Knowbot first commit

.env.example

@@ -0,0 +1,4 @@
+HF_TOKEN=
+OPENAI_API_KEY=
+ALPHAVANTAGE_API_KEY=
+PERPLEXITY_API_KEY=

.gitignore

@@ -0,0 +1,126 @@
+# ===============================
+# Python
+# ===============================
+__pycache__/
+*.py[cod]
+*.pyo
+*.pyd
+*.so
+*.egg-info/
+.eggs/
+dist/
+build/
+
+# Virtual environments
+.env
+.venv
+venv/
+env/
+myenv/
+ENV/
+
+# ===============================
+# Environment & Secrets
+# ===============================
+.env.local
+.env.*.local
+.env.production
+.env.development
+.env.test
+*.key
+*.pem
+
+# API keys / credentials
+secrets/
+credentials/
+config/secrets.yaml
+config/secrets.json
+
+# ===============================
+# Jupyter / Data Science
+# ===============================
+.ipynb_checkpoints/
+*.ipynb
+
+# ===============================
+# ML / AI Artifacts
+# ===============================
+models/
+checkpoints/
+weights/
+*.pt
+*.pth
+*.onnx
+*.joblib
+*.pkl
+
+# Vector stores / RAG indexes
+faiss_index/
+chroma/
+vectorstore/
+embeddings/
+
+# ===============================
+# Logs & Runtime Files
+# ===============================
+logs/
+*.log
+*.out
+*.err
+
+# ===============================
+# Gradio / FastAPI
+# ===============================
+gradio_cached_examples/
+.gradio/
+tmp/
+uploads/
+
+# ===============================
+# Cache / Temp
+# ===============================
+.cache/
+.mypy_cache/
+.pytest_cache/
+ruff_cache/
+coverage/
+htmlcov/
+
+# ===============================
+# OS / Editor
+# ===============================
+.DS_Store
+Thumbs.db
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# ===============================
+# Docker
+# ===============================
+docker-data/
+*.tar
+
+# ===============================
+# Deployment
+# ===============================
+*.local
+*.tfstate
+*.tfstate.backup
+.envrc
+
+# ===============================
+# Reports / Generated Content
+# ===============================
+reports/
+outputs/
+generated_images/
+charts/
+visualizations/
+
+# ===============================
+# Misc
+# ===============================
+*.bak
+*.tmp

README.md

@@ -11,3 +11,37 @@ short_description: ' Designed to be an intelligent assistant '
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+
+
+# 🤖 KnowBot AI — Voice Transcription with Whisper (Gradio + OpenAI)
+
+KnowBot AI is a simple **voice-to-text transcription app** built with **Gradio** and **OpenAI Whisper API (`whisper-1`)**.  
+It allows users to record their voice using a microphone and instantly get the transcription output.
+
+---
+
+## 🚀 Features
+
+- 🎤 Record voice directly from the browser (microphone input)
+- 🧠 Transcribe speech using **OpenAI Whisper (`whisper-1`)**
+- 🌍 Supports accents and multiple languages
+- 🖥️ Clean and simple Gradio interface
+
+---
+
+## 🛠️ Tech Stack
+
+- Python 3.9+
+- Gradio
+- OpenAI API (Whisper-1)
+- Whisper
+
+---
+
+## 📂 Project Structure
+
+```bash
+KnowBotAI/
+│── app.py
+│── requirements.txt
+│── README.md

app.py

@@ -0,0 +1,14 @@
+from ui.gradio import launch_ui
+from whisper_singleton import get_embedding,get_whisper
+from utils.central_logging import setup_logging
+
+
+setup_logging()
+
+def warmup():
+    get_whisper()
+    get_embedding()
+
+if __name__ == "__main__":
+    warmup()
+    launch_ui()

