Add readme

Author: adribaeza

.gitignore

@@ -2,4 +2,10 @@
 **/__pycache__/
 
 # pytest cache
-**/.pytest_cache/
+**/.pytest_cache/
+
+# Exclude .coverage files
+**/.coverage
+
+# Exclude .vscode files
+.vscode/

README.md

@@ -1,6 +1,6 @@
 '''
-#####################  FastAPI + TinyLlama + Docker #########################################
-Autor: Adrián Baeza Prieto
+#####################  FastAPI + TinyLlama Backend API #########################################
+Author: Adrián Baeza Prieto
 Github: @adribaeza
 Python 3.10+
 '''
@@ -74,7 +74,6 @@ async def get_current_user(token: str = Depends(oauth2_scheme)):
 # Load the model with the TinyLlama model
 pipe = pipeline("text-generation", model=LLM_MODEL, torch_dtype=torch.bfloat16, device_map="auto")
 
-
 class Message(BaseModel):
     role: str
     content: str
@@ -158,9 +157,4 @@ async def chat(request: ChatRequest, user: dict = Depends(get_current_user)):
         raise HTTPException(status_code=500, detail="Internal Server Error")
 
 # Include main router in the API
-api.include_router(api_router)
-
-# Execute the API with Uvicorn only if the script is executed directly in the local environment
-#if __name__ == '__main__':
-#    import uvicorn
-#    uvicorn.run(api)
+api.include_router(api_router)

backend/api/main.py

@@ -9,5 +9,4 @@ spec:
     - protocol: TCP
       port: 8000
       targetPort: 8000
-      nodePort: 30000  # Choose between 30000-32767
-  type: NodePort
+  type: LoadBalancer

backend/k8s/service.yaml

@@ -1,30 +1,16 @@
+'''
+#####################  Backend Unit Testing  #########################################
+Author: Adrián Baeza Prieto
+Github: @adribaeza
+Python 3.10+
+'''
 import logging
 from fastapi import HTTPException
 import pytest
-#from fastapi.testclient import TestClient
-#need import api and static token from main.py, need upload main.py to the test folder
-from backend.api.main import api, STATIC_TOKEN, verify_token
+from backend.api.main import STATIC_TOKEN, verify_token
 
-# Configurar logging
 logging.basicConfig(level=logging.DEBUG)
-# Configurar el cliente de prueba
-#client = TestClient(api)
 
-#def test_verify_token():
-#    logging.debug("Init test_verify_token")
-#    # Test with valid token
-#    logging.debug("Test with valid token")
-#    response = client.post("/api/v1/chat", headers={"Authorization": f"Bearer {STATIC_TOKEN}"}, json={"text": "Hello"})
-#    assert response.status_code == 200
-
-    # Test with invalid token
-#    logging.debug("Test with invalid token")
-#    response = client.post("/api/v1/chat", headers={"Authorization": "Bearer invalidtoken"}, json={"text": "Hello"})
-#    assert response.status_code == 401
-#    assert response.json() == {"detail": "Invalid authentication credentials"}
-
-
-#create test to check function verify_token in main.py without client
 def test_verify_token():
     logging.debug("Test with valid token")
     valid_token = STATIC_TOKEN

backend/tests/test_main.py

@@ -1,3 +1,9 @@
+'''
+#####################  Streamlit Chat With LLM Model   #########################################
+Author: Adrián Baeza Prieto
+Github: @adribaeza
+Python 3.10+
+'''
 import streamlit as st
 import requests, logging, os
 from dotenv import load_dotenv
@@ -20,30 +26,29 @@
 DEFAULT_TOP_K = 50
 DEFAULT_TOP_P = 0.9
 
-
-# Función para limpiar el historial de mensajes
+# Function to clear the chat history
 def clear_chat():
     st.session_state.messages = []
 
 def main():
 
-    # Configuración de la página
+    # Page configuration
     st.set_page_config(
         page_title="Chat with TinyLlama",
-        page_icon=":robot_face:",  # Puedes usar un emoji o una URL a un favicon específico
+        page_icon=":robot_face:",
         layout="centered",
         initial_sidebar_state="auto",
     )
 
-    # Configuración de la interfaz
+    # Interface title
     st.title("Chat with TinnyLLama LLM model")
     st.write("Simple chat interface to interact with TinyLlama LLM model")
 
-    # Añadir un botón para iniciar un nuevo chat
+    # Add a button to clear the chat history
     if st.button("➕ New Chat", help="Click to start a new chat and clear the current conversation history"):
         clear_chat()
 
-    # Additional params with help text
+    # Additional params with help text to adjust the LLM model behavior
     with st.expander("Config params", expanded=False):
         max_new_tokens = st.number_input(
             "Max New Tokens", 
@@ -75,28 +80,28 @@ def main():
             help="The cumulative probability of parameter highest probability vocabulary tokens to keep for nucleus sampling."
         )
 
-
+    # Check if the session state has the messages attribute to initialize it
     if "messages" not in st.session_state:
         st.session_state.messages = []
 
+    # Iterate over the messages in the session state to display them in the chat
     for message in st.session_state.messages:
         with st.chat_message(message["role"]):
             st.markdown(message["content"])
 
-
+    # Add a chat input to interact with the assistant
     if prompt := st.chat_input("What is up?"):
+
+        # Add the user message to the chat history
         st.session_state.messages.append({"role": "user", "content": prompt})
         with st.chat_message("user"):
             st.markdown(prompt)
-        
         headers = {
             "Authorization": f"Bearer {STATIC_TOKEN}",
             "Content-Type": "application/json"
         }
-        # Construir el historial de la conversación
+        # Build the data payload for the API request
         conversation_history = [{"role": msg["role"], "content": msg["content"]} for msg in st.session_state.messages]
-    
-
         data = {
             "messages": conversation_history,
             "max_new_tokens": max_new_tokens,
@@ -106,6 +111,8 @@ def main():
             "top_p": top_p
         }
         logging.info(f"Request data: {data}")
+
+        # Make a request to the API
         try:
             with st.spinner("The assistant is thinking..."):
                 response = requests.post("http://host.docker.internal:8000/api/v1/chat", headers=headers, json=data)
@@ -123,7 +130,7 @@ def main():
             st.error("Failed to connect to the API")
             logging.error(f"Failed to connect to the API: {e}")
 
-    # Añadir un footer con el texto deseado
+    # Add a footer with the app information
     st.markdown(
         """
         <style>
@@ -145,15 +152,7 @@ def main():
         unsafe_allow_html=True
     )
 
+# Run the main function
 if __name__ == "__main__":
     main()
-
-#'''
-####  Run the Streamlit app
-#To run the Streamlit app, execute the following command in the terminal:
-#    
-#    ```bash
-#    streamlit run frontend/app/main.py
-#    ```
-#'''
 

frontend/Readme.md

@@ -1,4 +1,9 @@
-# test_main.py
+'''
+#####################  Frontend Unit Testing  #########################################
+Author: Adrián Baeza Prieto
+Github: @adribaeza
+Python 3.10+
+'''
 import unittest
 from unittest.mock import patch, MagicMock
 from frontend.app.main import clear_chat