Upload 5 files
Browse files- Dockerfile +33 -0
- README.md +4 -10
- app.py +101 -0
- packages.txt +0 -0
- requirements.txt +9 -0
Dockerfile
ADDED
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
FROM python:3.13
|
2 |
+
|
3 |
+
WORKDIR /app
|
4 |
+
|
5 |
+
COPY ./requirements.txt /app/requirements.txt
|
6 |
+
COPY ./packages.txt /app/packages.txt
|
7 |
+
|
8 |
+
# Install curl and other dependencies
|
9 |
+
RUN apt-get update && apt-get install -y curl && apt-get clean && rm -rf /var/lib/apt/lists/*
|
10 |
+
|
11 |
+
# Install Ollama
|
12 |
+
RUN curl -fsSL https://ollama.ai/install.sh | sh
|
13 |
+
|
14 |
+
RUN apt-get update && xargs -r -a /app/packages.txt apt-get install -y && rm -rf /var/lib/apt/lists/*
|
15 |
+
RUN pip3 install --no-cache-dir -r /app/requirements.txt
|
16 |
+
|
17 |
+
# User
|
18 |
+
RUN useradd -m -u 1000 user
|
19 |
+
USER user
|
20 |
+
ENV HOME /home/user
|
21 |
+
ENV PATH $HOME/.local/bin:$PATH
|
22 |
+
|
23 |
+
WORKDIR $HOME
|
24 |
+
RUN mkdir app
|
25 |
+
WORKDIR $HOME/app
|
26 |
+
COPY . $HOME/app
|
27 |
+
|
28 |
+
EXPOSE 8501
|
29 |
+
CMD ollama serve & sleep 5 && ollama pull llama3.2 && streamlit run app.py \
|
30 |
+
--server.headless true \
|
31 |
+
--server.enableCORS false \
|
32 |
+
--server.enableXsrfProtection false \
|
33 |
+
--server.fileWatcherType none
|
README.md
CHANGED
@@ -1,11 +1,5 @@
|
|
1 |
-
|
2 |
-
title: Projectresilience Assistant
|
3 |
-
emoji: 📚
|
4 |
-
colorFrom: blue
|
5 |
-
colorTo: indigo
|
6 |
-
sdk: docker
|
7 |
-
pinned: false
|
8 |
-
license: mit
|
9 |
-
---
|
10 |
|
11 |
-
|
|
|
|
|
|
1 |
+
# QA Bot for Project Resilience
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
+
## Overview
|
4 |
+
|
5 |
+
The QA Bot for [Project Resilience](https://www.itu.int/en/ITU-T/extcoop/ai-data-commons/Pages/project-resilience.aspx) is designed to assist public by answering queries related to Project Resilience. This bot leverages Generative AI to provide contextual responses to questions regarding sustainable development goals (SDGs), project proposals, contributions, and platform navigation.
|
app.py
ADDED
@@ -0,0 +1,101 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
from langchain.embeddings import HuggingFaceEmbeddings
|
3 |
+
from langchain.vectorstores import FAISS
|
4 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
5 |
+
from langchain_community.llms import Ollama
|
6 |
+
import requests
|
7 |
+
|
8 |
+
# Function to fetch GitHub repo data
|
9 |
+
def fetch_github_data(repo_url):
|
10 |
+
parts = repo_url.split('/')
|
11 |
+
owner, repo = parts[-2], parts[-1]
|
12 |
+
|
13 |
+
headers = {'Accept': 'application/vnd.github.v3+json'}
|
14 |
+
base_url = 'https://api.github.com'
|
15 |
+
|
16 |
+
content = ""
|
17 |
+
repo_response = requests.get(f"{base_url}/repos/{owner}/{repo}", headers=headers)
|
18 |
+
if repo_response.status_code == 200:
|
19 |
+
repo_data = repo_response.json()
|
20 |
+
content += f"Description: {repo_data.get('description', '')}\n"
|
21 |
+
|
22 |
+
readme_response = requests.get(f"{base_url}/repos/{owner}/{repo}/readme", headers=headers)
|
23 |
+
if readme_response.status_code == 200:
|
24 |
+
import base64
|
25 |
+
readme_data = readme_response.json()
|
26 |
+
content += base64.b64decode(readme_data['content']).decode('utf-8') + "\n"
|
27 |
+
|
28 |
+
return content
|
29 |
+
|
30 |
+
# Function to create vector store
|
31 |
+
def create_vector_store(text_data):
|
32 |
+
text_splitter = RecursiveCharacterTextSplitter(
|
33 |
+
chunk_size=1000,
|
34 |
+
chunk_overlap=200
|
35 |
+
)
|
36 |
+
chunks = text_splitter.split_text(text_data)
|
37 |
+
|
38 |
+
embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
|
39 |
+
vector_store = FAISS.from_texts(chunks, embeddings)
|
40 |
+
return vector_store
|
41 |
+
|
42 |
+
# Streamlit app
|
43 |
+
def main():
|
44 |
+
st.title("Project Resilience Q&A Assistant")
|
45 |
+
st.write("Ask anything about Project Resilience - answers always come from repo data!")
