Upload 40 files

.env

@@ -0,0 +1,11 @@
+FASTAPI_CONFIG=production
+SECRET=zu=*nck@&r26rsa$2qv8a7g-p$slw79ym81mylj6s**w(da6&#
+OPENAI_API_KEY=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
+
+DATABASE_USER=hectool_ai_filter
+DATABASE_PASSWORD=nixtz3orhepndjw4m1D4
+DATABASE_HOST=localhost
+DATABASE_PORT=5432
+DATABASE_NAME=javea
+
+CONFIDENCE_PARAMETER=50

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+faiss_javea.index filter=lfs diff=lfs merge=lfs -text

Dockerfile

@@ -0,0 +1,21 @@
+FROM python:3.10
+
+ENV PYTHONDONTWRITEBYCODE 1
+ENV PYTHONBUFFERED 1
+ENV TRANSFORMERS_CACHE=/istvanai/.cache/huggingface/transformers
+
+RUN pip install --upgrade pip
+
+WORKDIR /code
+
+ADD . /code
+
+RUN chmod -R 777 /code
+
+RUN pip install -r requirements.txt
+
+COPY . .
+
+EXPOSE 7860
+
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

faiss_javea.index

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:983785b39b3e03719744c884620b11e0b64975ae4388157865abe0b205c02993
+size 3947565

main.py

@@ -0,0 +1,3 @@
+from project import create_app
+
+app = create_app()

project/__init__.py

@@ -0,0 +1,30 @@
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.staticfiles import StaticFiles
+
+from project.admin import AdminCustom, MessagePairAdmin
+from project.bot.auth import authentication_backend_admin
+from project.database import engine
+
+
+def create_app() -> FastAPI:
+    app = FastAPI()
+
+    from project.bot import bot_router
+    app.include_router(bot_router, tags=['bot'])
+
+    from project.ws import ws_router
+    app.include_router(ws_router, tags=['ws'])
+
+    app.add_middleware(
+        CORSMiddleware,
+        allow_origins=["*"],
+        allow_methods=["*"],
+        allow_headers=["*"],
+    )
+
+    app.mount('/static', StaticFiles(directory="static"), name="static")
+
+    admin = AdminCustom(app, engine, authentication_backend=authentication_backend_admin)
+    admin.add_view(MessagePairAdmin)
+    return app

project/admin.py

@@ -0,0 +1,51 @@
+from sqladmin import ModelView, action
+from sqladmin import Admin
+
+from fastapi import Request, Response
+from fastapi.responses import RedirectResponse
+from fastapi.templating import Jinja2Templates
+
+import logging
+
+from project.bot.models import MessagePair
+
+template = Jinja2Templates(directory='templates')
+
+logger = logging.getLogger(__name__)
+
+
+class MessagePairAdmin(ModelView, model=MessagePair):
+    category = 'Message History'
+    column_list = [MessagePair.user_message, MessagePair.bot_response, MessagePair.country]
+    is_async = True
+    icon = "fa-solid fa-message"
+    name = "Message Pair"
+    name_plural = "Message pairs"
+    column_searchable_list = [MessagePair.user_message, MessagePair.bot_response, MessagePair.country]
+    column_export_list = ['user_message', 'bot_response', 'country']
+    can_edit = False
+    can_create = False
+    column_default_sort = ('id', True)
+    column_formatters = {
+        MessagePair.bot_response: lambda m, a: m.bot_response[:50] + '...' if len(m.bot_response) > 50
+        else m.bot_response}
+
+
+class AdminCustom(Admin):
+
+    async def login(self, request: Request) -> Response:
+        assert self.authentication_backend is not None
+
+        context = {"request": request, "error": ""}
+
+        if request.method == "GET":
+            return template.TemplateResponse("login_admin.html", context)
+
+        ok = await self.authentication_backend.login(request)
+        if not ok:
+            context["error"] = "Invalid credentials."
+            return template.TemplateResponse(
+                "login_admin.html", context, status_code=400
+            )
+
+        return RedirectResponse(request.url_for("admin:index"), status_code=302)

project/asgi.py

@@ -0,0 +1,4 @@
+from project import create_app
+
+app = create_app()
+

project/bot/__init__.py

@@ -0,0 +1,7 @@
+from fastapi import APIRouter
+
+bot_router = APIRouter(
+    prefix=''
+)
+
+from project.bot import views, models

project/bot/auth.py

@@ -0,0 +1,28 @@
+import os
+import uuid
+
+from fastapi import Request
+from sqladmin.authentication import AuthenticationBackend as AuthBackendAdmin
+
+
+class AdminAuth(AuthBackendAdmin):
+    async def login(self, request: Request) -> bool:
+        form = await request.form()
+        username, password = form["username"], form["password"]
+        if username == 'hectool24' and password == 'hectoolshopify2024@':
+            request.session.update({"session": str(uuid.uuid4())})
+            return True
+        return False
+
+    async def logout(self, request: Request) -> bool:
+        request.session.clear()
+        return True
+
+    async def authenticate(self, request: Request) -> bool:
+        token = request.session.get("session")
+        if not token:
+            return False
+        return True
+
+
+authentication_backend_admin = AdminAuth(secret_key=os.getenv('SECRET'))

