Update app.py
Browse files
app.py
CHANGED
@@ -2,7 +2,6 @@ import gradio as gr
|
|
2 |
import os
|
3 |
api_token = os.getenv("HF_TOKEN")
|
4 |
|
5 |
-
|
6 |
from langchain_community.vectorstores import FAISS
|
7 |
from langchain_community.document_loaders import PyPDFLoader
|
8 |
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
@@ -18,11 +17,7 @@ import torch
|
|
18 |
list_llm = ["meta-llama/Meta-Llama-3-8B-Instruct", "mistralai/Mistral-7B-Instruct-v0.2"]
|
19 |
list_llm_simple = [os.path.basename(llm) for llm in list_llm]
|
20 |
|
21 |
-
# Load and split PDF document
|
22 |
def load_doc(list_file_path):
|
23 |
-
# Processing for one document only
|
24 |
-
# loader = PyPDFLoader(file_path)
|
25 |
-
# pages = loader.load()
|
26 |
loaders = [PyPDFLoader(x) for x in list_file_path]
|
27 |
pages = []
|
28 |
for loader in loaders:
|
@@ -34,14 +29,11 @@ def load_doc(list_file_path):
|
|
34 |
doc_splits = text_splitter.split_documents(pages)
|
35 |
return doc_splits
|
36 |
|
37 |
-
# Create vector database
|
38 |
def create_db(splits):
|
39 |
embeddings = HuggingFaceEmbeddings()
|
40 |
vectordb = FAISS.from_documents(splits, embeddings)
|
41 |
return vectordb
|
42 |
|
43 |
-
|
44 |
-
# Initialize langchain LLM chain
|
45 |
def initialize_llmchain(llm_model, temperature, max_tokens, top_k, vector_db, progress=gr.Progress()):
|
46 |
if llm_model == "meta-llama/Meta-Llama-3-8B-Instruct":
|
47 |
llm = HuggingFaceEndpoint(
|
@@ -77,36 +69,27 @@ def initialize_llmchain(llm_model, temperature, max_tokens, top_k, vector_db, pr
|
|
77 |
)
|
78 |
return qa_chain
|
79 |
|
80 |
-
# Initialize database
|
81 |
def initialize_database(list_file_obj, progress=gr.Progress()):
|
82 |
-
# Create a list of documents (when valid)
|
83 |
list_file_path = [x.name for x in list_file_obj if x is not None]
|
84 |
-
# Load document and create splits
|
85 |
doc_splits = load_doc(list_file_path)
|
86 |
-
# Create or load vector database
|
87 |
vector_db = create_db(doc_splits)
|
88 |
return vector_db, "Database created!"
|
89 |
|
90 |
-
# Initialize LLM
|
91 |
def initialize_LLM(llm_option, llm_temperature, max_tokens, top_k, vector_db, progress=gr.Progress()):
|
92 |
-
# print("llm_option",llm_option)
|
93 |
llm_name = list_llm[llm_option]
|
94 |
print("llm_name: ",llm_name)
|
95 |
qa_chain = initialize_llmchain(llm_name, llm_temperature, max_tokens, top_k, vector_db, progress)
|
96 |
return qa_chain, "QA chain initialized. Chatbot is ready!"