|
46 |
+
|
47 |
+
# Hardcoded GitHub URL
|
48 |
+
github_url = 'https://github.com/Project-Resilience/platform'
|
49 |
+
repo_data = fetch_github_data(github_url)
|
50 |
+
|
51 |
+
# Initialize session state
|
52 |
+
if 'vector_store' not in st.session_state:
|
53 |
+
st.session_state.vector_store = create_vector_store(repo_data)
|
54 |
+
st.session_state.llm = Ollama(model="llama3.2", temperature=0.7)
|
55 |
+
|
56 |
+
# Question input
|
57 |
+
question = st.text_input("Ask a question about the project")
|
58 |
+
|
59 |
+
# Get and display answer
|
60 |
+
if question:
|
61 |
+
with st.spinner("Generating answer..."):
|
62 |
+
# Retrieve top-k documents
|
63 |
+
k = 3
|
64 |
+
docs = st.session_state.vector_store.similarity_search(question, k=k)
|
65 |
+
|
66 |
+
# Extract the text from the documents
|
67 |
+
context = "\n\n".join([doc.page_content for doc in docs])
|
68 |
+
|
69 |
+
# Create the custom prompt for zero-shot prompting
|
70 |
+
prompt = (
|
71 |
+
f"You are a helpful assistant answers to the question about project Project Resilience.\n"
|
72 |
+
f"Based on the following context, answer the question:\n\n"
|
73 |
+
f"Context:\n{context}\n\n"
|
74 |
+
f"Question: {question}\n\n"
|
75 |
+
f"If the question cannot be answered by the document, say so.\n\n"
|
76 |
+
f"Answer:"
|
77 |
+
)
|
78 |
+
|
79 |
+
# Generate the answer using the language model
|
80 |
+
answer_container = st.empty()
|
81 |
+
stream = st.session_state.llm.stream(prompt)
|
82 |
+
answer = ""
|
83 |
+
for chunk in stream:
|
84 |
+
answer += chunk
|
85 |
+
answer_container.write(answer)
|
86 |
+
|
87 |
+
|
88 |
+
|
89 |
+
|
90 |
+
# Sidebar with additional info
|
91 |
+
st.sidebar.header("Project Resilience Assistant")
|
92 |
+
st.sidebar.write("""
|
93 |
+
Project Resilience's platform for decision makers, data scientists and the public.
|
94 |
+
|
95 |
+
Project Resilience, initiated under the Global Initiative on AI and Data Commons, is a collaborative effort to build a public AI utility that could inform and help address global decision-augmentation challenges.
|
96 |
+
|
97 |
+
The project empowers a global community of innovators, thought leaders, and the public to enhance and use a shared collection of data and AI tools, improving preparedness, intervention, and response to environmental, health, information, or economic threats in our communities. It also supports broader efforts toward achieving the Sustainable Development Goals (SDGs).
|
98 |
+
""")
|
99 |
+
|
100 |
+
if __name__ == "__main__":
|
101 |
+
main()
|
packages.txt
ADDED
File without changes
|
requirements.txt
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
faiss-cpu==1.10.0
|
2 |
+
langchain==0.3.20
|
3 |
+
langchain-community==0.3.19
|
4 |
+
langchain-ollama==0.2.3
|
5 |
+
langgraph==0.3.5
|
6 |
+
ollama==0.4.7
|
7 |
+
requests==2.32.3
|
8 |
+
sentence-transformers==3.4.1
|
9 |
+
streamlit==1.43.0
|