project/bot/documents.py

@@ -0,0 +1,56 @@
+import json
+import pandas as pd
+
+with open('../../data.json', 'r') as f:
+    data = json.load(f)
+
+chunks = []
+for post in data:
+    post_text = post['text']
+    comments: list[dict] = post['comments']
+    comments_str = ''
+    for i, comment in enumerate(comments):
+        comment_text = list(comment.keys())[0]
+        replies = comment[comment_text]
+        reply_str = 'Replies:\n'
+        for j, reply in enumerate(replies):
+            if j + 1 == len(replies):
+                reply_str += f'     • {reply}'
+            else:
+                reply_str += f'     • {reply}\n'
+        comments_str += f'{i + 1}. {comment_text}\n'
+        if replies:
+            comments_str += f'{reply_str}\n'
+
+    chunk = f"Post: {post_text}\n"
+    if comments:
+        chunk += f'Comments:\n{comments_str}'
+    chunks.append(chunk)
+#
+df = pd.DataFrame({"chunks": chunks})
+df.to_csv('chunks_javea.csv', index=False)
+
+# for post in data:
+#     post_text = post['text']
+#     comments: list[dict] = post['comments']
+#     comments_str = ''
+#     for i, comment in enumerate(comments):
+#         comment_text = list(comment.keys())[0]
+#         replies = comment[comment_text]
+#         reply_str = '\n'
+#         for j, reply in enumerate(replies):
+#             if j + 1 == len(replies):
+#                 reply_str += f'{reply}'
+#             else:
+#                 reply_str += f'{reply}\n'
+#         comments_str += f'{comment_text}\n'
+#         if replies:
+#             comments_str += f'{reply_str}\n'
+#
+#     chunk = f"{post_text}\n"
+#     if comments:
+#         chunk += f'\n{comments_str}'
+#     chunks.append(chunk)
+
+# df = pd.DataFrame({"chunks": chunks})
+# df.to_csv('chunks_javea_raw.csv', index=False)

project/bot/models.py

@@ -0,0 +1,16 @@
+from sqlalchemy import Column, String, Integer
+
+from project.database import Base
+
+
+class MessagePair(Base):
+    __tablename__ = 'message_pair'
+    id = Column(Integer, primary_key=True, index=True, autoincrement=True)
+    user_message = Column(String)
+    bot_response = Column(String)
+    country = Column(String, default='Undefined')
+
+    def __init__(self, user_message, bot_response, country):
+        self.user_message = user_message
+        self.bot_response = bot_response
+        self.country = country

project/bot/openai_backend.py

@@ -0,0 +1,132 @@
+import asyncio
+from typing import List, Dict
+import faiss
+import numpy as np
+import pandas as pd
+from sqlalchemy.ext.asyncio import AsyncSession
+from starlette.websockets import WebSocket
+
+from project.bot.models import MessagePair
+from project.config import settings
+
+
+class SearchBot:
+    chat_history = []
+    # is_unknown = False
+    # unknown_counter = 0
+
+    def __init__(self, memory=None):
+        if memory is None:
+            memory = []
+        self.chat_history = memory
+
+    async def _summarize_user_intent(self, user_query: str) -> str:
+        chat_history_str = ''
+        chat_history = self.chat_history[-self.unknown_counter * 2:]
+        for i in chat_history:
+            if i['role'] == 'user':
+                chat_history_str += f"{i['role']}: {i['content']}\n"
+        messages = [
+            {
+                'role': 'system',
+                'content': f"{settings.SUMMARIZE_PROMPT}\n"
+                           f"Chat history: ```{chat_history_str}```\n"
+                           f"User query: ```{user_query}```"
+            }
+        ]
+        response = await settings.OPENAI_CLIENT.chat.completions.create(
+            messages=messages,
+            temperature=0.1,
+            n=1,
+            model="gpt-3.5-turbo-0125"
+        )
+        user_intent = response.choices[0].message.content
+        return user_intent
+
+    @staticmethod
+    def _cls_pooling(model_output):
+        return model_output.last_hidden_state[:, 0]
+
+    async def _convert_to_embeddings(self, text_list):
+        encoded_input = settings.INFO_TOKENIZER(
+            text_list, padding=True, truncation=True, return_tensors="pt"
+        )
+        encoded_input = {k: v.to(settings.device) for k, v in encoded_input.items()}
+        model_output = settings.INFO_MODEL(**encoded_input)
+        return self._cls_pooling(model_output).cpu().detach().numpy().astype('float32')
+
+    @staticmethod
+    async def _get_context_data(user_query: list[float]) -> list[dict]:
+        radius = 30
+        _, distances, indices = settings.FAISS_INDEX.range_search(user_query, radius)
+        indices_distances_df = pd.DataFrame({'index': indices, 'distance': distances})
+        filtered_data_df = settings.products_dataset.iloc[indices].copy()
+        filtered_data_df.loc[:, 'distance'] = indices_distances_df['distance'].values
+        sorted_data_df: pd.DataFrame = filtered_data_df.sort_values(by='distance').reset_index(drop=True)
+        sorted_data_df = sorted_data_df.drop('distance', axis=1)
+        data = sorted_data_df.head(3).to_dict(orient='records')
+        return data
+
+    @staticmethod
+    async def create_context_str(context: List[Dict]) -> str:
+        context_str = ''
+        for i, chunk in enumerate(context):
+            context_str += f'{i + 1}) {chunk["chunks"]}'
+        return context_str
+
+    async def _rag(self, context: List[Dict], query: str, session: AsyncSession, country: str):
+        if context:
+            context_str = await self.create_context_str(context)
+            assistant_message = {"role": 'assistant', "content": context_str}
+            self.chat_history.append(assistant_message)
+            content = settings.PROMPT
+        else:
+            content = settings.EMPTY_PROMPT
+        user_message = {"role": 'user', "content": query}
+
+        self.chat_history.append(user_message)
+        messages = [
+            {
+                'role': 'system',
+                'content': content
+            },
+        ]
+        messages = messages + self.chat_history
+
+        stream = await settings.OPENAI_CLIENT.chat.completions.create(
+            messages=messages,
+            temperature=0.1,
+            n=1,
+            model="gpt-3.5-turbo",
+            stream=True
+        )
+        response = ''
+        async for chunk in stream:
+            if chunk.choices[0].delta.content is not None:
+                chunk_content = chunk.choices[0].delta.content
+                response += chunk_content
+                yield response
+                await asyncio.sleep(0.02)
+        assistant_message = {"role": 'assistant', "content": response}
+        self.chat_history.append(assistant_message)
+        try:
+            session.add(MessagePair(user_message=query, bot_response=response, country=country))
+        except Exception as e:
+            print(e)
+
+    async def ask_and_send(self, data: Dict, websocket: WebSocket, session: AsyncSession):
+        query = data['query']
+        country = data['country']
+        transformed_query = await self._convert_to_embeddings(query)
+        context = await self._get_context_data(transformed_query)
+        try:
+            async for chunk in self._rag(context, query, session, country):
+                await websocket.send_text(chunk)
+            # await websocket.send_text('finish')
+        except Exception:
+            await self.emergency_db_saving(session)
+
+    @staticmethod
+    async def emergency_db_saving(session: AsyncSession):
+        await session.commit()
+        await session.close()

