From d53d2bcc45649f4239c9c7f9d1a302a6a7d74aff Mon Sep 17 00:00:00 2001
From: somebody <onesome01@protonmail.com>
Date: Sat, 25 Feb 2023 18:19:46 -0600
Subject: [PATCH] Model: Fix crash on full GPU load

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

diff --git a/model.py b/model.py
index 2cf50c0c..600c646c 100644
--- a/model.py
+++ b/model.py
@@ -2274,7 +2274,7 @@ class GenericHFTorchInferenceModel(HFTorchInferenceModel):
         if utils.koboldai_vars.hascuda:
             if utils.koboldai_vars.usegpu:
                 # Use just VRAM
-                model = model.half().to(utils.koboldai_vars.gpu_device)
+                self.model = self.model.half().to(utils.koboldai_vars.gpu_device)
             elif utils.koboldai_vars.breakmodel:
                 # Use both RAM and VRAM (breakmodel)
                 if not utils.koboldai_vars.lazy_load: