fix cast fp8 (#140)

Author: tenderness-git

diffsynth_engine/pipelines/base.py

@@ -229,7 +229,7 @@ def eval(self):
                 model.eval()
         return self
 
-    def enable_cpu_offload(self, offload_mode: str | None, offload_to_disk:bool = False):
+    def enable_cpu_offload(self, offload_mode: str | None, offload_to_disk: bool = False):
         valid_offload_mode = ("cpu_offload", "sequential_cpu_offload", "disable", None)
         if offload_mode not in valid_offload_mode:
             raise ValueError(f"offload_mode must be one of {valid_offload_mode}, but got {offload_mode}")
@@ -244,8 +244,7 @@ def enable_cpu_offload(self, offload_mode: str | None, offload_to_disk:bool = Fa
             self._enable_sequential_cpu_offload()
         self.offload_to_disk = offload_to_disk
 
-
-    def _enable_model_cpu_offload(self):        
+    def _enable_model_cpu_offload(self):
         for model_name in self.model_names:
             model = getattr(self, model_name)
             if model is not None:
@@ -258,23 +257,22 @@ def _enable_sequential_cpu_offload(self):
             if model is not None:
                 enable_sequential_cpu_offload(model, self.device)
         self.offload_mode = "sequential_cpu_offload"
-    
+
     def _disable_offload(self):
-        self.offload_mode = None                
-        self._offload_param_dict = {}        
+        self.offload_mode = None
+        self._offload_param_dict = {}
         for model_name in self.model_names:
             model = getattr(self, model_name)
             if model is not None:
                 model.to(self.device)
 
-
     def enable_fp8_autocast(
         self, model_names: List[str], compute_dtype: torch.dtype = torch.bfloat16, use_fp8_linear: bool = False
     ):
         for model_name in model_names:
             model = getattr(self, model_name)
             if model is not None:
-                model.to(device=self.device, dtype=torch.float8_e4m3fn)                
+                model.to(dtype=torch.float8_e4m3fn)
                 enable_fp8_autocast(model, compute_dtype, use_fp8_linear)
         self.fp8_autocast_enabled = True
 
@@ -298,15 +296,17 @@ def load_models_to_device(self, load_model_names: List[str] | None = None):
         for model_name in load_model_names:
             model = getattr(self, model_name)
             if model is None:
-                raise ValueError(f"model {model_name} is not loaded, maybe this model has been destroyed by model_lifecycle_finish function with offload_to_disk=True")
+                raise ValueError(
+                    f"model {model_name} is not loaded, maybe this model has been destroyed by model_lifecycle_finish function with offload_to_disk=True"
+                )
             if model is not None and (p := next(model.parameters(), None)) is not None and p.device.type != self.device:
                 model.to(self.device)
         # fresh the cuda cache
         empty_cache()
 
     def model_lifecycle_finish(self, model_names: List[str] | None = None):
         if not self.offload_to_disk or self.offload_mode is None:
-            return 
+            return
         for model_name in model_names:
             model = getattr(self, model_name)
             del model
@@ -316,7 +316,6 @@ def model_lifecycle_finish(self, model_names: List[str] | None = None):
             print(f"model {model_name} has been deleted from memory")
             logger.info(f"model {model_name} has been deleted from memory")
             empty_cache()
-        
-        
+
     def compile(self):
         raise NotImplementedError(f"{self.__class__.__name__} does not support compile")