project/bot/schemas.py

@@ -0,0 +1,5 @@
+# from pydantic import BaseModel
+#
+#
+# class UserQuery(BaseModel):
+#     query: str

project/bot/templates/home.html

@@ -0,0 +1,82 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1">
+    <title>Hector AI</title>
+    <!-- Fonts -->
+    <link rel="preconnect" href="https://fonts.googleapis.com">
+    <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
+    <link href="https://fonts.googleapis.com/css2?family=Montserrat:wght@300;400;500&display=swap" rel="stylesheet">
+    <!-- Bootstrap css -->
+    <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/css/bootstrap.min.css" rel="stylesheet"
+          integrity="sha384-rbsA2VBKQhggwzxH7pPCaAqO46MgnOM80zW1RWuH61DGLwZJEdK2Kadq2F9CUG65" crossorigin="anonymous">
+    <!--     Connect style.css -->
+    <link rel="stylesheet" href="{{ url_for('static', path='/css/style.css') }}">
+<!--    <link rel="stylesheet" href="../../../static/css/style.css">-->
+</head>
+
+<div class="container-fluid px-0">
+    <div class="row mx-0">
+        <div class="app px-0">
+            <div id="toggleButton" class="rounded-4">
+                <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 50 50" width="60px" height="60px">
+                    <path d="M 25 4.5 C 15.204 4.5 5.9439688 11.985969 3.9179688 21.542969 C 3.9119687 21.571969 3.9200156 21.599906 3.9160156 21.628906 C 1.5620156 23.233906 -0.04296875 26.383 -0.04296875 30 C -0.04296875 35.238 3.3210312 39.5 7.4570312 39.5 C 7.7850313 39.5 8.0913438 39.339313 8.2773438 39.070312 C 8.4643437 38.800312 8.5065781 38.456438 8.3925781 38.148438 C 8.3775781 38.110438 6.9550781 34.244 6.9550781 29.5 C 6.9550781 24.506 8.3091719 22.022187 8.3261719 21.992188 C 8.5011719 21.683187 8.4983125 21.305047 8.3203125 20.998047 C 8.1433125 20.689047 7.8130313 20.5 7.4570312 20.5 C 7.0350313 20.5 6.62275 20.554625 6.21875 20.640625 C 8.58675 12.613625 16.57 6.5 25 6.5 C 32.992 6.5 40.688641 12.044172 43.431641 19.576172 C 43.133641 19.530172 42.831438 19.5 42.523438 19.5 C 42.169438 19.5 41.841109 19.689094 41.662109 19.996094 C 41.482109 20.302094 41.481297 20.683187 41.654297 20.992188 C 41.668297 21.016188 43.023437 23.5 43.023438 28.5 C 43.023438 32.44 42.045078 35.767641 41.705078 36.806641 C 40.558078 37.740641 38.815344 39.034297 36.777344 40.154297 C 36.016344 39.305297 34.839391 38.873437 33.650391 39.148438 L 31.867188 39.558594 C 31.024188 39.751594 30.308609 40.262094 29.849609 40.996094 C 29.391609 41.728094 29.245453 42.5965 29.439453 43.4375 C 29.783453 44.9335 31.11975 45.949219 32.59375 45.949219 C 32.83275 45.949219 33.074359 45.923187 33.318359 45.867188 L 35.103516 45.455078 C 35.945516 45.262078 36.661141 44.752531 37.119141 44.019531 C 37.503141 43.406531 37.653984 42.698234 37.583984 41.990234 C 39.728984 40.828234 41.570453 39.481469 42.814453 38.480469 C 46.814453 38.285469 50.023438 34.114 50.023438 29 C 50.023438 25.237 48.284437 21.989172 45.773438 20.451172 C 45.769438 20.376172 45.777859 20.301563 45.755859 20.226562 C 43.152859 11.113563 34.423 4.5 25 4.5 z M 12 19 C 11.447 19 11 19.447 11 20 L 11 32 C 11 32.553 11.447 33 12 33 L 28.044922 33 C 27.540922 34.057 26.743578 35.482375 26.142578 36.484375 C 25.941578 36.819375 25.954828 37.2405 26.173828 37.5625 C 26.360828 37.8395 26.673 38 27 38 C 27.055 38 27.109063 37.995328 27.164062 37.986328 C 33.351062 36.955328 38.412 32.95125 38.625 32.78125 C 38.862 32.59125 39 32.304 39 32 L 39 20 C 39 19.447 38.553 19 38 19 L 12 19 z M 13 21 L 37 21 L 37 31.501953 C 35.952 32.266953 32.821953 34.393672 29.001953 35.513672 C 29.643953 34.334672 30.328469 32.955266 30.480469 32.197266 C 30.539469 31.903266 30.462438 31.598187 30.273438 31.367188 C 30.082438 31.135188 29.8 31 29.5 31 L 13 31 L 13 21 z M 44.121094 21.822266 C 46.378094 22.758266 48.023437 25.622 48.023438 29 C 48.023438 32.456 46.299891 35.373281 43.962891 36.238281 C 44.420891 34.565281 45.023438 31.747 45.023438 28.5 C 45.023438 25.445 44.556094 23.226266 44.121094 21.822266 z M 5.859375 22.822266 C 5.423375 24.225266 4.9570313 26.445 4.9570312 29.5 C 4.9570312 32.747 5.5595781 35.565281 6.0175781 37.238281 C 3.6805781 36.373281 1.9570312 33.456 1.9570312 30 C 1.9570312 26.622 3.602375 23.758266 5.859375 22.822266 z M 18.5 23 C 17.098 23 16 24.317 16 26 C 16 27.683 17.098 29 18.5 29 C 19.902 29 