extract_text.py

@@ -0,0 +1,66 @@
+from pypdf import PdfReader
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_core.documents import Document
+from langchain_openai import OpenAIEmbeddings
+from langchain_chroma import Chroma
+import re
+import os
+
+def extract_text_from_pdf(file_path:str) -> str:
+    reader = PdfReader(file_path)
+    text = ""
+    for page in reader.pages:
+        text += page.extract_text() or ""
+    return text
+
+def pdf_to_documents(file_path:str,database_name:str,collection_name:str,embeddings:OpenAIEmbeddings,chunk_size=1000,chunk_overlap=200,metadata:dict=None):
+    text = extract_text_from_pdf(file_path)
+    text = re.sub(r"[^a-zA-Z0-9.,!?;:'\"()\s]", "", text)
+    if not text.strip():
+        return []
+
+    splitter = RecursiveCharacterTextSplitter(
+            chunk_size=chunk_size,
+            chunk_overlap=chunk_overlap)
+
+    chunks = splitter.split_text(text)
+    docs = []
+    for i,chunk in enumerate(chunks):
+        #print(f"index: {i} , {chunk}")
+        meta = metadata.copy() if metadata else {}
+        meta.update({"chunk":i})
+        docs.append(Document(page_content=chunk, metadata=meta))
+
+    if os.path.exists(database_name):
+        Chroma(persist_directory=database_name, embedding_function=embeddings,collection_name=collection_name).delete_collection()
+
+    vectorstore = Chroma.from_documents(documents=docs, embedding=embeddings, persist_directory=database_name,collection_name=collection_name)
+
+    return docs,vectorstore
+
+
+
+def store_data(text:str,database_name:str,collection_name:str,embeddings:OpenAIEmbeddings):
+    
+    text_splitter = RecursiveCharacterTextSplitter(
+        chunk_size = 1000,
+        chunk_overlap  = 0,
+        separators = [" ", ",", "\n"]
+    )
+
+    #with open(file_path) as f:
+    #    text = f.read()
+
+    texts = text_splitter.split_text(text)
+
+    #print(f"split: {texts}")
+    docs = [Document(page_content=t) for t in texts]
+
+
+    if os.path.exists(database_name):
+        Chroma(persist_directory=database_name, embedding_function=embeddings,collection_name=collection_name).delete_collection()
+
+    vectorstore = Chroma.from_documents(documents=docs, embedding=embeddings, persist_directory=database_name,collection_name=collection_name)
+    return vectorstore
+
+

prompt.py

@@ -0,0 +1,24 @@
+from langchain_core.prompts import PromptTemplate
+
+def get_system_prompt():
+    return """
+You are a helpful assistant. Only answer questions based on the context provided.
+Do not make assumptions. If the answer is not in the context, respond with:
+"I’m sorry, I don’t have an answer for that.
+
+Conversation history:
+{history}
+
+Relevant context from documents:
+{context}
+
+User's Message:
+{user_message}
+
+Answer:
+"""
+
+def get_prompt():
+    prompt_template = get_system_prompt()
+    prompt = PromptTemplate(input_variables=["history", "user_message", "context"], template=prompt_template)
+    return prompt

store.py

@@ -0,0 +1,6 @@
+from langchain_chroma import Chroma
+from langchain_text_splitters import RecursiveCharacterTextSplitter, CharacterTextSplitter
+from langchain_openai import OpenAIEmbeddings
+from langchain_core.documents import Document
+import os
+

style.py

@@ -0,0 +1,50 @@
+
+def get_gradio_style():
+    return """
+body {
+    background-color: #1e1e1e;
+    color: white;
+}
+.gradio-container {
+    background-color: #1e1e1e;
+}
+.gr-chat-message.user, .gr-chat-message.assistant {
+    background-color: #2b2b2b;
+    color: white;
+    border-radius: 8px;
+    padding: 5px 10px;
+    margin: 5px 0;
+}
+.gr-button {
+    background-color: #444444;
+    color: white;
+}
+.gr-textbox textarea {
+    background-color: #2b2b2b;
+    color: white;
+}
+span.md h2{
+  color:white;
+}
+
+#component-279{
+  height: 150px;
+}
+
+span.svelte-7ddecg p{
+  color:white;
+}
+
+span.chatbot p{
+  color: black;
+  font-weight: bold;
+  font-style: italic;
+  font-family: "Arial", sans-serif;
+}
+
+textarea.svelte-1ae7ssi{
+	background: whitesmoke;
+	font-weight: bold;
+}
+
+"""

ui/chat_handler.py

@@ -0,0 +1,148 @@
+import os
+from utils.central_logging import setup_logging,get_logger
+import textwrap
+from langchain_openai import OpenAI
+from langchain_chroma import Chroma
+#from langchain_community.document_loaders import SeleniumURLLoader
+from dotenv import load_dotenv
+import os
+import openai
+
+
+from langchain_openai import ChatOpenAI
+from langchain_core.runnables import RunnableLambda
+import chromadb
+
+import gradio as gr
+import time
+import asyncio
+import nest_asyncio
+import threading
+import re
+from openai import OpenAI
+#import streamlit as st
+
+from whisper_singleton import get_embedding,save_file,transcribe_content
+from extract_text import pdf_to_documents,store_data
+from prompt import get_prompt,get_system_prompt
+
+
+load_dotenv("./.env")
+
+setup_logging()
+logger = get_logger("chat")
+
+
+_embedding = None
+_retriever = None
+_vectore_store = None
+
+openai_api_key  = os.getenv("OPENAI_API_KEY")
+
+if openai_api_key:
+    logger.info("Open ai api key has been set")
+else:
+    logger.error("No open ai api key has been found")
+
+
+
+
+try:
+    llm_openai = ChatOpenAI(model='gpt-3.5-turbo',temperature=0)
+    client = OpenAI()
+    logger.info("Clients has been initialized")
+except Exception  as e:
+    logger.exception(f"An exception occured: {e}")
+
+
+
+def handle_upload(file_path):
+    global _embedding
+    global _retriever
+    _embedding = get_embedding()
+    text_content = ""
+    status_message = ""
+    file_name = "./transcribe.txt"
+    try:
+        if file_path.lower().endswith(".pdf"):
+            
+            collection_name = "pdffiles"
+            pdf_docs,_vectore_store =  pdf_to_documents(file_path,"transcribe_db",collection_name,_embedding)
+            text_content = "\n\n".join([doc.page_content for doc in pdf_docs])
+            status_message = "📄 PDF file uploaded — extraction implemented."
+            logger.info(status_message)
+            #save_file(file_name,text_content)
+        elif file_path.lower().endswith(".mp3") or file_path.lower().endswith('.mp4'):
+            print(f"path:{file_path}")
+            if file_path.lower().endswith(".mp3"):
+                collection_name = "audios"
+                status_message = "🎧 MP3 uploaded — transcription implemented."
+                logger.info(status_message)
+            else:
+                collection_name = "videos"
+                status_message = "🎬 MP4 uploaded — video transcription implemented."
+                logger.info(status_message)
+            
+            text_content = transcribe_content(file_path)
+            _vectore_store = store_data(text_content,"transcribe_db",collection_name,_embedding)
+            #save_file(file_name,text_content)
+        else:
+            status_message = "Invalid file format"
+    except Exception as e:
+        status_message = f"❌ Error processing file: {e}"
+        logger.exception(status_message)
+    _retriever = _vectore_store.as_retriever()
+    return status_message,text_content
+
+
+
+def stream_response(user_input,history):
+    
+    history = history or []
+
+    history.append({"role": "user", "content": user_input})
+    history.append({"role": "assistant", "content": ""})
+    
+    context = ""
+    if _retriever is not None:
+        docs = _retriever.invoke(user_input)
+        context = "\n\n".join([d.page_content for d in docs])
+
+    formatted_history = "\n".join(
+        f"{m['role'].capitalize()}: {m['content']}"
+        for m in history
+    )
+
+    
+
+    system_prompt = get_system_prompt().format(
+        history=formatted_history,
+        context=context,
+        user_message=user_input
+    )
+
+    messages = [
+        {"role": "system", "content": system_prompt},
+        {"role": "user", "content": user_input},
+    ]
+
+    partial_reply = ""
+
+    stream = client.chat.completions.create(
+        model="gpt-4o-mini",
+        messages=messages,
+        stream=True,
+        temperature = 0
+    )
+
+    for event in stream:
+        delta = event.choices[0].delta
+        if delta and delta.content:
+            token = delta.content
+            partial_reply += token
+            history[-1]["content"] = partial_reply 
+            yield history, history, "" 
+
+    history[-1]["content"] = partial_reply
+    yield history, history, ""
+

