Update app.py

app.py

@@ -1,9 +1,28 @@
+import os
+import requests
 import gradio as gr
 from llama_cpp import Llama
 
-model_path = "https://huggingface.co/vislupus/bulgarian-joke-master-SmolLM2-135M-Instruct-bnb-4bit-gguf/blob/main/unsloth.Q4_K_M.gguf"
-llm = Llama(model_path=model_path)
+# Download the model if not present
+def download_model(model_url, local_path):
+    if not os.path.exists(local_path):
+        print(f"Downloading model from {model_url}...")
+        response = requests.get(model_url)
+        with open(local_path, 'wb') as f:
+            f.write(response.content)
+        print("Model downloaded.")
+    else:
+        print("Model already exists.")
 
+model_url = "https://huggingface.co/vislupus/bulgarian-joke-master-SmolLM2-135M-Instruct-bnb-4bit-gguf/blob/main/unsloth.Q4_K_M.gguf"
+local_model_path = "model.gguf"
+
+download_model(model_url, local_model_path)
+
+# Load the model
+llm = Llama(model_path=local_model_path)
+
+# Gradio interface
 def generate_response(prompt):
     """Generate a response using the GGUF model."""
     output = llm(prompt, max_tokens=200)
@@ -21,3 +40,4 @@ with gr.Blocks() as demo:
     submit_button.click(fn=generate_response, inputs=prompt_input, outputs=output_box)
 
 demo.launch()
+