21 27.683 21 26 C 21 24.317 19.902 23 18.5 23 z M 31.5 23 C 30.098 23 29 24.317 29 26 C 29 27.683 30.098 29 31.5 29 C 32.902 29 34 27.683 34 26 C 34 24.317 32.902 23 31.5 23 z M 18.5 25 C 18.677 25 19 25.38 19 26 C 19 26.62 18.677 27 18.5 27 C 18.323 27 18 26.62 18 26 C 18 25.38 18.323 25 18.5 25 z M 31.5 25 C 31.677 25 32 25.38 32 26 C 32 26.62 31.677 27 31.5 27 C 31.323 27 31 26.62 31 26 C 31 25.38 31.323 25 31.5 25 z M 34.376953 41.064453 C 34.605953 41.064453 34.83225 41.128906 35.03125 41.253906 C 35.31025 41.428906 35.504125 41.702391 35.578125 42.025391 C 35.652125 42.348391 35.598828 42.678984 35.423828 42.958984 C 35.248828 43.237984 34.976297 43.433812 34.654297 43.507812 L 34.652344 43.507812 L 32.869141 43.917969 C 32.208141 44.071969 31.540672 43.654234 31.388672 42.990234 C 31.314672 42.668234 31.369922 42.337641 31.544922 42.056641 C 31.719922 41.777641 31.992453 41.581813 32.314453 41.507812 L 34.097656 41.097656 C 34.190656 41.076656 34.284953 41.064453 34.376953 41.064453 z"
+                          fill="#008dff"/>
+                </svg>
+            </div>
+            <div class="chatbot-window" id="chatbotWindow" style="visibility: hidden;">
+                <div class="chat-header d-flex align-items-center justify-content-between py-2" id="chatHeader">
+                    <div class="d-flex align-items-center">
+                        <div style="color: #008dff; font-size: 30px; letter-spacing: 7px" class="fw-bold me-3 ms-3">
+                            Hector AI
+                        </div>
+                        <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 50 50" width="50px" height="50px"
+                             class="">
+                            <path d="M 25 4.5 C 15.204 4.5 5.9439688 11.985969 3.9179688 21.542969 C 3.9119687 21.571969 3.9200156 21.599906 3.9160156 21.628906 C 1.5620156 23.233906 -0.04296875 26.383 -0.04296875 30 C -0.04296875 35.238 3.3210312 39.5 7.4570312 39.5 C 7.7850313 39.5 8.0913438 39.339313 8.2773438 39.070312 C 8.4643437 38.800312 8.5065781 38.456438 8.3925781 38.148438 C 8.3775781 38.110438 6.9550781 34.244 6.9550781 29.5 C 6.9550781 24.506 8.3091719 22.022187 8.3261719 21.992188 C 8.5011719 21.683187 8.4983125 21.305047 8.3203125 20.998047 C 8.1433125 20.689047 7.8130313 20.5 7.4570312 20.5 C 7.0350313 20.5 6.62275 20.554625 6.21875 20.640625 C 8.58675 12.613625 16.57 6.5 25 6.5 C 32.992 6.5 40.688641 12.044172 43.431641 19.576172 C 43.133641 19.530172 42.831438 19.5 42.523438 19.5 C 42.169438 19.5 41.841109 19.689094 41.662109 19.996094 C 41.482109 20.302094 41.481297 20.683187 41.654297 20.992188 C 41.668297 21.016188 43.023437 23.5 43.023438 28.5 C 43.023438 32.44 42.045078 35.767641 41.705078 36.806641 C 40.558078 37.740641 38.815344 39.034297 36.777344 40.154297 C 36.016344 39.305297 34.839391 38.873437 33.650391 39.148438 L 31.867188 39.558594 C 31.024188 39.751594 30.308609 40.262094 29.849609 40.996094 C 29.391609 41.728094 29.245453 42.5965 29.439453 43.4375 C 29.783453 44.9335 31.11975 45.949219 32.59375 45.949219 C 32.83275 45.949219 33.074359 45.923187 33.318359 45.867188 L 35.103516 45.455078 C 35.945516 45.262078 36.661141 44.752531 37.119141 44.019531 C 37.503141 43.406531 37.653984 42.698234 37.583984 41.990234 C 39.728984 40.828234 41.570453 39.481469 42.814453 38.480469 C 46.814453 38.285469 50.023438 34.114 50.023438 29 C 50.023438 25.237 48.284437 21.989172 45.773438 20.451172 C 45.769438 20.376172 45.777859 20.301563 45.755859 20.226562 C 43.152859 11.113563 34.423 4.5 25 4.5 z M 12 19 C 11.447 19 11 19.447 11 20 L 11 32 C 11 32.553 11.447 33 12 33 L 28.044922 33 C 27.540922 34.057 26.743578 35.482375 26.142578 36.484375 C 25.941578 36.819375 25.954828 37.2405 26.173828 37.5625 C 26.360828 37.8395 26.673 38 27 38 C 27.055 38 27.109063 37.995328 27.164062 37.986328 C 33.351062 36.955328 38.412 32.95125 38.625 32.78125 C 38.862 32.59125 39 32.304 39 32 L 39 20 C 39 19.447 38.553 19 38 19 L 12 19 z M 13 21 L 37 21 L 37 31.501953 C 35.952 32.266953 32.821953 34.393672 29.001953 35.513672 C 29.643953 34.334672 30.328469 32.955266 30.480469 32.197266 C 30.539469 31.903266 30.462438 31.598187 30.273438 31.367188 C 30.082438 31.135188 29.8 31 29.5 31 L 13 31 L 13 21 z M 44.121094 21.822266 C 46.378094 22.758266 48.023437 25.622 48.023438 29 C 48.023438 32.456 