From d6791790ed2e967f968b528e3d7660acc53fa940 Mon Sep 17 00:00:00 2001
From: John Smith <yfshi123@163.com>
Date: Wed, 26 Apr 2023 13:13:54 +0800
Subject: [PATCH] fix bug

---
 server/server.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/server/server.py b/server/server.py
index ef3dae8..de40fc5 100644
--- a/server/server.py
+++ b/server/server.py
@@ -88,7 +88,7 @@ class ModelServer:
             print('Quantized attention applied.')
 
             if self.lora_path is not None:
-                inject_lora_layers(model, self.lora_path, device='cuda', torch_dtype=torch.float16)
+                inject_lora_layers(model, self.lora_path, device='cuda', dtype=torch.float16)
         
         self.model, self.tokenizer = model, tokenizer
         print("Loaded in {:.2f} seconds.".format(time.time() - t0))