|
97 |
|
98 |
-
|
99 |
def format_chat_history(message, chat_history):
|
100 |
formatted_chat_history = []
|
101 |
for user_message, bot_message in chat_history:
|
102 |
formatted_chat_history.append(f"User: {user_message}")
|
103 |
formatted_chat_history.append(f"Assistant: {bot_message}")
|
104 |
return formatted_chat_history
|
105 |
-
|
106 |
|
107 |
def conversation(qa_chain, message, history):
|
108 |
formatted_chat_history = format_chat_history(message, history)
|
109 |
-
# Generate response using QA chain
|
110 |
response = qa_chain.invoke({"question": message, "chat_history": formatted_chat_history})
|
111 |
response_answer = response["answer"]
|
112 |
if response_answer.find("Helpful Answer:") != -1:
|
@@ -115,14 +98,11 @@ def conversation(qa_chain, message, history):
|
|
115 |
response_source1 = response_sources[0].page_content.strip()
|
116 |
response_source2 = response_sources[1].page_content.strip()
|
117 |
response_source3 = response_sources[2].page_content.strip()
|
118 |
-
# Langchain sources are zero-based
|
119 |
response_source1_page = response_sources[0].metadata["page"] + 1
|
120 |
response_source2_page = response_sources[1].metadata["page"] + 1
|
121 |
response_source3_page = response_sources[2].metadata["page"] + 1
|
122 |
-
# Append user message and response to chat history
|
123 |
new_history = history + [(message, response_answer)]
|
124 |
return qa_chain, gr.update(value=""), new_history, response_source1, response_source1_page, response_source2, response_source2_page, response_source3, response_source3_page
|
125 |
-
|
126 |
|
127 |
def upload_file(file_obj):
|
128 |
list_file_path = []
|
@@ -133,24 +113,23 @@ def upload_file(file_obj):
|
|
133 |
|
134 |
def demo():
|
135 |
custom_css = """
|
136 |
-
|
137 |
-
display: flex
|
138 |
-
flex-direction: row
|
139 |
-
flex-wrap: nowrap
|
140 |
-
width: 100% !important;
|
141 |
}
|
142 |
-
|
143 |
-
min-width:
|
144 |
-
max-width: 35
|
145 |
-
|
146 |
}
|
147 |
-
|
148 |
-
min-width: 500px
|
149 |
-
flex:
|
150 |
}
|
151 |
-
@media (max-width:
|
152 |
-
|
153 |
-
|
154 |
}
|
155 |
"""
|
156 |
|
@@ -158,55 +137,41 @@ def demo():
|
|
158 |
vector_db = gr.State()
|
159 |
qa_chain = gr.State()
|
160 |
gr.HTML("<center><h1>RAG PDF chatbot</h1><center>")
|
161 |
-
gr.Markdown("""<b>Query your PDF documents!</b> This AI agent is designed to perform retrieval augmented generation (RAG) on PDF documents.
|
162 |
-
<b>Please do not upload confidential documents.</b>
|
163 |
-
""")
|
164 |
|
165 |
-
with gr.Row(
|
166 |
-
with gr.Column(
|
167 |
gr.Markdown("<b>Step 1 - Upload PDF documents and Initialize RAG pipeline</b>")
|
168 |
-
|
169 |
-
|
170 |
-
|
171 |
-
|
172 |
-
|
173 |
-
|
174 |
-
|
175 |
-
|
176 |
-
|
177 |
-
|
178 |
-
|
179 |
-
|
180 |
-
|
181 |
-
with gr.Row():
|
182 |
-
slider_maxtokens = gr.Slider(minimum=128, maximum=9192, value=4096, step=128, label="Max New Tokens", info="Maximum number of tokens to be generated", interactive=True)
|
183 |
-
with gr.Row():
|
184 |
-
slider_topk = gr.Slider(minimum=1, maximum=10, value=3, step=1, label="top-k", info="Number of tokens to select the next token from", interactive=True)
|
185 |
-
with gr.Row():
|
186 |
-
qachain_btn = gr.Button("Initialize Question Answering Chatbot")
|
187 |
-
with gr.Row():
|
188 |
-
llm_progress = gr.Textbox(value="Not initialized", show_label=False)
|
189 |
-
|
190 |
-
with gr.Column(elem_classes="column-2"):
|
191 |
gr.Markdown("<b>Step 2 - Chat with your Document</b>")
|
192 |
chatbot = gr.Chatbot(height=505)
|
193 |
with gr.Accordion("Relevant context from the source document", open=False):
|
194 |
with gr.Row():
|
195 |
-
doc_source1 = gr.Textbox(label="Reference 1", lines=2,
|
196 |
source1_page = gr.Number(label="Page", scale=1)
|
197 |
with gr.Row():
|
198 |
-
doc_source2 = gr.Textbox(label="Reference 2", lines=2,
|
199 |
source2_page = gr.Number(label="Page", scale=1)
|
200 |
with gr.Row():
|
201 |
-
doc_source3 = gr.Textbox(label="Reference 3", lines=2,
|
202 |
source3_page = gr.Number(label="Page", scale=1)
|
203 |
-
|
204 |
-
msg = gr.Textbox(placeholder="Ask a question", container=True)
|
205 |
with gr.Row():
|
206 |
submit_btn = gr.Button("Submit")
|
207 |
clear_btn = gr.ClearButton([msg, chatbot], value="Clear")
