Spaces:
Runtime error
Runtime error
uploaded the model and changed the inference library
Browse files
app.py
ADDED
@@ -0,0 +1,228 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
from gradio_pdf import PDF
|
3 |
+
from qdrant_client import models, QdrantClient
|
4 |
+
from sentence_transformers import SentenceTransformer
|
5 |
+
from PyPDF2 import PdfReader
|
6 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
7 |
+
from langchain.callbacks.manager import CallbackManager
|
8 |
+
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
|
9 |
+
# from langchain.llms import LlamaCpp
|
10 |
+
from langchain.vectorstores import Qdrant
|
11 |
+
from qdrant_client.http import models
|
12 |
+
# from langchain.llms import CTransformers
|
13 |
+
from ctransformers import AutoModelForCausalLM
|
14 |
+
|
15 |
+
|
16 |
+
# loading the embedding model -
|
17 |
+
|
18 |
+
encoder = SentenceTransformer('jinaai/jina-embedding-b-en-v1')
|
19 |
+
|
20 |
+
print("embedding model loaded.............................")
|
21 |
+
print("####################################################")
|
22 |
+
|
23 |
+
# loading the LLM
|
24 |
+
|
25 |
+
callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])
|
26 |
+
|
27 |
+
print("loading the LLM......................................")
|
28 |
+
|
29 |
+
llm = LlamaCpp(
|
30 |
+
model_path="./llama-2-7b-chat.Q3_K_S.gguf",
|
31 |
+
n_ctx=2048,
|
32 |
+
f16_kv=True, # MUST set to True, otherwise you will run into problem after a couple of calls
|
33 |
+
callback_manager=callback_manager,
|
34 |
+
verbose=True,
|
35 |
+
)
|
36 |
+
|
37 |
+
# llm = AutoModelForCausalLM.from_pretrained("TheBloke/Llama-2-7B-Chat-GGUF",
|
38 |
+
# model_file="llama-2-7b-chat.Q3_K_S.gguf",
|
39 |
+
# model_type="llama",
|
40 |
+
# temperature = 0.2,
|
41 |
+
# repetition_penalty = 1.5,
|
42 |
+
# max_new_tokens = 300,
|
43 |
+
# )
|
44 |
+
|
45 |
+
|
46 |
+
|
47 |
+
print("LLM loaded........................................")
|
48 |
+
print("################################################################")
|
49 |
+
|
50 |
+
# def get_chunks(text):
|
51 |
+
# text_splitter = RecursiveCharacterTextSplitter(
|
52 |
+
# # seperator = "\n",
|
53 |
+
# chunk_size = 250,
|
54 |
+
# chunk_overlap = 50,
|
55 |
+
# length_function = len,
|
56 |
+
# )
|
57 |
+
|
58 |
+
# chunks = text_splitter.split_text(text)
|
59 |
+
# return chunks
|
60 |
+
|
61 |
+
|
62 |
+
# pdf_path = './100 Weird Facts About the Human Body.pdf'
|
63 |
+
|
64 |
+
|
65 |
+
# reader = PdfReader(pdf_path)
|
66 |
+
# text = ""
|
67 |
+
# num_of_pages = len(reader.pages)
|
68 |
+
|
69 |
+
# for page in range(num_of_pages):
|
70 |
+
# current_page = reader.pages[page]
|
71 |
+
# text += current_page.extract_text()
|
72 |
+
|
73 |
+
|
74 |
+
# chunks = get_chunks(text)
|
75 |
+
# print(chunks)
|
76 |
+
# print("Chunks are ready.....................................")
|
77 |
+
# print("######################################################")
|
78 |
+
|
79 |
+
# client = QdrantClient(path = "./db")
|
80 |
+
# print("db created................................................")
|
81 |
+
# print("#####################################################################")
|
82 |
+
|
83 |
+
# client.recreate_collection(
|
84 |
+
# collection_name="my_facts",
|
85 |
+
# vectors_config=models.VectorParams(
|
86 |
+
# size=encoder.get_sentence_embedding_dimension(), # Vector size is defined by used model
|
87 |
+
# distance=models.Distance.COSINE,
|
88 |
+
# ),
|
89 |
+
# )
|
90 |
+
|
91 |
+
# print("Collection created........................................")
|
92 |
+
# print("#########################################################")
|
93 |
+
|
94 |
+
|
95 |
+
|
96 |
+
# li = []
|
97 |
+
# for i in range(len(chunks)):
|
98 |
+
# li.append(i)
|
99 |
+
|
100 |
+
# dic = zip(li, chunks)
|
101 |
+
# dic= dict(dic)
|
102 |
+
|
103 |
+
# client.upload_records(
|
104 |
+
# collection_name="my_facts",
|
105 |
+
# records=[
|
106 |
+
# models.Record(
|
107 |
+
# id=idx,
|
108 |
+
# vector=encoder.encode(dic[idx]).tolist(),
|
109 |
+
# payload= {dic[idx][:5] : dic[idx]}
|
110 |
+
# ) for idx in dic.keys()
|
111 |
+
# ],
|
112 |
+
# )
|
113 |
+
|
114 |
+
# print("Records uploaded........................................")
|
115 |
+
# print("###########################################################")
|
116 |
+
|
117 |
+
def chat(file, question):
|
118 |
+
def get_chunks(text):
|
119 |
+
text_splitter = RecursiveCharacterTextSplitter(
|
120 |
+
# seperator = "\n",
|
121 |
+
chunk_size = 250,
|
122 |
+
chunk_overlap = 50,
|
123 |
+
length_function = len,
|
124 |
+
)
|
125 |
+
|
126 |
+
chunks = text_splitter.split_text(text)
|
127 |
+
return chunks
|
128 |
+
|
129 |
+
|
130 |
+
pdf_path = file
|
131 |
+
|
132 |
+
|
133 |
+
reader = PdfReader(pdf_path)
|
134 |
+
text = ""
|
135 |
+
num_of_pages = len(reader.pages)
|
136 |
+
|
137 |
+
for page in range(num_of_pages):
|
138 |
+
current_page = reader.pages[page]
|
139 |
+
text += current_page.extract_text()
|
140 |
+
|
141 |
+
|
142 |
+
chunks = get_chunks(text)
|
143 |
+
# print(chunks)
|
144 |
+
# print("Chunks are ready.....................................")
|
145 |
+
# print("######################################################")
|
146 |
+
|
147 |
+
client = QdrantClient(path = "./db")
|
148 |
+
# print("db created................................................")
|
149 |
+
# print("#####################################################################")
|
150 |
+
|
151 |
+
client.recreate_collection(
|
152 |
+
collection_name="my_facts",
|
153 |
+
vectors_config=models.VectorParams(
|
154 |
+
size=encoder.get_sentence_embedding_dimension(), # Vector size is defined by used model
|
155 |
+
distance=models.Distance.COSINE,
|
156 |
+
),
|
157 |
+
)
|
158 |
+
|
159 |
+
# print("Collection created........................................")
|
160 |
+
# print("#########################################################")
|
161 |
+
|
162 |
+
|
163 |
+
|
164 |
+
li = []
|
165 |
+
for i in range(len(chunks)):
|
166 |
+
li.append(i)
|
167 |
+
|
168 |
+
dic = zip(li, chunks)
|
169 |
+
dic= dict(dic)
|
170 |
+
|
171 |
+
client.upload_records(
|
172 |
+
collection_name="my_facts",
|
173 |
+
records=[
|
174 |
+
models.Record(
|
175 |
+
id=idx,
|
176 |
+
vector=encoder.encode(dic[idx]).tolist(),
|
177 |
+
payload= {dic[idx][:5] : dic[idx]}
|
178 |
+
) for idx in dic.keys()
|
179 |
+
],
|
180 |
+
)
|
181 |
+
|
182 |
+
# print("Records uploaded........................................")
|
183 |
+
# print("###########################################################")
|
184 |
+
|
185 |
+
|
186 |
+
hits = client.search(
|
187 |
+
collection_name="my_facts",
|
188 |
+
query_vector=encoder.encode(question).tolist(),
|
189 |
+
limit=3
|
190 |
+
)
|
191 |
+
context = []
|
192 |
+
for hit in hits:
|
193 |
+
context.append(list(hit.payload.values())[0])
|
194 |
+
|
195 |
+
context = context[0] + context[1] + context[2]
|
196 |
+
|
197 |
+
system_prompt = """You are a helpful assistant, you will use the provided context to answer user questions.
|
198 |
+
Read the given context before answering questions and think step by step. If you can not answer a user question based on
|
199 |
+
the provided context, inform the user. Do not use any other information for answering user. Provide a detailed answer to the question."""
|
200 |
+
|
201 |
+
|
202 |
+
B_INST, E_INST = "[INST]", "[/INST]"
|
203 |
+
|
204 |
+
B_SYS, E_SYS = "<<SYS>>\n", "\n<</SYS>>\n\n"
|
205 |
+
|
206 |
+
SYSTEM_PROMPT = B_SYS + system_prompt + E_SYS
|
207 |
+
|
208 |
+
instruction = f"""
|
209 |
+
Context: {context}
|
210 |
+
User: {question}"""
|
211 |
+
|
212 |
+
prompt_template = B_INST + SYSTEM_PROMPT + instruction + E_INST
|
213 |
+
|
214 |
+
result = llm(prompt_template)
|
215 |
+
return result
|
216 |
+
|
217 |
+
|
218 |
+
screen = gr.Interface(
|
219 |
+
fn = chat,
|
220 |
+
inputs = [PDF(label="Upload a PDF", interactive=True), gr.Textbox(lines = 10, placeholder = "Enter your question here π")],
|
221 |
+
outputs = gr.Textbox(lines = 10, placeholder = "Your answer will be here soon π"),
|
222 |
+
title="Q&A with PDF π©π»βπ»πβπ»π‘",
|
223 |
+
description="This app facilitates a conversation with PDFs available on https://www.delo.si/assets/media/other/20110728/100%20Weird%20Facts%20About%20the%20Human%20Body.pdfπ‘",
|
224 |
+
theme="soft",
|
225 |
+
# examples=["Hello", "what is the speed of human nerve impulses?"],
|
226 |
+
)
|
227 |
+
|
228 |
+
screen.launch()
|