Spaces:
Build error
Build error
Update rag.py
Browse files
rag.py
CHANGED
@@ -31,8 +31,12 @@ MONGODB_DB_NAME = "langchain_db"
|
|
31 |
MONGODB_COLLECTION_NAME = "gpt-4"
|
32 |
MONGODB_INDEX_NAME = "default"
|
33 |
|
34 |
-
LLM_CHAIN_PROMPT = PromptTemplate(
|
35 |
-
|
|
|
|
|
|
|
|
|
36 |
|
37 |
client = MongoClient(MONGODB_ATLAS_CLUSTER_URI)
|
38 |
collection = client[MONGODB_DB_NAME][MONGODB_COLLECTION_NAME]
|
@@ -49,28 +53,34 @@ def load_documents():
|
|
49 |
docs.extend(loader.load())
|
50 |
|
51 |
# YouTube
|
52 |
-
loader = GenericLoader(
|
53 |
-
|
|
|
|
|
|
|
54 |
docs.extend(loader.load())
|
55 |
|
56 |
return docs
|
57 |
|
58 |
def split_documents(config, docs):
|
59 |
-
text_splitter = RecursiveCharacterTextSplitter(
|
60 |
-
|
|
|
61 |
|
62 |
return text_splitter.split_documents(docs)
|
63 |
|
64 |
def store_chroma(chunks):
|
65 |
-
Chroma.from_documents(
|
66 |
-
|
67 |
-
|
|
|
68 |
|
69 |
def store_mongodb(chunks):
|
70 |
-
MongoDBAtlasVectorSearch.from_documents(
|
71 |
-
|
72 |
-
|
73 |
-
|
|
|
74 |
|
75 |
def rag_ingestion(config):
|
76 |
docs = load_documents()
|
@@ -81,22 +91,26 @@ def rag_ingestion(config):
|
|
81 |
store_mongodb(chunks)
|
82 |
|
83 |
def retrieve_chroma():
|
84 |
-
return Chroma(
|
85 |
-
|
|
|
86 |
|
87 |
def retrieve_mongodb():
|
88 |
-
return MongoDBAtlasVectorSearch.from_connection_string(
|
89 |
-
|
90 |
-
|
91 |
-
|
|
|
92 |
|
93 |
def get_llm(config):
|
94 |
-
return ChatOpenAI(
|
95 |
-
|
|
|
96 |
|
97 |
def llm_chain(config, prompt):
|
98 |
-
llm_chain = LLMChain(
|
99 |
-
|
|
|
100 |
|
101 |
with get_openai_callback() as cb:
|
102 |
completion = llm_chain.generate([{"question": prompt}])
|
@@ -111,11 +125,12 @@ def rag_chain(config, rag_option, prompt):
|
|
111 |
elif (rag_option == RAG_MONGODB):
|
112 |
db = retrieve_mongodb()
|
113 |
|
114 |
-
rag_chain = RetrievalQA.from_chain_type(
|
115 |
-
|
116 |
-
|
117 |
-
|
118 |
-
|
|
|
119 |
|
120 |
with get_openai_callback() as cb:
|
121 |
completion = rag_chain({"query": prompt})
|
|
|
31 |
MONGODB_COLLECTION_NAME = "gpt-4"
|
32 |
MONGODB_INDEX_NAME = "default"
|
33 |
|
34 |
+
LLM_CHAIN_PROMPT = PromptTemplate(
|
35 |
+
input_variables = ["question"],
|
36 |
+
template = os.environ["LLM_TEMPLATE"])
|
37 |
+
RAG_CHAIN_PROMPT = PromptTemplate(
|
38 |
+
input_variables = ["context", "question"],
|
39 |
+
template = os.environ["RAG_TEMPLATE"])
|
40 |
|
41 |
client = MongoClient(MONGODB_ATLAS_CLUSTER_URI)
|
42 |
collection = client[MONGODB_DB_NAME][MONGODB_COLLECTION_NAME]
|
|
|
53 |
docs.extend(loader.load())
|
54 |
|
55 |
# YouTube
|
56 |
+
loader = GenericLoader(
|
57 |
+
YoutubeAudioLoader(
|
58 |
+
[YOUTUBE_URL_1, YOUTUBE_URL_2],
|
59 |
+
YOUTUBE_DIR),
|
60 |
+
OpenAIWhisperParser())
|
61 |
docs.extend(loader.load())
|
62 |
|
63 |
return docs
|
64 |
|
65 |
def split_documents(config, docs):
|
66 |
+
text_splitter = RecursiveCharacterTextSplitter(
|
67 |
+
chunk_overlap = config["chunk_overlap"],
|
68 |
+
chunk_size = config["chunk_size"])
|
69 |
|
70 |
return text_splitter.split_documents(docs)
|
71 |
|
72 |
def store_chroma(chunks):
|
73 |
+
Chroma.from_documents(
|
74 |
+
documents = chunks,
|
75 |
+
embedding = OpenAIEmbeddings(disallowed_special = ()),
|
76 |
+
persist_directory = CHROMA_DIR)
|
77 |
|
78 |
def store_mongodb(chunks):
|
79 |
+
MongoDBAtlasVectorSearch.from_documents(
|
80 |
+
documents = chunks,
|
81 |
+
embedding = OpenAIEmbeddings(disallowed_special = ()),
|
82 |
+
collection = collection,
|
83 |
+
index_name = MONGODB_INDEX_NAME)
|
84 |
|
85 |
def rag_ingestion(config):
|
86 |
docs = load_documents()
|
|
|
91 |
store_mongodb(chunks)
|
92 |
|
93 |
def retrieve_chroma():
|
94 |
+
return Chroma(
|
95 |
+
embedding_function = OpenAIEmbeddings(disallowed_special = ()),
|
96 |
+
persist_directory = CHROMA_DIR)
|
97 |
|
98 |
def retrieve_mongodb():
|
99 |
+
return MongoDBAtlasVectorSearch.from_connection_string(
|
100 |
+
MONGODB_ATLAS_CLUSTER_URI,
|
101 |
+
MONGODB_DB_NAME + "." + MONGODB_COLLECTION_NAME,
|
102 |
+
OpenAIEmbeddings(disallowed_special = ()),
|
103 |
+
index_name = MONGODB_INDEX_NAME)
|
104 |
|
105 |
def get_llm(config):
|
106 |
+
return ChatOpenAI(
|
107 |
+
model_name = config["model_name"],
|
108 |
+
temperature = config["temperature"])
|
109 |
|
110 |
def llm_chain(config, prompt):
|
111 |
+
llm_chain = LLMChain(
|
112 |
+
llm = get_llm(config),
|
113 |
+
prompt = LLM_CHAIN_PROMPT)
|
114 |
|
115 |
with get_openai_callback() as cb:
|
116 |
completion = llm_chain.generate([{"question": prompt}])
|
|
|
125 |
elif (rag_option == RAG_MONGODB):
|
126 |
db = retrieve_mongodb()
|
127 |
|
128 |
+
rag_chain = RetrievalQA.from_chain_type(
|
129 |
+
llm,
|
130 |
+
chain_type_kwargs = {"prompt": RAG_CHAIN_PROMPT,
|
131 |
+
"verbose": True},
|
132 |
+
retriever = db.as_retriever(search_kwargs = {"k": config["k"]}),
|
133 |
+
return_source_documents = True)
|
134 |
|
135 |
with get_openai_callback() as cb:
|
136 |
completion = rag_chain({"query": prompt})
|