost:11434", "options": {"num_ctx": 8192}},
+        llm_model_kwargs={
+            "host": "http://localhost:11434",
+            "options": {"num_ctx": 8192},
+        },
         embedding_func=EmbeddingFunc(
             embedding_dim=768,
             max_token_size=8192,
@@ -64,6 +67,8 @@ async def lifespan(app: FastAPI):
 app = FastAPI(
     title="LightRAG API", description="API for RAG operations", lifespan=lifespan
 )
+
+
 # Data models
 class QueryRequest(BaseModel):
     query: str

examples/lightrag_api_openai_compatible_demo.py

@@ -75,7 +75,7 @@ async def get_embedding_dim():
 # Initialize RAG instance
 async def init():
     embedding_dimension = await get_embedding_dim()
-    
+
     rag = LightRAG(
         working_dir=WORKING_DIR,
         llm_model_func=llm_model_func,
@@ -88,9 +88,10 @@ async def init():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 @asynccontextmanager
 async def lifespan(app: FastAPI):
     global rag

examples/lightrag_bedrock_demo.py

@@ -21,6 +21,7 @@ WORKING_DIR = "./dickens"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -33,9 +34,10 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 def main():
     rag = asyncio.run(initialize_rag())
 
@@ -47,5 +49,7 @@ def main():
         print(f"| {mode.capitalize()} |")
         print("+-" + "-" * len(mode) + "-+\n")
         print(
-            rag.query("What are the top themes in this story?", param=QueryParam(mode=mode))
+            rag.query(
+                "What are the top themes in this story?", param=QueryParam(mode=mode)
+            )
         )

examples/lightrag_gemini_demo.py

@@ -12,6 +12,7 @@ from lightrag.kg.shared_storage import initialize_pipeline_status
 
 import asyncio
 import nest_asyncio
+
 # Apply nest_asyncio to solve event loop issues
 nest_asyncio.apply()
 
@@ -79,9 +80,10 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -98,5 +100,6 @@ def main():
 
     print(response)
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_hf_demo.py

@@ -16,6 +16,7 @@ WORKING_DIR = "./dickens"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -41,6 +42,7 @@ async def initialize_rag():
 
     return rag
 
+
 def main():
     rag = asyncio.run(initialize_rag())
 
@@ -49,23 +51,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_llamaindex_direct_demo.py

@@ -83,7 +83,7 @@ async def get_embedding_dim():
 
 async def initialize_rag():
     embedding_dimension = await get_embedding_dim()
-    
+
     rag = LightRAG(
         working_dir=WORKING_DIR,
         llm_model_func=llm_model_func,
@@ -96,7 +96,7 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
 
@@ -111,23 +111,32 @@ def main():
     # Test different query modes
     print("\nNaive Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     print("\nLocal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     print("\nGlobal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     print("\nHybrid Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_llamaindex_litellm_demo.py

@@ -86,7 +86,7 @@ async def get_embedding_dim():
 
 async def initialize_rag():
     embedding_dimension = await get_embedding_dim()
-    
+
     rag = LightRAG(
         working_dir=WORKING_DIR,
         llm_model_func=llm_model_func,
@@ -99,7 +99,7 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
 
@@ -114,23 +114,32 @@ def main():
     # Test different query modes
     print("\nNaive Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     print("\nLocal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     print("\nGlobal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     print("\nHybrid Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_lmdeploy_demo.py

@@ -41,6 +41,7 @@ async def lmdeploy_model_complete(
         **kwargs,
     )
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -63,9 +64,10 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -77,23 +79,32 @@ def main():
     # Test different query modes
     print("\nNaive Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     print("\nLocal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     print("\nGlobal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     print("\nHybrid Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
-    main()
+    main()

examples/lightrag_nvidia_demo.py

@@ -97,6 +97,7 @@ async def test_funcs():
 
 # asyncio.run(test_funcs())
 
+
 async def initialize_rag():
     embedding_dimension = await get_embedding_dim()
     print(f"Detected embedding dimension: {embedding_dimension}")
@@ -117,8 +118,10 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
+
+
 async def main():
     try:
         # Initialize RAG instance

examples/lightrag_ollama_age_demo.py

@@ -27,6 +27,7 @@ os.environ["AGE_POSTGRES_HOST"] = "localhost"
 os.environ["AGE_POSTGRES_PORT"] = "5455"
 os.environ["AGE_GRAPH_NAME"] = "dickens"
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -34,7 +35,10 @@ async def initialize_rag():
         llm_model_name="llama3.1:8b",
         llm_model_max_async=4,
         llm_model_max_token_size=32768,
-        llm_model_kwargs={"host": "http://localhost:11434", "options": {"num_ctx": 32768}},
+        llm_model_kwargs={
+            "host": "http://localhost:11434",
+            "options": {"num_ctx": 32768},
+        },
         embedding_func=EmbeddingFunc(
             embedding_dim=768,
             max_token_size=8192,
@@ -47,13 +51,15 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 async def print_stream(stream):
     async for chunk in stream:
         print(chunk, end="", flush=True)
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -65,22 +71,30 @@ def main():
     # Test different query modes
     print("\nNaive Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     print("\nLocal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     print("\nGlobal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     print("\nHybrid Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
     # stream response
@@ -94,5 +108,6 @@ def main():
     else:
         print(resp)
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_ollama_demo.py

@@ -17,6 +17,7 @@ logging.basicConfig(format="%(levelname)s:%(message)s", level=logging.INFO)
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -24,7 +25,10 @@ async def initialize_rag():
         llm_model_name="gemma2:2b",
         llm_model_max_async=4,
         llm_model_max_token_size=32768,
-        llm_model_kwargs={"host": "http://localhost:11434", "options": {"num_ctx": 32768}},
+        llm_model_kwargs={
+            "host": "http://localhost:11434",
+            "options": {"num_ctx": 32768},
+        },
         embedding_func=EmbeddingFunc(
             embedding_dim=768,
             max_token_size=8192,
@@ -36,13 +40,15 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 async def print_stream(stream):
     async for chunk in stream:
         print(chunk, end="", flush=True)
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -54,22 +60,30 @@ def main():
     # Test different query modes
     print("\nNaive Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     print("\nLocal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     print("\nGlobal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     print("\nHybrid Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
     # stream response
@@ -83,5 +97,6 @@ def main():
     else:
         print(resp)
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_ollama_gremlin_demo.py

