Upgrading the Ollama Hosted Phi3 Offline Chat Program

 In my previous post, "Chat with Ollama Phi3 in Python Offline," I introduced a basic program that allowed you to chat with Ollama's Phi3 language model offline using Python. Today, I’m excited to share an upgraded version of this program. This new version incorporates a graphical user interface (GUI) using PyQt6 and includes a text-to-speech (TTS) engine to read aloud Phi3's responses. The chat logs are displayed in a QTextEdit widget, making the interaction more user-friendly and accessible. This program is created in Python 3.10 same experience I encountered when I first create a chat program using Langchain and Openai. also a good pc/mac will deliver a better experience but it will still run on a Intel M3-100y but it will be very slow considering that I have downloaded the phi3 4b  quantized version.

Upgraded Features

1. Graphical User Interface (GUI): A chat window created using PyQt6.
2. Text-to-Speech: Utilizes pyttsx3 to read aloud Phi3's responses.
3. Chat Logs: Displays the conversation history in a QTextEdit widget.

Code Explanation

Below is the complete code for the upgraded program, followed by detailed explanations of each part.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

import pyttsx3 from PyQt6.QtGui import * from PyQt6.QtCore import * from PyQt6.QtWidgets import * from langchain_community.llms import Ollama class Window(QMainWindow): def __init__(self): super(Window, self).__init__() # Set up the layout layout = QVBoxLayout() hlayout = QHBoxLayout() central_widget = QWidget() central_widget.setLayout(layout) self.setCentralWidget(central_widget) # Initialize the language model self.llm = Ollama(model="phi3") # Add widgets to the layout layout.addLayout(hlayout) self.te = QLineEdit(self) self.le = QTextEdit(self) self.le.setReadOnly(True) self.btnStart = QPushButton("Ask", self) hlayout.addWidget(self.te) hlayout.addWidget(self.btnStart) layout.addWidget(self.le) # Connect button click to function self.btnStart.clicked.connect(self.onClickedStart) # Set window properties self.setGeometry(25, 45, 900, 500) self.setWindowTitle('Speak') def onClickedStart(self): global text # Get user input and display it in the chat log user_input = self.te.text() self.le.append(f'You: {user_input}') # Get response from the language model response = self.llm.invoke(user_input) # Display the response in the chat log self.le.append(f'Phi3: {response}') # Prepare text for text-to-speech text = response self.worker = WorkerThread() self.worker.start() text = '' engine = [] error = '' i = 0 class WorkerThread(QThread): def run(self): global text, engine, error, i try: i += 1 engine.append(pyttsx3.init()) engine[i-1].setProperty('rate', 120) engine[i-1].say(text) engine[i-1].startLoop() except Exception as err: error = str(err) if __name__ == "__main__": import sys app = QApplication(sys.argv) w = Window() w.show() sys.exit(app.exec())

Detailed Explanation

Imports

• pyttsx3: A text-to-speech conversion library in Python.
• PyQt6: A set of Python bindings for the Qt application framework, used to create the GUI.
• Ollama: A module from the langchain_community package for interacting with the Phi3 language model.

Window Class

• Initialization (__init__ method):

  • Sets up the main layout and widgets (QLineEdit, QTextEdit, QPushButton).
  • Initializes the Phi3 language model using Ollama.
  • Connects the "Ask" button to the onClickedStart method.
  • Configures the window size and title.

• onClickedStart Method:

  • Gets the user's input from the QLineEdit widget.
  • Appends the user's input to the QTextEdit widget for display.
  • Sends the input to the Phi3 language model and gets the response.
  • Appends the response to the QTextEdit widget.
  • Initiates a text-to-speech conversion of the response using the WorkerThread class.

WorkerThread Class

• A QThread subclass that handles the text-to-speech functionality.
• run Method:
  • Initializes a new pyttsx3 engine for each response.
  • Sets the speech rate and converts the response text to speech.
  • Starts the speech engine loop to read the response aloud.

Main Block

• Creates a QApplication instance and a Window instance.
• Displays the Window and starts the event loop.

Conclusion

This upgraded version of the Ollama Phi3 offline chat program provides a more interactive and user-friendly experience. The integration of a GUI and text-to-speech functionality makes it easier to interact with the Phi3 model without needing to modify the underlying code. Give it a try and see how these enhancements improve your interactions with Phi3!