Update app.py

app.py

@@ -50,13 +50,17 @@ def load_model():
         snapshot_download(
             repo_id=model_name,
             use_auth_token=hf_token,
-            allow_patterns=["config.json", "*.safetensors", "model.safetensors.index.json"],
-            ignore_patterns=["optimizer.pt", "pytorch_model.bin", "training_args.bin", "scheduler.pt", "tokenizer.*"]
+            allow_patterns=["config.json", "*.safetensors", "model.safetensors.index.json", "vocab.json", "merges.txt", "tokenizer.json"],
+            ignore_patterns=["optimizer.pt", "pytorch_model.bin", "training_args.bin", "scheduler.pt"]
         )
 
         model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16)
         model.to(device)
+        
         tokenizer = AutoTokenizer.from_pretrained(model_name)
+        if tokenizer is None:
+            raise ValueError("Failed to load tokenizer")
+        
         logger.info(f"Orpheus model and tokenizer loaded to {device}")
     except Exception as e:
         logger.error(f"Error loading model: {str(e)}")