@@ -32,6 +32,7 @@ os.environ["GREMLIN_TRAVERSE_SOURCE"] = "g"
 os.environ["GREMLIN_USER"] = ""
 os.environ["GREMLIN_PASSWORD"] = ""
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -39,7 +40,10 @@ async def initialize_rag():
         llm_model_name="llama3.1:8b",
         llm_model_max_async=4,
         llm_model_max_token_size=32768,
-        llm_model_kwargs={"host": "http://localhost:11434", "options": {"num_ctx": 32768}},
+        llm_model_kwargs={
+            "host": "http://localhost:11434",
+            "options": {"num_ctx": 32768},
+        },
         embedding_func=EmbeddingFunc(
             embedding_dim=768,
             max_token_size=8192,
@@ -52,13 +56,15 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 async def print_stream(stream):
     async for chunk in stream:
         print(chunk, end="", flush=True)
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -70,22 +76,30 @@ def main():
     # Test different query modes
     print("\nNaive Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     print("\nLocal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     print("\nGlobal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     print("\nHybrid Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
     # stream response
@@ -99,5 +113,6 @@ def main():
     else:
         print(resp)
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_ollama_neo4j_milvus_mongo_demo.py

@@ -32,6 +32,7 @@ os.environ["MILVUS_USER"] = "root"
 os.environ["MILVUS_PASSWORD"] = "root"
 os.environ["MILVUS_DB_NAME"] = "lightrag"
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -39,7 +40,10 @@ async def initialize_rag():
         llm_model_name="qwen2.5:14b",
         llm_model_max_async=4,
         llm_model_max_token_size=32768,
-        llm_model_kwargs={"host": "http://127.0.0.1:11434", "options": {"num_ctx": 32768}},
+        llm_model_kwargs={
+            "host": "http://127.0.0.1:11434",
+            "options": {"num_ctx": 32768},
+        },
         embedding_func=EmbeddingFunc(
             embedding_dim=1024,
             max_token_size=8192,
@@ -54,9 +58,10 @@ async def initialize_rag():
 
     await rag.initialize_storages()
     await initialize_pipeline_status()
-    
+
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -68,23 +73,32 @@ def main():
     # Test different query modes
     print("\nNaive Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     print("\nLocal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     print("\nGlobal Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     print("\nHybrid Search:")
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_openai_compatible_demo.py

@@ -53,6 +53,7 @@ async def test_funcs():
 
 # asyncio.run(test_funcs())
 
+
 async def initialize_rag():
     embedding_dimension = await get_embedding_dim()
     print(f"Detected embedding dimension: {embedding_dimension}")
@@ -71,6 +72,8 @@ async def initialize_rag():
     await initialize_pipeline_status()
 
     return rag
+
+
 async def main():
     try:
         # Initialize RAG instance

examples/lightrag_openai_compatible_demo_embedding_cache.py

@@ -53,6 +53,7 @@ async def test_funcs():
 
 # asyncio.run(test_funcs())
 
+
 async def initialize_rag():
     embedding_dimension = await get_embedding_dim()
     print(f"Detected embedding dimension: {embedding_dimension}")
@@ -76,6 +77,7 @@ async def initialize_rag():
 
     return rag
 
+
 async def main():
     try:
         # Initialize RAG instance

examples/lightrag_openai_compatible_stream_demo.py

@@ -15,6 +15,8 @@ if not os.path.exists(WORKING_DIR):
 print(f"WorkingDir: {WORKING_DIR}")
 
 api_key = "empty"
+
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -40,11 +42,13 @@ async def initialize_rag():
 
     return rag
 
+
 async def print_stream(stream):
     async for chunk in stream:
         if chunk:
             print(chunk, end="", flush=True)
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -63,6 +67,6 @@ def main():
     else:
         print(resp)
 
+
 if __name__ == "__main__":
     main()
-

examples/lightrag_openai_demo.py

@@ -9,6 +9,7 @@ WORKING_DIR = "./dickens"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -22,6 +23,7 @@ async def initialize_rag():
 
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -31,24 +33,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()
-

examples/lightrag_openai_mongodb_graph_demo.py

@@ -76,23 +76,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
-    main()
+    main()

examples/lightrag_openai_neo4j_milvus_redis_demo.py

@@ -50,6 +50,8 @@ embedding_func = EmbeddingFunc(
         texts, embed_model="shaw/dmeta-embedding-zh", host="http://117.50.173.35:11434"
     ),
 )
+
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -79,23 +81,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_oracle_demo.py

@@ -64,6 +64,7 @@ async def get_embedding_dim():
     embedding_dim = embedding.shape[1]
     return embedding_dim
 
+
 async def initialize_rag():
     # Detect embedding dimension
     embedding_dimension = await get_embedding_dim()
@@ -102,6 +103,7 @@ async def initialize_rag():
 
     return rag
 
+
 async def main():
     try:
         # Initialize RAG instance

examples/lightrag_siliconcloud_demo.py

@@ -47,6 +47,7 @@ async def test_funcs():
 
 asyncio.run(test_funcs())
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -71,24 +72,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()
-

examples/lightrag_tidb_demo.py

@@ -55,6 +55,7 @@ async def get_embedding_dim():
     embedding_dim = embedding.shape[1]
     return embedding_dim
 
+
 async def initialize_rag():
     # Detect embedding dimension
     embedding_dimension = await get_embedding_dim()
@@ -82,6 +83,7 @@ async def initialize_rag():
 
     return rag
 
+
 async def main():
     try:
         # Initialize RAG instance

examples/lightrag_zhipu_demo.py

@@ -19,6 +19,7 @@ api_key = os.environ.get("ZHIPUAI_API_KEY")
 if api_key is None:
     raise Exception("Please set ZHIPU_API_KEY in your environment")
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -38,6 +39,7 @@ async def initialize_rag():
 