|
208 |
-
|
209 |
-
# Rest of your event handlers remain the same...
|
210 |
db_btn.click(initialize_database,
|
211 |
inputs=[document],
|
212 |
outputs=[vector_db, db_progress])
|
@@ -216,7 +181,6 @@ def demo():
|
|
216 |
inputs=None,
|
217 |
outputs=[chatbot, doc_source1, source1_page, doc_source2, source2_page, doc_source3, source3_page],
|
218 |
queue=False)
|
219 |
-
|
220 |
msg.submit(conversation,
|
221 |
inputs=[qa_chain, msg, chatbot],
|
222 |
outputs=[qa_chain, msg, chatbot, doc_source1, source1_page, doc_source2, source2_page, doc_source3, source3_page],
|
@@ -231,4 +195,6 @@ def demo():
|
|
231 |
queue=False)
|
232 |
|
233 |
demo.queue().launch(debug=True)
|
234 |
-
|
|
|
|
|
|
2 |
import os
|
3 |
api_token = os.getenv("HF_TOKEN")
|
4 |
|
|
|
5 |
from langchain_community.vectorstores import FAISS
|
6 |
from langchain_community.document_loaders import PyPDFLoader
|
7 |
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
|
|
17 |
list_llm = ["meta-llama/Meta-Llama-3-8B-Instruct", "mistralai/Mistral-7B-Instruct-v0.2"]
|
18 |
list_llm_simple = [os.path.basename(llm) for llm in list_llm]
|
19 |
|
|
|
20 |
def load_doc(list_file_path):
|
|
|
|
|
|
|
21 |
loaders = [PyPDFLoader(x) for x in list_file_path]
|
22 |
pages = []
|
23 |
for loader in loaders:
|
|
|
29 |
doc_splits = text_splitter.split_documents(pages)
|
30 |
return doc_splits
|
31 |
|
|
|
32 |
def create_db(splits):
|
33 |
embeddings = HuggingFaceEmbeddings()
|
34 |
vectordb = FAISS.from_documents(splits, embeddings)
|
35 |
return vectordb
|
36 |
|
|
|
|
|
37 |
def initialize_llmchain(llm_model, temperature, max_tokens, top_k, vector_db, progress=gr.Progress()):
|
38 |
if llm_model == "meta-llama/Meta-Llama-3-8B-Instruct":
|
39 |
llm = HuggingFaceEndpoint(
|
|
|
69 |
)
|
70 |
return qa_chain
|
71 |
|
|
|
72 |
def initialize_database(list_file_obj, progress=gr.Progress()):
|
|
|
73 |
list_file_path = [x.name for x in list_file_obj if x is not None]
|
|
|
74 |
doc_splits = load_doc(list_file_path)
|
|
|
75 |
vector_db = create_db(doc_splits)
|
76 |
return vector_db, "Database created!"
|
77 |
|
|
|
78 |
def initialize_LLM(llm_option, llm_temperature, max_tokens, top_k, vector_db, progress=gr.Progress()):
|
|
|
79 |
llm_name = list_llm[llm_option]
|
80 |
print("llm_name: ",llm_name)
|
81 |
qa_chain = initialize_llmchain(llm_name, llm_temperature, max_tokens, top_k, vector_db, progress)
|
82 |
return qa_chain, "QA chain initialized. Chatbot is ready!"
|
83 |
|
|
|
84 |
def format_chat_history(message, chat_history):
|
85 |
formatted_chat_history = []
|
86 |
for user_message, bot_message in chat_history:
|
87 |
formatted_chat_history.append(f"User: {user_message}")
|
88 |
formatted_chat_history.append(f"Assistant: {bot_message}")
|
89 |
return formatted_chat_history
|
|
|
90 |
|
91 |
def conversation(qa_chain, message, history):
|
92 |
formatted_chat_history = format_chat_history(message, history)
|
|
|
93 |
response = qa_chain.invoke({"question": message, "chat_history": formatted_chat_history})
|
94 |
response_answer = response["answer"]
|
95 |
if response_answer.find("Helpful Answer:") != -1:
|
|
|
98 |
response_source1 = response_sources[0].page_content.strip()
|
99 |
response_source2 = response_sources[1].page_content.strip()
|
100 |
response_source3 = response_sources[2].page_content.strip()
|
|
|
101 |
response_source1_page = response_sources[0].metadata["page"] + 1
|
102 |
response_source2_page = response_sources[1].metadata["page"] + 1
|
103 |
response_source3_page = response_sources[2].metadata["page"] + 1
|
|
|
104 |
new_history = history + [(message, response_answer)]
|
105 |
return qa_chain, gr.update(value=""), new_history, response_source1, response_source1_page, response_source2, response_source2_page, response_source3, response_source3_page
|
|
|
106 |
|
107 |
def upload_file(file_obj):
|
108 |
list_file_path = []
|
|
|
113 |
|
114 |
def demo():
|
115 |
custom_css = """
|
116 |
+
#column-container {
|
117 |
+
display: flex;
|
118 |
+
flex-direction: row;
|
119 |
+
flex-wrap: nowrap;
|
|
|
120 |
}
|
121 |
+
#column-left {
|
122 |
+
min-width: 350px;
|
123 |
+
max-width: 35%;
|
124 |
+
margin-right: 20px;
|
125 |
}
|
126 |
+
#column-right {
|
127 |
+
min-width: 500px;
|
128 |
+
flex-grow: 1;
|
129 |
}
|
130 |
+
@media (max-width: 1200px) {
|
131 |
+
#column-left { min-width: 300px; }
|
132 |
+
#column-right { min-width: 400px; }
|
133 |
}
|
134 |
"""
|
135 |
|
|
|
137 |
vector_db = gr.State()
|
138 |
qa_chain = gr.State()
|
139 |
gr.HTML("<center><h1>RAG PDF chatbot</h1><center>")
|
140 |
+
gr.Markdown("""<b>Query your PDF documents!</b> This AI agent is designed to perform retrieval augmented generation (RAG) on PDF documents.""")
