-
Notifications
You must be signed in to change notification settings - Fork 0
/
try_with_newPDF5.py
201 lines (162 loc) · 8.25 KB
/
try_with_newPDF5.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
"""
Created by: Urveshkumar Koshti
Purpose of the code: To create the Chatbot for the Question-Answer based on the PDF.
Attributes of the Local Chatbot:
--> Input to the Code (via Frontend interface): PDF file
Output from the code (on the Frontend interface): Text Generation
--> Saves the Chat History
--> Creates the Vector Database using SQL
--> Uses the same Vector Database For the Questions from the same PDF until new PDF is not uploaded by User
To create the Docker container from the Docker Image - docker run --network="host" -d -p 8501:8501 DOCKER_IMAGE_NAME
To run the File in Docker container (after creating a docker container):
--> Pull the image from the Docker Hub and then Run the Image.
--> Open the Chatbot by typing 'streamlit run chatbot_app.py'.
"""
import os
import time
import streamlit as st
from langchain.chains import RetrievalQA
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
from langchain.callbacks.manager import CallbackManager
from langchain_community.llms import Ollama
from langchain_community.embeddings.ollama import OllamaEmbeddings
from langchain_community.vectorstores import Chroma
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.document_loaders import PyPDFLoader
from langchain.prompts import PromptTemplate
from langchain.memory import ConversationBufferMemory
# Define models as constants for easy configuration
EMBEDDING_MODEL = "nomic-embed-text:latest "
# LLM_MODEL = "llama3.1:70b"
LLM_MODEL = "deepseek-coder-v2:16b"
# LLM_MODEL = "mistral:latest"
BASE_URL = "http://localhost:11434"
class PDFChatbot:
def __init__(self):
self.setup_directories()
self.setup_session_state()
self.display_title()
def setup_directories(self):
if not os.path.exists('files'):
os.mkdir('files')
if not os.path.exists('vector_database'):
os.mkdir('vector_database')
def setup_session_state(self):
if 'template' not in st.session_state:
st.session_state.template = """You are a knowledgeable chatbot, here to help with questions of the user. Your tone should be professional and informative.
Context: {context}
History: {history}
User: {question}
Chatbot:"""
if 'prompt' not in st.session_state:
st.session_state.prompt = PromptTemplate(
input_variables=["history", "context", "question"],
template=st.session_state.template,
)
if 'memory' not in st.session_state:
st.session_state.memory = ConversationBufferMemory(
memory_key="history",
return_messages=True,
input_key="question"
)
if 'vectorstore' not in st.session_state:
st.session_state.vectorstore = Chroma(persist_directory='vector_database',
embedding_function=OllamaEmbeddings(base_url=BASE_URL,
model=EMBEDDING_MODEL)
)
if 'llm' not in st.session_state:
st.session_state.llm = Ollama(base_url=BASE_URL,
model=LLM_MODEL,
verbose=True,
callback_manager=CallbackManager(
[StreamingStdOutCallbackHandler()]),
)
if 'chat_history' not in st.session_state:
st.session_state.chat_history = []
def display_title(self):
st.title("PDF Chatbot")
def clear_vectorstore(self):
if 'vectorstore' in st.session_state:
del st.session_state['vectorstore']
if 'retriever' in st.session_state:
del st.session_state['retriever']
if 'qa_chain' in st.session_state:
del st.session_state['qa_chain']
if 'chat_history' in st.session_state:
st.session_state.chat_history = []
def handle_upload(self, uploaded_file):
if uploaded_file is not None:
if 'current_pdf' not in st.session_state or st.session_state.current_pdf != uploaded_file.name:
self.clear_vectorstore()
st.session_state.current_pdf = uploaded_file.name
st.info("Uploading and processing your PDF...")
start_time = time.time()
bytes_data = uploaded_file.read()
with open("files/" + uploaded_file.name + ".pdf", "wb") as f:
f.write(bytes_data)
upload_time = time.time() - start_time
st.success(f"PDF uploaded and saved in {upload_time:.2f} seconds.")
st.info("Analyzing your document...")
start_time = time.time()
loader = PyPDFLoader("files/" + uploaded_file.name + ".pdf")
data = loader.load()
text_splitter = RecursiveCharacterTextSplitter(
chunk_size=1500,
chunk_overlap=200,
length_function=len
)
all_splits = text_splitter.split_documents(data)
st.session_state.vectorstore = Chroma.from_documents(
documents=all_splits,
embedding=OllamaEmbeddings(model=EMBEDDING_MODEL)
)
st.session_state.vectorstore.persist()
processing_time = time.time() - start_time
st.success(f"Document analyzed and embeddings created in {processing_time:.2f} seconds.")
st.session_state.retriever = st.session_state.vectorstore.as_retriever()
st.session_state.qa_chain = RetrievalQA.from_chain_type(
llm=st.session_state.llm,
chain_type='stuff',
retriever=st.session_state.retriever,
verbose=True,
chain_type_kwargs={
"verbose": True,
"prompt": st.session_state.prompt,
"memory": st.session_state.memory,
}
)
def display_chat_history(self):
for message in st.session_state.chat_history:
with st.chat_message(message["role"]):
st.markdown(message["message"])
def handle_user_input(self):
user_input = st.chat_input("You:", key="user_input")
if user_input:
user_message = {"role": "user", "message": user_input}
st.session_state.chat_history.append(user_message)
with st.chat_message("user"):
st.markdown(user_input)
with st.chat_message("assistant"):
with st.spinner("Assistant is typing..."):
start_time = time.time()
response = st.session_state.qa_chain(user_input)
response_time = time.time() - start_time
message_placeholder = st.empty()
full_response = ""
words = response['result'].split()
for word in words:
full_response += word + " "
message_placeholder.markdown(full_response + "▌")
time.sleep(0.05)
message_placeholder.markdown(full_response)
st.success(f"Response generated in {response_time:.2f} seconds.")
chatbot_message = {"role": "assistant", "message": response['result']}
st.session_state.chat_history.append(chatbot_message)
def run(self):
uploaded_file = st.file_uploader("Upload your PDF", type='pdf')
self.handle_upload(uploaded_file)
self.display_chat_history()
self.handle_user_input()
if __name__ == "__main__":
chatbot = PDFChatbot()
chatbot.run()