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -47,23 +49,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/lightrag_zhipu_postgres_demo.py

@@ -28,6 +28,7 @@ os.environ["POSTGRES_USER"] = "rag"
 os.environ["POSTGRES_PASSWORD"] = "rag"
 os.environ["POSTGRES_DATABASE"] = "rag"
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -55,8 +56,9 @@ async def initialize_rag():
 
     return rag
 
+
 async def main():
-        # Initialize RAG instance
+    # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
 
     # add embedding_func for graph database, it's deleted in commit 5661d76860436f7bf5aef2e50d9ee4a59660146c

examples/query_keyword_separation_example.py

@@ -80,6 +80,8 @@ async def test_funcs():
 asyncio.run(test_funcs())
 
 embedding_dimension = 3072
+
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -101,7 +103,7 @@ async def initialize_rag():
 async def run_example():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
-    
+
     book1 = open("./book_1.txt", encoding="utf-8")
     book2 = open("./book_2.txt", encoding="utf-8")
 

examples/test.py

@@ -1,4 +1,5 @@
 import os
+import asyncio
 from lightrag import LightRAG, QueryParam
 from lightrag.llm.openai import gpt_4o_mini_complete
 from lightrag.kg.shared_storage import initialize_pipeline_status
@@ -13,6 +14,7 @@ WORKING_DIR = "./dickens"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -35,23 +37,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
-    main()
+    main()

examples/test_chromadb.py

@@ -112,12 +112,13 @@ async def initialize_rag():
             },
         )
 
-
     await rag.initialize_storages()
     await initialize_pipeline_status()
 
     return rag
 
+
+def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
 
@@ -126,23 +127,32 @@ async def initialize_rag():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/test_faiss.py

@@ -58,6 +58,7 @@ async def embedding_func(texts: list[str]) -> np.ndarray:
     embeddings = model.encode(texts, convert_to_numpy=True)
     return embeddings
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -78,8 +79,8 @@ async def initialize_rag():
 
     return rag
 
+
 def main():
-    
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
     # Insert the custom chunks into LightRAG

examples/test_neo4j.py

@@ -15,6 +15,7 @@ WORKING_DIR = "./local_neo4jWorkDir"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -29,6 +30,7 @@ async def initialize_rag():
 
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())
@@ -38,23 +40,32 @@ def main():
 
     # Perform naive search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="naive"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="naive")
+        )
     )
 
     # Perform local search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="local"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="local")
+        )
     )
 
     # Perform global search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="global"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="global")
+        )
     )
 
     # Perform hybrid search
     print(
-        rag.query("What are the top themes in this story?", param=QueryParam(mode="hybrid"))
+        rag.query(
+            "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+        )
     )
 
+
 if __name__ == "__main__":
     main()

