Update app.py

app.py

@@ -1,38 +1,56 @@
 import gradio as gr
 from transformers import pipeline, set_seed
 
-# Load Hugging Face model (adjust as needed)
-generator = pipeline("text-generation", model="gpt2")
+# Preload models into a dictionary for selection
+AVAILABLE_MODELS = {
+    "GPT-2 (small, fast)": "gpt2",
+    "Falcon (TII UAE)": "tiiuae/falcon-7b-instruct",
+    "Mistral (OpenAccess)": "mistralai/Mistral-7B-v0.1"
+}
+
 set_seed(42)
 
-chat_history = []
+# Cache loaded models to avoid reloading
+model_cache = {}
+
+def load_model(model_name):
+    if model_name not in model_cache:
+        model_id = AVAILABLE_MODELS[model_name]
+        model_cache[model_name] = pipeline("text-generation", model=model_id)
+    return model_cache[model_name]
+
+# Chat history buffer
+chat_memory = {}
+
+# Terminal response function
+def codette_terminal(user_input, model_name, session_id):
+    if session_id not in chat_memory:
+        chat_memory[session_id] = []
 
-# Text generation function
-def codette_terminal(user_input):
-    global chat_history
-    if user_input.lower() in ["exit", "quit"]:
-        chat_history = []
-        return "🧠 Codette signing off. Type again to restart."
+    if user_input.lower() in ['exit', 'quit']:
+        chat_memory[session_id] = []
+        return "🧠 Codette signing off... Session reset."
 
-    output = generator(user_input, max_length=100, num_return_sequences=1)
+    generator = load_model(model_name)
+    output = generator(user_input, max_length=100, num_return_sequences=1, do_sample=True)
     response = output[0]['generated_text'].strip()
 
-    # Update terminal-style chat log
-    chat_history.append(f"🖋️ You > {user_input}")
-    chat_history.append(f"🧠 Codette > {response}")
-    return "\n".join(chat_history[-10:])  # Keep last 10 entries for brevity
+    chat_memory[session_id].append(f"🖋️ You > {user_input}")
+    chat_memory[session_id].append(f"🧠 Codette > {response}")
+    return "\n".join(chat_memory[session_id][-10:])  # limit memory output
 
-# Gradio Interface
-with gr.Blocks(title="Codette Terminal") as demo:
-    gr.Markdown("## 🧬 Codette Terminal Interface (Hugging Face Edition)")
-    gr.Markdown("Type your message below. Type `'exit'` to reset conversation.\n")
+# UI
+with gr.Blocks(title="Codette Terminal - Multi-Model AI") as demo:
+    gr.Markdown("## 🧬 Codette Terminal (Multi-Model, Hugging Face Edition)")
+    gr.Markdown("Type something below. Use the dropdown to switch models. Type `'exit'` to reset session.")
 
-    with gr.Row():
-        input_box = gr.Textbox(label="Your input", placeholder="Ask me anything...", lines=1)
-    output_box = gr.Textbox(label="Codette Output", lines=15, interactive=False)
+    session_id = gr.Textbox(label="Session ID (hidden)", value="session_default", visible=False)
+    model_select = gr.Dropdown(choices=list(AVAILABLE_MODELS.keys()), value="GPT-2 (small, fast)", label="Choose AI Model")
+    user_input = gr.Textbox(label="Your Prompt", placeholder="Ask me anything...")
+    output_box = gr.Textbox(label="Terminal Output", lines=20, interactive=False)
 
-    input_box.submit(fn=codette_terminal, inputs=input_box, outputs=output_box)
+    user_input.submit(fn=codette_terminal, inputs=[user_input, model_select, session_id], outputs=output_box)
 
-# Launch in HF Space
+# For Hugging Face Spaces
 if __name__ == "__main__":
     demo.launch()