Nahyunho commited on
Commit
4873748
1 Parent(s): 4cab92b

Upload 3 files

Browse files
Files changed (3) hide show
  1. app.py +171 -0
  2. htmlTemplates.py +44 -0
  3. requirements.txt +14 -0
app.py ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import streamlit as st
2
+ from dotenv import load_dotenv
3
+ from PyPDF2 import PdfReader
4
+ from langchain.text_splitter import CharacterTextSplitter, RecursiveCharacterTextSplitter
5
+ from langchain.embeddings import OpenAIEmbeddings, HuggingFaceInstructEmbeddings
6
+ from langchain.vectorstores import FAISS, Chroma
7
+ from langchain.embeddings import HuggingFaceEmbeddings # General embeddings from HuggingFace models.
8
+ from langchain.chat_models import ChatOpenAI
9
+ from langchain.memory import ConversationBufferMemory
10
+ from langchain.chains import ConversationalRetrievalChain
11
+ from htmlTemplates import css, bot_template, user_template
12
+ from langchain.llms import HuggingFaceHub, LlamaCpp, CTransformers # For loading transformer models.
13
+ from langchain.document_loaders import PyPDFLoader, TextLoader, JSONLoader, CSVLoader
14
+ import tempfile # 임시 파일을 생성하기 위한 라이브러리입니다.
15
+ import os
16
+
17
+
18
+ # PDF 문서로부터 텍스트를 추출하는 함수입니다.
19
+ def get_pdf_text(pdf_docs):
20
+ temp_dir = tempfile.TemporaryDirectory() # 임시 디렉토리를 생성합니다.
21
+ temp_filepath = os.path.join(temp_dir.name, pdf_docs.name) # 임시 파일 경로를 생성합니다.
22
+ with open(temp_filepath, "wb") as f: # 임시 파일을 바이너리 쓰기 모드로 엽니다.
23
+ f.write(pdf_docs.getvalue()) # PDF 문서의 내용을 임시 파일에 씁니다.
24
+ pdf_loader = PyPDFLoader(temp_filepath) # PyPDFLoader를 사용해 PDF를 로드합니다.
25
+ pdf_doc = pdf_loader.load() # 텍스트를 추출합니다.
26
+ return pdf_doc # 추출한 텍스트를 반환합니다.
27
+
28
+ # 과제
29
+ # 아래 텍스트 추출 함수를 작성
30
+
31
+ def get_text_file(docs):
32
+ with NamedTemporaryFile() as temp_file:
33
+ temp_file.write(docs.getvalue())
34
+ temp_file.seek(0)
35
+ text_loader = TextLoader(temp_file.name)
36
+ text_doc = text_loader.load()
37
+
38
+ return text_doc
39
+
40
+
41
+ def get_csv_file(docs):
42
+ with NamedTemporaryFile() as temp_file:
43
+ temp_file.write(docs.getvalue())
44
+ temp_file.seek(0)
45
+ text_loader = CSVLoader(temp_file.name)
46
+ text_doc = text_loader.load()
47
+
48
+ return text_docs
49
+
50
+ def get_json_file(docs):
51
+ with NamedTemporaryFile() as temp_file:
52
+ temp_file.write(docs.getvalue())
53
+ temp_file.seek(0)
54
+ json_loader = JSONLoader(temp_file.name,
55
+ jq_schema='.scans[].relationships',
56
+ text_content=False)
57
+ json_doc = json_loader.load()
58
+
59
+ return json_doc
60
+
61
+
62
+ # 문서들을 처리하여 텍스트 청크로 나누는 함수입니다.
63
+ def get_text_chunks(documents):
64
+ text_splitter = RecursiveCharacterTextSplitter(
65
+ chunk_size=1000, # 청크의 크기를 지정합니다.
66
+ chunk_overlap=200, # 청크 사이의 중복을 지정합니다.
67
+ length_function=len # 텍스트의 길이를 측정하는 함수를 지정합니다.
68
+ )
69
+
70
+ documents = text_splitter.split_documents(documents) # 문서들을 청크로 나눕니다
71
+ return documents # 나눈 청크를 반환합니다.
72
+
73
+
74
+ # 텍스트 청크들로부터 벡터 스토어를 생성하는 함수입니다.
75
+ def get_vectorstore(text_chunks):
76
+ # OpenAI 임베딩 모델을 로드합니다. (Embedding models - Ada v2)
77
+
78
+ embeddings = OpenAIEmbeddings()
79
+ vectorstore = FAISS.from_documents(text_chunks, embeddings) # FAISS 벡터 스토어를 생성합니다.
80
+
81
+ return vectorstore # 생성된 벡터 스토어를 반환합니다.
82
+
83
+
84
+ def get_conversation_chain(vectorstore):
85
+ gpt_model_name = 'gpt-3.5-turbo'
86
+ llm = ChatOpenAI(model_name = gpt_model_name) #gpt-3.5 모델 로드
87
+
88
+ # 대화 기록을 저장하기 위한 메모리를 생성합니다.
89
+ memory = ConversationBufferMemory(
90
+ memory_key='chat_history', return_messages=True)
91
+ # 대화 검색 체인을 생성합니다.
92
+ conversation_chain = ConversationalRetrievalChain.from_llm(
93
+ llm=llm,
94
+ retriever=vectorstore.as_retriever(),
95
+ memory=memory
96
+ )
97
+ return conversation_chain
98
+
99
+ # 사용자 입력을 처리하는 함수입니다.
100
+ def handle_userinput(user_question):
101
+ # 대화 체인을 사용하여 사용자 질문에 대한 응답을 생성합니다.
102
+ response = st.session_state.conversation({'question': user_question})
103
+ # 대화 기록을 저장합니다.
104
+ st.session_state.chat_history = response['chat_history']
105
+
106
+ for i, message in enumerate(st.session_state.chat_history):
107
+ if i % 2 == 0:
108
+ st.write(user_template.replace(
109
+ "{{MSG}}", message.content), unsafe_allow_html=True)
110
+ else:
111
+ st.write(bot_template.replace(
112
+ "{{MSG}}", message.content), unsafe_allow_html=True)
113
+
114
+
115
+ def main():
116
+ load_dotenv()
117
+ st.set_page_config(page_title="Chat with multiple Files",
118
+ page_icon=":books:")
119
+ st.write(css, unsafe_allow_html=True)
120
+
121
+ if "conversation" not in st.session_state:
122
+ st.session_state.conversation = None
123
+ if "chat_history" not in st.session_state:
124
+ st.session_state.chat_history = None
125
+
126
+ st.header("Chat with multiple Files :")
127
+ user_question = st.text_input("Ask a question about your documents:")
128
+ if user_question:
129
+ handle_userinput(user_question)
130
+
131
+ with st.sidebar:
132
+ openai_key = st.text_input("Paste your OpenAI API key (sk-...)")
133
+ if openai_key:
134
+ os.environ["OPENAI_API_KEY"] = openai_key
135
+
136
+ st.subheader("Your documents")
137
+ docs = st.file_uploader(
138
+ "Upload your PDFs here and click on 'Process'", accept_multiple_files=True)
139
+ if st.button("Process"):
140
+ with st.spinner("Processing"):
141
+ # get pdf text
142
+ doc_list = []
143
+
144
+ for file in docs:
145
+ print('file - type : ', file.type)
146
+ if file.type == 'text/plain':
147
+ # file is .txt
148
+ doc_list.extend(get_text_file(file))
149
+ elif file.type in ['application/octet-stream', 'application/pdf']:
150
+ # file is .pdf
151
+ doc_list.extend(get_pdf_text(file))
152
+ elif file.type == 'text/csv':
153
+ # file is .csv
154
+ doc_list.extend(get_csv_file(file))
155
+ elif file.type == 'application/json':
156
+ # file is .json
157
+ doc_list.extend(get_json_file(file))
158
+
159
+ # get the text chunks
160
+ text_chunks = get_text_chunks(doc_list)
161
+
162
+ # create vector store
163
+ vectorstore = get_vectorstore(text_chunks)
164
+
165
+ # create conversation chain
166
+ st.session_state.conversation = get_conversation_chain(
167
+ vectorstore)
168
+
169
+
170
+ if __name__ == '__main__':
171
+ main()
htmlTemplates.py ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ css = '''
2
+ <style>
3
+ .chat-message {
4
+ padding: 1.5rem; border-radius: 0.5rem; margin-bottom: 1rem; display: flex
5
+ }
6
+ .chat-message.user {
7
+ background-color: #2b313e
8
+ }
9
+ .chat-message.bot {
10
+ background-color: #475063
11
+ }
12
+ .chat-message .avatar {
13
+ width: 20%;
14
+ }
15
+ .chat-message .avatar img {
16
+ max-width: 78px;
17
+ max-height: 78px;
18
+ border-radius: 50%;
19
+ object-fit: cover;
20
+ }
21
+ .chat-message .message {
22
+ width: 80%;
23
+ padding: 0 1.5rem;
24
+ color: #fff;
25
+ }
26
+ '''
27
+
28
+ bot_template = '''
29
+ <div class="chat-message bot">
30
+ <div class="avatar">
31
+ <img src="https://i.ibb.co/cN0nmSj/Screenshot-2023-05-28-at-02-37-21.png" style="max-height: 78px; max-width: 78px; border-radius: 50%; object-fit: cover;">
32
+ </div>
33
+ <div class="message">{{MSG}}</div>
34
+ </div>
35
+ '''
36
+
37
+ user_template = '''
38
+ <div class="chat-message user">
39
+ <div class="avatar">
40
+ <img src="https://i.ibb.co/rdZC7LZ/Photo-logo-1.png">
41
+ </div>
42
+ <div class="message">{{MSG}}</div>
43
+ </div>
44
+ '''
requirements.txt ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ langchain
2
+ llama-cpp-python
3
+ PyPDF2==3.0.1
4
+ faiss-cpu==1.7.4
5
+ ctransformers
6
+ pypdf
7
+ chromadb
8
+ tiktoken
9
+ pysqlite3-binary
10
+ streamlit-extras
11
+ InstructorEmbedding
12
+ sentence-transformers
13
+ jq
14
+ openai