46.299891 35.373281 43.962891 36.238281 C 44.420891 34.565281 45.023438 31.747 45.023438 28.5 C 45.023438 25.445 44.556094 23.226266 44.121094 21.822266 z M 5.859375 22.822266 C 5.423375 24.225266 4.9570313 26.445 4.9570312 29.5 C 4.9570312 32.747 5.5595781 35.565281 6.0175781 37.238281 C 3.6805781 36.373281 1.9570312 33.456 1.9570312 30 C 1.9570312 26.622 3.602375 23.758266 5.859375 22.822266 z M 18.5 23 C 17.098 23 16 24.317 16 26 C 16 27.683 17.098 29 18.5 29 C 19.902 29 21 27.683 21 26 C 21 24.317 19.902 23 18.5 23 z M 31.5 23 C 30.098 23 29 24.317 29 26 C 29 27.683 30.098 29 31.5 29 C 32.902 29 34 27.683 34 26 C 34 24.317 32.902 23 31.5 23 z M 18.5 25 C 18.677 25 19 25.38 19 26 C 19 26.62 18.677 27 18.5 27 C 18.323 27 18 26.62 18 26 C 18 25.38 18.323 25 18.5 25 z M 31.5 25 C 31.677 25 32 25.38 32 26 C 32 26.62 31.677 27 31.5 27 C 31.323 27 31 26.62 31 26 C 31 25.38 31.323 25 31.5 25 z M 34.376953 41.064453 C 34.605953 41.064453 34.83225 41.128906 35.03125 41.253906 C 35.31025 41.428906 35.504125 41.702391 35.578125 42.025391 C 35.652125 42.348391 35.598828 42.678984 35.423828 42.958984 C 35.248828 43.237984 34.976297 43.433812 34.654297 43.507812 L 34.652344 43.507812 L 32.869141 43.917969 C 32.208141 44.071969 31.540672 43.654234 31.388672 42.990234 C 31.314672 42.668234 31.369922 42.337641 31.544922 42.056641 C 31.719922 41.777641 31.992453 41.581813 32.314453 41.507812 L 34.097656 41.097656 C 34.190656 41.076656 34.284953 41.064453 34.376953 41.064453 z"
+                                  fill="#008dff"/>
+                        </svg>
+                    </div>
+                    <div style="color: #008dff; cursor: pointer" onclick="closeChatBotWindow()">
+                        <i class="fa-solid fa-xmark fs-1 me-3"></i>
+                    </div>
+                </div>
+                <div class="chat-body" id="chatBody">
+                    <div class="message">
+                        <div class="bot_message mt-4 py-3 px-4 rounded-4 w-75 ms-3">
+                            Hallo! Ik ben een virtuele assistent, een helper voor Nederlanders die naar Javea zijn
+                            gemigreerd. Ik weet alles over deze regio en zal je helpen om je aan te passen in je nieuwe
+                            thuis. Waarmee kan ik je helpen?
+                        </div>
+                    </div>
+                </div>
+                <div class="chat-footer px-3 mb-2 bg-white pt-3" id="chatFooter">
+                    <input type="text" class="form-control rounded-5" id="textInput"
+                           placeholder="Stellen Sie eine Frage" style="height: 50px">
+                    <div style="cursor: pointer; background-color: #105ead; height: 50px"
+                         class="rounded-5 mt-3 d-flex text-white justify-content-center align-items-center fs-4"
+                         id="sendButton">
+                        <span>Send <img src="../../../static/images/send.png" alt="" height="35px"></span>
+                    </div>
+                </div>
+            </div>
+        </div>
+    </div>
+</div>
+
+</html>
+<script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.bundle.min.js"
+        integrity="sha384-kenU1KFdBIe4zVF0s0G1M5b4hcpxyD9F7jL+jjXkk+Q2h455rYXK/7HAuoJl+0I4"
+        crossorigin="anonymous"></script>
+<script src="https://code.jquery.com/jquery-3.6.3.min.js"
+        integrity="sha256-pvPw+upLPUjgMXY0G+8O0xUf+/Im1MZjXxxgOcBQBXU=" crossorigin="anonymous"></script>
+<link rel="stylesheet" href="https://code.jquery.com/ui/1.12.1/themes/base/jquery-ui.css">
+<script src="https://code.jquery.com/ui/1.12.1/jquery-ui.js"></script>
+<script src="https://kit.fontawesome.com/d4ffd37f75.js" crossorigin="anonymous"></script>
+<script src="https://cdn.jsdelivr.net/npm/marked/marked.min.js"></script>
+<!--<script type="text/javascript" src="../../../static/js/main.js"></script>-->
+<!--<script type="text/javascript" src="../../../static/js/utils.js"></script>-->
+<!--<script type="text/javascript" src="../../../static/js/ws.js"></script>-->
+<script type="text/javascript" src="{{ url_for('static', path='/js/main.js') }}"></script>
+<script type="text/javascript" src="{{ url_for('static', path='/js/utils.js') }}"></script>
+<script type="text/javascript" src="{{ url_for('static', path='/js/ws.js') }}"></script>