examples/test_split_by_character.ipynb

@@ -1,1313 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "id": "4b5690db12e34685",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:40:58.307102Z",
-     "start_time": "2025-01-09T03:40:51.935233Z"
-    }
-   },
-   "outputs": [],
-   "source": [
-    "import os\n",
-    "import logging\n",
-    "import numpy as np\n",
-    "from lightrag import LightRAG, QueryParam\n",
-    "from lightrag.llm.openai import openai_complete_if_cache, openai_embed\n",
-    "from lightrag.utils import EmbeddingFunc\n",
-    "from lightrag.kg.shared_storage import initialize_pipeline_status\n",
-    "import nest_asyncio"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "dd17956ec322b361",
-   "metadata": {},
-   "source": [
-    "#### split by character"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "id": "8c8ee7c061bf9159",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:41:13.961167Z",
-     "start_time": "2025-01-09T03:41:13.958357Z"
-    }
-   },
-   "outputs": [],
-   "source": [
-    "nest_asyncio.apply()\n",
-    "WORKING_DIR = \"../../llm_rag/paper_db/R000088_test1\"\n",
-    "logging.basicConfig(format=\"%(levelname)s:%(message)s\", level=logging.INFO)\n",
-    "if not os.path.exists(WORKING_DIR):\n",
-    "    os.mkdir(WORKING_DIR)\n",
-    "API = os.environ.get(\"DOUBAO_API_KEY\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "id": "a5009d16e0851dca",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:41:16.862036Z",
-     "start_time": "2025-01-09T03:41:16.859306Z"
-    }
-   },
-   "outputs": [],
-   "source": [
-    "async def llm_model_func(\n",
-    "    prompt, system_prompt=None, history_messages=[], keyword_extraction=False, **kwargs\n",
-    ") -> str:\n",
-    "    return await openai_complete_if_cache(\n",
-    "        \"ep-20241218114828-2tlww\",\n",
-    "        prompt,\n",
-    "        system_prompt=system_prompt,\n",
-    "        history_messages=history_messages,\n",
-    "        api_key=API,\n",
-    "        base_url=\"https://ark.cn-beijing.volces.com/api/v3\",\n",
-    "        **kwargs,\n",
-    "    )\n",
-    "\n",
-    "\n",
-    "async def embedding_func(texts: list[str]) -> np.ndarray:\n",
-    "    return await openai_embed(\n",
-    "        texts,\n",
-    "        model=\"ep-20241231173413-pgjmk\",\n",
-    "        api_key=API,\n",
-    "        base_url=\"https://ark.cn-beijing.volces.com/api/v3\",\n",
-    "    )"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 5,
-   "id": "397fcad24ce4d0ed",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:41:24.950307Z",
-     "start_time": "2025-01-09T03:41:24.940353Z"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "INFO:lightrag:Logger initialized for working directory: ../../llm_rag/paper_db/R000088_test1\n",
-      "INFO:lightrag:Load KV llm_response_cache with 0 data\n",
-      "INFO:lightrag:Load KV full_docs with 0 data\n",
-      "INFO:lightrag:Load KV text_chunks with 0 data\n",
-      "INFO:nano-vectordb:Init {'embedding_dim': 4096, 'metric': 'cosine', 'storage_file': '../../llm_rag/paper_db/R000088_test1/vdb_entities.json'} 0 data\n",
-      "INFO:nano-vectordb:Init {'embedding_dim': 4096, 'metric': 'cosine', 'storage_file': '../../llm_rag/paper_db/R000088_test1/vdb_relationships.json'} 0 data\n",
-      "INFO:nano-vectordb:Init {'embedding_dim': 4096, 'metric': 'cosine', 'storage_file': '../../llm_rag/paper_db/R000088_test1/vdb_chunks.json'} 0 data\n",
-      "INFO:lightrag:Loaded document status storage with 0 records\n"
-     ]
-    }
-   ],
-   "source": [
-    "import asyncio\n",
-    "import nest_asyncio\n",
-    "\n",
-    "nest_asyncio.apply()\n",
-    "\n",
-    "async def initialize_rag():\n",
-    "    rag = LightRAG(\n",
-    "        working_dir=WORKING_DIR,\n",
-    "        llm_model_func=llm_model_func,\n",
-    "        embedding_func=EmbeddingFunc(\n",
-    "            embedding_dim=4096, max_token_size=8192, func=embedding_func\n",
-    "        ),\n",
-    "        chunk_token_size=512,\n",
-    "    )\n",
-    "    await rag.initialize_storages()\n",
-    "    await initialize_pipeline_status()\n",
-    "\n",
-    "    return rag\n",
-    "\n",
-    "rag = asyncio.run(initialize_rag())"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "id": "1dc3603677f7484d",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:41:37.947456Z",
-     "start_time": "2025-01-09T03:41:37.941901Z"
-    }
-   },
-   "outputs": [],
-   "source": [
-    "with open(\n",
-    "    \"../../llm_rag/example/R000088/auto/R000088_full_txt.md\", \"r\", encoding=\"utf-8\"\n",
-    ") as f:\n",
-    "    content = f.read()\n",
-    "\n",
-    "\n",
-    "async def embedding_func(texts: list[str]) -> np.ndarray:\n",
-    "    return await openai_embed(\n",
-    "        texts,\n",
-    "        model=\"ep-20241231173413-pgjmk\",\n",
-    "        api_key=API,\n",
-    "        base_url=\"https://ark.cn-beijing.volces.com/api/v3\",\n",
-    "    )\n",
-    "\n",
-    "\n",
-    "async def get_embedding_dim():\n",
-    "    test_text = [\"This is a test sentence.\"]\n",
-    "    embedding = await embedding_func(test_text)\n",
-    "    embedding_dim = embedding.shape[1]\n",
-    "    return embedding_dim"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 7,
-   "id": "6844202606acfbe5",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:41:39.608541Z",
-     "start_time": "2025-01-09T03:41:39.165057Z"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n"
-     ]
-    }
-   ],
-   "source": [
-    "embedding_dimension = await get_embedding_dim()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 8,
-   "id": "d6273839d9681403",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:44:34.295345Z",
-     "start_time": "2025-01-09T03:41:48.324171Z"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "INFO:lightrag:Processing 1 new unique documents\n",
-      "Processing batch 1:   0%|          | 0/1 [00:00<?, ?it/s]INFO:lightrag:Inserting 35 vectors to chunks\n",
-      "\n",
-      "Generating embeddings:   0%|          | 0/2 [00:00<?, ?batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  50%|█████     | 1/2 [00:00<00:00,  1.36batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings: 100%|██████████| 2/2 [00:04<00:00,  2.25s/batch]\u001b[A\n",
-      "\n",
-      "Extracting entities from chunks:   0%|          | 0/35 [00:00<?, ?chunk/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠙ Processed 1 chunks, 1 entities(duplicated), 0 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:   3%|▎         | 1/35 [00:04<02:47,  4.93s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠹ Processed 2 chunks, 2 entities(duplicated), 0 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:   6%|▌         | 2/35 [00:05<01:18,  2.37s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠸ Processed 3 chunks, 9 entities(duplicated), 5 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:   9%|▊         | 3/35 [00:26<05:43, 10.73s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠼ Processed 4 chunks, 16 entities(duplicated), 11 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  11%|█▏        | 4/35 [00:26<03:24,  6.60s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠴ Processed 5 chunks, 24 entities(duplicated), 18 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  14%|█▍        | 5/35 [00:33<03:24,  6.82s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠦ Processed 6 chunks, 35 entities(duplicated), 28 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  17%|█▋        | 6/35 [00:42<03:38,  7.53s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠧ Processed 7 chunks, 47 entities(duplicated), 36 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  20%|██        | 7/35 [00:43<02:28,  5.31s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠇ Processed 8 chunks, 61 entities(duplicated), 49 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  23%|██▎       | 8/35 [00:45<01:52,  4.16s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠏ Processed 9 chunks, 81 entities(duplicated), 49 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠋ Processed 10 chunks, 90 entities(duplicated), 62 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  29%|██▊       | 10/35 [00:46<01:06,  2.64s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠙ Processed 11 chunks, 101 entities(duplicated), 80 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  31%|███▏      | 11/35 [00:52<01:19,  3.31s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠹ Processed 12 chunks, 108 entities(duplicated), 85 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  34%|███▍      | 12/35 [00:54<01:11,  3.12s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠸ Processed 13 chunks, 120 entities(duplicated), 100 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  37%|███▋      | 13/35 [00:59<01:18,  3.55s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠼ Processed 14 chunks, 131 entities(duplicated), 110 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  40%|████      | 14/35 [01:00<00:59,  2.82s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠴ Processed 15 chunks, 143 entities(duplicated), 110 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  43%|████▎     | 15/35 [01:02<00:52,  2.64s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠦ Processed 16 chunks, 162 entities(duplicated), 124 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  46%|████▌     | 16/35 [01:05<00:53,  2.80s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠧ Processed 17 chunks, 174 entities(duplicated), 132 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  49%|████▊     | 17/35 [01:06<00:39,  2.19s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠇ Processed 18 chunks, 185 entities(duplicated), 137 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  51%|█████▏    | 18/35 [01:12<00:53,  3.15s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠏ Processed 19 chunks, 193 entities(duplicated), 149 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  54%|█████▍    | 19/35 [01:18<01:06,  4.14s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠋ Processed 20 chunks, 205 entities(duplicated), 158 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  57%|█████▋    | 20/35 [01:19<00:50,  3.35s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠙ Processed 21 chunks, 220 entities(duplicated), 187 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  60%|██████    | 21/35 [01:27<01:02,  4.47s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠹ Processed 22 chunks, 247 entities(duplicated), 216 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  63%|██████▎   | 22/35 [01:30<00:54,  4.16s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠸ Processed 23 chunks, 260 entities(duplicated), 230 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  66%|██████▌   | 23/35 [01:34<00:48,  4.05s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠼ Processed 24 chunks, 291 entities(duplicated), 253 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  69%|██████▊   | 24/35 [01:38<00:44,  4.03s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠴ Processed 25 chunks, 304 entities(duplicated), 262 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  71%|███████▏  | 25/35 [01:41<00:36,  3.67s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠦ Processed 26 chunks, 313 entities(duplicated), 271 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  74%|███████▍  | 26/35 [01:41<00:24,  2.76s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠧ Processed 27 chunks, 321 entities(duplicated), 283 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  77%|███████▋  | 27/35 [01:47<00:28,  3.52s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠇ Processed 28 chunks, 333 entities(duplicated), 290 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  80%|████████  | 28/35 [01:52<00:28,  4.08s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠏ Processed 29 chunks, 348 entities(duplicated), 307 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  83%|████████▎ | 29/35 [01:59<00:29,  4.88s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠋ Processed 30 chunks, 362 entities(duplicated), 329 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  86%|████████▌ | 30/35 [02:02<00:21,  4.29s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠙ Processed 31 chunks, 373 entities(duplicated), 337 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  89%|████████▊ | 31/35 [02:03<00:13,  3.28s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠹ Processed 32 chunks, 390 entities(duplicated), 369 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  91%|█████████▏| 32/35 [02:03<00:07,  2.55s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠸ Processed 33 chunks, 405 entities(duplicated), 378 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  94%|█████████▍| 33/35 [02:07<00:05,  2.84s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠼ Processed 34 chunks, 435 entities(duplicated), 395 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  97%|█████████▋| 34/35 [02:10<00:02,  2.94s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠴ Processed 35 chunks, 456 entities(duplicated), 440 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks: 100%|██████████| 35/35 [02:23<00:00,  4.10s/chunk]\u001b[A\n",
-      "INFO:lightrag:Inserting entities into storage...\n",
-      "\n",
-      "Inserting entities: 100%|██████████| 324/324 [00:00<00:00, 17456.96entity/s]\n",
-      "INFO:lightrag:Inserting relationships into storage...\n",
-      "\n",
-      "Inserting relationships: 100%|██████████| 427/427 [00:00<00:00, 29956.31relationship/s]\n",
-      "INFO:lightrag:Inserting 324 vectors to entities\n",
-      "\n",
-      "Generating embeddings:   0%|          | 0/11 [00:00<?, ?batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:   9%|▉         | 1/11 [00:00<00:06,  1.48batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  18%|█▊        | 2/11 [00:02<00:11,  1.25s/batch]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  27%|██▋       | 3/11 [00:02<00:06,  1.17batch/s]\u001b[A\n",
-      "Generating embeddings:  36%|███▋      | 4/11 [00:03<00:04,  1.50batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  45%|████▌     | 5/11 [00:03<00:03,  1.78batch/s]\u001b[A\n",
-      "Generating embeddings:  55%|█████▍    | 6/11 [00:03<00:02,  2.01batch/s]\u001b[A\n",
-      "Generating embeddings:  64%|██████▎   | 7/11 [00:04<00:01,  2.19batch/s]\u001b[A\n",
-      "Generating embeddings:  73%|███████▎  | 8/11 [00:04<00:01,  2.31batch/s]\u001b[A\n",
-      "Generating embeddings:  82%|████████▏ | 9/11 [00:04<00:00,  2.41batch/s]\u001b[A\n",
-      "Generating embeddings:  91%|█████████ | 10/11 [00:05<00:00,  2.48batch/s]\u001b[A\n",
-      "Generating embeddings: 100%|██████████| 11/11 [00:05<00:00,  1.91batch/s]\u001b[A\n",
-      "INFO:lightrag:Inserting 427 vectors to relationships\n",
-      "\n",