|
|
|
|
|
141 |
|
142 |
+
with gr.Row(elem_id="column-container"):
|
143 |
+
with gr.Column(elem_id="column-left"):
|
144 |
gr.Markdown("<b>Step 1 - Upload PDF documents and Initialize RAG pipeline</b>")
|
145 |
+
document = gr.Files(height=300, file_count="multiple", file_types=[".pdf"], interactive=True, label="Upload PDF documents")
|
146 |
+
db_btn = gr.Button("Create vector database")
|
147 |
+
db_progress = gr.Textbox(value="Not initialized", show_label=False)
|
148 |
+
gr.Markdown("<b>Select Large Language Model (LLM) and input parameters</b>")
|
149 |
+
llm_btn = gr.Radio(list_llm_simple, label="Available LLMs", value=list_llm_simple[0], type="index")
|
150 |
+
with gr.Accordion("LLM input parameters", open=False):
|
151 |
+
slider_temperature = gr.Slider(0.01, 1.0, value=0.5, step=0.1, label="Temperature")
|
152 |
+
slider_maxtokens = gr.Slider(128, 9192, value=4096, step=128, label="Max New Tokens")
|
153 |
+
slider_topk = gr.Slider(1, 10, value=3, step=1, label="top-k")
|
154 |
+
qachain_btn = gr.Button("Initialize Question Answering Chatbot")
|
155 |
+
llm_progress = gr.Textbox(value="Not initialized", show_label=False)
|
156 |
+
|
157 |
+
with gr.Column(elem_id="column-right"):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
158 |
gr.Markdown("<b>Step 2 - Chat with your Document</b>")
|
159 |
chatbot = gr.Chatbot(height=505)
|
160 |
with gr.Accordion("Relevant context from the source document", open=False):
|
161 |
with gr.Row():
|
162 |
+
doc_source1 = gr.Textbox(label="Reference 1", lines=2, scale=20)
|
163 |
source1_page = gr.Number(label="Page", scale=1)
|
164 |
with gr.Row():
|
165 |
+
doc_source2 = gr.Textbox(label="Reference 2", lines=2, scale=20)
|
166 |
source2_page = gr.Number(label="Page", scale=1)
|
167 |
with gr.Row():
|
168 |
+
doc_source3 = gr.Textbox(label="Reference 3", lines=2, scale=20)
|
169 |
source3_page = gr.Number(label="Page", scale=1)
|
170 |
+
msg = gr.Textbox(placeholder="Ask a question")
|
|
|
171 |
with gr.Row():
|
172 |
submit_btn = gr.Button("Submit")
|
173 |
clear_btn = gr.ClearButton([msg, chatbot], value="Clear")
|
174 |
+
|
|
|
175 |
db_btn.click(initialize_database,
|
176 |
inputs=[document],
|
177 |
outputs=[vector_db, db_progress])
|
|
|
181 |
inputs=None,
|
182 |
outputs=[chatbot, doc_source1, source1_page, doc_source2, source2_page, doc_source3, source3_page],
|
183 |
queue=False)
|
|
|
184 |
msg.submit(conversation,
|
185 |
inputs=[qa_chain, msg, chatbot],
|
186 |
outputs=[qa_chain, msg, chatbot, doc_source1, source1_page, doc_source2, source2_page, doc_source3, source3_page],
|
|
|
195 |
queue=False)
|
196 |
|
197 |
demo.queue().launch(debug=True)
|
198 |
+
|
199 |
+
if __name__ == "__main__":
|
200 |
+
demo()
|