project/bot/views.py

@@ -0,0 +1,11 @@
+from fastapi.templating import Jinja2Templates
+from fastapi.requests import Request
+
+from project.bot import bot_router
+
+template = Jinja2Templates(directory='project/bot/templates')
+
+
+@bot_router.get('/', name='main')
+async def main(request: Request):
+    return template.TemplateResponse("home.html", {'request': request})

project/config.py

@@ -0,0 +1,77 @@
+import os
+
+import faiss
+import pandas as pd
+from openai import AsyncOpenAI
+import pathlib
+from functools import lru_cache
+from environs import Env
+from transformers import AutoModel, AutoTokenizer
+import torch
+
+env = Env()
+env.read_env()
+
+
+class BaseConfig:
+    BASE_DIR: pathlib.Path = pathlib.Path(__file__).parent.parent
+    DATA_DIR: pathlib.Path = BASE_DIR / 'project' / 'data'
+    MODEL_NAME = 'sentence-transformers/paraphrase-multilingual-mpnet-base-v2'
+    INFO_MODEL = AutoModel.from_pretrained(MODEL_NAME)
+    INFO_TOKENIZER = AutoTokenizer.from_pretrained(MODEL_NAME)
+    OPENAI_CLIENT = AsyncOpenAI(api_key=os.getenv('OPENAI_API_KEY'))
+    FAISS_INDEX = faiss.read_index(str(BASE_DIR / 'faiss_javea.index'))
+
+
+class DevelopmentConfig(BaseConfig):
+    pass
+
+
+class ProductionConfig(BaseConfig):
+    DATABASE_URL = f"postgresql+asyncpg://{env('DATABASE_USER')}:" \
+                   f"{env('DATABASE_PASSWORD')}@" \
+                   f"{env('DATABASE_HOST')}:" \
+                   f"{env('DATABASE_PORT')}/" \
+                   f"{env('DATABASE_NAME')}"
+    PROMPT = "Je bent een expert in de regio Javea in Italië, die alles weet om mensen te helpen die migreren van " \
+             "Nederland naar Spanje. Je taak is om mensen te helpen zich te vestigen in de nieuwe stad. Gebruik " \
+             "kennis uit je vorige antwoord (voornamelijk uit opmerkingen) om een informatief antwoord " \
+             "te geven op de gebruikersvraag. Vermeld nooit dat je kennis haalt uit posts of opmerkingen. Spreek " \
+             "vanuit jezelf."
+    EMPTY_PROMPT = "Je bent een expert in Javea aan de Costa Blanca in Spanje, met uitgebreide kennis om Nederlanders " \
+                   "te helpen die naar deze regio verhuizen. Je taak is om mensen te helpen zich thuis te voelen in " \
+                   "hun nieuwe stad. Gebruik je kennis over deze regio maar informatieve antwoorden te " \
+                   "geven op de vragen van gebruikers."
+    SUMMARIZE_PROMPT = "Study the user's requests, paying special attention to the specific mentioned wishes when " \
+                       "choosing a house. Combine these details into a single query that reflects all the user's " \
+                       "needs. Formulate your answer as if you were a user, clearly and concisely stating the " \
+                       "requirements. Make sure that all relevant user wishes are indicated in your response. "
+
+    def __init__(self):
+        if torch.cuda.is_available():
+            device = torch.device("cuda")
+
+        else:
+            device = torch.device("cpu")
+        self.device = device
+        self.INFO_MODEL.to(device)
+        self.products_dataset = pd.read_csv(self.BASE_DIR / 'chunks_javea.csv')
+
+
+class TestConfig(BaseConfig):
+    pass
+
+
+@lru_cache()
+def get_settings() -> DevelopmentConfig | ProductionConfig | TestConfig:
+    config_cls_dict = {
+        'development': DevelopmentConfig,
+        'production': ProductionConfig,
+        'testing': TestConfig
+    }
+    config_name = env('FASTAPI_CONFIG', default='development')
+    config_cls = config_cls_dict[config_name]
+    return config_cls()
+
+
+settings = get_settings()

