Spaces:
Runtime error
Runtime error
Rams901
commited on
Commit
•
33a7c5b
0
Parent(s):
Duplicate from Rams901/rent-qa
Browse files- .gitattributes +36 -0
- README.md +13 -0
- app.py +278 -0
- rent_data/index.faiss +3 -0
- rent_data/index.pkl +3 -0
- requirements.txt +10 -0
.gitattributes
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
5 |
+
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
+
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
12 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
13 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
14 |
+
*.npy filter=lfs diff=lfs merge=lfs -text
|
15 |
+
*.npz filter=lfs diff=lfs merge=lfs -text
|
16 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
17 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
18 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
19 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
20 |
+
*.pickle filter=lfs diff=lfs merge=lfs -text
|
21 |
+
*.pkl filter=lfs diff=lfs merge=lfs -text
|
22 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
23 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
24 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
25 |
+
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
26 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
27 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
28 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
29 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
30 |
+
*.wasm filter=lfs diff=lfs merge=lfs -text
|
31 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
32 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
33 |
+
*.zst filter=lfs diff=lfs merge=lfs -text
|
34 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
35 |
+
availability.json filter=lfs diff=lfs merge=lfs -text
|
36 |
+
rent_data/index.faiss filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
title: Rent QA
|
3 |
+
emoji: null
|
4 |
+
colorFrom: blue
|
5 |
+
colorTo: gray
|
6 |
+
sdk: gradio
|
7 |
+
sdk_version: 3.27.0
|
8 |
+
app_file: app.py
|
9 |
+
pinned: false
|
10 |
+
duplicated_from: Rams901/rent-qa
|
11 |
+
---
|
12 |
+
|
13 |
+
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
app.py
ADDED
@@ -0,0 +1,278 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
import numpy as np
|
3 |
+
from langchain.document_loaders import UnstructuredPDFLoader
|
4 |
+
from langchain.indexes import VectorstoreIndexCreator
|
5 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
6 |
+
from langchain.text_splitter import CharacterTextSplitter
|
7 |
+
from langchain.chains import LLMChain
|
8 |
+
from langchain import PromptTemplate
|
9 |
+
import re
|
10 |
+
import pandas as pd
|
11 |
+
from langchain.vectorstores import FAISS
|
12 |
+
import requests
|
13 |
+
from bs4 import BeautifulSoup
|
14 |
+
from typing import List
|
15 |
+
from langchain.document_loaders import YoutubeLoader
|
16 |
+
from langchain.schema import (
|
17 |
+
SystemMessage,
|
18 |
+
HumanMessage,
|
19 |
+
AIMessage
|
20 |
+
)
|
21 |
+
from langchain.embeddings import HuggingFaceEmbeddings
|
22 |
+
from langchain.chat_models import ChatOpenAI
|
23 |
+
|
24 |
+
|
25 |
+
CHARACTER_CUT_OFF = 20000
|
26 |
+
text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=100)
|
27 |
+
|
28 |
+
embeddings = HuggingFaceEmbeddings()
|
29 |
+
db = FAISS.load_local('rent_data', embeddings)
|
30 |
+
|
31 |
+
llm = ChatOpenAI(
|
32 |
+
temperature=0,
|
33 |
+
model='gpt-3.5-turbo'
|
34 |
+
)
|
35 |
+
|
36 |
+
def remove_tags(soup: BeautifulSoup) -> str:
|
37 |
+
for data in soup(["style", "script"]):
|
38 |
+
# Remove tags
|
39 |
+
data.decompose()
|
40 |
+
# return data by retrieving the tag content
|
41 |
+
return " ".join(soup.stripped_strings)
|
42 |
+
|
43 |
+
def read_webpage(url: str) -> str:
|
44 |
+
print(f"Getting the response from url : {url})")
|
45 |
+
response = requests.get(url)
|
46 |
+
html_content = response.content
|
47 |
+
|
48 |
+
# Parse the HTML content using BeautifulSoup
|
49 |
+
soup = BeautifulSoup(html_content, "html.parser")
|
50 |
+
|
51 |
+
# Get all the text content from the relevant HTML tags
|
52 |
+
text_content = remove_tags(soup)
|
53 |
+
|
54 |
+
# for tag in ["p", "h1", "h2", "h3", "h4", "h5", "h6", "li", "div"]:
|
55 |
+
# for element in soup.find_all(tag):
|
56 |
+
# text_content += element.get_text() + " "
|
57 |
+
|
58 |
+
print(text_content)
|
59 |
+
return text_content
|
60 |
+
|
61 |
+
def grab_transcript(url):
|
62 |
+
loader = YoutubeLoader.from_youtube_url(url)
|
63 |
+
transcript = loader.load()
|
64 |
+
return transcript[0].page_content
|
65 |
+
|
66 |
+
def process_webpages(urls: List[str]):
|
67 |
+
# A set to keep track of visited pages
|
68 |
+
visited_pages = set()
|
69 |
+
content = []
|
70 |
+
|
71 |
+
for url in urls:
|
72 |
+
aggregated_text = ""
|
73 |
+
|
74 |
+
try:
|
75 |
+
|
76 |
+
if 'youtube' not in url:
|
77 |
+
|
78 |
+
visited_pages.add(url)
|
79 |
+
aggregated_text += f"\nGetting the content of {url}:\n"
|
80 |
+
|
81 |
+
try:
|
82 |
+
aggregated_text += read_webpage(url)
|
83 |
+
except e as Exception:
|
84 |
+
|
85 |
+
print(read_webpage(url))
|
86 |
+
aggregated_text += "No Transcript found"
|
87 |
+
|
88 |
+
else:
|
89 |
+
# Youtube work:
|
90 |
+
aggregated_text += f"\nGetting the transcription of {url}:\n"
|
91 |
+
aggregated_text += grab_transcript(url)
|
92 |
+
|
93 |
+
except Exception as e:
|
94 |
+
print(e)
|
95 |
+
|
96 |
+
content.append(aggregated_text)
|
97 |
+
|
98 |
+
return content
|
99 |
+
|
100 |
+
def extract_urls(text):
|
101 |
+
url_regex = r"(https?://\S+)"
|
102 |
+
urls = re.findall(url_regex, text)
|
103 |
+
return urls
|
104 |
+
|
105 |
+
def add_text(history, text):
|
106 |
+
|
107 |
+
print(history)
|
108 |
+
history = history + [(text, None)]
|
109 |
+
|
110 |
+
return history, ""
|
111 |
+
|
112 |
+
def create_db_from_urls(urls: str) -> FAISS:
|
113 |
+
|
114 |
+
content = process_webpages(urls)
|
115 |
+
# 1K CHUNK SIZE
|
116 |
+
text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=100)
|
117 |
+
print(len(content), len(urls))
|
118 |
+
docs = text_splitter.create_documents(content)
|
119 |
+
|
120 |
+
# print(docs[0])
|
121 |
+
global db
|
122 |
+
if(type(db) != str):
|
123 |
+
# local_db =
|
124 |
+
|
125 |
+
docs += list(db.docstore._dict.values())
|
126 |
+
|
127 |
+
print(docs)
|
128 |
+
|
129 |
+
# db = FAISS.from_documents(docs, embeddings)
|
130 |
+
# print(db.docstore())
|
131 |
+
# print(docs[0])
|
132 |
+
db = FAISS.from_documents(docs, embeddings)
|
133 |
+
|
134 |
+
return db
|
135 |
+
|
136 |
+
# How to access the file? Where is it saved?
|
137 |
+
def add_file(history, files):
|
138 |
+
history = []
|
139 |
+
files = files[0]
|
140 |
+
docs = []
|
141 |
+
for file in files:
|
142 |
+
|
143 |
+
loader = UnstructuredPDFLoader(file.name)
|
144 |
+
text = loader.load()
|
145 |
+
# pdf_content = pdf2text(file.name)
|
146 |
+
|
147 |
+
docs += text_splitter.split_documents(text)
|
148 |
+
# print(docs[0])
|
149 |
+
|
150 |
+
global db
|
151 |
+
if(type(db) != str):
|
152 |
+
# local_db =
|
153 |
+
|
154 |
+
docs += list(db.docstore._dict.values())
|
155 |
+
|
156 |
+
print(docs)
|
157 |
+
history = history + [(f"{len(files)} PDF(s) Uploaded", None),]
|
158 |
+
|
159 |
+
|
160 |
+
db = FAISS.from_documents(docs, embeddings)
|
161 |
+
print(f"History in add file: {history}")
|
162 |
+
# print(db.docstore())
|
163 |
+
print(type(history), type(history[0]))
|
164 |
+
return ([history,],)
|
165 |
+
|
166 |
+
def qa_retrieve(chatlog, ):
|
167 |
+
|
168 |
+
print(f"Chatlog qa: {chatlog}")
|
169 |
+
query = chatlog[-1][0]
|
170 |
+
docs = ""
|
171 |
+
# Extracting urls from query
|
172 |
+
# urls = None
|
173 |
+
|
174 |
+
# if (urls):
|
175 |
+
# create_db_from_urls(urls)
|
176 |
+
# global db
|
177 |
+
# if(type(db) != str):
|
178 |
+
# # local_db =
|
179 |
+
# docs = list(db.docstore._dict.values())
|
180 |
+
# print(docs)
|
181 |
+
# db = FAISS.from_documents(docs, embeddings)
|
182 |
+
# # db.merge_from(local_db)
|
183 |
+
# else:
|
184 |
+
# db = FAISS.from_documents(docs, embeddings)
|
185 |
+
global db
|
186 |
+
print(db)
|
187 |
+
# if (type(db) == str):
|
188 |
+
# discussion = [j for i in chatlog for j in i]
|
189 |
+
# messages = [[SystemMessage(content = "You are Wikibot. You are a WikiPedia assistant that that can digest articles and answer questions based on your library of content.")]]
|
190 |
+
# messages += [([HumanMessage(content = x[0])],AIMessage(content = x[-1])) for x in chatlog[:-1] ]
|
191 |
+
|
192 |
+
# messages = [j for i in messages for j in i]
|
193 |
+
# print(messages)
|
194 |
+
|
195 |
+
# messages.append(HumanMessage(content = chatlog[-1][0]))
|
196 |
+
# print(messages)
|
197 |
+
|
198 |
+
# response = llm(messages=messages
|
199 |
+
# ).content
|
200 |
+
|
201 |
+
# else:
|
202 |
+
|
203 |
+
docs = db.similarity_search(query, k=4)
|
204 |
+
|
205 |
+
docs_page_content = " ".join([d.page_content for d in docs])
|
206 |
+
|
207 |
+
|
208 |
+
prompt = PromptTemplate(
|
209 |
+
input_variables=["question", "docs"],
|
210 |
+
template="""
|
211 |
+
As a consultant, your role is to assist the user in analyzing different cases of lord and tenants behaviours.
|
212 |
+
You will help the user with questions related to any of the information provided by the documents. You will make sure to give as much help as possible
|
213 |
+
even if sometimes the information seeked does not exist. Your priority is to find the information asked by the user and if it doesn't exist you will try
|
214 |
+
to think of how to answer using your own thoughts.
|
215 |
+
|
216 |
+
Answer the following question: {question}
|
217 |
+
Use the following documents: {docs}
|
218 |
+
|
219 |
+
If you feel like you don't have enough information to answer the question, say "I don't know".
|
220 |
+
|
221 |
+
""",
|
222 |
+
)
|
223 |
+
|
224 |
+
|
225 |
+
# llm = BardLLM()
|
226 |
+
chain = LLMChain(llm=llm, prompt = prompt)
|
227 |
+
|
228 |
+
response = chain.run(question=query, docs=docs_page_content)
|
229 |
+
|
230 |
+
chatlog[-1][1] = response
|
231 |
+
return chatlog
|
232 |
+
|
233 |
+
def flush():
|
234 |
+
global db
|
235 |
+
db = ""
|
236 |
+
return None
|
237 |
+
|
238 |
+
with gr.Blocks() as demo:
|
239 |
+
chatbot = gr.Chatbot([], elem_id="chatbot").style(height=750)
|
240 |
+
|
241 |
+
with gr.Row():
|
242 |
+
with gr.Column(scale=0.65):
|
243 |
+
txt = gr.components.Textbox(
|
244 |
+
placeholder="Ask me anything",show_label=False
|
245 |
+
)
|
246 |
+
|
247 |
+
with gr.Column(scale=0.15, min_width=0):
|
248 |
+
btn = gr.UploadButton("📁", file_types=["text"], file_count = 'multiple')
|
249 |
+
|
250 |
+
# with gr.Row():
|
251 |
+
|
252 |
+
# with gr.Column(scale=0.85):
|
253 |
+
|
254 |
+
# url = gr.components.Textbox(
|
255 |
+
|
256 |
+
# label="Website URLs",
|
257 |
+
# placeholder="https://www.example.org/ https://www.example.com/",
|
258 |
+
# )
|
259 |
+
|
260 |
+
with gr.Column(scale=0.15, min_width = 0):
|
261 |
+
send_btn = gr.Button("📨")
|
262 |
+
|
263 |
+
with gr.Row():
|
264 |
+
with gr.Column():
|
265 |
+
clear = gr.Button("Clear")
|
266 |
+
pdf_content = gr.Textbox("", visible = False)
|
267 |
+
txt.submit(add_text, [chatbot, txt], [chatbot, txt]).then(
|
268 |
+
qa_retrieve, [chatbot], chatbot
|
269 |
+
).then(lambda : (None), outputs = [ pdf_content])
|
270 |
+
btn.upload(add_file, [chatbot, btn], [chatbot,], batch = True).then(qa_retrieve, [chatbot], chatbot)
|
271 |
+
|
272 |
+
send_btn.click(add_text, [chatbot, txt, ], [chatbot, txt]).then(
|
273 |
+
qa_retrieve, [chatbot, ], chatbot).then(lambda : None, outputs = [ pdf_content])
|
274 |
+
|
275 |
+
clear.click(flush, None, outputs = chatbot, queue=False)
|
276 |
+
|
277 |
+
demo.queue(concurrency_count = 4)
|
278 |
+
demo.launch()
|
rent_data/index.faiss
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b2944eec8a548a3d4c97a9f26d2a87054ede0b063f9139f2f788c37888f32b02
|
3 |
+
size 171764781
|
rent_data/index.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:76630ed8bbc38dd10e113c088bca6a0531cd321469a3bcec99f427920468a35d
|
3 |
+
size 60863161
|
requirements.txt
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
unstructured
|
3 |
+
pdf2image
|
4 |
+
langchain
|
5 |
+
gradio
|
6 |
+
openai
|
7 |
+
sentence_transformers
|
8 |
+
youtube-transcript-api
|
9 |
+
FAISS-gpu
|
10 |
+
BeautifulSoup4
|