lint

bitsandbytes/nn/modules.py

@@ -360,12 +360,12 @@ class Params4bit(torch.nn.Parameter):
     def __torch_function__(cls, func, types, args=(), kwargs=None):
         if kwargs is None:
             kwargs = {}
-    
+
         if func in [torch.chunk, torch.split]:
             tensor = args[0]
-            
+
             result = super().__torch_function__(func, types, args, kwargs)
-    
+
             if isinstance(result, tuple):
                 return tuple(
                     cls(
@@ -393,9 +393,9 @@ class Params4bit(torch.nn.Parameter):
                     module=tensor.module,
                     bnb_quantized=tensor.bnb_quantized,
                 )
-        
+
         return super().__torch_function__(func, types, args, kwargs)
-        
+
 
 def fix_4bit_weight_quant_state_from_module(module: Union["Embedding4bit", "Linear4bit"]):
     if getattr(module.weight, "quant_state", None) is not None: