space update

app.py

@@ -28,6 +28,7 @@ def load_model():
             context_length=context_length, 
             num_layers=num_layers
         )        
+        
         # Load the trained weights if available
         model_path = "v1/u_model.pth"
 
@@ -37,18 +38,29 @@ def load_model():
             try:
                 print("📥 Downloading model weights from GitHub...")
                 import requests
-                url = "https://github.com/malibayram/llm-from-scratch/raw/main/u_model.pth"
-                response = requests.get(url)
+                url = "https://github.com/malibayram/llm-from-scratch/raw/main/v1/u_model.pth"
+                
+                headers = {
+                    'Accept': 'application/octet-stream',
+                    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
+                }
+                
+                response = requests.get(url, headers=headers)
                 response.raise_for_status()  # Raise an exception for bad status codes
-                print("✅ Model weights downloaded successfully!")
+                
+                # Check if we got a proper binary file (PyTorch files start with specific bytes)
+                if response.content[:4] != b'PK\x03\x04' and b'<html' in response.content[:100].lower():
+                    raise Exception("Downloaded HTML instead of binary file - check URL")
+                
+                print(f"📦 Downloaded {len(response.content)} bytes")
                 
                 # Create v1 directory if it doesn't exist
                 os.makedirs("v1", exist_ok=True)
-
+                
                 # Save the model weights to the local file system
                 with open(model_path, "wb") as f:
                     f.write(response.content)
-                print("✅ Model weights downloaded successfully!")
+                print("✅ Model weights saved successfully!")
             except Exception as e:
                 print(f"❌ Failed to download model weights: {e}")
                 print("Using random initialization.")
@@ -153,14 +165,7 @@ demo = gr.ChatInterface(
         ),
     ],
     title="🤖 Usta Model Chat",
-    description="Chat with a custom transformer language model built from scratch! This model specializes in geographical knowledge including countries, capitals, and cities.",
-    examples=[
-        "the capital of france",
-        "tell me about spain", 
-        "what is the capital of united states",
-        "paris is in",
-        "germany and its capital"
-    ]
+    description="Chat with a custom transformer language model built from scratch! This model specializes in geographical knowledge including countries, capitals, and cities."
 )
 
 if __name__ == "__main__":