fix rag documentation

LazyAGI · Nov 29, 2024 · 5bde304 · 5bde304
1 parent ad0e9c5
commit 5bde304
Show file tree

Hide file tree

Showing 3 changed files with 22 additions and 21 deletions.
diff --git a/docs/en/Cookbook/rag.md b/docs/en/Cookbook/rag.md
@@ -120,7 +120,7 @@ Now that we have two different retrieval functions returning results in differen
 # Part8
 
 reranker = Reranker(name="ModuleReranker",
-                    model="bge-reranker-large",
+                    model=lazyllm.OnlineEmbeddingModule(type="rerank"),
                     topk=1)
 ```
 
@@ -179,7 +179,7 @@ retriever2 = lazyllm.Retriever(doc=documents,
 # Part8
 
 reranker = lazyllm.Reranker(name='ModuleReranker',
-                            model="bge-reranker-large",
+                            model=lazyllm.OnlineEmbeddingModule(type="rerank"),
                             topk=1)
 
 # Part3
@@ -251,7 +251,7 @@ with lazyllm.pipeline() as ppl:
                                            topk=3)
 
     ppl.reranker = lazyllm.Reranker(name='ModuleReranker',
-                                    model="bge-reranker-large",
+                                    model=lazyllm.OnlineEmbeddingModule(type="rerank"),
                                     topk=1) | bind(query=ppl.input)
 
     ppl.formatter = (

diff --git a/docs/zh/Cookbook/rag.md b/docs/zh/Cookbook/rag.md
@@ -179,7 +179,7 @@ retriever2 = lazyllm.Retriever(doc=documents,
 # Part8
 
 reranker = lazyllm.Reranker(name='ModuleReranker',
-                            model="bge-reranker-large",
+                            model=lazyllm.OnlineEmbeddingModule(type="rerank"),
                             topk=1)
 
 # Part3
@@ -250,7 +250,7 @@ with lazyllm.pipeline() as ppl:
                                            topk=3)
 
     ppl.reranker = lazyllm.Reranker(name='ModuleReranker',
-                                    model="bge-reranker-large",
+                                    model=lazyllm.OnlineEmbeddingModule(type="rerank"),
                                     topk=1) | bind(query=ppl.input)
 
     ppl.formatter = (

diff --git a/lazyllm/docs/tools.py b/lazyllm/docs/tools.py
@@ -290,12 +290,12 @@
 >>> import lazyllm
 >>> from lazyllm.tools import Document, Reranker, Retriever
 >>> m = lazyllm.OnlineEmbeddingModule()
->>> documents = Document(dataset_path='rag_master', embed=m, manager=False)
+>>> documents = Document(dataset_path='/path/to/user/data', embed=m, manager=False)
 >>> retriever = Retriever(documents, group_name='CoarseChunk', similarity='bm25', similarity_cut_off=0.01, topk=6)
->>> reranker = Reranker(name='ModuleReranker', model='bg-reranker-large', topk=1)
+>>> reranker = Reranker(name='ModuleReranker', model='bge-reranker-large', topk=1)
 >>> ppl = lazyllm.ActionModule(retriever, reranker)
 >>> ppl.start()
->>> print(ppl("query"))
+>>> print(ppl("user query"))
 ''')
 
 # ---------------------------------------------------------------------------- #
@@ -346,19 +346,19 @@
 >>> import lazyllm
 >>> from lazyllm.tools import Retriever, Document, SentenceSplitter
 >>> m = lazyllm.OnlineEmbeddingModule()
->>> documents = Document(dataset_path='your_doc_path', embed=m, manager=False)
+>>> documents = Document(dataset_path='/path/to/user/data', embed=m, manager=False)
 >>> rm = Retriever(documents, group_name='CoarseChunk', similarity='bm25', similarity_cut_off=0.01, topk=6)
 >>> rm.start()
->>> print(rm("query"))
+>>> print(rm("user query"))
 >>> m1 = lazyllm.TrainableModule('bge-large-zh-v1.5').start()
->>> document1 = Document(dataset_path='your_doc_path', embed={'online':m , 'local': m1}, manager=False)
+>>> document1 = Document(dataset_path='/path/to/user/data', embed={'online':m , 'local': m1}, manager=False)
 >>> document1.create_node_group(name='sentences', transform=SentenceSplitter, chunk_size=1024, chunk_overlap=100)
 >>> retriever = Retriever(document1, group_name='sentences', similarity='cosine', similarity_cut_off=0.4, embed_keys=['local'], topk=3)
->>> print(retriever("query"))
->>> document2 = Document(dataset_path='your_doc_path', embed={'online':m , 'local': m1}, manager=False)
+>>> print(retriever("user query"))
+>>> document2 = Document(dataset_path='/path/to/user/data', embed={'online':m , 'local': m1}, manager=False)
 >>> document2.create_node_group(name='sentences', transform=SentenceSplitter, chunk_size=512, chunk_overlap=50)
 >>> retriever2 = Retriever([document1, document2], group_name='sentences', similarity='cosine', similarity_cut_off=0.4, embed_keys=['local'], topk=3)
->>> print(retriever2("query"))
+>>> print(retriever2("user query"))
 ''')
 
 # ---------------------------------------------------------------------------- #
@@ -430,15 +430,16 @@
 
 add_example('LLMParser.transform', '''
 >>> import lazyllm
->>> from lazyllm.tools import LLMParser, TrainableModule
->>> llm = TrainableModule("internlm2-chat-7b")
->>> m = lazyllm.TrainableModule("bge-large-zh-v1.5")
+>>> from lazyllm.tools import LLMParser
+>>> llm = lazyllm.TrainableModule("internlm2-chat-7b").start()
+>>> m = lazyllm.TrainableModule("bge-large-zh-v1.5").start()
 >>> summary_parser = LLMParser(llm, language="en", task_type="summary")
 >>> keywords_parser = LLMParser(llm, language="en", task_type="keywords")
->>> documents = Document(dataset_path='your_doc_path', embed=m, manager=False)
->>> rm = Retriever(documents, group_name='CoarseChunk', similarity='bm25', similarity_cut_off=0.01, topk=6)
->>> summary_result = summary_parser.transform(rm[0])
->>> keywords_result = keywords_parser.transform(rm[0])
+>>> documents = lazyllm.Document(dataset_path="/path/to/your/data", embed=m, manager=False)
+>>> rm = lazyllm.Retriever(documents, group_name='CoarseChunk', similarity='bm25', topk=6)
+>>> doc_nodes = rm("test")
+>>> summary_result = summary_parser.transform(doc_nodes[0])
+>>> keywords_result = keywords_parser.transform(doc_nodes[0])
 ''')
 
 # ---------------------------------------------------------------------------- #