Update app.py

app.py

@@ -34,6 +34,7 @@ def start_worker(model_path: str, bits=16):
     assert bits in [4, 8, 16], "It can be only loaded with 16-bit, 8-bit, and 4-bit."
     if bits != 16:
         model_name += f"-{bits}bit"
+    model_name += "-lora"
     worker_command = [
         sys.executable,
         "-m",
@@ -46,6 +47,8 @@ def start_worker(model_path: str, bits=16):
         model_path,
         "--model-name",
         model_name,
+        "--model-base",
+        "llava-hf/llava-1.5-7b-hf"
         "--use-flash-attn",
     ]
     if bits != 16: