Delete app.py
Browse files
app.py
DELETED
@@ -1,127 +0,0 @@
|
|
1 |
-
import gradio as gr
|
2 |
-
from huggingface_hub import InferenceClient
|
3 |
-
from typing import List, Tuple
|
4 |
-
import fitz # PyMuPDF
|
5 |
-
from sentence_transformers import SentenceTransformer, util
|
6 |
-
import numpy as np
|
7 |
-
import faiss
|
8 |
-
|
9 |
-
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
10 |
-
|
11 |
-
# Placeholder for the app's state
|
12 |
-
class MyApp:
|
13 |
-
def __init__(self) -> None:
|
14 |
-
self.documents = []
|
15 |
-
self.embeddings = None
|
16 |
-
self.index = None
|
17 |
-
self.load_pdf("THEDIA1.pdf")
|
18 |
-
self.build_vector_db()
|
19 |
-
|
20 |
-
def load_pdf(self, file_path: str) -> None:
|
21 |
-
"""Extracts text from a PDF file and stores it in the app's documents."""
|
22 |
-
doc = fitz.open(file_path)
|
23 |
-
self.documents = []
|
24 |
-
for page_num in range(len(doc)):
|
25 |
-
page = doc[page_num]
|
26 |
-
text = page.get_text()
|
27 |
-
self.documents.append({"page": page_num + 1, "content": text})
|
28 |
-
print("PDF processed successfully!")
|
29 |
-
|
30 |
-
def build_vector_db(self) -> None:
|
31 |
-
"""Builds a vector database using the content of the PDF."""
|
32 |
-
model = SentenceTransformer('all-MiniLM-L6-v2')
|
33 |
-
self.embeddings = model.encode([doc["content"] for doc in self.documents])
|
34 |
-
self.index = faiss.IndexFlatL2(self.embeddings.shape[1])
|
35 |
-
self.index.add(np.array(self.embeddings))
|
36 |
-
print("Vector database built successfully!")
|
37 |
-
|
38 |
-
def search_documents(self, query: str, k: int = 3) -> List[str]:
|
39 |
-
"""Searches for relevant documents using vector similarity."""
|
40 |
-
model = SentenceTransformer('all-MiniLM-L6-v2')
|
41 |
-
query_embedding = model.encode([query])
|
42 |
-
D, I = self.index.search(np.array(query_embedding), k)
|
43 |
-
results = [self.documents[i]["content"] for i in I[0]]
|
44 |
-
return results if results else ["No relevant documents found."]
|
45 |
-
|
46 |
-
app = MyApp()
|
47 |
-
|
48 |
-
def respond(
|
49 |
-
message: str,
|
50 |
-
history: List[Tuple[str, str]],
|
51 |
-
system_message: str,
|
52 |
-
max_tokens: int,
|
53 |
-
temperature: float,
|
54 |
-
top_p: float,
|
55 |
-
):
|
56 |
-
system_message = (
|
57 |
-
"You are a knowledgeable DBT (Dialectical Behavior Therapy) coach. You greet users warmly and ask questions like a real counselor. "
|
58 |
-
"You are concise, respectful, and a good listener. You use the DBT book to guide users through DBT exercises and provide helpful information. "
|
59 |
-
"When needed, you ask one follow-up question at a time to guide the user to ask appropriate questions. "
|
60 |
-
"You avoid giving suggestions if any dangerous act is mentioned by the user and refer them to call someone or emergency services. "
|
61 |
-
"Your responses are accurate and concise, and you maintain a professional and supportive tone throughout."
|
62 |
-
)
|
63 |
-
|
64 |
-
messages = [{"role": "system", "content": system_message}]
|
65 |
-
|
66 |
-
for val in history:
|
67 |
-
if val[0]:
|
68 |
-
messages.append({"role": "user", "content": val[0]})
|
69 |
-
if val[1]:
|
70 |
-
messages.append({"role": "assistant", "content": val[1]})
|
71 |
-
|
72 |
-
messages.append({"role": "user", "content": message})
|
73 |
-
|
74 |
-
# RAG - Retrieve relevant documents
|
75 |
-
try:
|
76 |
-
retrieved_docs = app.search_documents(message)
|
77 |
-
context = "\n".join(retrieved_docs)
|
78 |
-
messages.append({"role": "system", "content": "Relevant documents: " + context})
|
79 |
-
|
80 |
-
response = ""
|
81 |
-
response_buffer = []
|
82 |
-
for message in client.chat_completion(
|
83 |
-
messages,
|
84 |
-
max_tokens=max_tokens,
|
85 |
-
stream=True,
|
86 |
-
temperature=temperature,
|
87 |
-
top_p=top_p,
|
88 |
-
):
|
89 |
-
token = message.choices[0].delta.content
|
90 |
-
response += token
|
91 |
-
response_buffer.append(token)
|
92 |
-
if token.endswith('.') or token.endswith('?'):
|
93 |
-
yield ''.join(response_buffer)
|
94 |
-
response_buffer = []
|
95 |
-
|
96 |
-
if response_buffer:
|
97 |
-
yield ''.join(response_buffer)
|
98 |
-
|
99 |
-
except Exception as e:
|
100 |
-
yield f"An error occurred: {str(e)}"
|
101 |
-
|
102 |
-
demo = gr.Blocks()
|
103 |
-
|
104 |
-
with demo:
|
105 |
-
gr.Markdown("🧘♀️ **Dialectical Behaviour Therapy**")
|
106 |
-
gr.Markdown(
|
107 |
-
"‼️Disclaimer: This chatbot is based on a DBT exercise book that is publicly available. "
|
108 |
-
"We are not medical practitioners, and the use of this chatbot is at your own responsibility.‼️"
|
109 |
-
)
|
110 |
-
|
111 |
-
chatbot = gr.ChatInterface(
|
112 |
-
respond,
|
113 |
-
examples=[
|
114 |
-
["I feel overwhelmed with work."],
|
115 |
-
["Can you guide me through a quick meditation?"],
|
116 |
-
["How do I stop worrying about things I can't control?"],
|
117 |
-
["What are some DBT skills for managing anxiety?"],
|
118 |
-
["Can you explain mindfulness in DBT?"],
|
119 |
-
["I am interested in DBT exercises"],
|
120 |
-
["I feel restless. Please help me."],
|
121 |
-
["I have destructive thoughts coming to my mind repetitively."]
|
122 |
-
],
|
123 |
-
title='Dialectical Behaviour Therapy Assistant👩⚕️'
|
124 |
-
)
|
125 |
-
|
126 |
-
if __name__ == "__main__":
|
127 |
-
demo.launch()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|