add comment for #4407 and remove seemingly unnecessary cudnn.enabled

modules/devices.py

@@ -53,9 +53,11 @@ def torch_gc():
 
 def enable_tf32():
     if torch.cuda.is_available():
+
+        # enabling benchmark option seems to enable a range of cards to do fp16 when they otherwise can't
+        # see https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/4407
         if any([torch.cuda.get_device_capability(devid) == (7, 5) for devid in range(0, torch.cuda.device_count())]):
             torch.backends.cudnn.benchmark = True
-            torch.backends.cudnn.enabled = True
 
         torch.backends.cuda.matmul.allow_tf32 = True
         torch.backends.cudnn.allow_tf32 = True