-      "Generating embeddings:   0%|          | 0/14 [00:00<?, ?batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:   7%|▋         | 1/14 [00:01<00:14,  1.11s/batch]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  14%|█▍        | 2/14 [00:02<00:14,  1.18s/batch]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  21%|██▏       | 3/14 [00:02<00:08,  1.23batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  29%|██▊       | 4/14 [00:03<00:06,  1.56batch/s]\u001b[A\n",
-      "Generating embeddings:  36%|███▌      | 5/14 [00:03<00:04,  1.85batch/s]\u001b[A\n",
-      "Generating embeddings:  43%|████▎     | 6/14 [00:03<00:03,  2.05batch/s]\u001b[A\n",
-      "Generating embeddings:  50%|█████     | 7/14 [00:04<00:03,  2.23batch/s]\u001b[A\n",
-      "Generating embeddings:  57%|█████▋    | 8/14 [00:04<00:02,  2.37batch/s]\u001b[A\n",
-      "Generating embeddings:  64%|██████▍   | 9/14 [00:04<00:02,  2.46batch/s]\u001b[A\n",
-      "Generating embeddings:  71%|███████▏  | 10/14 [00:05<00:01,  2.54batch/s]\u001b[A\n",
-      "Generating embeddings:  79%|███████▊  | 11/14 [00:05<00:01,  2.59batch/s]\u001b[A\n",
-      "Generating embeddings:  86%|████████▌ | 12/14 [00:06<00:00,  2.64batch/s]\u001b[A\n",
-      "Generating embeddings:  93%|█████████▎| 13/14 [00:06<00:00,  2.65batch/s]\u001b[A\n",
-      "Generating embeddings: 100%|██████████| 14/14 [00:06<00:00,  2.05batch/s]\u001b[A\n",
-      "INFO:lightrag:Writing graph with 333 nodes, 427 edges\n",
-      "Processing batch 1: 100%|██████████| 1/1 [02:45<00:00, 165.90s/it]\n"
-     ]
-    }
-   ],
-   "source": [
-    "# rag.insert(content)\n",
-    "rag.insert(content, split_by_character=\"\\n#\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 9,
-   "id": "c4f9ae517151a01d",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:45:11.668987Z",
-     "start_time": "2025-01-09T03:45:11.664744Z"
-    }
-   },
-   "outputs": [],
-   "source": [
-    "prompt1 = \"\"\"你是一名经验丰富的论文分析科学家，你的任务是对一篇英文学术研究论文进行关键信息提取并深入分析。\n",
-    "请按照以下步骤进行分析：\n",
-    "1. 该文献主要研究的问题是什么？\n",
-    "2. 该文献采用什么方法进行分析？\n",
-    "3. 该文献的主要结论是什么？\n",
-    "首先在<分析>标签中，针对每个问题详细分析你的思考过程。然后在<回答>标签中给出所有问题的最终答案。\"\"\""
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 10,
-   "id": "7a6491385b050095",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:45:40.829111Z",
-     "start_time": "2025-01-09T03:45:13.530298Z"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:lightrag:Local query uses 5 entites, 12 relations, 3 text units\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:lightrag:Global query uses 8 entites, 5 relations, 4 text units\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "<分析>\n",
-      "1. **该文献主要研究的问题是什么？**\n",
-      "    - 思考过程：通过浏览论文内容，查找作者明确阐述研究目的的部分。文中多处提及“Our study was performed to explore whether folic acid treatment was associated with cancer outcomes and all-cause mortality after extended follow-up”，表明作者旨在探究叶酸治疗与癌症结局及全因死亡率之间的关系，尤其是在经过长期随访后。\n",
-      "2. **该文献采用什么方法进行分析？**\n",
-      "    - 思考过程：寻找描述研究方法和数据分析过程的段落。文中提到“Survival curves were constructed using the Kaplan-Meier method and differences in survival between groups were analyzed using the log-rank test. Estimates of hazard ratios (HRs) with 95% CIs were obtained by using Cox proportional hazards regression models stratified by trial”，可以看出作者使用了Kaplan-Meier法构建生存曲线、log-rank检验分析组间生存差异以及Cox比例风险回归模型估计风险比等方法。\n",
-      "3. **该文献的主要结论是什么？**\n",
-      "    - 思考过程：定位到论文中总结结论的部分，如“Conclusion Treatment with folic acid plus vitamin $\\mathsf{B}_{12}$ was associated with increased cancer outcomes and all-cause mortality in patients with ischemic heart disease in Norway, where there is no folic acid fortification of foods”，可知作者得出叶酸加维生素$\\mathsf{B}_{12}$治疗与癌症结局和全因死亡率增加有关的结论。\n",
-      "<回答>\n",
-      "1. 该文献主要研究的问题是：叶酸治疗与癌症结局及全因死亡率之间的关系，尤其是在经过长期随访后，叶酸治疗是否与癌症结局和全因死亡率相关。\n",
-      "2. 该文献采用的分析方法包括：使用Kaplan-Meier法构建生存曲线、log-rank检验分析组间生存差异、Cox比例风险回归模型估计风险比等。\n",
-      "3. 该文献的主要结论是：在挪威没有叶酸强化食品的情况下，叶酸加维生素$\\mathsf{B}_{12}$治疗与缺血性心脏病患者的癌症结局和全因死亡率增加有关。\n",
-      "\n",
-      "**参考文献**\n",
-      "- [VD] In2Norwegianhomocysteine-lowering trialsamongpatientswithischemicheart disease, there was a statistically nonsignificantincreaseincancerincidenceinthe groupsassignedtofolicacidtreatment.15,16 Our study was performed to explore whetherfolicacidtreatmentwasassociatedwithcanceroutcomesandall-cause mortality after extended follow-up.\n",
-      "- [VD] Survivalcurveswereconstructedusing theKaplan-Meiermethodanddifferences insurvivalbetweengroupswereanalyzed usingthelog-ranktest.Estimatesofhazard ratios (HRs) with $95\\%$ CIs were obtainedbyusingCoxproportionalhazards regressionmodelsstratifiedbytrial.\n",
-      "- [VD] Conclusion Treatment with folic acid plus vitamin $\\mathsf{B}_{12}$ was associated with increased cancer outcomes and all-cause mortality in patients with ischemic heart disease in Norway, where there is no folic acid fortification of foods.\n"
-     ]
-    }
-   ],
-   "source": [
-    "resp = rag.query(prompt1, param=QueryParam(mode=\"mix\", top_k=5))\n",
-    "print(resp)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "4e5bfad24cb721a8",
-   "metadata": {},
-   "source": [
-    "#### split by character only"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 11,
-   "id": "44e2992dc95f8ce0",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:47:40.988796Z",
-     "start_time": "2025-01-09T03:47:40.982648Z"
-    }
-   },
-   "outputs": [],
-   "source": [
-    "WORKING_DIR = \"../../llm_rag/paper_db/R000088_test2\"\n",
-    "if not os.path.exists(WORKING_DIR):\n",
-    "    os.mkdir(WORKING_DIR)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 12,
-   "id": "62c63385d2d973d5",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:51:39.951329Z",
-     "start_time": "2025-01-09T03:49:15.218976Z"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "INFO:lightrag:Logger initialized for working directory: ../../llm_rag/paper_db/R000088_test2\n",
-      "INFO:lightrag:Load KV llm_response_cache with 0 data\n",
-      "INFO:lightrag:Load KV full_docs with 0 data\n",
-      "INFO:lightrag:Load KV text_chunks with 0 data\n",
-      "INFO:nano-vectordb:Init {'embedding_dim': 4096, 'metric': 'cosine', 'storage_file': '../../llm_rag/paper_db/R000088_test2/vdb_entities.json'} 0 data\n",
-      "INFO:nano-vectordb:Init {'embedding_dim': 4096, 'metric': 'cosine', 'storage_file': '../../llm_rag/paper_db/R000088_test2/vdb_relationships.json'} 0 data\n",
-      "INFO:nano-vectordb:Init {'embedding_dim': 4096, 'metric': 'cosine', 'storage_file': '../../llm_rag/paper_db/R000088_test2/vdb_chunks.json'} 0 data\n",
-      "INFO:lightrag:Loaded document status storage with 0 records\n",
-      "INFO:lightrag:Processing 1 new unique documents\n",
-      "Processing batch 1:   0%|          | 0/1 [00:00<?, ?it/s]INFO:lightrag:Inserting 12 vectors to chunks\n",
-      "\n",
-      "Generating embeddings:   0%|          | 0/1 [00:00<?, ?batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings: 100%|██████████| 1/1 [00:02<00:00,  2.95s/batch]\u001b[A\n",
-      "\n",