project/database.py

@@ -0,0 +1,27 @@
+from typing import AsyncGenerator
+
+from sqlalchemy.ext.asyncio import AsyncEngine, create_async_engine, async_sessionmaker, AsyncSession
+from sqlalchemy.orm import declarative_base
+
+from project.config import settings
+
+Base = declarative_base()
+
+
+def get_async_engine(url: str) -> AsyncEngine:
+    return create_async_engine(url=url, echo=True, pool_pre_ping=True)
+
+
+def get_async_sessionmaker(engine: AsyncEngine) -> async_sessionmaker:
+    return async_sessionmaker(bind=engine, class_=AsyncSession)
+
+
+engine = get_async_engine(
+    settings.DATABASE_URL
+)
+async_session_maker = get_async_sessionmaker(engine)
+
+
+async def get_async_session() -> AsyncSession:
+    async with async_session_maker() as session:
+        return session

project/ws/__init__.py

@@ -0,0 +1,5 @@
+from fastapi import APIRouter
+
+ws_router = APIRouter()
+
+from . import views

project/ws/views.py

@@ -0,0 +1,19 @@
+from fastapi import WebSocket, WebSocketDisconnect
+
+from . import ws_router
+from ..bot.openai_backend import SearchBot
+from ..database import get_async_session
+
+
+@ws_router.websocket("/ws/{client_id}")
+async def websocket_endpoint(websocket: WebSocket, client_id: str):
+    await websocket.accept()
+    chatbot = SearchBot()
+    session = await get_async_session()
+    try:
+        while True:
+            data = await websocket.receive_json()
+            await chatbot.ask_and_send(data, websocket, session)
+    except WebSocketDisconnect:
+        await session.commit()
+        await session.close()

requirements.txt

@@ -0,0 +1,72 @@
+annotated-types==0.6.0
+anyio==4.3.0
+async-timeout==4.0.3
+asyncpg==0.29.0
+certifi==2024.2.2
+charset-normalizer==3.3.2
+click==8.1.7
+distro==1.9.0
+environs==11.0.0
+exceptiongroup==1.2.0
+faiss-gpu==1.7.2
+fastapi==0.110.0
+filelock==3.13.4
+fsspec==2024.3.1
+greenlet==3.0.3
+h11==0.14.0
+httpcore==1.0.5
+httptools==0.6.1
+httpx==0.27.0
+huggingface-hub==0.22.2
+idna==3.6
+itsdangerous==2.1.2
+Jinja2==3.1.3
+MarkupSafe==2.1.5
+marshmallow==3.21.1
+mpmath==1.3.0
+networkx==3.3
+numpy==1.26.4
+nvidia-cublas-cu12==12.1.3.1
+nvidia-cuda-cupti-cu12==12.1.105
+nvidia-cuda-nvrtc-cu12==12.1.105
+nvidia-cuda-runtime-cu12==12.1.105
+nvidia-cudnn-cu12==8.9.2.26
+nvidia-cufft-cu12==11.0.2.54
+nvidia-curand-cu12==10.3.2.106
+nvidia-cusolver-cu12==11.4.5.107
+nvidia-cusparse-cu12==12.1.0.106
+nvidia-nccl-cu12==2.19.3
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nvtx-cu12==12.1.105
+openai==1.17.1
+packaging==24.0
+pandas==2.2.2
+pydantic==2.6.4
+pydantic_core==2.16.3
+python-dateutil==2.9.0.post0
+python-dotenv==1.0.1
+python-multipart==0.0.9
+pytz==2024.1
+PyYAML==6.0.1
+regex==2023.12.25
+requests==2.31.0
+safetensors==0.4.2
+six==1.16.0
+sniffio==1.3.1
+sqladmin==0.16.1
+SQLAlchemy==2.0.29
+starlette==0.36.3
+sympy==1.12
+tokenizers==0.15.2
+torch==2.2.2
+tqdm==4.66.2
+transformers==4.39.3
+triton==2.2.0
+typing_extensions==4.10.0
+tzdata==2024.1
+urllib3==2.2.1
+uvicorn==0.29.0
+uvloop==0.19.0
+watchfiles==0.21.0
+websockets==12.0
+WTForms==3.1.2