ui/gradio.py

@@ -0,0 +1,52 @@
+from style import get_gradio_style
+from .chat_handler import stream_response,handle_upload
+import gradio as gr
+
+
+
+
+
+def show_button(file):
+  title = ""
+  content = ""
+  return title,content,gr.update(visible=bool(file))
+     
+
+
+def launch_ui():
+  with gr.Blocks(css=get_gradio_style) as demo:
+    gr.Markdown("## 🤖 💬 KnowBot AI — Document-Aware Chat Assistant")
+
+    with gr.Row():
+
+      with gr.Column(scale=1):
+        upload_file = gr.File(
+              label="Upload a PDF, MP4, or MP3 file",
+              file_types=[".pdf", ".mp4", ".mp3"],
+              type="filepath"
+          )
+        upload_button = gr.Button("Upload and Process", visible=False)
+        upload_status = gr.Markdown()
+        pdf_text_area = gr.Textbox(
+                  label="PDF Text Content",
+                  lines=15,
+                  interactive=False,
+                  placeholder="Extracted text will appear here...")
+        upload_file.change(fn=show_button,inputs=upload_file,outputs=[upload_status,pdf_text_area,upload_button])
+
+
+      with gr.Column(scale=3):
+        chatbot = gr.Chatbot(height=400,show_label=False, render_markdown=True)
+        #audio_input = gr.Audio(label="🎤 Record your message",type="filepath",sources=["microphone"],interactive=True)
+        msg = gr.Textbox(label="Your message")
+        clear = gr.Button("Clear Conversation")
+
+    state = gr.State([])
+    msg.submit(stream_response, [msg, state], [chatbot, state,msg])
+    clear.click(lambda: ([], [],""), None, [chatbot, state,msg])
+    upload_button.click(handle_upload,inputs=upload_file,outputs=[upload_status, pdf_text_area])
+    #audio_input.change(stream_response,inputs=[msg, state],outputs=[chatbot, state, msg])
+    demo.queue(default_concurrency_limit=64)
+    demo.launch(debug=True, share=False)
+
+

utils/central_logging.py

@@ -0,0 +1,51 @@
+import logging
+import logging.handlers
+from pathlib import Path
+
+LOG_DIR = Path("logs")
+LOG_DIR.mkdir(exist_ok=True)
+
+LOG_FILE = LOG_DIR / "advisor.log"
+
+LOG_FORMAT = (
+    "%(asctime)s | %(levelname)s | %(name)s | "
+    "%(funcName)s:%(lineno)d | %(message)s"
+)
+
+DATE_FORMAT = "%Y-%m-%d %H:%M:%S"
+
+
+def setup_logging(log_level=logging.INFO):
+    """Global logging configuration"""
+
+    root_logger = logging.getLogger()
+    root_logger.setLevel(log_level)
+
+    # Prevent duplicate logs in notebooks / reloads
+    if root_logger.handlers:
+        return
+
+    formatter = logging.Formatter(LOG_FORMAT, DATE_FORMAT)
+
+    # ---- File Handler (advisor.log) ----
+    file_handler = logging.handlers.RotatingFileHandler(
+        LOG_FILE,
+        maxBytes=10 * 1024 * 1024,  # 10 MB
+        backupCount=5,
+        encoding="utf-8",
+    )
+    file_handler.setFormatter(formatter)
+    file_handler.setLevel(log_level)
+
+    # ---- Console Handler ----
+    console_handler = logging.StreamHandler()
+    console_handler.setFormatter(formatter)
+    console_handler.setLevel(log_level)
+
+    root_logger.addHandler(file_handler)
+    root_logger.addHandler(console_handler)
+
+
+def get_logger(name: str) -> logging.Logger:
+    """Get a named logger"""
+    return logging.getLogger(name)

whisper_singleton.py

@@ -0,0 +1,47 @@
+from utils.central_logging import get_logger
+from langchain_openai import OpenAIEmbeddings
+from pathlib import Path
+import whisper
+import threading
+import os 
+
+
+logger =  get_logger("whisper")
+
+_whisper_model = None
+_lock = threading.Lock()
+_embedding = None
+_embedding_lock = threading.Lock()
+
+def get_whisper():
+    global _whisper_model 
+
+    if _whisper_model is  None:
+        with _lock:
+            if _whisper_model is None:
+                _whisper_model = whisper.load_model("base")
+                logger.info("Whisper model has been loaded")
+    return _whisper_model
+
+def get_embedding():
+    global _embedding
+
+    if _embedding is None:
+        with _embedding_lock:
+            if _embedding is None:
+                _embedding = OpenAIEmbeddings(model="text-embedding-ada-002")
+                logger.info("Openai  embedding has been initialized")
+    return _embedding
+
+
+def transcribe_content(url_path:str) -> str:
+    safe_path = Path(url_path).resolve().as_posix()
+    model = get_whisper()
+    result = model.transcribe(url_path)
+    return  result["text"]
+
+
+def save_file(file_name,result):
+    with open(file_name,'w') as file:
+        file.write(result)
+