-      "Extracting entities from chunks:   0%|          | 0/12 [00:00<?, ?chunk/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠙ Processed 1 chunks, 0 entities(duplicated), 0 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:   8%|▊         | 1/12 [00:03<00:43,  3.93s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠹ Processed 2 chunks, 8 entities(duplicated), 8 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  17%|█▋        | 2/12 [00:29<02:44, 16.46s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠸ Processed 3 chunks, 17 entities(duplicated), 15 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  25%|██▌       | 3/12 [00:30<01:25,  9.45s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠼ Processed 4 chunks, 27 entities(duplicated), 22 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  33%|███▎      | 4/12 [00:39<01:16,  9.52s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠴ Processed 5 chunks, 36 entities(duplicated), 33 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  42%|████▏     | 5/12 [00:40<00:43,  6.24s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠦ Processed 6 chunks, 49 entities(duplicated), 42 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  50%|█████     | 6/12 [00:49<00:43,  7.33s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠧ Processed 7 chunks, 62 entities(duplicated), 65 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  58%|█████▊    | 7/12 [01:05<00:50, 10.05s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠇ Processed 8 chunks, 81 entities(duplicated), 90 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  67%|██████▋   | 8/12 [01:23<00:50, 12.69s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠏ Processed 9 chunks, 99 entities(duplicated), 117 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  75%|███████▌  | 9/12 [01:32<00:34, 11.54s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠋ Processed 10 chunks, 123 entities(duplicated), 140 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  83%|████████▎ | 10/12 [01:48<00:25, 12.79s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠙ Processed 11 chunks, 158 entities(duplicated), 174 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks:  92%|█████████▏| 11/12 [02:03<00:13, 13.50s/chunk]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "⠹ Processed 12 chunks, 194 entities(duplicated), 221 relations(duplicated)\r"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "Extracting entities from chunks: 100%|██████████| 12/12 [02:13<00:00, 11.15s/chunk]\u001b[A\n",
-      "INFO:lightrag:Inserting entities into storage...\n",
-      "\n",
-      "Inserting entities: 100%|██████████| 170/170 [00:00<00:00, 11610.25entity/s]\n",
-      "INFO:lightrag:Inserting relationships into storage...\n",
-      "\n",
-      "Inserting relationships: 100%|██████████| 218/218 [00:00<00:00, 15913.51relationship/s]\n",
-      "INFO:lightrag:Inserting 170 vectors to entities\n",
-      "\n",
-      "Generating embeddings:   0%|          | 0/6 [00:00<?, ?batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  17%|█▋        | 1/6 [00:01<00:05,  1.10s/batch]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  33%|███▎      | 2/6 [00:02<00:04,  1.07s/batch]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  50%|█████     | 3/6 [00:02<00:02,  1.33batch/s]\u001b[A\n",
-      "Generating embeddings:  67%|██████▋   | 4/6 [00:02<00:01,  1.67batch/s]\u001b[A\n",
-      "Generating embeddings:  83%|████████▎ | 5/6 [00:03<00:00,  1.95batch/s]\u001b[A\n",
-      "Generating embeddings: 100%|██████████| 6/6 [00:03<00:00,  1.66batch/s]\u001b[A\n",
-      "INFO:lightrag:Inserting 218 vectors to relationships\n",
-      "\n",
-      "Generating embeddings:   0%|          | 0/7 [00:00<?, ?batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  14%|█▍        | 1/7 [00:01<00:10,  1.74s/batch]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  29%|██▊       | 2/7 [00:02<00:05,  1.04s/batch]\u001b[A\n",
-      "Generating embeddings:  43%|████▎     | 3/7 [00:02<00:02,  1.35batch/s]\u001b[A\n",
-      "Generating embeddings:  57%|█████▋    | 4/7 [00:03<00:01,  1.69batch/s]\u001b[AINFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "\n",
-      "Generating embeddings:  71%|███████▏  | 5/7 [00:03<00:01,  1.96batch/s]\u001b[A\n",
-      "Generating embeddings:  86%|████████▌ | 6/7 [00:03<00:00,  2.17batch/s]\u001b[A\n",
-      "Generating embeddings: 100%|██████████| 7/7 [00:04<00:00,  1.68batch/s]\u001b[A\n",
-      "INFO:lightrag:Writing graph with 174 nodes, 218 edges\n",
-      "Processing batch 1: 100%|██████████| 1/1 [02:24<00:00, 144.69s/it]\n"
-     ]
-    }
-   ],
-   "source": [
-    "rag = LightRAG(\n",
-    "    working_dir=WORKING_DIR,\n",
-    "    llm_model_func=llm_model_func,\n",
-    "    embedding_func=EmbeddingFunc(\n",
-    "        embedding_dim=4096, max_token_size=8192, func=embedding_func\n",
-    "    ),\n",
-    "    chunk_token_size=512,\n",
-    ")\n",
-    "\n",
-    "# rag.insert(content)\n",
-    "rag.insert(content, split_by_character=\"\\n#\", split_by_character_only=True)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 13,
-   "id": "3c7aa9836d8d43c7",
-   "metadata": {
-    "ExecuteTime": {
-     "end_time": "2025-01-09T03:52:37.000418Z",
-     "start_time": "2025-01-09T03:52:09.933584Z"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:lightrag:Local query uses 5 entites, 3 relations, 2 text units\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/embeddings \"HTTP/1.1 200 OK\"\n",
-      "INFO:lightrag:Global query uses 9 entites, 5 relations, 4 text units\n",
-      "INFO:httpx:HTTP Request: POST https://ark.cn-beijing.volces.com/api/v3/chat/completions \"HTTP/1.1 200 OK\"\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "<分析>\n",
-      "- **该文献主要研究的问题是什么？**\n",
-      "    - **思考过程**：通过浏览论文的标题、摘要、引言等部分，寻找关于研究目的和问题的描述。论文标题为“Cancer Incidence and Mortality After Treatment With Folic Acid and Vitamin B12”，摘要中的“Objective”部分明确指出研究目的是“To evaluate effects of treatment with B vitamins on cancer outcomes and all-cause mortality in 2 randomized controlled trials”。因此，可以确定该文献主要研究的问题是评估B族维生素治疗对两项随机对照试验中癌症结局和全因死亡率的影响。\n",
-      "- **该文献采用什么方法进行分析？**\n",
-      "    - **思考过程**：在论文的“METHODS”部分详细描述了研究方法。文中提到这是一个对两项随机、双盲、安慰剂对照临床试验（Norwegian Vitamin [NORVIT] trial和Western Norway B Vitamin Intervention Trial [WENBIT]）数据的联合分析，并进行了观察性的试验后随访。具体包括对参与者进行分组干预（不同剂量的叶酸、维生素B12、维生素B6或安慰剂），收集临床信息和血样，分析循环B族维生素、同型半胱氨酸和可替宁等指标，并进行基因分型等，还涉及到多种统计分析方法，如计算预期癌症发生率、构建生存曲线、进行Cox比例风险回归模型分析等。\n",
-      "- **该文献的主要结论是什么？**\n",
-      "    - **思考过程**：在论文的“Results”和“Conclusion”部分寻找主要结论。研究结果表明，在治疗期间，接受叶酸加维生素B12治疗的参与者血清叶酸浓度显著增加，且在后续随访中，该组癌症发病率、癌症死亡率和全因死亡率均有所上升，主要是肺癌发病率增加，而维生素B6治疗未显示出显著影响。结论部分明确指出“Treatment with folic acid plus vitamin $\\mathsf{B}_{12}$ was associated with increased cancer outcomes and all-cause mortality in patients with ischemic heart disease in Norway, where there is no folic acid fortification of foods”。\n",
-      "</分析>\n",
-      "\n",
-      "<回答>\n",
-      "- **主要研究问题**：评估B族维生素治疗对两项随机对照试验中癌症结局和全因死亡率的影响。\n",
-      "- **研究方法**：采用对两项随机、双盲、安慰剂对照临床试验（Norwegian Vitamin [NORVIT] trial和Western Norway B Vitamin Intervention Trial [WENBIT]）数据的联合分析，并进行观察性的试验后随访，涉及分组干预、多种指标检测以及多种统计分析方法。\n",
-      "- **主要结论**：在挪威（食品中未添加叶酸），对于缺血性心脏病患者，叶酸加维生素B12治疗与癌症结局和全因死亡率的增加有关，而维生素B6治疗未显示出显著影响。\n",
-      "\n",
-      "**参考文献**\n",
-      "- [VD] Cancer Incidence and Mortality After Treatment With Folic Acid and Vitamin B12\n",
-      "- [VD] METHODS Study Design, Participants, and Study Intervention\n",
-      "- [VD] RESULTS\n",
-      "- [VD] Conclusion\n",
-      "- [VD] Objective To evaluate effects of treatment with B vitamins on cancer outcomes and all-cause mortality in 2 randomized controlled trials.\n"
-     ]
-    }
-   ],
-   "source": [
-    "resp = rag.query(prompt1, param=QueryParam(mode=\"mix\", top_k=5))\n",
-    "print(resp)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "7ba6fa79a2550d10",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 2
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython2",
-   "version": "2.7.6"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 5
-}

