dfasd commited on
Commit
30932f4
·
verified ·
1 Parent(s): 95cf17b

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +5 -46
app.py CHANGED
@@ -1,11 +1,6 @@
1
- from flask import Flask, render_template, jsonify, request, redirect, url_for
2
- from flask_wtf.csrf import CSRFProtect
3
-
4
- # from tavily import TavilyClient
5
 
6
  from dotenv import load_dotenv
7
  import os
8
-
9
  from langchain_community.document_loaders import TextLoader
10
  from langchain_community.vectorstores import Chroma
11
  from langchain_text_splitters import CharacterTextSplitter
@@ -23,13 +18,7 @@ from langchain.prompts import PromptTemplate
23
  import time
24
  load_dotenv()
25
 
26
- # TAVILY_API_KEY = os.getenv("TAVILY_API_KEY")
27
  OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
28
- # tavily = TavilyClient(api_key=TAVILY_API_KEY)
29
-
30
- app = Flask(__name__, static_folder='static')
31
- app.config['SECRET_KEY'] = 'secret'
32
- csrf = CSRFProtect(app)
33
 
34
  text_splitter = CharacterTextSplitter(separator = "\n", chunk_size=1000, chunk_overlap=200, length_function = len)
35
  embeddings = OpenAIEmbeddings(openai_api_key=OPENAI_API_KEY)
@@ -38,20 +27,6 @@ llm = ChatOpenAI(api_key=OPENAI_API_KEY)
38
 
39
  vectordb_path = "./vector_db"
40
 
41
- @app.route('/')
42
- def home():
43
- return redirect(url_for('search_view'))
44
-
45
- @app.route('/search_view')
46
- def search_view():
47
- return render_template('search.html')
48
-
49
- @app.route('/rag_view')
50
- def rag_view():
51
- dbs = [f.name for f in os.scandir(vectordb_path) if f.is_dir()]
52
- return render_template('rag.html', dbs = dbs)
53
-
54
- @app.route('/query', methods=['POST'])
55
  def query():
56
  if request.method == "POST":
57
  prompt = request.get_json().get("prompt")
@@ -91,14 +66,6 @@ def query():
91
  output = stuff_chain({"input_documents": docs, "human_input": prompt}, return_only_outputs=False)
92
 
93
  final_answer = output["output_text"]
94
- # prompt = ChatPromptTemplate.from_messages(
95
- # [("system", "Please answer to user's query based on following context.\n\nContext: {context}")]
96
- # )
97
-
98
-
99
- # chain = create_stuff_documents_chain(llm, prompt)
100
-
101
- # answer = chain.invoke({"context": docs, "prompt": prompt})
102
 
103
  data = {"success": "ok", "response": final_answer}
104
 
@@ -108,12 +75,10 @@ def query():
108
 
109
  return jsonify(data)
110
 
111
- @app.route('/uploadDocuments', methods=['POST'])
112
- @csrf.exempt
113
  def uploadDocuments():
114
  # uploaded_files = request.files.getlist('files[]')
 
115
  dbname = request.form.get('dbname')
116
- uploaded_files = ['https://www.airbus.com/sites/g/files/jlcbta136/files/2024-03/Airbus-Annual-Report-2023.pdf', 'https://www.singaporeair.com/saar5/pdf/Investor-Relations/Annual-Report/annualreport2223.pdf']
117
  if dbname == "":
118
  return {"success": "db"}
119
 
@@ -136,8 +101,6 @@ def uploadDocuments():
136
  else:
137
  return {"success": "bad"}
138
 
139
- @app.route('/dbcreate', methods=['POST'])
140
- @csrf.exempt
141
  def dbcreate():
142
  dbname = request.get_json().get("dbname")
143
 
@@ -147,11 +110,7 @@ def dbcreate():
147
  else:
148
  return {'success': 'bad'}
149
 
150
- # import gradio as gr
151
- # chatbot = gr.Chatbot(avatar_images=["user.png", "bot.jpg"], height=600)
152
- # clear_but = gr.Button(value="Clear Chat")
153
- # demo = gr.ChatInterface(fn=search, title="Mediate.com Chatbot Prototype", multimodal=False, retry_btn=None, undo_btn=None, clear_btn=clear_but, chatbot=chatbot)
154
-
155
-
156
- if __name__ == '__main__':
157
- app.run(debug=True)
 
 
 
 
 
1
 
2
  from dotenv import load_dotenv
3
  import os
 
4
  from langchain_community.document_loaders import TextLoader
5
  from langchain_community.vectorstores import Chroma
6
  from langchain_text_splitters import CharacterTextSplitter
 
18
  import time
19
  load_dotenv()
20
 
 
21
  OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
 
 
 
 
 
22
 
23
  text_splitter = CharacterTextSplitter(separator = "\n", chunk_size=1000, chunk_overlap=200, length_function = len)
24
  embeddings = OpenAIEmbeddings(openai_api_key=OPENAI_API_KEY)
 
27
 
28
  vectordb_path = "./vector_db"
29
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30
  def query():
31
  if request.method == "POST":
32
  prompt = request.get_json().get("prompt")
 
66
  output = stuff_chain({"input_documents": docs, "human_input": prompt}, return_only_outputs=False)
67
 
68
  final_answer = output["output_text"]
 
 
 
 
 
 
 
 
69
 
70
  data = {"success": "ok", "response": final_answer}
71
 
 
75
 
76
  return jsonify(data)
77
 
 
 
78
  def uploadDocuments():
79
  # uploaded_files = request.files.getlist('files[]')
80
+ uploaded_files = ['annualreport2223.pdf', 'Airbus-Annual-Report-2023.pdf']
81
  dbname = request.form.get('dbname')
 
82
  if dbname == "":
83
  return {"success": "db"}
84
 
 
101
  else:
102
  return {"success": "bad"}
103
 
 
 
104
  def dbcreate():
105
  dbname = request.get_json().get("dbname")
106
 
 
110
  else:
111
  return {'success': 'bad'}
112
 
113
+ import gradio as gr
114
+ chatbot = gr.Chatbot(avatar_images=["user.png", "bot.jpg"], height=600)
115
+ clear_but = gr.Button(value="Clear Chat")
116
+ demo = gr.ChatInterface(fn=search, title="Mediate.com Chatbot Prototype", multimodal=False, retry_btn=None, undo_btn=None, clear_btn=clear_but, chatbot=chatbot)