py/api.py

@@ -4,6 +4,7 @@ from transformers import AutoTokenizer, LlamaForCausalLM
 
 
 class API:
+    # This method processes a message via transformers. (NOT FINISHED!)
     @staticmethod
     def process_text_transformers(prompt, model):
         model = LlamaForCausalLM.from_pretrained(model)
@@ -14,6 +15,7 @@ class API:
         generate_ids = model.generate(inputs.input_ids, max_length=30)
         return tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
 
+    # This method processes a message via ollama
     @staticmethod
     def process_text_local(prompt, model):
         ollama_url = "http://localhost:11434"
@@ -34,6 +36,7 @@ class API:
         else:
             return "Error: " + response.text
 
+    # This method sends a message to a certain AI.
     def send_message(self, message, model):
         if model == 1:
             answer = self.process_text_local(message, "phi3.5")

py/templates/index.html

@@ -71,19 +71,22 @@
         <!-- Output section: Simulating a conversation with AI -->
         <div class="output">
             <div class="conversation">
-                <div class="user-message">User: What is the weather today?</div>
-                <div class="ai-message">AI: It's sunny with a slight breeze.</div>
-                <div class="user-message">User: Great! Thank you!</div>
-                <div class="ai-message">AI: You're welcome!</div>
+                {% for message in messages %}
+                    {% if message.startswith('User:') %}
+                        <div class="user-message">{{ message }}</div>
+                    {% else %}
+                        <div class="ai-message">{{ message }}</div>
+                    {% endif %}
+                {% endfor %}
             </div>
         </div>
         
         <!-- Input section: Where user input is provided -->
-        <div class="input">
-            <input type="text" placeholder="Type your message here..." />
-            <button><img src="/static/img/microphone.svg" alt="microphone"></button>
-            <button>Send</button>
-        </div>
+        <form class="input" method="POST" action="">
+            <input type="text" name="user_message" placeholder="Type your message here..." />
+            <button type="submit" name="option" value="voice"><img src="/static/img/microphone.svg" alt="microphone"></button>
+            <button type="submit" name="option" value="chat">Send</button>
+        </form>
     </div>
 </body>
 

py/venv.sh

@@ -5,3 +5,6 @@ source venv/bin/activate
 pip install transformers
 pip install torch
 pip install flask
+pip install SpeechRecognition
+pip install pyaudio
+pip install pocketsphinx

py/web_flask.py

@@ -1,14 +1,35 @@
-from flask import Flask, send_from_directory
+from flask import Flask, request, render_template
+from api import API
+from voice_recognition import Voice
+
+APP = Flask(__name__)
+api = API()
+voice = Voice()
+
+messages = []
 
 
-class WebHost:
-    @staticmethod
-    def main_page():
-        app = Flask(__name__)
+# The following method shows the user the GUI and does the backend connections to the API.
+@APP.route('/', methods=['GET', 'POST'])
+def index():
+    global messages
 
-        @app.route('/')
-        def index():
-            return app.send_static_file('index.html')
+    if request.method == 'POST':
+        option = request.form['option']
 
-        if __name__ == '__main__':
-            app.run(debug=True)
+        user_message = request.form['user_message']
+
+        if option == "voice":
+            user_message = voice.listen()
+            messages.append(f"User: {user_message}")
+        elif option == "chat":
+            messages.append(f"User: {user_message}")
+
+        ai_response = "AI: " + api.send_message(user_message, 1)
+        messages.append(ai_response)
+
+    return render_template('index.html', messages=messages)
+
+
+if __name__ == '__main__':
+    APP.run(debug=True)