static/css/style.css

@@ -0,0 +1,81 @@
+:root {
+    --main-blue-color: #0e88df;
+    --main-blue-color-hover: #f51958;
+    --main-bg-color: #e8edf3;
+}
+
+body {
+    font-family: 'Montserrat', sans-serif;
+}
+
+.chat-header {
+    background-color: #f1f1f1;
+    border-bottom: 1px solid #bdbdbd;
+}
+
+.chat-body {
+    background-image: url('../../static/images/bg.png');
+    background-size: cover;
+    padding-top: 1px;
+    border-top: 1px solid transparent;
+    display: flex;
+    flex-direction: column;
+    overflow-y: auto;
+}
+
+.chatbot-window {
+    border: 1px solid #bdbdbd;
+}
+
+/*::placeholder {*/
+/*    font-size: 30px;*/
+/*}*/
+
+.message {
+    max-width: 100%;
+    display: flex;
+    justify-content: flex-start;
+    margin-top: 5px;
+    font-size: 14px;
+}
+
+.user_message {
+    background-color: #f3f2f2;
+    margin-left: auto;
+    border: 1px solid #d7d7d7;
+    /*font-size: 20px;*/
+}
+
+.bot_message {
+    margin-right: auto;
+    background-color: #d1e8ff;
+    border: 1px solid #71beff;
+}
+
+.bot_message p {
+    margin-bottom: 0;
+}
+
+.bot_message ol {
+    margin-bottom: 0;
+}
+
+.bot_message ul {
+    margin-bottom: 0;
+}
+
+.bot_message h3 {
+    margin-top: 0.5rem;
+    font-size: 20px;
+}
+
+#toggleButton {
+    cursor: pointer;
+    position: fixed;
+    right: 15px;
+    bottom: 15px;
+    background-color: #f7f9fc;
+    border: 1px solid #ddd;
+    padding: 20px;
+    box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
+}

static/js/main.js

@@ -0,0 +1,64 @@
+const chatBody = document.getElementById('chatBody')
+const openButton = document.getElementById('toggleButton')
+const chatbotWindow = document.getElementById('chatbotWindow');
+
+let socket
+let isFirstWord = false
+let botMessage
+
+window.addEventListener('load', adjustChatBodyHeight);
+window.addEventListener('resize', adjustChatBodyHeight);
+
+function getTotalHeight(element) {
+    const styles = window.getComputedStyle(element);
+    const margins = ['marginTop', 'marginBottom', 'borderTopWidth', 'borderBottomWidth']
+        .reduce((acc, style) => acc + parseFloat(styles[style]), 0);
+    return element.offsetHeight + margins;
+}
+
+function adjustChatBodyHeight() {
+    const chatFooter = document.getElementById('chatFooter')
+    const chatHeader = document.getElementById('chatHeader')
+    const chatFooterHeight = getTotalHeight(chatFooter);
+    const chatHeaderHeight = getTotalHeight(chatHeader);
+    const openButtonWindow = getTotalHeight(openButton)
+    const viewportHeight = window.innerHeight - chatHeaderHeight - chatFooterHeight;
+    chatBody.style.height = viewportHeight + 'px';
+}
+
+function openChatBotWindow() {
+    let lastScrollHeight = chatBody.scrollHeight;
+    const uuid = generateUUID()
+    socket = new WebSocket(`ws://127.0.0.1:8000/ws/${uuid}`);
+    socket.onclose = (event) => console.log('WebSocket disconnected', event);
+    socket.onerror = (error) => {
+        alert('Something was wrong. Try again later.')
+        window.location.reload()
+    };
+    socket.onmessage = (event) => {
+        if (chatBody.scrollHeight > lastScrollHeight) {
+            chatBody.scrollTop = chatBody.scrollHeight;
+            lastScrollHeight = chatBody.scrollHeight;
+        }
+        if (!isFirstWord) {
+            isFirstWord = true
+            createNewMessage(event.data, 'bot')
+            const botMessages = document.querySelectorAll('.bot_message');
+            botMessage = botMessages[botMessages.length - 1];
+        } else {
+            botMessage.innerHTML = marked.parse(event.data)
+        }
+    }
+}
+
+function closeChatBotWindow() {
+    socket.close()
+    chatbotWindow.style.visibility = 'hidden'
+    openButton.style.display = 'block'
+}
+
+openButton.addEventListener('click', function () {
+    chatbotWindow.style.visibility = "visible";
+    openButton.style.display = 'none'
+    openChatBotWindow();
+});

static/js/utils.js

@@ -0,0 +1,48 @@
+const sendButton = document.getElementById('sendButton')
+const textInput = document.getElementById('textInput')
+
+function createNewMessage(text, type) {
+    const message = document.createElement('div')
+    message.className = 'message'
+    let content = document.createElement('div')
+    if (type === 'bot') {
+        content.className = 'bot_message mt-4 py-3 px-4 rounded-4 w-75 ms-3'
+    } else {
+        content.className = 'user_message mt-4 py-3 px-4 rounded-4 w-75 me-3'
+    }
+    content.innerText = text
+    message.appendChild(content)
+    chatBody.appendChild(message)
+}
+
+function sendMessageToServer() {
+    const userQuery = textInput.value;
+    if (userQuery.length === 0) {
+        return
+    }
+    createNewMessage(userQuery, 'user')
+    socket.send(JSON.stringify({'query': userQuery, 'country': "Undefined"}));
+    textInput.value = ''
+    isFirstWord = false
+    chatBody.scrollTop = chatBody.scrollHeight;
+}
+
+function generateUUID() {
+    const arr = new Uint8Array(16);
+    window.crypto.getRandomValues(arr);
+
+    arr[6] = (arr[6] & 0x0f) | 0x40;
+    arr[8] = (arr[8] & 0x3f) | 0x80;
+
+    return ([...arr].map((b, i) =>
+        (i === 4 || i === 6 || i === 8 || i === 10 ? "-" : "") + b.toString(16).padStart(2, "0")
+    ).join(""));
+}
+
+textInput.addEventListener("keydown", function (event) {
+    if (event.key === "Enter") {
+        sendMessageToServer();
+    }
+});
+
+sendButton.addEventListener("click", sendMessageToServer);