Update app.py

app.py

@@ -1,13 +1,17 @@
 import gradio as gr
 from llama_cpp import Llama
+from huggingface_hub import hf_hub_download
 
-# Path to your GGUF model inside the space
-MODEL_PATH = "Futuresony/gemma2-2b-gguf-q4_k_m"
+# Download GGUF model from Hugging Face Hub
+MODEL_REPO = "Futuresony/gemma2-2b-gguf-q4_k_m"
+MODEL_FILENAME = "gemma-2b-it-q4_k_m.gguf"  # Or check exact filename on the repo
+
+model_path = hf_hub_download(repo_id=MODEL_REPO, filename=MODEL_FILENAME)
 
 # Load model
-llm = Llama(model_path=MODEL_PATH, n_ctx=2048, n_threads=4, verbose=True)
+llm = Llama(model_path=model_path, n_ctx=2048, n_threads=4, verbose=True)
 
-# Function to format the prompt
+# Format prompt as Alpaca-style
 def format_prompt(user_message):
     return f"""### Instruction:
 {user_message}
@@ -24,7 +28,7 @@ def respond(user_message, chat_history):
 
 # Gradio UI
 with gr.Blocks() as demo:
-    gr.Markdown("## 🤖 DStv AI Assistant (Offline - GGUF)")
+    gr.Markdown("## 🤖 DStv AI Assistant - Powered by Gemma 2B GGUF")
     chatbot = gr.Chatbot()
     msg = gr.Textbox(placeholder="Ask your question...")
     state = gr.State([])
@@ -32,4 +36,3 @@ with gr.Blocks() as demo:
     msg.submit(respond, [msg, state], [msg, chatbot])
 
 demo.launch()
-