examples/vram_management_demo.py

@@ -14,6 +14,7 @@ TEXT_FILES_DIR = "/llm/mt"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     # Initialize LightRAG
     rag = LightRAG(
@@ -31,6 +32,7 @@ async def initialize_rag():
 
     return rag
 
+
 # Read all .txt files from the TEXT_FILES_DIR directory
 texts = []
 for filename in os.listdir(TEXT_FILES_DIR):
@@ -82,7 +84,8 @@ def main():
     try:
         print(
             rag.query(
-                "What are the top themes in this story?", param=QueryParam(mode="global")
+                "What are the top themes in this story?",
+                param=QueryParam(mode="global"),
             )
         )
     except Exception as e:
@@ -91,18 +94,17 @@ def main():
     try:
         print(
             rag.query(
-                "What are the top themes in this story?", param=QueryParam(mode="hybrid")
+                "What are the top themes in this story?",
+                param=QueryParam(mode="hybrid"),
             )
         )
     except Exception as e:
         print(f"Error performing hybrid search: {e}")
 
-
     # Function to clear VRAM resources
     def clear_vram():
         os.system("sudo nvidia-smi --gpu-reset")
 
-
     # Regularly clear VRAM to prevent overflow
     clear_vram_interval = 3600  # Clear once every hour
     start_time = time.time()
@@ -114,5 +116,6 @@ def main():
             start_time = current_time
         time.sleep(60)  # Check the time every minute
 
+
 if __name__ == "__main__":
     main()

reproduce/Step_1.py

@@ -31,6 +31,7 @@ WORKING_DIR = f"../{cls}"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(working_dir=WORKING_DIR)
 
@@ -39,6 +40,7 @@ async def initialize_rag():
 
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())

reproduce/Step_1_openai_compatible.py

@@ -62,6 +62,7 @@ WORKING_DIR = f"../{cls}"
 if not os.path.exists(WORKING_DIR):
     os.mkdir(WORKING_DIR)
 
+
 async def initialize_rag():
     rag = LightRAG(
         working_dir=WORKING_DIR,
@@ -76,6 +77,7 @@ async def initialize_rag():
 
     return rag
 
+
 def main():
     # Initialize RAG instance
     rag = asyncio.run(initialize_rag())